The post Premier League Draws appeared first on The Church of Betting.
]]>A draw betting strategy has been making the rounds after having been covered by David Sumpter in his book Soccermatics: Mathematical Adventures in the Beautiful Game. The strategy resonated around the betting world and the book author has recently published an article about it on Pinnacle. So what exactly is this money maker? The strategy suggests backing the draw in games between equally matched teams in Premier League.
It turns out for that for years in all tiers of English football the market was pricing draws in games between equally matched teams too generously. In contrast, draws in games with large difference between the strengths of the opponents were overpriced. The distribution (from the article above) looked something like this:
The founder of this pricing anomaly Mr Sumpter has decided to place some bets on it achieving a pretty decent return…
…which convinced him to write a book and tell the world about it. An edge shared is an edge halved I hear you say. Surely, Mr Sumpter has been having second thoughts regarding his decision as he himself admits in the article. In fact, since the book has been published (May 5, 2016) the edge has not only been halved but has completely disappeared, as the article reports. Or has it really?
I decided to check that for myself. Having collected the data from the great football-data.co.uk I have analysed the profitability of backing draws in all tiers of English football using Pinnacle closing prices since Season 2012/13 (the first one with Pinnacle closing prices on record). Here is the season-long profitability of a strategy of backing draws for all games in all leagues (blue line) compared to backing the draw where the difference between the winning probabilities of the two teams pre-game was less than 10%, 7% and 5% respectively:
The sample is indeed short and the returns volatile but one does see a downward trend for the betting strategy, whereas the blue line (all games) remains relatively flat. The premium of the strategy seems to have disappeared as we see the rest of the lines closing on and crossing the blue one in season 2017/18. That seems consistent with the findings of Mr Sumpter from his article and supports the intuitive notion that an edge so popular would have by now disappeared.
It makes sense to double check that by having a look at the opposite strategy, namely backing the draw in games with clear favourite and outsider. According to Soccermatics the draws for this subset were previously overpriced. I have compared the profitability of all draws with draws for games where the pre-game winning probability difference between the teams was at least 30%, 40% and 50%.
The results are consistent with the earlier ones in that the draws for games with clear favourites seem to have been underperforming all draws for some time. Later on the rest of the lines caught up with the blue one and in season 2017/2018 actually delivered a premium and a profit. It is perhaps too early to conclude that the trend has reversed but that is certainly worth further investigation.
Before closing the case I decided to have a look at the results per league. After all the original strategy has focused on the Premier League in particular. I was surprised to find out that the Premier League still seems to deliver a premium for backing draws of evenly matched teams. The graph for backing draws between evenly matched teams for the Premier League alone looks like this:
Here again there seems to be a downward trend, however for most seasons the lines (excluding the blue one) remain in the profit zone. The premium compared to the blue line does not seem to have disappeared. Backing the draw for evenly matched teams in the Premier League continues to be a profitable strategy. Here are the numbers for backing the draw where there is less than 10% difference in winning probabilities (with level stakes):
Season | Bets | Implied Probability | Actual Probability | Total Return | Edge | Delta Edge (comp. to base case) |
2017/18 | 24 | 31.04% | 37.50% | 5.46 | 6.46% | 7.43% |
2016/17 | 60 | 30.45% | 36.67% | 11.59 | 6.21% | 8.97% |
2015/16 | 79 | 29.47% | 34.18% | 13.27 | 4.70% | 2.60% |
2014/15 | 68 | 30.01% | 32.35% | 5.31 | 2.34% | 3.62% |
2013/14 | 52 | 29.74% | 36.54% | 12.51 | 6.80% | 11.29% |
2012/13 | 58 | 29.43% | 39.66% | 20.52 | 10.22% | 7.31% |
How does the graph look like for the opposite strategy – backing the draw for games with clear favourites in the Premier League?
Not too optimistic, since the gap between the games with favourites and all games seems to have closed. Yet, compared to the graph for all leagues the trend at least has not reversed but at most evened up with the general case. This might be attributed to the larger initial gap as compared to the all-leagues-graph.
By the way, I have checked other leagues as well for the same trend and have discovered similar (although less pronounced) results for La Liga (again, level stakes, <10% difference in winning probability):
Season | Bets | Implied Probability | Actual Probability | Total Return | Edge | Delta Edge (comp. to base case) |
2017/18 | 28 | 29.93% | 21.43% | -8.01 | -8.50% | -7.37% |
2016/17 | 64 | 30.21% | 34.38% | 9.31 | 4.17% | 5.05% |
2015/16 | 56 | 29.76% | 30.36% | 1.52 | 0.60% | 0.82% |
2014/15 | 67 | 30.05% | 34.33% | 9.62 | 4.28% | 4.53% |
2013/14 | 58 | 29.47% | 34.48% | 9.62 | 5.02% | 6.35% |
2012/13 | 56 | 28.61% | 33.93% | 10.09 | 5.32% | 8.05% |
Here too you will find steady premiums for backing the draw between equally matched teams as compared to all draws. The current year is a notable exception, however this can be attributed to the abnormally low frequency of draws in this bracket compared to previous seasons so reversion to the mean should be expected.
The analysis did also find out that in the second tiers of Spanish and Italian football as well as in Eredivisie backing the draw generally appeared to be a profitable strategy for the last 6 seasons. Serie B also shows some premium for backing the draw between evenly matched teams but it can mostly be attributed to the last season, which had too few such games so far so no conclusions can be drawn here.
To summarize, I do believe that backing the draw for Premier League games between equally matched teams remains a strategy with strong potential. Yet caution is warranted, since the counter-strategy is pretty much in line with the general draw profitability. Furthermore it doesn’t make intuitive sense that the bias has disappeared everywhere but in the Premier League. The Premier League is supposed to be the market with the highest volume and the smartest money in it so if anything you would expect the least market inefficiencies to be found there. And finally, we shouldn’t forget that the edge has been shared within a well-selling book so we shouldn’t expect whatever is left from it to be around much longer.
However, this surely continues to look like a profitable strategy so with the above concerns in mind I am currently putting small money on backing the draw on games where the difference of winning probability between the two teams is less than 10% in the Premier League. If you are interested in how the strategy is going just send me an email anytime during the season and I’ll let you know.
Some news from the betting services I’m partnering with:
RebelBetting has introduced a one-week trial for their arbitrage software. Try it out for free and if you are not happy with the product, just unsubscribe before the trial is over at no cost. There is not much to lose!
Furthermore, PremiumTradings is offering free deposits with Skrill, Neteller and Ecopayz for the month of December so you still have a few days left to make use of this offer.
Finally, I want to wish to all of you a Merry Christmas and a Happy New Year. 2017 has been a great year for me and I am looking back at it with a smile. A big Thank You to everyone for the nice mails and comments you have sent me throughout the year or even just for sticking around and reading my articles – this is what makes the experience of running The Church of Betting worth it. I hope you all have a great holiday season and see you soon in 2018!
The post Premier League Draws appeared first on The Church of Betting.
]]>The post Crypto Coin Betting appeared first on The Church of Betting.
]]>
In 2017 Bitcoin has taken the world by storm. During the last year the leading crypto currency has appreciated nearly tenfold as measured in most major currencies. The ones who bought some of the precious crypto coins early enough have made a killing. Stories made the rounds about unassuming crypto enthusiasts who bought/mined a few coins some years ago to forget them on their hard drives and come back years later to find out their coins are now worth a fortune. But only few of us realise that many of those who cashed in on the latest price increase were fellow bettors.
If you look around at some of the Bitcoin communities around the web you will find out that the bitcoin economy currently consist almost entirely of gambling websites. It only make sense – in such a heavily regulated industry there are a number of extremely lucrative markets that are basically closed to the books due to legal restrictions on betting. Finding a way around those regulations is guaranteed to make any sports book a fortune provided they could get away with it.
This is indeed what is happening since online gambling has been banned in the USA and bitcoin has made the news. Bitcoin turned out to be one of the few (and probably the easiest) way for US gamblers to place an online bet. Furthermore, a number of Blockchain enthusiasts (Blockchain is the technology behind Bitcoin and all the other crypto coins) are just looking for a marketplace where they can turnover their favourite crypto currency.
This is how the first bitcoin bookmakers have been founded. The likes of Nitrogen Sports and Sportsbet are already on the market since a few years and have established themselves as brand names in the crypto community. After their initial success a number of bitcoin gambling websites started popping up around them. We already have the bitcoin betting exchange BitBTC and the bitcoin prediction markets BetMoose and BitBet, with Augur and Gnosis being in development after having attracted large investment capital (I found out about the last two companies thanks to a great post by Steve from Daily25 on the same topic).
At this point surely many of you are already wondering – are crypto coin betting websites worth it? What are the benefits compared to traditional books and are there some money-making opportunities in this space? Here I will try to shortly answer these questions.
When I first started my research to this article I was really pumped up about this new blockchain technology and its applications in betting. Blockchain is being marketed as a technology that brings trust into a trust-less world. And I can hardly thing of an industry with a lower level of trust between operators and customers than gambling. I was already imagining the Blockchain bookmakers of the future where you keep no balance on the book but every bet is a separate transaction that gets settled by independent miners. No verifications, no limits, no frozen accounts and no forfeited funds.
I was in for a let-down. Even though in theory Blockchain technology could offer the means to settle bets by an independent party bitcoin books are not nearly there. What you get is simply the same old softbooks with Bitcoin (or another crypto coins) as an offered account currency instead of (or next to) the traditional (referred to by the crypto community as ‘fiat’) currencies. Can the bookie run away with your money – no doubt. Arguably it would even be easier as the Bitcoin books are even less regulated than the traditional books and since the technology is relatively knew there are not many established names in the field so the risk of stumbling upon a scam website is higher.
That discovery has surely somewhat cooled my initial enthusiasm about crypto coin betting. However, digging deeper I managed to find out there is more to it than I initially thought. Swinging between excitement and disappointment my dive into the world of crypto coin betting has revealed to me some interesting innovations that crypto currencies are introducing into the betting markets. Here I will shortly list the main advantages and disadvantages of crypto coin betting that I have identified.
That’s a big one. I want to say it upfront because that might just sound like a boring detail – it is not. Remember when I said you are still at the mercy of the bookmaker as far as your balance is concerned? Well… there’s a bit more to it. Even though you are still required to hold a balance with the bitcoin book in order to place a bet and bets are still settled by the book alone and not in a decentralised manner, the Blockchain transaction technology is changing the game quite a bit.
Every time you deposit or withdraw via a credit card the merchant (a bookie) will pay something like a 2-3% transaction fee to the credit card company. Therefore bookies hate transactions and will make whatever necessary to restrict them to an absolute minimum. The main measure to achieve that is passing over the fees to the customer after a certain number of withdrawals within a time period. Not to mention that due to the costs involved, a transaction always puts your account under scrutiny and might take quite some time to get processed. All of this usually results in a large player balance held at the bookmaker in question, which is always asking for trouble.
Not anymore. The fees for a crypto coin transaction are many times lower than the ones for a credit card transaction. Furthermore, the transactions are processed quickly and reliably. Therefore most bitcoin gamblers typically make a significantly larger number of transactions compared to the traditional punter. It is not unusual to deposit the required coins just before you place a bet and withdraw immediately after the bet has been settled. The result is a much lower balance in the book which, save for really large bets, gives less leverage to the bookmaker in case of a dispute. That in itself makes quite a big difference to how bookmakers traditionally operate and is one of the biggest advantages of crypto coin betting.
…however, the crypto currencies (lead by Bitcoin) are extremely volatile. Given that you can’t buy your groceries in Bitcoin just yet sooner or later you will need to exchange your coins for local currency so the wild swings in the price might hurt you. Sure, they might benefit you as well but my gut feeling is telling me that the more crypto currencies raise in price the less likely it becomes that the volatility will work in your favour. Of course, I have little to no clue on where the prices of crypto currencies are going so, as always, make sure to do your own research.
The other thing that grabbed my attention on the bitcoin betting platforms was the prediction markets. The most developed one as of now seems to be BetMoose. Prediction markets function a little bit like a betting exchange with the difference that anyone can open their own markets. Furthermore the markets don’t need to be on a sports event (although this is a possibility), but can be based upon just about anything. The future price of Bitcoin seems to be a popular bet but there are even platforms betting on the weather. Given that some weather patterns are so random that they are sometimes being used for random number generators it is not a market I would like to be involved in but apparently there are people who see it differently.
I believe the prediction markets represent an interesting opportunity. An experienced punter, armed with his knowledge about probabilities and some basic maths can give those markets a try and eventually turn a profit. Not only that, but creating an interesting market and attracting some volume there could earn you a fee for “managing” and settling the market when the time comes.
Unfortunately as it turns out the current liquidity on those prediction markets is quite low. The markets also normally last for a long time (a couple of months on average) which means you won’t be able to withdraw your money for much longer then you are used to. Needless to say, in the meantime you need to account for various risks connected to the platforms reliability. Therefore I must conclude I see some good potential in those markets, however they need some further growth in order to become attractive for the average punter.
Some people like to stay under the government’s radar and there is nothing wrong with that in itself. Being able to initiate direct transactions to the bookie with no intermediary payment provider, a player is able to maintain anonymity to a higher degree, even though bookmakers still might request verification. I believe everyone is well aware of the regulation in their own countries, so use this option at your own risk.
However, the lack of regulation on the one hand and of established brands on the other inevitably leads to a lower quality of service. Many of the bitcoin betting websites that are popping up recently have significant downtimes, change of ownership leading to operating halt or what’s worse, a scammer deliberately closing it all down to run away with the punters’ money. Needless to say, those risks exist with the traditional bookmakers as well, however due to regulation and the reputation of the books (or whatever’s left from it) it is less likely.
There is definitely great potential in Blockchain-based gambling but I am not too excited about the developments in the area just yet. I will personally adopt a ‘wait and see’ approach as it comes to crypto coin betting exchanges and prediction markets. Perhaps they develop into something interesting but the liquidity does not seem to be there yet. Furthermore, the Bitcoin books might be a good target for arbing, but given the taste of arbers for safe and steady returns the wild swings in any crypto coin’s exchange rate might give them a second thought.
But mostly, the big question for me remains whether the time will come when bets will be settled by an independent third party making bookmakers merely a counterpart in a bet as opposed to a self-elected guardian of customer funds and arbiter of bet outcomes. That would certainly ignite a revolution on the betting markets and the Blockchain technology has the potential to make it happen.
If you want to chat about crypto books and exchanges feel free to drop a comment below. In any case, thanks for reading and stay tuned for the incoming articles on DFS, betting strategies and arbitrage.
The post Crypto Coin Betting appeared first on The Church of Betting.
]]>The post Winning at DFS & How to Optimize Your Lineup appeared first on The Church of Betting.
]]>My journey in the DFS universe continues with my next article from the series. In my introductory article I briefly covered calculating expecting points and the different DFS providers available. Today I will talk about which tournaments are best to compete in, my biggest win yet, how to optimize your line up and some other lessons I learned while making my first steps in the football fantasy world.
As already pointed out in the previous article, I am currently playing DFS on the FanTeam website where I focus on football tournaments. The biggest tournament by far on FanTeam (currently ~800 weekly players) is the weekly Premier League (PL) challenge, which is also the one I regularly participate in. Since preparing the weekly Premier League line up takes quite a lot of effort (it takes some time to analyze 20 teams worth of players) I was drawn to some of the smaller tournaments during the week with considerably less games. The Champions League (CL) tournament had only 7 games this week and strong participation (about a quarter of the PL tournament participation) and there were also batches of 1-4 international games per day the week before during the World Cup qualifiers.
My thinking was the following: fewer games per tournament mean less time spent doing player analysis. Meanwhile my model should still be able to identify the optimal team, thus giving me some positive return. In fact, it should be even easier to identify the optimal team since a lower total number of games means less variance and a smaller pool of players to choose from. Therefore, even though the prize pool is smaller I will need to invest much less time so my return per hour spent might actually end up being higher.
I decided to take that theory to the extreme and joined a bunch of 1-game-tournaments that were available during the week. Calculating the expected points of the 22 players took no time at all and assembling the optimal team was an easy task. I was already excited about how much faster and easier that was compared to picking my weakly PL team. There was only one problem.
Picking a team from a pool of 22 players turned out to be exactly as easy for my opponents as it was for me. In one instance the tournament running on the game England-Lithuania had 15 participants and almost all of us seemed to have picked more or less the same team. At the end a few impact bonuses were given away to the players, which have left me 1 point behind the first place. And 1 point behind the first place in this case translated to 9th place out of 15 and leaving the tournament with empty hands!
I have learned my lesson. In DFS you should avoid the low-hanging fruit. If it seems too easy it probably is, meaning everyone can do it and whatever advantage you had quickly disappears. From now on I will avoid the 1-game tournaments and I’d advise you to do the same.
The Premier League and Champions League tournaments are a different story altogether. The 7 games from the last Champions League round already required some time investment to pick the team, plus all games started at the same time meaning no guess work for the starting elevens was needed. I have finished my only CL tournament so far at place 39 from 183 players, which was a few points short of the prize pool, but still showed some profit-potential.
And what about the Premier League challenge? At the end of the day for me it is the one offering the best value-for-the-money from all tournaments. Yes, here you would need to invest the most time in player analysis. Yes, you will also get the highest variance meaning you could (and often would) end up with nothing to show for it.
However, you will also get the highest number of participants, meaning the average skill of your opponents will be the lowest. In poker speak, you will find the most fish in this pond. This will increase your average rate of return and in the long term that is all that counts. After all, remember you need to recover your 10% rake and earn something on top of it, so 12% return-before-rake is two times 11%, and with 9% you are still in the red! Therefore, in DFS you should never sacrifice expected return for easier lineup preparation, lower variance or whatever. Your opponents will get to use the exact same benefits and you will probably fail to recover your rake. The only viable options are to go big or go home.
This brings me to my favourite part of today’s article – the one where I get to brag about my biggest win yet in DFS! Last week I’ve finished 20th from 784 players in the weekly PL challenge. This has earned me a prize of €100.08, beating my previous record of €58.44 from an international round. Here are the guys who brought home the sweet prize:
I was lucky to have Raheem Sterling playing a great game for Man City who smashed Stoke 7-2 with 1 goal and 2 assists from my captain. Overall I had variance on my side this day which lead to this pleasing result. Yet you would notice I wasted quite some budget on Danilo who didn’t start at all, which again underlines the importance of guessing the starting eleven. This has cost me dearly this week when two players of my twelve picks did not play, which has effectively thrown me out of the competition despite all the others earning at least as much points as my model expected them to. I make another note here: guessing the starting eleven is a big part of winning at DFS and I have to work further on improving this part of my model.
However, there is one thing my model does really well and it is optimizing my lineup conditional upon the expected points I’ve calculated. I have written in the introductory article that this would be something complicated to do as it includes optimizing a linear system of equations under a number of conditions. I was partly right – the optimization indeed required solving a linear system – only doing it turned out much easier than I expected! Searching for a solution of my optimization problem, I have first stumbled upon some articles featuring pages of complicated R and Python codes, until finally arriving at a Youtube Video about the Solver Excel Add-in which, apparently, solves linear systems just fine. Watch the video and thank me (or rather its author) later.
Alright, that was my latest fantasy football report, the second of what I hope to be a long series of articles on DFS. Let me know of your experience with DFS or any questions you might have in the comments. You might have also noticed that I have slightly upgraded the design of the page. I hope you like it.
Thanks for reading and till the next time!
The post Winning at DFS & How to Optimize Your Lineup appeared first on The Church of Betting.
]]>The post The World of Fantasy Football appeared first on The Church of Betting.
]]>I must have made the longest break since the start of this blog with my last article being from the 13th of June. In case any of you wondered, I am not dead, but merely moved to a new home, which in German translates to 3 months without internet. Just a few weeks ago I had the great honour of being paid a visit by the Telecom technician, so here I am again. Despite not having posted anything new in the last months I used every bit of kilobytes I got my hands on further exploring the new trends in the betting markets. In fact, I was lucky to get this no-Internet period in the middle of the summer when most gamblers are on vacation anyway. But not me! Since the bookmakers offered nothing of interest to me in July and most of August I decided to look further. A few innovative betting products caught my eye – I tried my luck at spread betting and also dived into the world of fantasy football. The later one I found pretty exciting so I decided to dedicate my comeback article to it. I will share with you my experience with fantasy football and will give you some tips for optimizing the performance of your team.
The concept of fantasy football was not entirely new to me. The official fantasy game of the Premier League has been running since a long time ago and attracted millions of players. The game was basically a marketing tool for the league and for the players it has always been just for fun.
Than the fantasy world took an interesting turn on the other side of the ocean, where two big players emerged offering fantasy on a large scale, but with a bit of a gambling taste. Draft Kings and FanDuel were growing at an unprecedented pace and were striking huge sponsorship deals in NFL, MLB and elsewhere. The two have recently attempted a merger, which was blocked from the federal trade commission on anti-monopoly grounds. A political debate has been raging in the recent years whether what they offered, namely playing fantasy sports with money, should be considered gambling or not. Every state has its own opinion on that, but the fact remains that the growth of those businesses has put any European bookmaker to shame. Yet another instance of European high-street bookies missing an important (and incredibly profitable) trend. But hey, at least we got FOBTs!
Anyway, when this new fantasy trend emerged I thought it might be a great business idea to launch something similar for Europe, with the option to open accounts in local currency instead of only USD and a focus on football. Unfortunately I neither have the money nor the time to start anything like that, plus I recently found out someone got to it first.
FanTeam is on the market since 2013 and has meanwhile grown its base of active players significantly. The platform is running regular tournaments in several sports, but football attracts the most players by far. To my knowledge FanTeam are the biggest DFS (daily fantasy sports) provider based in Europe and offering accounts in EUR. Sadly, currently they don’t offer accounts in GBP, which I think is a pity. Since those of you based in UK would still need to exchange currency to play DFS, you might also want to consider the aforementioned Draft Kings and FanDuel. They also have football (soccer) on offer and you can open an account in USD there. The two sites have a different scoring system, which I have found a bit more complicated and more difficult for modelling. On the other hand they are more established and a lot bigger than FanTeam, which has its benefits.
But back to FanTeam. The average weekly Premier League tournament attracts around 700-800 participants. How does it work? You built a team of 12 players with a restricted budget of 100 points, where every player has a certain price arbitrarily determined by the platform. There are tournaments with different team size and budget but currently this seems to be the most popular format. How the players’ price is determined is an interesting topic in itself but not the focus of this article (I assume it has something to do with the expected points for each player as calculated by the platform).
Once you build your team (you have time until the start of the first game of the round) your players start collecting points and your team gets ranked according to your total points. The points are awarded based on a scoring system used by FanTeam (which does not differ much by the classical point system of the Premier League’s fantasy game). You get in the game by paying a rake amounting to 10% of your buy-in. The rake is where the FanTeam’s revenue comes from, while the buy-in goes towards the prize pool.
The more savvy punters among you probably already see the potential here. If you get to develop a system that picks the perfect team and you manage to score above-average in the weekly rankings you have a good chance of winning your rake back or even turning a profit. If you know a thing or two about probabilities in sports you could make a fairly accurate prediction on a number of events determining the point-outcome of a game. Of course, one also needs to factor in the expected variance, which in my experience is nothing short of what you are used to from the betting markets. There is a way to get around that and I will come to it later.
Another benefit of playing fantasy football as opposed to betting with a bookie is the business model. The platform makes its profit from the rake-in and the prize pool is formed by the entry fees of all the players. This means that the platform doesn’t really care about who is winning and how much, they only need to make sure everything is up and running. Stark contrast to the high-street “losers welcome” bookmakers who will kick you out at the moment they catch you making a cent of profit.
Furthermore, since the profit of the platform is independent from the outcome of the game there is much less risk involved for the company itself, so your money is arguably safer here than with a bookmaker. Again and again we hear stories about high-street bookmakers confiscating a player’s balance for whatever outrageous reason just to compensate for the losses they have made due to their terrible odds. I don’t see why that would happen here.
Getting started is fairly easy. You just need to make a deposit and verify your account. This could be done at a later stage but I always recommend doing it before depositing any money if the option is available. There are a few deposit options to choose from, including Skrill, Neteller and credit card deposit, which are all free of charge.
You get a 200% bonus, which you only get to release in the form of a 25% discount from your rake-ins. In other words, if you pay EUR 10 towards the prize pool, you would normally need to pay a rake of EUR 1 to FanTeam to participate in the tournament. If you have a bonus you will only need to pay EUR 0.75 and the EUR 0.25 discount will be released from your bonus balance. As you see it would take you quite some time to make a full use of your bonus. On the other hand there is no possibility for bonus abuse, which certainly improves the reliability of the provider. So all in all I find the bonus satisfactory.
So far I have chosen to participate in the weekly tournaments. I find them more exciting since you get the outcome only in a few days. There are full-season leagues available too, but I wouldn’t like to have my money locked in there for almost a year. On the other hand, a full-season tournament has the advantage that in the long term it evens out much of the variance, whereas in the single-round tournaments you are strongly dependent on the outcome of this and that game even if you make the perfect team.
As an example, a few weeks ago I have picked quite a few Man City and Tottenham players in my team. I have calculated a high expected score for all of them largely due to the fact that the betting markets gave those two teams the highest probability for winning their games and expected them to score a lot of goals. However, City only managed to win 1-2 with a late goal and less than stellar performance, while Tottenham disappointed with a home draw against Burnley. Meanwhile Liverpool smashed Arsenal 4-0, which came as quite a surprise to everyone.
You should expect things like that to happen all the time. The variance is big and especially in the weekly tournaments it might take some time for skill to pay off. One way players deal with this is to play with several teams in one week’s tournament. It is certainly an option, but first, it is sub-optimal since you don’t only run the team with the highest expected score and second, it requires even bigger time investment in analysis on your side, so for now I choose not to go this way and try to live with the variance.
With this in mind, you could start building your first team. The approach I chose was to pick the players who would bring me the highest expected points. Calculating expected points for players is tricky and requires a lot of work. I must say upfront, if you are hoping to make easy money with fantasy football you should give up, since I have found the analysis is quite time-consuming. However, what I liked about it was that most of the information you need is out there so if you are willing to invest the time you can come up with a fairly accurate forecast.
The team format used in the weekly tournament includes 12 players, 11 of whom play and the last one sits on the bench and comes in in case any of the 11 doesn’t start. As you already see, it is pretty important that all of your chosen players start and guessing the starting line-up for each team is an important winning factor in fantasy football.
You have two restrictions in building your team. One is the budget, the other is the formation.
You have a certain budget available to buy players. For the weekly tournament it is 100 points. The price of players changes every week mostly in accordance to whom the team is playing against. The algorithm for calculating the weekly price of players is not disclosed by FanTeam (at least not to my knowledge). What is left for you is to determine a “fair price”, compare it to the ones offered by the platform and buy the players you consider a bargain.
The second restriction is the formation. Players have one of four positions, which are relevant for the scoring as well as for the formation (GK, MID, DEF, FWD). There are only certain formations that are allowed (5-4-1, 5-3-2, 4-5-1, 4-4-2, 4-3-3, 3-5-2, 3-4-3). If your 12^{th} player needs to come in, the resulting formation also needs to be an eligible formation. Otherwise your sub stays on the bench even if someone from your starting eleven didn’t participate in the game.
Finding players who are a bargain and optimizing your team based on the restrictions above is the exercise you need to do in order to make the perfect fantasy team. To find out which players are sold at a discount you must first determine what amount of points you could expect to get from every one of them. To do so you must carefully study the scoring rules. As I wrote earlier, the scoring rules would be familiar to anyone who has played the Premier League fantasy game as they are not much different. What needs to be done at this point is to go through all the factors that may give points to (or take points from) a player, calculate how much points you expect that player to earn from each factor and finally sum it all up.
I do this by having a look at the betting markets for this game and making use of as many of them as possible. As an example, a defender who has played 90 minutes in every game of his team since the beginning of the season, would have a fairly high chance of starting and playing a full game this week. The market for a clean sheet for his team is at odds (say 1.9 for a clean sheet and 1.9 for no clean sheet). That means the team has a 50% chance for a clean sheet, which would add 4 points to the score of the player, or 50%*4 = 2 expected clean sheet points for this player. You repeat the exercise for all factors and all players, sum it all up and arrive at an expected score for each player for this round.
Once you have calculated the expected points of all players the question comes: Who do you need to buy? The best “value for the money” team would be one composed of the 12 players with the highest expected points/price ratio. The two restrictions , the set budget and the allowed formation, add additional difficulty to the calculation. You need to maximize your team’s expected points based on those two limitations. What complicates the function further is the “Captain” option. The Captain is one player from your team picked by you, who scores double. With this option high-scoring players who would normally be less of a bargain compared to mid-scoring ones could end up being more valuable to your team.
At this point I haven’t figured out how to mathematically optimize this function. I have a feeling you have to use some sort of linear algebra but I must dig into that further. The approach I am using at the moment is to arrange my team based on intuition. It is not perfect but it is still a working solution.
So at the end of the day you follow the above rules and arrive at the perfect team, in expectation at least.
That makes sense, so what results did you get?
Frankly, the results by now have been mediocre. I have played five rounds so far. In the first one I underestimated the required time for analysis, spent too much time calculating the players’ expected points and left too little for building the team. I ended up picking up a bunch of players in the last minutes before the deadline, who contributed for a nice team but also one with a total value of only around 80 points of 100 available. I was pretty pissed at my poor time management until I found out at the end of the round that I was only a few points short of the prize pool (15% of participating players receive some prize, of course the higher you rank the more you get).
Being super excited about the promising score of my team of underdogs I was sure the next round I will make it big. I didn’t know at the time I was up for a huge disappointment! Having a bunch of expensive players from two teams who were heavy favorites for the week I was hit by the variance with both teams playing well below expectation. As you have probably guessed, the two teams were the above mentioned Man City and Tottenham. I had a below average score placing me 500^{th} from 775 teams. Quite some cool-off.
What happened? Obviously, in the first round the luck was with me and in the second one against me. I saw in practice what I should have already anticipated in theory – variance is an important factor here and even if you have the perfect model it would take quite some time to see some persistence in returns.
Then recently I joined a round of international games where luck was again on my side and I managed to score 3rd, winning EUR 56 on an entry fee of 10, evening out my FanTeam balance. So armed with my knewly found knowledge and quite a lot of patience I continue to compete in the weekly Premier League challenges. I am still pretty excited about DFS so I will continue working on this and will let you know how it is going.
Conclusion
In the meantime, I’d be more than happy to read about your experience with football fantasy. Did you already give a try to any paid fantasy football leagues and how did it go? Or maybe you developed a cool system for the official Premier League fantasy game? Let me know in the comments.
Also, a huge thank you to everyone who keeps coming back here even after such a long period with no new material from me. I really appreciate your interest and support and the thought that you are still interested in what I have to say keeps me and this blog going. Thanks again, you’re awesome. And see you soon!
The post The World of Fantasy Football appeared first on The Church of Betting.
]]>The post League of Betting Legends appeared first on The Church of Betting.
]]>And today I want to dive into one of the hottest topics in the betting world at the moment: e-Sports betting. More specifically, I will look into the most popular e-Sport nowadays, the League of Legends (LoL). Every month more and more bookmakers are offering new League of Legends betting markets and if until today you were scrolling past those weird sections with a raised eyebrow it is perhaps time to pay them the attention they deserve, since there might be some money to be made.
Of all the exotic betting markets, why did I pick exactly this one, you might ask. I might as well have written an article about darts, chess, the Eurovision song contest or the latest UK election. While these are all interesting topics in themselves, what caught my eye about e-Sports and LoL in particular was the recent availability of great arbing opportunities on those markets. I have had more than one case where I was able to place an arb between two sharp books on an e-Sports event. Those of you with some arbing experience would know how rarely an arb involving only sharp books comes along. This got me thinking: if even the sharp books cannot adjust their markets in a timely manner, it is possible that the traders are not that far ahead from the rest of us in this new market.
What’s more, it seems that the market is here to stay. The LoL story can by no means be compared to the short-lived Pokemon Go craze from a few months ago. LoL is on the market since a few years already and has been on a growing path ever since. As of September 2016 the game had reached a milestone of 100 million monthly players. There are numerous international tournaments being hosted with players from all over the world flying in to play in front of huge crowds in packed sports stadiums. The events are also broadcasted online where the bigger games attract hundreds of thousands and sometimes millions of viewers. Some of the teams are being owned and funded by notable sports investors who don’t like to waste their time with short-lived spells. Many of the tournaments and players have secured contracts with large corporate sponsors.
Soda drinks and Doritos included
In brief, we have a young market where a lot of books are still testing the ground as it comes to setting the right odds, but also want a piece of the action as the market is expected to grow further in the future. So I think that is a pretty good place to be at as a punter. What is left for us is to take the challenge and start building our LoL betting skills. What you should know about the sport, where to place your bets and some betting strategies, I will cover it all here.
LoL is a tactical team-game played by two rival teams of 5 players each. Teams start from two opposing corners of the map, each player picks a champion and the champions fight each other with the ultimate goal of reaching and destroying the other team’s base. In the process of doing so the two teams are destroying vital buildings but also killing each other’s champions. A killed champion is placed off the map for a certain time before reappearing in his team’s base.
The most basic LoL betting market is the win outright market. In the common 2 out of 3 games format the game also does not have the option for a draw, so it is a two-way market. However there are a few more markets on offer such as a total number of kills, a kill handicap (notice that you don’t necessarily require more kills to win, so this market is more similar to a game handicap in tennis than a point handicap in basketball), which team will make the first kill (aka “First blood”) and a few more.
Whether you’d go for the more established markets or the more exotic ones is a matter of taste, however I do think you have to pick your niche and stick to it, since those markets are driven by very different factors. I will just share a few thoughts on the main ones.
This is the most usual LoL market and therefore it could be said it is the most efficient one. However, I still think it is one of the best ones to get into. Being the most basic market, you will find selections for every featured event, even the most obscure ones. And arguably the smaller leagues are where you could find the most opportunities. This is also the market where I have landed all my arbitrages so far.
It is a difficult and extremely unpredictable market. Depending on how the game goes or what strategy a team has chosen you could be surprised by the end result and trends are difficult to spot. However, if you get a solid knowledge about the teams and their preferred tactics against certain opponents you might get an edge here.
This is actually an interesting market with a great potential. The reason for this is that by design, during the game the winning team accumulates an advantage against the losing one, resulting in the greatest gap between the two right at the end of the game. So, except for games that are really close ties, you could expect that at the end of the game the winning team will lead with a handful of kills. Therefore, if you want to back the favourite, but are not happy with the odds, you might go with the handicap instead. By the same logic, a handicap bet on the outsider would of course be a poor bet.
This one is a very unpredictable market (think “who will have the first corner”), but is still far from a coin toss. Teams’ odds are usually closely matched in this market even when there is a clear favourite for winning the game. However, there is some rationale behind picking the winner that can give you the edge. I will talk about that in more detail in the betting strategies part
For big games you sometimes get an even more obscure markets such as “Which team will kill the Baron Nashor” (Baron Noshar is a strong neutral beast at the center of the map) and the like. In my opinion those markets are incredibly unpredictable and so far I could not identify the factors that can help you make an accurate prediction there. Still, you might look into those once you get deeper into the sport
As in every sport, knowing the game, the teams and the players pays off. Look for stats of the teams and players (1, 2, 3) and try to evaluate the relative player’s strength. Rosters (the LoL transfers) are a common occurrence and can influence the team’s performance significantly, so be up-to-date. To learn more about the players or just get a feeling of the whole thing, watch a few games on Twitch, live or recorded VOD (videos-on-demand). The games are broadcasted free of charge with a very professional commentary, often from other experienced players, so they could be a great source of knowledge.
How about Rating systems? There is plenty of data on LoL around there, so you could surely prepare some cool model. However, there are a few things you should keep in mind. For one, team performance is volatile and heavily influenced by, among all, the participating players who are switching teams constantly. So it makes much more sense to rank players as opposed to teams. Team chemistry is also a factor, but team performance from year to year or even from tournament to tournament can change dramatically. League compositions are also very dynamic so teams will often come and go from one season to the next.
Actually, player ratings have been around since the beginning. LoL was using Elo-ratings to rank players for some time before switching to a proprietary rating system (LP), which should act similarly, but some details are kept secret. You could use the LP or develop your own ratings.
A trader-like approach, you can aim at reacting to price-relevant information before everyone else does. As it is with other sports, social media is your friend. You can also join the big LoL communities such as the Reddit boards dedicated to LoL. A great tool for you for this strategy (and LoL betting in general, actually) are the Discord chats where the LoL community is discussing the games, teams and what’s new.
What sort of information you are looking for? There are quite a few surprises as it comes to the starting players for a team. For international contests Visa issues often pop up. Players are being substituted for this and other reasons, which would sometimes damage the winning chances of a team, so it might be a good idea to place a quick bet on the opponent.
You could develop your own thesis on certain trends for certain leagues, backtest them and then try them live. For example, Jordan Mackenzie, who is sharing his LoL betting experience in one of the websites dedicated to LoL, has observed that bookmakers offer too generous odds on the favourite for a few leagues. I recommend following his blog, I find it an interesting read.
Being based on a computer game, the rules of LoL as a sport tend to change quite frequently, especially compared to a traditional sport. Every new version (aka Patch) of the game brings something new to the gameplay and these come at least a couple of times a year. These are no fundamental changes to the rules, but they can give an advantage to certain champions, or roles, on the board.
Roles is a huge topic in itself so I would not get too deep into it, but in short, different champions play different roles on the field, just like in every other sport. You have a Jungler, who is for the most part hiding from the opponent players, trying to surprise them for a quick kill, and plays in a very mobile fashion in general. There is the Support who tries to improve the capabilities or save the lives of the other team players. Then you have the Tank, whose main goal is to attract the attention of the enemy players to himself due to his durability, while the rest of the team members use the time to destroy their opponents. And then there are many other roles and subroles, which are detrimental for the tactics of the game.
Why do you need to know that? Players usually specialize in a certain role. So a team might have a great Support player, but their Jungler might be less than impressive. If a patch that benefits the Support champions and reduces the role of the Junglers gets introduced, that might be a great boost for such a team and might influence their short term winning rates substantially.
Being aware of the new patches and their effect is another way to gain an edge in front of the bookie
And finally, an interesting and really fun approach is to try to guess which team will get the first blood. I have mentioned some of the roles above, and the one role that plays the most here is the Jungler. Again, that is the guy who is moving around the map and is waiting for an opportunity to surprise an opponent and make a killing. Often times the Jungler will be the one to take the first blood or at the very least he will be involved in it. The best strategy here is to wait for the players to pick their champions and immediately after that place your bet. You should not waste any time as the market closes shortly after the last pick is made.
At this point you should know which of the two Junglers is the better one, the more aggressive one, has the better champion and in general is more likely to go for the first kill and make it. Factors like how fragile are the other team’s champions in the early game can also influence your pick. Once you have taken all those factors into account, if the odds are right, you could try your luck. The bet will be settled relatively quickly as the first blood normally takes place during the first few minutes of the game. You could also analyse past games and the short time window for analysis allows taking a look at a larger number of games. In general I believe that is an interesting market and mastering it could ultimately bring you some nice profits.
Kind of like those gamer guys who just won 2 million dollars from a LoL tournament
League of Legends is a young and growing market where most bookmakers are still testing the ground, but there are already a lot of betting opportunities available. There are quite a few betting strategies you could practice and master, which can give you the edge over the rest of the market participants. But most importantly, LoL is a cool new sport which you can have a lot of fun betting on. And it is also available in the summer! So there is a great way to fill in the summer days when the big sports and leagues are taking a break.
If you have any experience to share from the e-sports markets I would love to hear from you. If you have any questions or comments feel free to post below or drop me an email at admin@churchofbetting.com. Finally, I hope the arbers among you had a great season so far and will enter the summer months in a happy mood! Thanks for reading and see you around!
The post League of Betting Legends appeared first on The Church of Betting.
]]>The post The Favourite-Longshot Bias appeared first on The Church of Betting.
]]>Today I want to expand further on the overround analysis from my last article. There I have demonstrated the relatively well-known fact that overround in the betting markets is declining in the recent years, but have also offered a more detailed look into the different developments within different markets/countries/leagues. The article was received well, showing that the topic is an interesting one for the betting community and is perhaps not getting the coverage it deserves.
So today I dive deeper into the topic of overround and more specifically the famous favourite-longshot bias. The favourite-longshot bias refers to the not-so-intuitive fact that bookmakers tend to assign disproportionally large overround to longshots. As a result the average price a bettor receives on a favourite is, relatively speaking, more favourable compared to the one for the longshot. There have been a few attempts to explain why that is the case, so I will not speculate on that here. But I have not come across a lot of material trying to quantify the bias. The question by how much exactly is the overround for longshots overweighed remains largely unanswered. So that is what I am looking at today.
The base case, or the simplest assumption one could make is that the margin is attributed equally to all outcomes. I will call this the linear case. So if you have a 5% margin you simply multiply the offered odds with 1.05 to get the fair odds, the formula being:
O_{f }= O * (1+M)
The alternative is a formula proposed by Joseph Buchdahl in football-data.co.uk. In summary, it assigns the margin to the different outcomes not equally, but in proportion to the size of their odds, or in a reverse proportion to the probability of the outcome. I call it the proportional case. You can see the exact formula in the link above. Unlike the linear formula, the proportional one takes the favourite-longshot bias into account, yet we could test if it quantifies it correctly. The formula:
O_{f }= n*O / (n – M*O)
In the two formulas above O_{f} stands for fair odds, O for offered odds, M for margin (aka overround) and n for number of outcomes (3 for a 3-way market).
I am using the names linear and proportional for the sake of simplicity, although one might argue that they are not factually correct, since the linear formula still adds the overround in proportion to the odds while the proportional one gives the odds even higher weight resulting in a sort of an “overproportional” attribution of overround. Anyway, I will stick to that convention for the rest of the article.
I test the two formulas with the odds data provided by football-data.co.uk. I have gathered results for the few soft books with the longest record (Interwetten, William Hill, Ladbrokes, bet365) and for Pinnacle.
First I simulate a betting strategy with the fair odds assumed by the two formulas. If we were able to bet level stakes at the margin-free fair odds as opposed to the ones the bookmakers offer, we would expect to have around zero profit in the long run.
I have arbitrarily determined 10 ranges of odds for each bookmaker so that I get approximately the same number of bets in each of them. The first bracket has the lowest odds, the second one has the second lowest and so on. Then I have averaged the returns of such a strategy for each range and for both formulas and have drawn a graph. Ideally the lines should be flat and close to 0. The results are below:
It is safe to say that the graphs confirm the favourite-longshot bias. Clearly, for most if not all of the bookmakers, the linear formula gives too high fair odds in the low odds range and too low in the short odds range. The proportional formula seems to do a better job.
The graphs give an interesting insight but are inconclusive for shorter sets of data and lower margins – like the Pinnacle set. In particular, the graphs suggest that the favorite-longshot bias is not present in Pinnacle odds to the same extent as in the soft books, but are insufficient to make a strong conclusion. So I started looking for a more scientific way to measure the correct distribution of margin in order to be able to draw more precise conclisions.
There are a number of so called scoring rules often applied for that purpose. The most common ones include RMSE (root mean squared error), MAE (mean absolute error) and LE (logarithmic error) – but I did not find any of them particularly useful. The problem I have encountered is that they do well in measuring wrong predictions in the low/high probability range (like 5% and 95%) but were essentially useless for probabilities close to 50%. I don’t want to bore you with the details, so just consider this: assigning a 50% probability to a series of 100 outcomes, whether you got 100 wins, 100 losses or 50 wins and 50 losses does not change any of those scores – while obviously you made a good prediction only in the third case.
Still, I decided to report the results for MAE as it is somewhat useful in detecting deviations for low/high probabilities. I have decided against RMSE and LE as they both overweight mistakes in the extremes (e.g. a very short odds bet lost or a very long odds bet won), which, variance aside, is all the same for the bettor, since he cares mostly about how much money he loses/wins on average and MAE is the best estimate for that. However, MAE is just as incapable as the other two at detecting bad predictions around 50% assigned probability so keep that in mind.
As I was a bit disappointed by the accuracy of the scoring rules I kept searching and found an interesting suggestion of using the series of binaries (1-happened, 0-did not happen) as the dependent variable and the percentage predictions based on the fair odds as the independent variable in a linear regression. In any case the correlation and r-squared are low because the dependent variable always ends up at one of the two extremes. However, with accurate predictions one would expect a slope close to 1 and an intercept close to 0.
You can see the summarized results for the linear regressions and the MAEs (the lower, the better) below:
For the linear margin distribution we consistently get slope higher than 1 and negative intercept. So to arrive at the correct probabilities we need to take the “fair probabilities” derived from the linear model, give proportional increase to all of our probabilities and then deduct a fixed amount from the result. Let me give you an example.
Let’s say our initial probabilities according to the calculation with the linear formula for two outcomes of an event are 80% (fair odds 1.25) and 20% (fair odds 5). We have a slope of 1.10 and an intercept of -0.05. So to arrive at the correct probabilities, we first multiply both by 1.1 (or 10%) to 88% and 22%, then deduct 5% from both arriving at the final probabilities of 83% and 17%. Now, these are not strictly the correct probabilities either, since depending on the coefficients they won’t necessarily add up to one, but in the context of the model those numbers do a better job at explaining the result of the event then the initial ones. Therefore we could safely say they are closer to the true probabilities.
Meaning we have underestimated the probability for the 80% (short odds) and overestimated it for the 20% (long odds). Why did we do that? Well, most probably we have assigned too big of a margin to the short odds, therefore the fair short odds we arrived at were too high and the respective fair probability was too low – and the other way around for the long odds.
So, for a slope higher than 1 and an intercept lower than 0, the bookmaker assigned a higher portion of the margin to the long odds and a lower one to the short odds than what the model accounted for. And the other way around for slope lower than 1 and an intercept higher than 0 (as we have it for some bookies with the proportional model).
Our table shows us again that the proportional formula accounts for margin attribution on average much better than the linear one. However, we also see that different bookmakers assign margins differently and there is no one-size-fits-all solution. For example, Interwetten assigns an even lower share of the margin to the short odds and an even higher one to the long odds than what the proportional formula would suggest. On the other hand William Hill seems to be using something between the linear and the proportional formula.
What does that mean in practice? First, if you bet randomly at all books (which I hope you don’t), at Interwetten you will get much better results betting on short odds than on long ones, at William Hill the difference will be smaller, but in all of them short odds will give you a better result.
Second, and more importantly, we see that if we are to use just one formula to estimate the fair odds across all bookmakers, we can hardly find a better contestant then the proportional formula suggested by Joseph Buchdahl at football-data.co.uk. This is important for sharp bettors or services like Trademate that try to estimate the fair odds a bookmaker has calculated based on their odds on offer. For them the proportional formula is a great choice and could be used as it is or fine-tuned a bit to fit the bookmaker in question more accurately. In any case the proportional formula beats the linear one by a mile and I would prefer it for any bookmaker out there.
So that’s it from me on the favourite-longshot bias and the bookmaker overround. I see that some of you like these articles a lot, yet others seem to be a bit bored by them. So I plan to follow with some lighter stuff like interesting stats from Trademate, some arbing wisdom and perhaps a few exotic betting markets. In any case, if there is anything in particular you would like to read more about just let me know. Good luck to everyone and till the next time!
On an unrelated note, Rebel Betting have let me know that they have started a major sale, offering large discounts on their long-term offers. It’s a big one-time payment but a great per-month value and if you are planning to arb long-term I think you can hardly find a better deal than that. If you are going for the 6 months, just make sure that the leagues you are aiming at are playing during summer, since it is usually a quiet season for arbers. But in any case I don’t think you can go wrong with the 32 month option.
Also, a big thank you to whoever purchased the software from my banner. If you need any support or advice just drop me an e-mail at admin@churchofbetting.com and I will do my best to help you.
The post The Favourite-Longshot Bias appeared first on The Church of Betting.
]]>The post Overround Decay appeared first on The Church of Betting.
]]>The bookmaker margin, overround or vig is the profit margin of a bookmaker from a bet a punter makes (assuming the bookmaker’s books are balanced). It is calculated by dividing one by the odds for every possible outcome and adding those fractions up. For example if a bookmaker offers you 1.92 for o2.5 goals and 1.89 for u2.5 that translates to (1/1.92+1/1.89) – 1 = 4.99% overround. An arbitrage is a group of bets with a negative overround. The better the odds a bookmaker is offering you the lower the overround and vice versa.
A higher overround takes money out of the punter’s pocket and puts it in the bookmaker’s. So in a sense, a decrease in the overround indicates an improvement of the quality of service a bookmaker offers. This means we as punters can look into historical data to see what treatment we receive from the bookies and how it has changed over the years.
I have taken odds data from football-data-co.uk for my analysis. The data includes odds from several bookmakers as well as aggregate average and maximum odds from BetBrain. Many European leagues are covered, but I will only look into the few main ones – England, Germany, Spain, Italy, France – to see how the margin varies across countries. I have gathered the data for the others as well so if there is any country in particular you are interested in just let me know.
First I had a look into the development of the overround for the different tiers within a country. I have taken England as the dataset covers the widest range of leagues there. I have worked with aggregated data, looking into the 3-way (home-draw-away), over/under 2.5 goals, as well as Asian handicap markets. For all three markets there is a graph for overround based on average odds (left) and on best odds (right).
This is where we see the most dramatic reduction of overround. The average overround got down from ~9.5% in 2005 to ~4% in 2016 for Premier League. The reduction was a bit smaller in relative terms for the lower tiers of English football, but still significant.
The case is similar for overround using best odds, which for Premier League is negative since 2013 and for lower leagues is in the 0-1% range, a substantial decrease since the beginning of the period. An interesting detail is that 2015 was the only year in which the overround using best odds increased across all leagues, while in 2016 the trend got unclear. Meanwhile, looking at average odds, the decrease of overround does not seem to be slowing down in 2015 and 2016. Was that due to the fact that margins approached the 0% level and arbitrage started clearing market inefficiencies? Or was there less divergence between odds of big bookmakers from 2015 onwards? Later on I will use some further evidence from the data to try to answer that question.
There is a decline here similar to the 3-way market but less pronounced. Across leagues the average overround has fallen from ~8% to ~6%. In recent year the gap between the overround of higher and lower leagues has increased as well, which is again somewhat similar to the 3-way market. And finally, there is the same jump in overround using best odds from 2014 to 2015. It seems this particular effect is apparent across markets.
Asian Handicap seems to be the only market where the average overround hasn’t declined significantly during the years but has rather remained stable. The Asian Handicap concept has been considered an innovation some 15 years ago and while today it is established and understood by most punters it is probably still not as popular as the 3-way and Totals markets. Perhaps this is the reason why we don’t see the same improvement in the average overround for this particular market.
We also see a spike in overround (both average and using best odds) in 2009, which is missing for the 3-way and the Totals. The overround has since returned to its previous levels. I am curious what happened in 2009 that caused this so if anyone has ideas please feel free to share.
The comparison between countries is interesting from two points of view. First, to see whether the effects in place for the English leagues have been replicated abroad and second, to check which countries are assigned lower and higher overrounds by the bookies. Here are the charts for the different markets (again using average odds to the left and best odds to the right):
For all of the markets there are no significant differences in overround development across countries. Again we see the steepest decline in average overround for the 3-way markets, followed by a milder decline for the Totals and a very small to none for Asian Handicap. Again as it comes to overround taking best odds there is an increase in overround in 2015 while the average for this year keeps going down, indicating a convergence of odds between different bookmakers.
In terms of overround per country, we see the lowest overround for England, which is to be expected as this is perhaps the most commercial and watched league as well as the one attracting the most betting action. For the most periods/markets Germany, Spain and Italy follow with the second smallest overround, followed by France with the highest one. The overround differences across leagues are probably caused by their relative popularity, but those differences are not too big so a change of ranks in the coming years does not seem unlikely.
Finally, a comparison between bookmakers offers a perspective on how the overrounds developed across different betting firms. If you use only one of those, perhaps you could find a bookmaker where you get better value for your money. Since we have already found out that differences between countries are pretty slim, I decided to go back again to the different English divisions and see how overrounds moved there during the years.
Again we see that the overrounds have been declining across the board. The ones who were the most dedicated to offering higher odds to their customers seem to have done fairly well in attracting turnover, like bet365 for example. We also see, unsurprisingly, that only a few soft bookies can match Pinnacle’s odds for any of the leagues. Bet365 and Betvictor offer the best odds in English football from the soft book group and William Hill and Ladbrokes seem to be catching up although with some delay. But of course, as long as those bookies are available at your jurisdiction, regardless of their overrounds it makes sense to shop for the best odds among them for any given event.
During the last 12 seasons overrounds in football were declining for virtually any bookmaker, any country and almost any market in the data set. The overround was declining the fastest for the 3-way market, slower for the Totals market and remained stable for the Asian Handicap market. The trend is pointing towards a further decline for the 3-way market and is rather unclear for the others.
The analysis shows that the online betting revolution has truly improved the prices the average punter has access to. It takes longer for the mug punter to blow his bank than it used to. And even though the bookies have worse profitability per bet than they had 12 years ago, their presence online has allowed them to take an unprecedented amount of bets. The increased turnover has more than compensated the reduced profitability per bet, so their profits continue to rise. Of course some betting companies have done better and some worse in this new environment but the general trend is clear and every betting company needs to adjust to it.
The overround using best odds on the other hand seems to have stabilized and even increased a bit in recent years. Is that a sign for some consolidation in odds calculation across bookmakers/odds calculating firms?
Some insiders seem to believe that this is what the future holds for the betting industry. All bookmakers offering the same or similar odds on all events will basically eliminate arbitrage opportunities on the betting markets even with decreasing overround. Would this bring the end of profitable sports betting for everyone except the bookmakers?
I personally don’t believe that would be the case. As the saying goes, if everyone is thinking the same, no one is thinking. Different methodologies for odds calculation and different perspectives on sports betting (even if some are on average less successful than others) are what makes the sports betting market efficient. I don’t think that a single odds calculator or betting company can take hold of the countless sports events worldwide which you can bet on nowadays. You have probably seen in my Trademate review that even Pinnacle’s odds are not perfectly efficient before closing. A week ago I have made a few e-games arbs between two sharp books, one of which was Pinnacle, which was another evidence for that.
So my believe is that regardless of what the future brings, the betting industry will always have place for different kinds of players, big and small. A single sharp punter will always have the “little guy” advantage over the big books, being able to specialize in a single market and get his hands on odds-moving information before anyone else. This hasn’t changed so far even with all the amazing technology bookmakers have on their disposal and I don’t see it changing in the future either.
The post Overround Decay appeared first on The Church of Betting.
]]>The post Review of Trademate (revised) appeared first on The Church of Betting.
]]>Luckily, as I was preparing to write this article I got a message from the Trademate team inviting everyone on the platform to a stream where they discussed the aggregate results of the tips the platform was producing. This was a great opportunity to compare the data I have gathered with the overall performance of all the platform’s clients. It has helped me confirm some of my observations, but also get a new perspective on my results.
An important point I would like to make is that after writing the first version of the article a major issue with the settlement of tennis bets was discovered, which severely affected my results. After settling the problematic bets manually I have updated the results and my comments on them.
Before I dive into the analysis, let me give you a short overview on what the review will be about. I will of course share my experience with using the product, the functions that the software is offering and the customer support. And while all that is certainly important, the million dollar question is whether the product delivers on its promise to find an edge on the betting markets and improve your betting returns. Therefore I will focus most of the article on answering this question. There were some things I liked and some things I believe could be worked on. I will make sure to go through all of these. Here we go.
Trademate is a web-based betting app that scrapes data from the betting markets in order to identify value bets and deliver them to its customers. How does it do that? It will take a base bookmaker offering the highest liquidity for a given betting market and compare all the others to it. Taking that difference and adding up the margin of the bookmaker, it will calculate the supposed edge for that selection and deliver it to you.
You can watch some videos from the Trademate Youtube channel to get a feeling of the platform. I have found it very intuitive and easy to use. You can build all kinds of presets to test different betting strategies. You can choose the bookmakers you want, the target edge, a desired range of time left till the start of the event and a few other things. The Core version gives you a bunch of soft bookmakers to choose from and with the Pro version you get three sharp books added to the list – SBO, IBC (Maxbet) and ISN. You get a section with your open and settled trades which you can use for bookkeeping purposes.
I have used the preset building option to build two main presets. The first one was the widest possible preset including all the books, odds, edges, etc. The goal there was to collect the biggest pool of data possible and cut it into smaller pieces later for a more detailed analysis. Then my second preset was one with only the sharp books, whereby I have added a sound alert to it so that I can catch as many of those as I could. I was particularly interested in the performance of this subset since this was basically what the Pro version offers on top of the Core one.
I have gathered a total of 11554 bets of which 3263 were with sharp books. Below are the overall results for the two subsets of bookmakers:
You will find a lot of data here so I will give you a quick rundown on what is what. Above you see two numbers for Total EV (Expected Value) and Closing EV. These are generated based on the assumed edge of your bet at the time of placing it and at the start of the event respectively. The closing edge is calculated based on the closing price at the base book. The colors seem to be a bit messed up so look at the labels.
You will see that the average closing EV always stands below the total EV. This means that the price at the base book on average moves closer towards the price we have taken as the event approaches. That is not all that surprising given that different books giving different odds represents market inefficiency and those tend to get corrected by the market. However it is worth noting that the gap closes from both sides – albeit surely not at the same rate – meaning that odds at the base book are not necessarily fully efficient. I have seen that in practice too in the form of many drifts at sharp books I have been victim of during arbitrage betting.
This brings me back to an exchange we’ve had with a reader in the comment section of my last article regarding what edge you have when betting the soft side of an arbitrage pair. As I said back then I think the base book is not fully efficient at the time the arbitrage is placed and I believe the numbers above confirm this.
Yet it is commonly stated that the closing price at the base book is the most efficient price you can get for an event. Therefore Trademate calculates a closing edge based on that. The closing edge is supposed to be the actual edge you have for this bet, not to be confused with the edge you see when placing the bet, as the closing price is supposed to be the most efficient one. Again, the margin of the base book is added in the calculation.
Therefore, if all is well, in the long run we would expect our return on capital to approximate our closing edge and our profits to stay around the Closing EV line.
In an earlier version of this article the profit line was running way below the closing EV line. The main reason behind this was that the data provider of Trademate seems to settle Over/Under Tennis bets wrongly. All of those bets were settled as losses whereas of course some of them were winners.
Overall the profit line runs closely to the closing EV line for a large number of bets. From this it can be concluded that the software does its job.
In my trading I have not aimed at maximizing my profit but at expanding the data set I have available for analysis. Using Trademate’s advice and my own intuition I have tested some more refined subsets in order to try and improve my results. In doing so I have noticed some interesting trends and also some issues with edge calculation within specific subsets. I will come to that again later.
Before going to the subsets where it doesn’t go that well, let’s first see where we can find the most profit potential. In the help section of the platform the Trademate team recommends to look for higher edges the further away the start of the event lays. The logic behind this is that the variance of your edge is proportional to the time left before the start of the event, therefore from risk/reward perspective you would want a higher return to take on a higher variance. That makes sense.
Furthermore, in general the team recommends focusing on lower odds and betting closer to the start of the event. I believe that suggestion comes from their analysis of the community results where they have noticed that lower odds & betting closer to kick-off deliver better results.
I have to admit I was a bit baffled by that suggestion. While I understand the risk/reward point, I don’t get why would anyone drop high odds and early bets out completely. Yes, the risk for both is higher but there you can surely find value bets too. In fact, as it comes to time before kickoff I would expect to find more value bets earlier while the market hasn’t matured yet. Furthermore, concerning odds, the higher risk of high odds is considered by the Kelly stake formula, which is used by Trademate, so this shouldn’t be an issue at all.
I decided to test those suggestions using the data I have gathered.
Let’s start with the most obvious one. Higher edge must bring higher returns – I guess that doesn’t need further explaining. Let’s see how the picture looks for different edge ranges. Below I have put the results for 5 different minimum edges in the range of 0-8%.
The profitability in terms of average ROI does seem to improve with increasing edge. Somewhat similar effect can be observed for flat ROI, although the effect is not that strong there, which is a bit concerning. In general it might be concluded that higher edge will give you better profitability. Still, you get a negative flat ROI even for some high edge ranges, which can raise some concerns. I believe there is a clear indicator of value in those edge ranges as for higher edges I beat the closing odds with a higher margin, so the increasing profits are not the result of chance. Then what’s the matter with the flat ROI?
Have a closer look at the average odds for the different edge ranges. As we increase the edge, the average odds of the tips the platform returns increase too. Does that mean that the longer odds contain more value? I don’t think so. I believe that the edge might be overestimated for the long odds. This leads to the fact that for the flat strategy you get long odds selections, the value of which is in fact not that high, if there is a value at all. The Kelly stakes mitigate that negative effect as they adjust the stake to the odds but one can still see the issue in the flat ROI.
That aside, betting with a high edge will give you better profitability, but of course also a lower total number of bets. The advantage you have against the closing odds at the base book is a clear evidence for that. However, I think some “false positives” slip into the data set. I will look further into these with an analysis of different odds ranges.
I believe the task for everyone using Trademate would be to find a balance between a high enough number of bets in order to turn over the whole bank, together with a satisfying expected return per bet. I will give my suggestion on that later.
As I said, the Trademate team recommends avoiding long odds bets. Yet, according to the analysis above, long odds bets on average give you a higher edge, which to me is strange. I decided to look into how long and short odds are performing within the data set. I have looked into the results with minimum odds of 1-6.
The results confirm the suspicion that something is wrong with the long odds bets. Once your odds go above 4 most profitability metrics start deteriorating. I have repeated the test for all possible edge ranges and with very few exceptions the trend holds. It is telling that the flat ROI goes down as well for the high odds ranges. For me this is another clue suggesting that the edge for high odds might be estimated incorrectly.
So far, so good. We have established that higher edges give us higher returns, showing that the software does identify value. We have also noticed, strangely enough, that we should rather drop the high odds as something’s wrong there.
The Trademate team suggested focusing on bets short before kickoff to minimize your variance. Let us see how the performance differs across the Time-Before-Kickoff ranges. I have looked into bets placed 0-6 hours, 6-12 hours and 12-48 hours before kickoff.
The suggestion does seem to have some merit. We get slightly better returns and lower variance for the range of up to 6 hours before kickoff. However, the difference is not that big and I believe excluding the early markets altogether would rob you of too many good bets. If you are aiming at a lower total number of bets it might be a good solution for you, otherwise I would not recommend it.
I still don’t believe that there is more value to be found closer to kickoff. Intuitively it just does not make sense – this is when the markets are supposed to be most efficient. It is indeed the case that betting further away from kickoff increases variance in the edge and the results prove this. I have talked with the Trademate team about that and at this point I am sure there is no error in the margin calculation for earlier bets. My belief is that even though value bets do exist in the early markets, it is harder for Trademate to identify them as Pinnacle is not that efficient before closing. I guess using fundamental knowledge instead of technical methods like the one applied by Trademate, the sharp bettor might find just as many if not more value bets in the early markets as opposed to the late ones.
I was thinking about how an optimal strategy with Trademate would look like. We have seen that the profitability improves for higher edge ranges, that long odds bets suggested by Trademate are generally less profitable and that lower time-before-kickoff does seem to improve results somewhat but not by much. As I would like to aim at the highest total number of bets at a decent profitability I would go with an edge of at least 4% and will avoid odds higher than 4. Here are the results of such strategy.
The profit line moves next to the Closing EV as it should be. You get a solid profitability and a total of 1790 bets for around a month and a half of betting. Very nice!
You have perhaps noticed I have also restricted the maximum edge at 10%. The reason I have done that is that apparently the majority of bets with edge higher than 10% were palpable errors. As you probably know these will not only fail to deliver any profit (as they will most probably be declared void by the bookie) but will also limit your account in no time, so it is best to avoid them. Of course, you can include the highest edges too and decide on the spot whether it is a palpable error or not. Odds comparison website like oddsportal will help you for that. In my experience the majority of them are just that so be careful.
As I mentioned earlier, the Pro version adds three additional sharp bookmakers to the list – SBObet, IBC (currently Maxbet) and ISNbet. You also get Dafabet but I don’t count it separately as it shares odds with Maxbet. I am running at a small profit so far after a total of 3263 bets. However, that does not tell much by itself as in the overall subset I don’t manage to beat the closing line on average so I am still a bit skeptical. Therefore I have had a look at what happens when I increase the edge. You can see the results for the overall sharp data set as well as a minimum edge set at 2, 4 and 6%.
The profitability does look better by some measures as the edge is being increased. The edge against the closing odds increases. However, one can also see that the closing EV line is further away from the EV line at placement. It seems beating the Asians is much harder task as the Pinny odds drift much more in this case. The edges are slimmer and the variance higher. My sample is not large enough to determine what is the actual edge, but the community seems to be running at a 0.6% long-term ROI in the sharp books, which could be taken as benchmark.
Of course, given that we are talking about sharp books here the amount of bets you get is certainly smaller than with the softs, however I don’t think this would be an issue since you do not get limited, so you don’t have to focus on a league or sport and could just bet on everything. The average edge is much smaller in comparison to the whole data set, so you should not expect profitability as high as with the softs, but that’s not unusual either.
All in all so far I think the Pro package would be a valuable tool for the high-roller. I would strongly advise you, given the reduced profitability compared to the softs and the higher price of the product, to consider carefully whether you have the needed capital to make it work. Furthermore, considering the tiny edge you will have pursuing such a strategy, be ready for some serious variance down the road. You can refer to my article on betting variance to get an idea.
Alright, we have established that the software works and you can make a solid profit from it. But I am also suspecting that there might be a problem with the edge calculation for higher odds. What could the problem be?
A common trait for long odds bets is that the bookmakers apply to them higher margins than usual. I have been assured that Trademate does apply the margin when calculating the odds. However how exactly it is applied is unclear to me.
What is the case in practice? Intuitively, one might expect the margin for an event to be applied to all betting outcomes proportionally to their implied probabilities. However, this is not the case. In fact, longer odds receive an over-proportional share of the margin. It could be speculated what is the reason for that, but perhaps bookmakers want to increase their expected profit per long odds bet due to the higher riskiness such a bet brings them compared to a short odds one. You can read more on that in Joseph Buchdahl’s Fixed Odds Sports Betting: Statistical Forecasting and Risk Management.
Furthermore, margins vary per sport, league and country. In general, the margin can be easily calculated for any given event but it must be split between the different outcomes carefully. I might be wrong, but I think a further look into this can significantly improve the performance of high-odds bets in the Trademate platform. This would bring value to the customers as it would expand the total range of viable betting opportunities.
I haven’t said much about the support from the team yet. I would like to thank Marius and the rest of the team as they have been very responsive with every question I had. They also organize periodic meetings to discuss the results of the whole community. I really like that as it shows great transparency, which is extremely important in the betting industry. At those events the customers get the opportunity to see how everyone else is doing and to share their thoughts and feelings about the product with the whole community. All in all I have found the customer support to be at a good level.
Trademate does deliver on its promise to identify profitable opportunities on the sports betting markets. The team has developed an innovative software solution which would be an asset for any serious bettor.
I believe there are some flaws with the edge calculation of certain subsets that I have outlined above, which I am sure will be addressed by the team soon enough. Keeping those in mind you should have no problem turning a profit with Trademate relatively quickly.
Another point of improvement is the collection of data – there is currently an issue with the settlement of tennis trades in the platform, which has seriously impacted my records. I needed to settle those manually, which for data set this big was quite a lot of work. I hope the team will take care of that in the near future.
That aside, the software is easy to use and offers a lot of nice features. The customer support is on a very good level. Overall I would recommend the software to anyone who follows value betting strategies and has the necessary budget.
Some final words
This is certainly my longest article so far, so thanks for reading to everyone who has made it to the end. Thanks to the Trademate team for giving me access to their product, which made this review possible. I would also like to use the opportunity to thank Steve from Daily25 for giving a great feedback to my previous article and helping me improve the count of my Twitter followers substantially. Steve is offering a product based on Trademate for the Australian market, so if you are an Australian bettor make sure to check it out.
Also, if you would like to see any further stats from my data set in Trademate or want to ask anything just drop a comment or get in touch. I am sure the Trademate team will also be around to answer your questions.
On an unrelated note, I have prepared a short e-book on Sports Arbitrage, which you can download from the ‘Sports Arbitrage Guide’ tab on the top of this page. In that way you will also be added to an email list where I will update you on any new content in my blog. Alternatively, you can also subscribe to my RSS feed below this article or in the upper right corner. Thanks for following!
EDIT (12.03.2017, 12:39 CET)
Just wanted to let you know that I have presented the review to Marius from Trademate and he had a few comments on it. I am posting them with some delay here.
Concerning the edge calculation, Marius shared my opinion that there is an inconsistency in the margin attribution for long odds bets. He has let me know that the team is working on developing a solution. I am glad that the issue is being addressed by Trademate and I am sure that solving this problem will further improve the quality of the product. Marius has also stressed that it is advised to focus on lower odds in order to reduce overall variance, since many customers find out they are uncomfortable with the risk that higher odds bets carry.
Regarding the time before kickoff, Marius raised the valid point that the margin could be (and is) easily calculated using the odds at time of placing the bet. Therefore it is safe to assume that the margin is calculated correctly there. That fact only makes the underpeformance of early bets more interesting – since the margin is not the reason, there must be some other factor depressing the returns. At this point I am struggling to see what that might be, so if you have any ideas I would be happy to hear them. Regardless, this might turn out to be quite a useful finding by itself. It is good that platforms like Trademate collect large amounts of betting data so we can notice such particularities and improve our betting knowledge.
It was confirmed that all the closing edges are calculated against the Pinnacle closing line.
I have also added the missing link to the picture in the Time-Before-Kickoff Analysis part.
EDIT2 (26.03.2017, 18:56 CET)
It was discovered an issue with wrongly resolved tennis results has badly affected my results. It turns out all Total Game Tennis bets were settled as losses, which depressed the returns of my data set and most subsets significantly. Since the effect of this is substantial I decided to resolve these bets manually and update the article accordingly. I apologize to all readers for this confusion.
The post Review of Trademate (revised) appeared first on The Church of Betting.
]]>The post Arbing without a Hedge: The Insanity of Betting Variance appeared first on The Church of Betting.
]]>So at the end it all comes down to whether you think the price of getting rid of this variance is worth it. It is hard to give an intuitive answer to this question if you don’t have a good feeling for how much variance you should expect. I have run some calculations to show you just that.
So I want to simulate a strategy of placing arbitrage bets on soft books without covering them in sharps. It would be similar to a value betting strategy with a certain edge.
I start with a bank of 100 units, placing 1 unit per bet.
For simplicity, it is assumed that all bets are placed at odds of 2. Since odds of 2 imply a probability of 50%, this must be the average odds of your bets in the long run.
I have assumed you are always getting an edge of 1%. In my experience this is approximately your average return on turnover when doing sports arbitrage, so I consider it to be a realistic number. This means for every bet with odds 2 and implied probability of 50%, your real chance of winning it will in fact be 1% higher, or 50.5%. Your chance of losing will be 49.5%, respectively.
I have run three simulations: one with 100 bets, one with 1000 bets and one with 5000 bets.
Using the above preconditions, I have run a 1000 Monte Carlo simulations for each number of bets to find out what is the chance to end up with a loss and what is the chance to blow your bank.
A “Monte Carlo simulation” might sound scary to those of you who haven’t dealt with it. In fact it is quite simple. I just simulate a thousand instances of 100/1000/5000 consecutive bets under the conditions described above, using the Excel random number generator (the RAND() function). Here is what I have found out.
You cannot blow your bank in 100 bets under the above conditions, as for this you would need to lose all 100 of them, which is extremely unlikely. However, what is the chance to end up with a zero or negative return after the 100 bets even though you have an edge of 1%? Higher then you might have thought, it turns out. When simulating 1000 instances of 100 bets and looking at how many of those instances failed to show a profit, the average of all simulations seems to lie around 50% and I very rarely get anything below 48%.
In fact, to get the exact probability of failing to turn a profit we don’t need a Monte Carlo simulation. Since the profits of such series would follow a binomial distribution we might simply calculate it. The actual formula is:
…but since I’m way too lazy for that, I just go to this binomial calculator, input p=0.505, number of trials = 100, number of successes = 51 (the minimum number required for a profit) and hit the button. It turns out in exactly 49.99% of the cases, 100 bets with an edge of 1% will fail to turn a profit.
Think about that for a second. In a single arbing session you might place anywhere from 20 to 100 bets. But if you have a few bad days, don’t have access to that many bookmakers and you are arbing only on the weekends 100 arbs might be your total for a couple of weeks. So you might go for a couple of weeks of arbing without hedging and there will still be a 50% chance that you will have nothing to show for it! That would be very disappointing and very likely too.
Let’s see what happens when we increase the number of bets to 1000.
As the number of bets increases, we would expect our edge to finally materialize in a nice profit. Indeed, the percentage of negative profits for series of 1000 bets looks better. Not much better though. Again using the binomial calculator, we find there is a 38.79% chance to end your 1000 bet adventure without a profit.
1000 bets can extend to one or a few months of arbing. You must keep in mind that in the meantime your money is locked and you need to invest in transaction fees and software subscription, which are not accounted for in the equation. And at the end of it, if you don’t hedge in sharps, you have a 38.79% chance to have absolutely no profit.
What’s more, there is even the slight chance that you blow your bank. The Monte Carlo simulation of 1000 instances shows a destroyed bank or two. The chance for this to happen seems to revolve around 0.1%. So what if we increase the number of bets even further to a 5000?
That must be much better, right? Well… kind of. The binomial calculator says there is still a 24.41% possibility that you fail to turn a profit. That is 1 out of 4, not exactly negligible.
5000 bets is easily half a year of arbing. 27 arbs per day if you arb every single day. Considering that summers are usually quite weak and you will have to go out of your cave to communicate with real people now and then, that probably translates to at least a full year of arbing. And you have a 1 in 4 chance to fail to make any profit from it. On top of the wasted time you will need to cover a year of arbitrage software fees. Ouch.
But it doesn’t end here. With such a long time span you get the added benefit of around 5% chance of blowing your entire bank, as the Monte Carlo simulations show.
I have run a simulation on a series of 10000 bets whereby the chance to blow your bank increases to around 10% and the chance of finishing in negative territory is at 16.10%. You can play with the numbers some more with the calculator linked above.
The numbers above are all for an edge of 1%, which I find to be a realistic estimate for most arbers. I have also run the calculations for an edge of 2% and 3%. Such a high edge is rare, but certainly possible to obtain. Intuitively, with a higher edge it must be less likely to blow your bank or finish in negative zone. The question is, how much less likely?
Below I have summarized the results for the different sample sizes and edges. You will find the probability of negative or zero return first and the probability for losing the whole bank second.
100 bets | 1000 bets | 5000 bets | |
---|---|---|---|
1% edge | 49.99% / 0% | 38.79% / ~0.1% | 24.41% / ~4.5% |
2% edge | 45.99% / 0% | 27.39% / ~0.01% | 8.07% / ~1.25% |
3% edge | 42.04% / 0% | 17.94% / ~0% | 1.75% / ~0.2% |
Given the above, should you skip hedging your arbitrage bets? Well, if you are fine with arbing for a year or two without turning a profit you could … I guess. Unfortunately I don’t know many people who would wait for that long. Moreover, you need to pay your arbitrage software and transaction fees in the meantime and they don’t really depend on how well your arbitrage business is running.
So for me, it’s a clear no. I always hedge my arbs and I believe this to be the superior strategy from a risk/reward perspective. But since everyone has their own risk appetite, feel free to decide for yourself. Just make sure to keep the numbers above in mind.
By the way, this simplified example can also give you an idea about the inherent variance in most value betting strategies. An edge of around 1% is what most successful value bettors and tipsters work with. You might make good money, but you would be in for a wild ride.
If anyone has some related experience, I would love to read from you in the comments. If you want to run your own Monte Carlo simulations you can check out this video or drop me an email at admin@churchofbetting.com and I will give you my Excel sheet.
Thanks for reading and see you around!
The post Arbing without a Hedge: The Insanity of Betting Variance appeared first on The Church of Betting.
]]>The post Maximising Profits – How to Improve Your Turnover appeared first on The Church of Betting.
]]>So you are already up and running. You got familiar with the main risks and pitfalls and even managed to score a few profitable arbs without getting limited (yet). It is time to think about maximising profits with the capital you have available.
Your arbing profit is in reality your employed capital times your arbing returns. Therefore, these are the two things you have to pay attention to. Maximizing your returns is pretty straightforward. Your arbitrage service will give you the return of every arb pair delivered. You should just strive to get the ones with the highest return (of course, keeping palpable errors in mind). On the other hand, how do you go about employing the most of your capital? In this chapter I want to give a few insights on that part.
When you are getting started with sports arbitrage money management is key, especially when you are tight on budget. You would like to avoid making too many transactions and it might possibly be you are betting on sharp bookmakers through a broker, so you need to be able to plan how much money you are going to need for each bookie or group of bookies.
Furthermore, in any given day you are arbing you will want to deploy your full bank as this is what will bring you the highest return on your capital. Imagine having a 10k budget split 5k in soft and 5k in sharp books. It runs well and at some point during the day you are out of money in the soft books but are still left with 2k in the sharp books. That means you have 10k available but you only used 8k for Sports Arbitrage and you cannot use more as a withdrawal of the other 2k might take days. That would be a pity. So what is your best approach?
Knowing that an arb normally involves a sharp and a soft book, splitting your budget 50:50 between softs and sharps would be a reasonable starting point. However, there are a few more factors to consider, as you are going to see.
First, since sharps are more competent in setting the right coefficients, in the long-term you will expect at least in theory to win in soft books and to lose in sharps. In my experience this is true in practice as well, meaning you would need a bit more money for the sharp books as your balance there tends to decrease.
Perhaps you are also thinking now, what would happen if you just do not bet in sharps at all? You expect to lose money there, so why not cover only the part of the arbitrage in the soft book, wouldn’t that increase your expected profit? These are very legitimate questions and are in fact central to the theory of sports arbitrage. See my in-depth answer to this question in the following article. In short, I always prefer to cover my arbs as not covering is going to increase volatility tremendously and this is something I prefer to avoid.
So we know that it is more likely to lose money in the sharps and win in the soft books. Now, let us look at the odds. If we are getting the same average odds in both soft and sharp books, they would not play a role in our budgeting decision. However, that is usually not the case.
In my experience, in an arbitrage pair you will more often get the higher odds in the sharp book and the lower ones in the soft one. Meaning you will have to use more capital in your soft book and it has to be better funded than the sharp one. There are of course exceptions but on average the rule holds. So this is one factor in favor of funding your soft book accounts better.
Finally, when in a certain book you’re all out of money, give preference to arbs involving higher odds in the same book, even when the returns are a bit smaller. You might have to wait a bit for those, but since you are out of capital it is not that urgent to place arbs right away anyway. Your total profitability would improve even when you take lower return arbs since this return will come on a larger pile of capital.
Taking all of the above factors in mind, you have to decide how to optimally split your capital. I cannot give you the exact numbers as that really depends on the markets you are arbing on, but you must aim at the optimal distribution in order to maximize your employed capital and improve your returns. You will get a good feeling for that with the passage of time. Just keep in mind that maximum capital * maximum return = maximum profit.
The article above is just another chapter from the Sports Arbitrage guide that I am preparing and will publish quite soon. Since I know this stuff is interesting to you I give it to you here as well so you don’t need to wait until I’m done with all the structure and formatting. But the material is growing bigger and bigger every week and I think it will make for a nice and comprehensible strategy guide.
I continue to work on the review of the Trademate Sports software. I have already collected a total record of around 2500 bets of which 500 in sharp books. Not bad in itself, but I am hoping to gather at least double that to have a reliable sample. As the guys from Trademate themselves say, you should compare your monthly and not your daily returns. I stand by that and continue to gather data in order to be able to present to you the full picture. So if you are looking forward to this, please be aware that it will take some time but you might be sure I will not compromise with the quality of the analysis. Just follow my Twitter, Facebook and Google+ pages you will be among the first to get it.
The post Maximising Profits – How to Improve Your Turnover appeared first on The Church of Betting.
]]>