A Look At The Books

While we eagerly await the final matches of the weekend and publication of this weekend's FTL table, I took a few minutes to review the odds used to determine the totals.

I had a little more time on my hands this weekend than usual due to a tactical change from FTL entrant Punters' Friend, aka Neil, who has decided to try a more filtered and rigorous approach to his selections.

As most readers will know, I chose to record this season prices for everyone in the table using those recorded for Pinnacle Sports taken from the incredibly useful Football Data site.

One thing I wasn't aware of until today is that for weekend games, the prices are collected on Friday afternoons. I had assumed that they were closing numbers, or at least close to lick-off prices, and it certainly goes some way to explaining why there is often a fairly sizeable discrepancy between the odds recorded, and the odds seen at kick-off.

This discovery came about while I was researching how good Pinnacle's prices really are for football. I continue to use the exchanges for most of my betting, and the debate on whether they really are usually the best prices after taking into account commission and possibly Premium Charges on Betfair continues, but the consensus was that Pinnacle were the best, and that was why they were chosen**.

"Best" can mean more than "best odds" of course. The ability to keep a winning account open is a rather important consideration.

Looking at the full English Premier league season of 2012-13, it may surprise many that if you were looking to bet on the favourite, Bet Victor were top-price in 107 of the 380 matches.

Of more interest to me is the draw price, and if this is your speciality, Pinnacle Sports come out as top in the most markets, but only 73. That means 75% of the time, you would find a better price elsewhere.

Football Data used 9 books for their recorded prices, although since Bet & Win and Gamebookers are absolutely identical, it's really only 8. The other seven are Bet365, Interwetten, Ladbrokes, Pinnacle, William Hill, Bet Victor and Stan James.

That may seem like a broad sample, but Football Data uses an even broader sample of books to calculate their average and maximum prices, last season anywhere between 33 and 41 with 39 the average number polled. More often than not, the 9 (or 8) books are not those offering the top price.

Of the named books, for backing the favourite or the home team, Bet Victor by quite a long way offer the top price most often, and they are close to being the best on away sides too, just one match behind Interwetten and Pinnacle Sports.

Unfortunately Bet Victor have a reputation for being very quick to close accounts, with reports of accounts being closed after as few as two bets, so there top prices may be of academic interest.

If we ignore the frankly awful Bet & Win / Gamebookers books, the worst for favourites are Bet365 followed closely by William Hill and Pinnacle, the worst for Home teams are William Hill followed closely by Bet365, for Aways the worst are Ladbrokes follwed by William Hill and for Draws, you should forget about Ladbrokes.

This is a limited study of course, and the results may be different across other leagues - certainly the Pinnacle draw price in France and Italy is often ridiculously short and much higher on Betfair, but it was an interesting exercise and revealing how often the big names are not competitive.

Note that top-price counts include joint-top prices, so the totals will not be 380.

Update: ** An example from tonight is the West Bromwich Albion v Everton game. This was an XX Draw Selection, and was backed at 3.65 on Betfair, and was 3.55 / 3.6 at kick-off. The price recorded at Pinnacle? 3.39 

Obfuscation

A few more thoughts from Ballabriggs on the topic of the week, and I'm not talking about ten-fold accumulators. These were originally posted on the Betfair forum, but since he didn't object to me stealing his previous entries, I'm hoping he will forgive me again.

One more glaring clanger in Sporting Data's statement is this

"Bets will be placed when the odds generated by the model are significantly out of line with the market." .

What they actually mean is

"Bets will be placed after every point, but will only be matched when the odds generated by the model are significantly out line with the market."

The courtsider will press a button saying Player A or Player B has just won the point. If they click Player A, bets are instantly put in with a lay of say 1.69 and a back of 1.74. [If they click Player B, bets are instantly put in with a lay of say 1.79 and a back of 1.84]. Bets will be put into the market every point, and matched if they are out of line with the spread of 1.69-1.74, or 1.79 to 1.84.

When SD say "There are plenty of times that they will send info back and nothing will be done with it., that means that on those particular points, they didn't manage to hoover any out of date money, and will come back to try again after the next point is won or lost.

When SD say "the odds and stake of the bets are all determined by the individuals in London and not by the employees on court"., if you were paying SD for this expensive information, you would set your model to instantly put in bets of 1.69-1.74 or 1.79-1.84, should the courtsider press button A or button B. There really is no point paying top dollar for SD information, if you're going to manually see a point has been won or lost, then decide a stake size, then type in your bets and increments, then faff a bit more etc etc. There is no conceivable way these SD related trades are not automated, the courtsider presses a button, and this triggers the computer in London to automatically put in 1.69 lay/1.74 backs, or 1.79 lay/1.84 backs, all done in nanoseconds.

I do not see any realistic end difference between taking a computer courtside and placing the bet, to having the SD tactic of having a courtsider press button A or button B, and in nanoseconds a signal is instantly sent to London, where trades are automatically smashed into the exchange markets. Yes the London SD customer (or SD themselves) have programmed the bot to calculate stakes/exposures instantly, yes the bot calculates the correct back/lay prices instantly, and yes the signal from the courtsider results directly in bets going into the markets, which may or may not be matched within 5.000001 seconds of button A or button B being pressed, taking advantage of any recreational gamblers who are unlucky enough to have either lost concentration, or just not known about the hoovering.

Betfair have a serious problem now, which isn't going to go away. Sporting Data have been operating since 2011, and to be fair, Betfair must have known both from social contacts between current staff and the three (ex-?) staff members who set it up, and from key account holders, what Sporting Data, and other organisations are doing. There is no way that this 22 year old chap in 2014 is the first person they have ever sent into an event, and it is extremely likely they have been doing this both at previous Australian Opens, and at other world tournaments, where for some of them there are supposed to be draconian legal restrictions on betting in-running.

The Australian regulator is much more aggressive and sharp-fanged than the British one. If they work out that Betfair have known there are former employees running companies and sending people to the Australian Open, pressing buttons which result in trades going into the market live after 5.000001 seconds of 22 year olds pressing Button A or Button B, there is no reasonable difference between doing that and actually betting courtside to get bets in live after 5.000000 seconds. This loophole is very, very ugly, and since Betfair could have done something about it in the 3 years in which Sporting Data have been running, and haven't done anything about, one day Betfair's luck will run out. It may or may not be technically illegal, but it is against the spirit of the laws banning in-running betting like this from the Australian Open, and it really wouldn't be a shock for betting licenses on events from Australia to be thoroughly reviewed across all sport. Betfair could have stopped all this, but I just hope the blind eye doesn't come back to haunt the 99% of punters who want to bet fairly, and Betfair themselves.

It dos seem somewhat implausible that the data is received in London, and then passed on to their clients, who would surely then all be competing against each other for the same money. How does one get to be a client anyway? The Sultan may know more about how they operate, but in his case he appears to have been a temporary employee working on a commission basis. He does state though that Sporting Data are an example of an "ordinary betting syndicate", which are "legitimate businesses that make their money from sports trading - just like the one I traded for a little over a year ago". 

Yet Sporting Data themselves say that they make their money from selling their data to clients - see 1) below:

We wish to make the following points and clarifications.:

1) Sporting Data is a company providing sporting data and services to certain individuals. Our clients use the information provided to place bets.

So the clients (plural) use the information to place bets.

2) We have employees on the courtside sending back information to London and that is then used to place bets on the outcome of the match. However, the odds and stake of the bets are all determined by the individuals in London and not by the employees on court. There are plenty of times that they will send info back and nothing will be done with it. In no way could they considered to be betting themselves.

No mention in this line item about who actually places the bets, other than "individuals in London", which is odd. Why would all their clients be based in London? And why would Sporting Data make a decision on whether or not to pass information on to their clients? Not all points in tennis are equal of course, but the decision on whether or not to make use of it should be that of the client.

3) We use mathematical models to assess the probability of a match outcome. Bets will be placed when the odds generated by the model are significantly out of line with the market. A lot of syndicates use a similar methodology. Clearly, we need the most up to date information to generate accurate match probabilities. We cannot rely on TV pictures as they are out of date.

Again the 'bets will be placed' statement is rather vague, but it is easy to infer now that it is now a syndicate placing the bets rather than "individuals", and the "we" implies that Sporting Data are doing more with the data then selling it.

4) Most of the bets placed are placed on betting exchanges - platforms devised specifically for punters to pit their information and simulation techniques against each other. It is the tightest and most competitive environment there is and we have invested a lot of time and effort to become competitive.

How would Sporting Data know how the data they sell to their clients is used, and where? If Sporting Data's business is "providing sporting data and services to certain individuals" their time and effort would be in the area of data transmission, and the environment the data is used in would be an irrelevance.

5) We have no interest in corrupting the outcome of a match. We want both players in any given match to perform as well as they possibly can. Our models fail to work if the result is any way compromised and we would take all steps to avoid betting on any match we suspect to be corrupted.

Again, why would a data provider have a model? The odds and stake of the bets in 2) are being determined by the individuals in London. 

For a statement that is meant to be a clarification, it really is quite confusing. It's more of an obfuscation than a clarification.

Either Sporting Data's business is selling data, or they are a betting syndicate making their money from placing bets themselves. Since the data only has value for a few seconds, they can't be doing both. Sloppy seconds doesn't work too well in the betting markets.

If they are placing bets themselves, then as Ballabriggs says, there are likely to be some awkward questions asked of Betfair and their integrity team turning a blind eye to something that has been happening for three years at least. 

Anonymous commented on my previous post that:

I'm not sure I agree with all points made here. In-play will inherently always be imbalanced from a time perspective, it's the nature of the beast. Radio beats TV, analog signals beat digital. The in-play delay should protect you in as far as you being able to cancel your bets? This is your protection. I don't get my fingers burnt any more in in-play markets, it was an easy choice. So don't bet/trade in them if you can't compete? Simple.

Agreed, but the in-play delay is not sufficient to protect customers, and the Gambling Commission state that:

This is done in order to ensure that the odds on offer reflect the progress of the event and that bets made are a true reflection of what is happening with an up-to-date market.

And agreed that if you can't compete, you should stay away. The problem is that not everyone is equally aware of the risks - as the Gambling Commission report stated:

Although betting with varying levels of knowledge and skill is not unfair in itself, there may be a problem where betting customers perceive their chances to be better than they actually are due to a lack of knowledge of the in-running betting in terms of the subject-matter itself or the actual market and how it operates.

If the same account is always first to act, and this move is beating the countdown clock, then there is a problem. Betfair may not have a legal duty to protect their clients, but they have a moral duty to do so, or they run the risk of being accused of treating some accounts favourably, something that is even more bad-tasting if it belongs to previous employees or their business associates.

It would also be in their own long-term interest to ensure as fair a playing-field as possible. Betfair's business model would surely be better served by limiting any unfairness to "knowledge and skill" and keeping players in the game, than by allowing privileged accounts to slowly suck the life out of them.

Yes, court-side is fastest, then radio, TV etc., but clearly the difference between first and most is often too large, and the Gambling Commission might reasonably expect Betfair, or any exchange for that matter, to close such a loophole. None of us want to see in-play trading banned, an action that would simply drive it underground or to more dubious jurisdictions. One of Betfair's advantages is that it is a legitimate UK company, trading on the London stock market, and well regulated, but an exchange surely has a duty to its customers to ensure that the infrastructure is as fair as possible. If someone is constantly taking an advantage of a loophole, the loophole should have been closed long before three years, whether the beneficiary is previous employees, their associates or an entrepreneurial individual and the Gambling Commission and whoever the Australian regulator is will know this.

Gambling Commission In-Play Integrity Concerns

The debate about Sporting Data and their business model continues in the blog and Twitter worlds, as well as mainstream media, although old media in general appear to have no clue about the issue, confusing it with match-fixing which is clearly nonsense as I have previously written.

Sporting Data are not doing anything legally wrong, but:

“Is it fair play and is it morally the right thing to do? I wouldn’t think so,” the founder of online betting company Sportsbet Matthew Tripp said. “But I wouldn’t call it illegal.”

And I wouldn't either. Sporting Data, and reportedly others, have clearly invested much time and money into exploiting this opportunity, as they are quite entitled to, but there are some serious questions that should be asked.

The media also seem obsessed with the idea that Sporting Data's clients are betting on points. They aren't - they are betting on the match outcome, but with the advantage of being a point or so ahead of many.

Hejik commented:

A hot topic and some interesting points raised. I don't think Betfair operations are really ever called into question quite as much as they should be. Considering this style of operation would run into "Super-PC" very quickly indeed, it certainly makes one wonder if (a)there is some kind of "agreement" in place (wouldn't be the only "agreement" Betfair have with a customer) or (b)indeed they are levied the full PC amounts but happy to pay it regardless as there is no risk involved whatsoever. If losing weeks are not an option then PC is in fact no problem at all. It's merely an inconvenience, like Income Tax. Also, I'd suggest this arrangement suits Betfair all ends up (tight markets, illusion of liquidity, regular income) Questions clearly need asking before they can be answered but I'm struggling to see Betfair's motivation for making them go away.

So one question to be asked is if there is a special arrangement in place for former employees or certain companies and their clients to somehow be exempt from the Super Premium Charge. As Hejik says, if there is minimal risk in their trading strategy, then perhaps Sporting Data's clients are happy to pay the charge. Perhaps, but another related question is this.

Many traders found themselves left high-and-dry when Betfair introduced their Premium Charge and again by the Super Premium Charge. Several had given up jobs to reap the rewards of full-time trading, only to find that overnight the rules had changed significantly. Would Sporting Data's directors, and others, have been so willing to take the risks involved in setting up their businesses without some kind of guarantee that the goalposts would not be moved again, or at least not for a minimum period? 

The £250,000 limit on winnings is one variable that could easily be changed in an instant to any lower figure, and while exchanges proudly claim not to ban winners, at least one exchange has been accused of doing so, as was discussed a little over a year ago, and there's no reason why others shouldn't do this in the future. It would be a huge risk to give up a career (or three) for an enterprise that could be stopped in its tracks at any moment.

In fact, it may well be in Betfair's interests to ban winners in some instances. Anonymous left this comment:

One also wonders if Betfair has any kind of a business relationship with Sporting Data?" Many courtsiders I talked to along the past years, they all have the same suspicion: someone out there is still faster than them. It has to be somebody on the court, of course, but also with some kind of technical advantage. There are two theories: it could be someone related to the ATP circuit, who enjoys a fast dsl connection wherever he is, while the other courtsiders rely on unstable 3g/4g connections. Or Betfair is getting their cut on the markets using a third part - somebody with a little plus that allows them to put their hands on the money a bit faster than anybody else.

I'm not feeling too much sympathy for court-siders being beaten by other court-siders, (and hard to see how any technological reason would explain a five second delay) but another, or possibly the same Anonymous, had this to say:

Superb article. Would not surprise me if Betfair are taking a cut of courtsiding profits . In effect Betfair are killing off the exchange as they have run out of people to bleed dry.

G added:

Great piece again, however it does leave me wondering if there is any chance for the ordinary punter in these markets where syndicates with models are operating? Talk about being up against it, cant see the field ever being level though.I suppose like sharks they'll always need the little fish to feed off- so, easy solution, all the small fish (casual punter)should leave Betfair now..:) I've also found the NBA in-play markets virtually impossible nowadays so back to pre-play markets for me as well- perhaps 10 team accas are the way to go......:)

The NBA markets are different this year. For example, if the true price is 2.0, the back and lay prices available are 1.8 and 2.2. I haven't traded them so much this season (getting old, need my beauty sleep) but when I have, this is what I see. Someone with access to real-time data appears to be mopping up anything of value. It would be interesting to know if liquidity is down on the NBA this season. As for the 10 team accumulators, they are certainly NOT the way to go :)

It's not in Betfair's interest to publicise that some users have, quite legally, access to price-changing data several seconds ahead of others, and are regularly beating the countdown. They have to put their disclaimer out there of course, and hope that like most small print, it is widely ignored.

Betfair want their customers, and also the Gambling Commission, to believe that the playing field is almost even. There will always be some ahead of others, but so long as the time between the first and last is less than the countdown, then the field is playable.

Some readers may remember the X-Factor 'scandal' of 2011, when the Gambling Commission found three Virgin Media Inc. employees had misused access to company data on telephone voting on last year’s edition of the television talent show to bet on which contestant would be eliminated. They were caught after Betfair flagged suspicious betting patterns. While the data from the Australian Open isn't technically inside information, it certainly isn't information that is (for a few more seconds at least) in the public domain. Again, nothing illegal about being the first one to see the news, but surely Betfair can see that allowing a few sharks to consistently clean up isn't a sustainable strategy.

There are some out there who claim that the presence in the markets of such organised opposition is just a minor inconvenience, but this is somewhat disingenuous. On any given day, anyone can of course win money in any market, but it should be obvious that if someone is consistently, say ten seconds ahead of you, any bets you are matched at will be bets being made at poor value. Poor value doesn't mean a loss in any one market, but it does mean a loss in the long run. No one can beat poor value over time - it's kind of what 'value' means. The exception does not prove the rule, and while Matt's (@puntdotcom of the now dormant punt.com) big win last night on Ana Ivanovic's win over Serena Williams was fantastic (although very disappointed to read on Twitter that he "did an Iverson" and locked in his profit at poor value...), it comes despite the presence of syndicates in the market. Incidentally, with the sums involved for Matt, doing an Iverson wasn't unreasonable. It may not have been optimal, but understandable.   

The Sultan is one proponent of this dangerous idea, but as a trader who has, by his own admission, worked with a set-up similar to that of Sporting Data in the past, we should be wary. Anyone with access to fast data of course has a reason to take this approach and do what they can to encourage 'normal' people to keep playing the markets. After all, if there are no tiddlers in the pool, the sharks will die, but to imply that Mr Normal has just as much chance of success as those with fast data is nonsense. In fact, one now wonders how much of the Sultan's turnaround in fortunes is due to this big advantage. Quite a lot, I suspect. There have certainly been several posts hinting at something a little secretive was going on.

January 2013:

I've had a very interesting offer put on the table for me. I've been given the opportunity to trade with a bigger bank. Basically, I've been approached through this blog by someone who is willing to risk their own money in order to take a share in my profits. I won't go into exact specifics of the deal but it is still being discussed before we come to an agreement on all aspects. Suffice to say, I'm a tad sceptical but at the same time, very excited. If this goes to plan, I will be able to make significantly larger profit than I have so far been able to accrue, due to my relatively small trading bank, even accounting for the shared profits. This was something I was not going to reveal on the blog but as the interested party shows no problem with it, I don't see why I should keep it a secret.

And now we know more. I'd be excited at trading with data ten seconds ahead of most people too! I hope this significant edge was revealed to prospective Sultan Academy members though, otherwise a few people might be claiming refunds due to a lack of full disclosure.

The nature of gamblers, as opposed to investors, is that they will bet despite the odds being against them. There's a reason that casinos and bookmakers thrive, but the arrival of exchanges a decade or so meant that the sharp-minded (Betfair's words, not mine) could make a profit. In-play was a revolutionary idea, and one that has been very kind to many of us, but the revelation that this kind of activity is taking place is sure to raise questions and if Betfair and other exchanges don't close the loophole, lawmakers likely will.

The Gambling Commission's mission statement is "Keeping gambling fair and safe for all" and their 2009 report into in-play betting wasn't wholly accepting of it in my opinion, but took more of a 'we'll let it go for now and keep an eye on it' approach. Much emphasis was on making sure that bettors were aware of the risks, and that not all data-feeds are born equal.

At the risk of ruining the ending form you all, their conclusion was this:

Although in the last 18 months we have received few specific complaints about particular aspects of in-running betting we have heard from those who have general concerns about in-running betting in the context of the three licensing objectives. However those concerned did not provide much by way of specific evidence or corroboration of those concerns. From the evidence we have received and considered to date, we have come to the conclusion that the current regulatory regime we have in place for in-running betting is sufficient and does not need further controls at this time. However we will continue to monitor in-running betting as part of the Commission’s compliance programme. Particularly we will focus on developing information sharing under licence condition 15.1. We will address in-running betting’s potential for facilitating cheating as part of the Commission’s wider efforts to maintain integrity in betting. 

Other highlights include:

In-running betting offers some increased potential for individuals to exploit this form of gambling illicitly for their own benefit. However, other forms of betting and gambling also have the potential for such exploitation. On the information currently available the Commission does not consider that in–running betting requires special regulatory treatment but it will be addressed as part of the Commission’s wider efforts to maintain integrity in sports betting.

Note the '"On the information currently available" qualifier. Five years is a long time, and it will be interesting to see how this week's revelations are received if they are to fulfil their duty to keep gambling fair. There's a difference between betting into a 110% over-round, and betting against someone who knows what just happened. As the Commission's report said:

Although betting with varying levels of knowledge and skill is not unfair in itself, there may be a problem where betting customers perceive their chances to be better than they actually are due to a lack of knowledge of the in-running betting in terms of the subject-matter itself or the actual market and how it operates. 

I think the scenario being discussed might fit under the 'may be a problem' section. The report continued:

At present, operators set a time delay so that when a customer places an order for a bet in-running they will have to wait a number of seconds between pressing the ‘place bet’ button and receiving confirmation that the bet has been made. This is done in order to ensure that the odds on offer reflect the progress of the event and that bets made are a true reflection of what is happening with an up-to-date market. It should be noted that there is a difference in reasoning between traditional bookmakers and betting exchanges – exchanges put in place a standard delay to protect customers, traditional bookmakers put in a delay to ensure their own prices are correct. This delay usually covers the time delay in the speed of feed that is being used to adjust the market and a small margin to enable the operator to react to a significant event.

Are bets being matched at odds that reflect the 'progress of the event'? Is the time delay in place enough to protect customers? And more:

It has been suggested to us that the ability to obtain information ahead of the rest of the market would constitute an unfair advantage. This could apply to betting customers with shorter time delays on their feed showing the event, or equally to the betting customer who is betting while actually at the live event. The counter-proposition is that provided that customers are fully informed of the risks that they run, they should be free to decide whether or not to bet.

 The Commission would not take action against technological advances per se unless the risk to the licensing objectives were proven or probable. On the basis of our information to date there is an insufficient case in relation to any threat to the licensing objectives for regulatory action over the use of technological advantages such as differential feeds or use of bots. 

Information to date. By allowing certain individuals or syndicates to profit from a timing loophole, even if currently totally legal (but 'morally questionable'), Betfair risk the regulatory implementation of a complete ban on in-play betting, and many sharp-minded traders, and indeed Betfair, would lose out. Their integrity team knows who the consistent winners are, but it appears that for some reason they are turning a blind-eye to it. If the Gambling Commission are doing their job, and it's hard to imagine they are not following the news from Melbourne last week, we should expect some changes. The status quo may not be illegal, but it is clearly unfair.

Update: Just noticed an interesting post by Ballabriggs on the Betfair Forum, pointing out something I had missed:

1) Sporting Data is a company providing sporting data and services to certain individuals. Our clients use the information provided to place bets.

3) We use mathematical models to assess the probability of a match outcome. Bets will be placed when the odds generated by the model are significantly out of line with the market. A lot of syndicates use a similar methodology. Clearly, we need the most up to date information to generate accurate match probabilities. We cannot rely on TV pictures as they are out of date.
------------------------------------------------
There's another rather large rat that has been lovingly dipped into the bog of eternal stench right there. In the first bit they claim that the company is one which is in the business of providing data and services TO its customers, and then saying it is their clients who use their data and services to influence their trading. The second bit says that Sporting Data itself is using mathematical models, and placing bets according to those models, and that "we" [sic] need the most up to date info to do what they're doing. Maybe in the 60 seconds between writing parts 1 and 3 to their statement, the goblin king had cheekily Oublietted them.

They've basically left Betfair having figured out what they thought was a way to make big money using a non-skill based edge from the Australian Open (amongst others). It really is time for Betfair to put these clowns to the sword.

It is the tightest and most competitive environment there is and we have invested a lot of time and effort to become competitive.

should read

"It is the tightest and most competitive environment there is, and we have invested a lot of time and effort paying for flights and hotels, sending young people around the world, and finding the best devices to hide on their persons, using the finest creative tailoring we could find, to take advantage of loopholes which we think will get us to be able to try to beat the clock from the Australian Open".

Increased Customer Confidence

Having drawn attention earlier in the week to the fact that only one team at sub 1.3 failed to win in the English Premier League last season (2012-13), right on cue comes the second such occurrence of this season, as Liverpool failed to beat Aston Villa and thus joined Manchester United who lost to West Bromwich Albion at 1.29 back in September. At least we should be spared the heart-wrenching Press Releases this weekend detailing just how tough the life of a modern day bookmaker is, words that in our minds are accompanied by the saddest music emanating from a tiny violin.

Ralph Topping, William Hill CEO, issued this Press Release last week:

Current trading

Online Sportsbook has continued to show strong wagering growth, up 48% in the first two weeks of 2014. However, football results in week 2 were highly unfavourable with an unusually high number of odds-on favourites winning. Driven by the impact of this on our otherwise very attractive accumulator business, we recorded a £13m loss in the week. There is no certainty that we can recoup this shortfall to internal expectations but based on previous experience of such customer-friendly outcomes, such as 'Dettori Day' in 1996, we anticipate a positive benefit from increased customer confidence, particularly with so much of the season ahead and with the 2014 World Cup to come.

Coral's Simon Clare released this yesterday, with similar expectations of increased betting as word spreads about how easy it is to win big money:

Britain’s bookmakers are expecting a massive increase in football betting turnover tomorrow after the most punter-friendly football results in recent history cost the betting industry over £30 million pounds last weekend.

Many thousands of customers across the country won sizeable four figure sums for stakes of less than £10, with twenty five Coral betting shop customers winning sums of between £10,000 and £89,000. Coral is anticipating that the huge national media and social media coverage of all these stories of successful punters will encourage more people than ever before to have a flutter on football on Saturday.

“We suffered record losses on football last weekend that will live in the memory for a very long time. However we expect that the tales of such an unprecedented number of successful punters and that have been shared across the media, the internet and through social media all week, will encourage anyone who’s ever had a flutter to have a small stakes accumulator on this weekend’s football to try and emulate their success,” said Simon Clare, Coral Spokesman.

“Last weekend’s football results were almost as bad as they could be, as 15 of the 16 most popular teams won, so we are looking forward to a huge upsurge in bets tomorrow and just crossing our fingers that lightning doesn’t strike twice”, added Clare.

To help anyone struggling to follow the yellow brick road to riches, Simon helpfully included the "Most Popular Top 15 Selections on Accumulators with Coral this weekend".

And how did they fare? Of the thirteen matches complete (Rangers play tomorrow, and Wolverhampton Wanderers match was postponed), six of the favourites lost, including Liverpool as mentioned above. The full list was:

Arsenal, Man City, Liverpool, Leicester, Burnley, Birmingham, QPR, Notts Forest, Leyton Orient, Brentford, Wolves, Chesterfield, Scunthorpe, Celtic, and Rangers

Business as normal in other words, but don't expect any Press Releases reporting on record profits.

Quantennis

From the Sydney Morning Herald:

For someone who says he bets millions of dollars on tennis a year, sports gambler Elihu Feustel doesn't watch many matches.

"Which one is Granollers?" Feustel says, referring to Marcel Granollers, a Spaniard ranked 35th in the world. "Is he the one that's good on clay courts?"

Feustel, from South Bend, Indiana, says he doesn't need to pay attention to who the players are on the men's ATP World Tour to double his money. He relies on an algorithm he created using data from 260,000 matches to make about 30 bets a day on Grand Slams such as the Australian Open, which started on January 13.Gamblers and investment funds are increasingly vying for profits from tennis by using computer models to win money from more casual bettors, according to Scott ­Ferguson, a former Betfair Group education officer. Such quantitative analysts, or so-called quants, are focusing on tennis in the same way their counterparts are employed by hedge funds to predict moves for stocks, bonds and other assets.

Betfair, a London-based company that enables bettors to wager against each other online, matched almost £50 million ($93 million) of bets on the 2012 final in which Novak Djokovic beat Rafael Nadal. Djokovic is an 8-11 favourite to win a fourth straight title in Melbourne with UK ­bookmaker William Hill, meaning a successful $11 wager would return $8 plus the original stake.

Granollers prefers clay courts, according to his men's tour profile, and lost his first-round match with Marin Cilic of Croatia in five sets on the second day of play on the hard courts of this year's Australian Open.

Tennis is an "attractive" sport for which to create an algorithm because there are only two players in a singles match and statistics are freely available, according to William ­Knottenbelt, an associate professor of computing at London's Imperial College. He ­co-wrote a tennis algorithm that he says would have made an average 3.8 per cent return on bets on 2173 ATP matches in 2011.

Feustel, who says he puts in a 60-hour week checking and improving his model, works with a computer programmer and trader. The programmer trawls the internet for data such as serve speed and break-point conversions. That's plugged into the model which comes up with "fair" betting prices for scheduled games.

If those odds diverge from market prices, Feustel says, his trader – who lives outside the US – will gamble as much as the market will allow at bookmakers including Pinnacle Sports, based on the Caribbean island of Curacao. That can be about $30,000 on a match result in later tournament rounds.

While bookies have used mathematics for decades to set odds, more gamblers have started making a living by applying computerised data to betting in the last 10 years, according to Warwick Bartlett, chief executive of Isle of Man-based Global Betting and Gaming Consultancy. There could be as many as 20 professional tennis gamblers ­betting on a similar scale to Feustel, according to Ferguson.

Tennis authorities would rather gambling not take place because of the risk of lower-ranked players "who travel around in a Volkswagen Kombi and stay in hostels" being lured into fixing games, Ferguson said. The International Tennis Federation, ATP and women's WTA Tour didn't respond to a request for comment on algorithm-driven tennis betting.

Spain's Guillermo Olaso, the world No. 254, was banned for five years for trying to manipulate the outcome of matches in 2010, the London-based Tennis Integrity Unit said in a December 23 statement.

The unit is a joint venture between the ITF, ATP, WTA and Grand Slam tournament organisers.

Feustel declined to say how much he has made using his computerised tennis model, which he has employed for 2½ years. He said he aims for a 2 per cent return on the total amount he bets each year, implying that wagers totalling $5 million would bring $100,000 in profit. The annual amount bet can greatly exceed the starting sum as money is constantly re-bet. The 43-year-old said it is possible to double his starting capital in a year through compounding.

After paying his partners their cut, ­Feustel still earns considerably more than in his previous job as a commercial lawyer, he says. A lawyer in Indiana might command a salary of between $50,000 and $100,000, he says.

It's "very possible" for Feustel to double his money in a year if his bankroll is less than $500,000 as it may be easier to take bigger risks with a small amount, according to Brendan Poots, CEO of Australia's Priomha Capital, a sports-betting investment fund.

Priomha wagers about $100 million a year, turning over $5 million of funds under management about twice a month, Poots said by phone. The Melbourne-based fund uses computer models to bet on soccer, cricket, horse racing, golf and tennis and had an annual return of about 28 per cent last year, Poots said. The fund's 2012 report showed a 7.56 per cent return. Priomha takes a more cautious approach than ­Feustel because it manages clients' money, hedging bets to offset potential losses, Poots added.

Quantitative funds operating in financial markets gained 11.5 per cent on average last year and have produced an annualised return of 6.6 per cent since the end of 2008, according to Chicago-based Hedge Fund Research.

For tennis betting it's "very hard" to get an annual return of more than 5 per cent on investment, according to Dan Weston, whose tennisratings.co.uk website charges subscribers for insight. He said his service using a model would have returned 6.1 per cent on the amount bet since August.

"The market is fairly efficient," Weston, 34, said. "Casual bettors aren't stupid."

Weston, a University of Kent accounting graduate based in Preston, England, said his model processes data such as the record players have of recovering from a break-point deficit, and their accumulated fatigue from matches.

That helps in situations like that of Peng Shuai, who was the overwhelming 1-5 favourite to beat Storm Sanders in the first round of an event on January 7 in Hobart, Australia, after a slog to a final against Li Na in Shenzhen, China, three days earlier. Peng lost 6-2, 6-2 to Sanders, who was only in the tournament because another player withdrew. Weston says his data show certain ­players tire more quickly than others as a match progresses. American Jack Sock has an "awful" record in the third set, Weston said, adding he advises against betting on the world No. 95 winning in five-set Grand Slam matches.

Feustel says he's "not really interested" in tennis, watching less than one hour a year, and would only recognise a handful of players, including Roger Federer and Maria Sharapova. He's devoted to his algorithm, though. "You never bet against the model," ­Feustel says.

Uneven Playing Field

I doubt that anyone reading this post will be unaware of the much publicised court-siding incident at the Australian Open this week, in which a certain Daniel Dobson was allegedly caught with an electronic device stitched into his shorts, allegedly for the purpose of transmitting real-time data back to the UK.

The accusations of match-fixing are clearly nonsense, but while the presence of court-siders in tennis has long been recognised, one interesting revelation is that Mr. Dobson wasn't an entrepreneurial individual or working with a partner, but was an employee of a Surrey company called Sporting Data, formed in 2011 and coincidentally based in my neck of the woods, the sleepy town of Oxted in Surrey where I previously worked, played football for the local Holland Sports club for many years, placed my first ever football bet at Corals and a town where my daughter sometimes works to this day.

Anyway, enough of memory lane, here are some words from their web site:

1) Sporting Data is a company providing sporting data and services to certain individuals. Our clients use the information provided to place bets.

2) We have employees on the courtside sending back information to London and that is then used to place bets on the outcome of the match. However, the odds and stake of the bets are all determined by the individuals in London and not by the employees on court. There are plenty of times that they will send info back and nothing will be done with it. In no way could they considered to be betting themselves.

3) We use mathematical models to assess the probability of a match outcome. Bets will be placed when the odds generated by the model are significantly out of line with the market. A lot of syndicates use a similar methodology. Clearly, we need the most up to date information to generate accurate match probabilities. We cannot rely on TV pictures as they are out of date.

4) Most of the bets placed are placed on betting exchanges - platforms devised specifically for punters to pit their information and simulation techniques against each other. It is the tightest and most competitive environment there is and we have invested a lot of time and effort to become competitive.

It always amuses me to read posters on the Betfair Forum talking about 'fast pictures' from race tracks. All pictures are delayed, some more than others, and if you bet based on what you see on your TV, you are probably at a disadvantage. It's the same with tennis. If you bet based on what you see on TV, Sporting Data's description of their business makes it pretty clear that you are fighting an uphill battle. Sporting Data are perhaps technically not doing anything illegal, although the Daily Telegraph states that:

UK police said earlier this week that courtsiders could be arrested under the Gambling Act 2005, while Britain's major sports all said the practice was banned under their own ground regulations.

Sporting Data are not trying to get one over on bookmakers, at least not usually, because they say "Most of the bets placed are placed on betting exchanges". The money they are going after is that of individuals who have "out of date" information, and they are doing nothing wrong - simply making the most of an opportunity that currently exists.

It has been suggested on the forums that this type of entrepreneurship would have a short shelf-life, because the Super Premium Charge will soon kick in, and that might be true for a regular Joe. 

However, Sporting Data's three directors are publicly listed as Steve High, Simon Allen and Martin Pendlebury, perhaps not the rarest of names, and it could certainly be coincidence that one Steve High was a Senior Product Manager at Betfair, a Simon Allen was previously a Software Engineer at Betfair and a Martin Pendlebury was at one time a Senior Sportsbook Odds Compiler/Trader at Betfair.

If anyone would know how to go under the radar and have multiple Betfair accounts and avoid the Premium Charge, is it unreasonable to think that a former employee (or three) would have the best idea of how to do this? One also wonders if Betfair has any kind of a business relationship with Sporting Data?   

A couple of other thoughts on this practice - one is that while not in violation of Betfair's terms and conditions, it is a practice that will, or certainly should, have the effect of drying up the in-play markets. Only a fool will keep playing under unfair conditions, and while there will always be newcomers trying their luck, at some point any sane person would realise the odds are tilted against you. Not everyone has the resources to send a friend to Australia after all.  

A second is to wonder what other sports and markets Sporting Data are active in. There are some big games this Sunday in the NFL, Conference Championships to be precise, and with most Americans blissfully unaware of in-play trading or exchange betting, any field-sider sending data back to the UK would probably avoid detection. NBA basketball attracts big money and it's been noticeable this season that the prices move quicker than ever. Baseball's been dead for in-play for a while though.

Perhaps the reduced liquidity I read about isn't all down to the Premium and Super Premium Charges. 

One truth about in-play trading is that the playing field is not a level one. Unfortunately there is no easy way to remedy this. Increase the countdown delay too much, and liquidity dries up. Several years ago the NHL markets were liquid enough, but the countdown delay was increased (reportedly due to fast data in Canada) and the in-play markets are now deserted. Limiting in-play trading to breaks in play is impractical, and Suspending an NFL game after every penalty, scoring play, turnover etc. would be impossible. Betfair have enough trouble training staff to recognise a significant event on football! 

The best advice I can offer is to be very wary when trading in-play, and recognise who you are competing against. It's a zero-sum game, and clearly some people have the resources to gain a big edge against you. As I wrote in the first post of this year, my enthusiasm for betting these days is not what it was, in big part because of the Super Premium Charge, and while trading has been where I have made most of my money, I am actually finding myself more challenged these days by identifying value pre-game and doing things the old fashioned way. I fear that unless Betfair take steps to stop the same accounts from consistently winning large amounts, (even if not technically against the current rules), and Premium Charges clearly aren't hitting the targets they should, then in-play betting will at some point be banned by law to protect the, admittedly willing, losers - although a ban on FOBTs has yet to come in, so it could be a while. The Stock Exchange would never allow an edge to be effectively bought like this. Would they?  

Odds-On In The EPL 2012-13

In the wake of last weekend's once-in-a-season success of all odds-on favourites in the EPL, I took a look back.

In the 2012-13 English Premier League season, 221 of the 380 matches (58%) saw one of the teams starting at odds-on (using the average odds from Football Data).

141 (64%) of these 'certainties' won, for a net loss of 1.51 points.

Had you combined such selections playing on the same day in multiples, you would have lost 14.43 points.

March 30th would have been a good day, with a five-fold earning 6.61 points, and with the round completing with games on the 31st and April 1st, all seven odds-on selections would have paid out at 25 to 1.

Round 37 was also a good one, with the six odds-on teams all winning, at 13.73, but as the laws of mathematics show, combining negative value bets into multiples only makes losses worse.

Only one team (of 30)  priced at sub 1.3 failed to win, (Chelsea at 1.23 v Southampton on January 16th) and as I have commented on before, and incorporated into the Bundeslayga rules, the 1.3 band seems to be the watershed between profitable backing and profitable laying in football.

Numbers Do Not Lie

Accumulators are much to the fore in the betting world this week, a natural consequence of the rare event whereby all eight odds-on Premier League teams won, a 40-1 chance. As I mentioned previously, only one odds-on favourite (using Pinnacle’s prices) failed to win in the three Football League divisions also.

I drew attention to the Coral's Press Release that came out on Sunday, and most readers will have seen through the intent of Coral's and understood that the reason they released their statement was because bookies love accumulator bets, and they must be wringing their hands at the thought of the extra income likely to be generated on this coming weekend’s fixtures. How much extra will be bet on accas this weekend as a result of the publicity?

Simon P comments:

Try telling all these folks and those that made life changing sums out of Frankie Dettori's magnificent 7 that accumulators and multiple bets are a waste of time. They give the ordinary punter a shot at a life changing win for a fairly insignificant stake. I once remember paying a chap out over £80k for his £5 bet on a football acc. in the 1990's - you can be sure that changed his life for the better! As for myself, I love placing 6 or 7 timers on the NFL and have had many decent pick ups down the years. I've also had my share of near misses - Cincinnati's failure to cover a 2 and a half point handicap at Arizona in 2003 costing me just under £51k! Sadly, the real mug punters (numbers do not lie) are those frequenting the ever growing number of betting shops and blowing weekly wages or benefit cheques on the vile FOBT machines. Will any politician ever have the bottle to ban these despicable milking machines? 

Interesting that this pro-accumulator comment comes from a bookie (or former bookie). It's exactly how bookies want you to feel, and it's not that they are a waste of time so much as they are of money. The fact that someone wins the lottery doesn’t mean that a lottery ticket is a value proposition, and there’s a reason why Mr. John Sinclair will find his account is left open despite winning an amount far greater than that won (or even lost) by more savvy punters. Mr. John Sinclair is a recreational punter who got lucky, and that’s great for him but most don’t, and the small stakes they lose week after week add up to far more than the amount ever won.

The mathematical fact is that on 110% over-round markets, the over-round on an eight-fold accumulator grows to 214.4%. [Fixed Odds Sports Betting by Joseph Buchdahl] One caveat is that if you can negate this over-round into an edge, then doubles etc. become your friend. For example, backing XX Draw selections with the same kick-off times as multiples would have made a slightly higher profit of 6.45 points.

Simon mentions Dettori’s seven winners and a couple of thoughts on this. One is that two of the big winners, Mary and John Bolton, should have received about £900,000, yet ‘only’ received £500,000 due to the limit Ladbrokes placed on their liability. The other comment is that the weight of money on Fujiyama Crest, the seventh ride that day for Dettori, pushed a 12-1 shot down to 2-1, and one bookmaker Gary Wiltshire famously took a bit hit going with value:

A hulk of a man, this on-course bookmaker faced several lean years after taking a dauntless stand against a flood of money from betting shop chains for Dettori's seventh mount, Fujiyama Crest. Wiltshire stood next to his pitch on the rails and bellowed: "9-4 Fujiyama! And 9-4 me for the Job Centre!" Later, he explained: "The odds were miles wrong. I was laying 2-1 about a horse I made a 10-1 shot."

He faced ruin, owing over £800,000. He sold his house and cars and scraped money together selling Christmas paper in Oxford Street. After that he "worked every hour God made", betting at six race meetings and six dog tracks every week, and "settled every bet". He vows he would do the same again.

While I disagree with him about ‘God making the hours’, although he likely didn’t mean those words literally, I do agree with him that laying Fujiyama Crest was clearly the correct decision. Huge value, but I’m not sure I agree with his level of exposure though!

I'd be interesting to hear how much Simon is up or down on his accumulators betting. He mentions 'many decent pick ups' over the years, but the nature of a gambler is to remember the wins and forget about the losses. It's one reason keeping accurate records is essential if you are serious about investing. As for the 'near misses' - they are of course what keep the acca fans coming back for more. Fully agree with the FOBT comment though. They have no place in the high streets in my opinion.

Accumulators offer poor value and serious investors should stay away from them but for the recreational punter, aka mug, so loved by the bookies, they do indeed offer a shot at a life changing win in the same way that the Football Pools or National Lottery do, and mugs will get lucky sometimes.

Fulltimebetting has a blog post related to this topic which some of you may find interesting.

On the subject of subscriber abuse, Peter Nordsted stopped by to say that:

Yes in regards to abusive emails etc. I can honestly say any member of either my personal service or Premier Betting has never been abusive in any shape or form.

I tend to find that people that sign up to these services generally know the risks involved. Indeed the only criticisms we receive is of the constructive type. Obviously as you point out there are those that vote with their feet which is completely understandable when you are having a bad run.

The only time that I have ever received abuse is when I have done the occasional free in play room and on Twitter if I give out a tip that has not won. I also find this abuse is almost 100% from people who remain faceless/anonymous and have never subscribed to any of the services.

Staying with Premier Betting, Anonymous writes:

If you thought Premier Betting were bad then there was a fee charging site that advised 8-0 9-0 and 10-0 in a game and it ended 2-0

As I wrote in yesterday’s post, “it happens”. No one can judge a service by one freak result. If Peter’s results for Premier Betting fail to improve though, it does occur to me that he will probably be renaming his service to the more appropriate “Championship Betting” for next season.

Even the XX Draws have the occasional result which, taken on its own looks awful, but it’s really only when you have a season or longer of results to review that you can form a meaningful opinion. 7-0 ? Not only Dettori’s score on 28.September.1996 but also Internazionale's score at Sassuolo on 22.September.2013.

Anoymous, a disgruntled Premier Betting subscriber perhaps, adds:

Hyde are winning 2-0 . Premier betting must have advised Welling . They could stop a train.

JC asks me if it is “free to join the sot forum?” I have no idea. What IS the sot forum? Forums are usually free though. What am I missing?

And finally, John Walsh offers his reasons for the change of tactic in the NHL I mentioned yesterday, and also, to his credit, addresses his poor NFL performance this season:

I can see how the results so far in this 'change in strategy' would justify this response. With the exception of last season (a shortened season) the second half of the season has consistently been the more profitable for me. When we are looking at eight bets it is a really small sample size, especially when dealing with bets that have odds in the 4.00 neighborhood. Based on history and the methods I use to find value I am really confident with the regulation tie betting. If I find a game that my model says has a 30% chance of being tied at the end of regulation and the price is 4.00 I will take that bet every time. The lower probability of striking the single bet does leave this strategy more prone to losing runs, but if I feel actual probability is higher than implied probability then I take the bet.

In the NFL the earlier part of the season is usually my better performing section. I have no excuses for this season. I know because of the poor performance I've lost some followers. I will be working hard over the summer to fix this. The NFL is usually where my most consistent profits come from.

Are the Coyotes or Jets moving to Los Angeles? If Gary Bettman moves the Jets again he will never be allowed back in Winnipeg.

No, the Coyotes or Jets are not moving to LA that I know of. For some reason, I was thinking the Ducks were included in the first bet, but clearly I was having a senior moment.  

Close Title Race

Arsenal continued the trend of odds-on favourites winning last night, making it 8 from 8 in the Premier league and a 40-1 accumulator for those who like these things.

More importantly, the game closed out the weekend's action, and the FTL sponsored by TFA is updated.

Of the entries in profit, the big loser was Fedslam who plummeted from top spot to sixth after firing six blanks. As discussed in previous posts, Skeeve also drew a blank dropping down a couple of places, and the beneficiaries are my own Cassini Value Selections, Hofs Hackers, Jamie A and Webbo.

As you can see, it is very tight at the top, with just 0.66 points separating the top three. Hofs Hackers are sitting on £250 after gaining 2.73 points from four selections, while Jamie A was up 4.76 points with four winners from five and is in line for £125. Skeeve is still in the money too, and Webbo moves up with 2.98 points from his ten Premier League selections. Webbo's been steadily climbing back from a nadir of -13.73 points in early November to now be up by 6.07 points. Then come my own XX Draws, who started the weekend poorly, ended the weekend poorly, but which hit a purple patch in the later games on Saturday and early Sunday finding five winners in six matches, including three consecutive perfect (0-0) draws, followed by a 1-1, a 0-1 and then another 0-0. A spell of 365 minutes of XX Draw action without a goal - that's how I like my football these days. Drawmaster Pete also made a small profit with one winner from three and rounds out the top nine.

Falling out of profit was sponsor The Football Analyst, whose sponsorship deal (bounty payment) is now at £150. Graeme is now down 0.53 points after just one winner (York City) from five selections, a list that included the only losing odds-on favourite in the entire Football League. I'm sure Graeme sees the funny side of that.

No action in Germany of course, and the Unders bets on the XX Draws came out at exactly zero despite seven winners and six losses. Fairfranco had one winner from five selections and lost 1.66 points.

There was very little change in positions for the entries in the red by double figures. Football Elite had another bad weekend with three winners from 11, as did Peter Nordsted's Premier Betting selections whose three selections all lost. Peter probably had a wry smile at the eight goals at the Britannia Stadium after recommending the Under 1.5 goal bet. It happens. Here are the laggards:

Punters' Friend Neil had another huge selection of matches, but lost another 15.8 points to fall further adrift at the bottom. It's not often that you can have the most wins and be bottom of the table!

A couple of comments to address. Jake asked how he might follow Skeeve, and the answer is to send an email to him for details. skeevepicks@gmail.com and there are more details on his service at http://www.skeevepicks.com/

Anonymous revealed that:

Premier Betting only have 2 clients .

Not sure about that, but they may be a few members reconsidering their subscription renewals next season. Anonymous goes on:

A serious point is what makes people follow one tipster over another . There are some great tipsters with solid results who have hardly any clients and some really poor ones that have loads. The big craze is accas so the football data analysts are not getting a look in.

Re Southport v Hyde I am interested in the thought process that there was expectation that Southport would win > if we toss a coin 1000000000000000 we may get 10 heads in a row and Hyde with only 3 draws this season surely would be considered to get a few more points this season ????

The price that tipsters charge is obviously a factor, as is their reputation and the type of selection that are offered. It is likely that a 'great tipster with solid results' is either too expensive or has yet to prove himself over the longer period, if they have hardly any clients. A poor tipster isn't likely to have many followers for too long. Profits talk, and while being a nice guy might prolong loyalty a little longer than results alone might warrant, no one in their right mind is going to throw good money after bad.

As for the Southport v Hyde bet, only Skeeve knows where he had Southport priced. With just three points to their name all season, I'm not sure I'd be going with the 'they must be due a win' theory myself.

Over to the US Sports and John Walsh had a small loss on his NFL play-off picks this weekend.

On the ice, John seems to have developed an unhealthy obsession with betting on the draw (or rather tie) and as hard as that is in football, it is even harder in ice-hockey. Seven of his last eight bets have been on the tie, including the two matches involving teams from Los Angeles last night, but with only one winner to date the change in strategy has yet to pay dividends.

I know John stops by from time to time, and it would be interesting to hear his thoughts on why he would change a winning formula at this stage of the season, without giving away any trade secrets of course.

Watford Gap

Further to the Perfect Storm post yesterday, which had us reaching for our hankies, I dug a little deeper and found that, using Pinnacle's prices, there were 19 Football League teams starting at odds-on over the weekend.

All seven Premier League teams won (Everton 1.39, Chelsea 1.63, Manchester United 1.58, Southampton 1.85, Tottenham Hotspur 1.38, Manchester City 1.67 and Liverpool 1.64 - a 25.03 accumulator), as did the five teams in League One (11.71) and one (Chesterfield 1.88) in League Two.

The Championship had six teams starting at odds-on, and only Watford (1.83) failed to win with the five others winning at 1.98, 1.48, 1.79, 1.93 and 1.85.

Quite a remarkable weekend, and Arsenal go off at around 1.6 tonight, with Rubicon, Fedslam and Forza Fizzer all hoping to see Arsenal become the second team of the weekend to fail at odds-on.

The trend was limited to England and Scotland. The Bundesliga was on its winter break, but odds-on teams failed to win in the other major leagues - AC Milan at 1.74, Monaco 1.93, Lille 1.5 and Real Betis 1.8. Villareal and Internazionale should both be odds-on tonight.

The three Scottish Premier League teams starting at odds-on also won this weekend - Aberdeen 1.83, Motherwell 1.5 and St Johnstone at 1.88.