Sunday, 19 October 2014

Price Pain

My approach to 'Conditional' bets prompted an email from one subscriber, who wrote:

Hope you are well. I've got another point to raise on your conditional selections, just an argument you might want to address. I must stress, I have no issues with the way you already do things and am happy with the service.
You say subscribers should be profitable on the conditional selections and if 3.15 can be obtained, then that is value. That may be true, but quite possibly, 3.15 would not be high enough to yield a profit on conditional picks long term. I'm sure you've got some evidence to back up that 3.15 is sufficient, though. It's quite possible for a subscriber of yours to look at your conditional picks, go out and secure prices of 3.15+, and end up with a lot worse P+L than your official results show. If I managed to back all your conditional picks at 3.15 this season, or even 3.2, along with your official picks, how different would my P+L be to your official results? These conditional picks could make a big loss, possibly due to variance. If a potential subscriber looked at your official results for the season, which are so much better than one of your subscribers who managed to secure 3.15 on conditional picks that made a loss, is that possibly misleading?
I am not sure if the conditional picks should be included in the email at all, that's just my opinion, not that it matters either way to me. I check Pinny on Friday afternoon and if the conditional picks are not at least 3.15, I don't back them at any price.
The subscriber is quite correct in saying that 3.15 may not be value. No price can ever guarantee a profit since as in a season such as the last (2013-14), draws were simply short in supply, so it is true that 3.15 is a best estimate of likely profitability based on historical results.

As I have mentioned before, I doubt that at the end of a season, any two subscribers will have exactly the same results. Having warned that 3.15 is borderline, I imagine that many people either omit selections anywhere near this price, or reduce their stakes. The idea of including them as Conditional picks is to let subscribers know that these matches historically are value draw selections, but historically draw prices have not been in the 3.0x range. Any subscriber backing at exactly, or close to, 3.15 does so in the knowledge that they are in a sense, on their own on this one, if Pinnacle's price doesn't meet that threshold.

As for "I am not sure if the conditional picks should be included in the email at all" my answer is really as simple as 'what harm does it do'? For Premium Charge payers, those 3.15 or close bets might be useful; for other more risk averse subscribers they may not be. at least by including them, the subscriber has a choice, which is generally a good thing!

Dmitri had some observations regarding the recording of prices, noting that some "use fair prices collected just BEFORE the email is sent". As he says, using old prices is not a fair way of recording prices, especially if they are in relatively illiquid markets. Dmitri writes:
To see the effect of using your method of recording (Pin prices on Friday) compare for example Combo’s official result of 9.84% with the 6.04% for Pin prices. Or Graeme's best system 31 with 8.40% down to 4.59%.

You are betting on the top divisions and are not influenced by odds movement, down from 11.95% to 11.46%.

I can see the problem from both sides, but as a punter I prefer your recording method.
Just a clarification here, that Dmitri is comparing my prices at the time of the email, versus my official price which is the independently taken Football Data Pinnacle price. My email price is given as a courtesy, along with my Cassini price and the current edge, but the recorded price is the minimum price that subscribers should be expected to be able to obtain, and is usually beatable. For example, this weekend the Newsletter was sent out on Tuesday morning, and gave the following for the Value Selections:
 
The official prices are yet to be published as I write, but subscribers have what I considered to be the 'true' price for the selections, and have several days to shop around for a price at which they are comfortable. The Pinny price was from Tuesday morning, and prices do move, even in the deep waters of the top leagues. It is mildly annoying when a (winning) selection is recorded at a significantly lower price than at the email time, but it's a small price to pay for the benefit of using an independently recorded price.

On the subject of the timing of releasing bets, my thoughts are that for a number of reasons, it is surely better to release them sooner rather than later. The more time subscribers have to place, or not place, bets the better. For anyone who is not full-time, it's also not always convenient to send out an email at a pre-determined time. Unless selections are based on individual player metrics and need to wait for injury / team selection news there is simply no need to wait. Send the bets out, with the price that is considered 'true' and give subscribers as much time as possible to find a price that they are comfortable with. Releasing selections at a set time late in the week, especially in soft markets, simply leads to crashes in price and little time for them to bounce back.

On the subject of 'true' prices, Dmitri observes:
If you as many punters do always use level staking this number is irrelevant. Then you are only interested in beating the bookie.

But if you are more advanced and use some sort of variable staking like fractional Kelly, then value is very important. On average these services are overestimating the value on their selections and this in turn will influence a bettor using variable staking negatively.
I do record my selections using level staking, but as a Kelly man, my real-world bets are not level. Dmitri continues, with a rather depressing conclusion:
Regarding minimum odds I would never consider signing up for a service that did not provide these prices. They represent the models estimate of fair prices.
Of course no model can get it right, but on a very large sample the profit of proofing to min odds should be about a round zero.
This is actually not the case. The profit ranges from -3.43% to -7.15%.

I can’t remember any service with a positive profit!
 We're all useless!  The true price hurts!

Saturday, 18 October 2014

Pinning The Price

Hopefully you have all had a chance to read the comments on recording prices from Football Elite’s Matt.

For services, and especially for full-time services, the subject is very sensitive. Their livelihood depends on having subscribers, and having subscribers is heavily influenced by previous results. As I mentioned in an earlier post, the service thus has a vested interest in making those results as attractive as possible, while the potential subscriber wants the results to realistically reflect the results he or she might obtain themselves were they to sign up. Anyway, here is Matt’s comment in full, with my responses:

Hi Cassini,
The ideal scenario would be if an independent party could provide a standard Football SP a la the racing SP.
How the SP is calculated and whether it is based on prices just before kick-off or at a specified point of the weekend doesn't particularly matter. Whatever it was it would be the same for every service and would provide an easy comparison between each services results.

Until that happens, in my opinion the only fair way of recording is to record at the general price that is available when the bets are sent out.
I’m not sure what ‘independent party’ would benefit from a standard football SP. There is no need for an industry SP in football (or other sports betting), because nobody bets on an outcome without knowing their potential return. That people still do this on horses in off-course betting shops says much about the social nature of the activity rather than as a serious attempt to make money. The only beneficiary of an ‘official’ football price would be punters, for whom the data would be useful in trialling systems, and potential subscribers to betting services. The solitary nature of these activities means that an official price will not happen. Matt continues:
No standout prices, no soft/dodgy books, ignore the top few prices, record at the price that is widely available with a good amount of firms and to decent liquidity on the exchanges after commission.

With all due respect Cassini, proofing to Pinnacle prices is one of those things that sounds better than it actually is. Pinnacle are generally among the top prices on most top flight football bets, in particular on the draw and the outsider. On occasions they will be the very best. Hence most of the time you will be recording at higher than the general odds and on occasions at the very top price in the market.
The first problem I have with Matt’s comments here is that the suggestion to exclude standout prices, soft or dodgy books, is vague and subjective, as is the term "widely available". Who decides if a book is "soft or dodgy"? What is a “good amount of firms”? What is "widely available"? When it comes to betting, not all firms are created equal. Not all members have accounts with all books, and if the service is any good, several accounts will soon be closed or limited. The idea of “general odds” or “good amount of firms” to me is nonsense. 

Firms need to be weighted based on accessibility and reputation. To suggest that a firm such as GoBetGo (established 2013, arguable reputation) should be considered the same as Pinnacle Sports (1998, excellent reputation) makes no sense. 

No standout prices”? "Ignore the top few prices"? It rather depends on who is offering the standout / top few prices. If it is Pinnacle Sports, Coral (1926), William Hill (1934), Bet365 (1974) or any established and known book, why should that price be excluded? A service can’t cater for every individual’s situation, but there aren’t many subscribers serious about making money who do not have access to Pinnacle’s prices. The fact that they are generally among the top prices is one reason why they are my book of choice for prices recording, but more important is their accessibility for the majority of punters, with no accounts closed or limited, and no concerns about their reputation.

Matt again:

My results in your FTL table are better than my official results as Pinnacle prices will almost always beat general prices. I would guess you could improve (or at the very least not harm) most reputable football services official results if they recorded purely at Pinnacle prices instead of their current recording.
I doubt that, because if that were the case, why would not all services be recording the way I do? It's a lot easier than most methods, but the service loses control, and most won't want to do that.

Pinnacle’s prices may be competitive and above average (which isn’t hard to achieve when you are averaged in with books who are working to a large over-round) but they can usually be beaten. A quick look at last season shows that Pinnacle Sports were the top price 23 times in the EL last season for odds-on selections, from 211 opportunities – 10.9%. At odds against, they were indeed the top price a little more often, 17% of the time. For me that reinforces my claim that Pinnacle’s prices can usually be beaten, and the Pinnacle line is the minimum a serious investor could expect to achieve. Matt:
E.g. if we take tonight's Lens v PSG game on oddsportal. Odds for the draw are:-
5.2 with GoBetGo 5.13 with Pinnacle 5.1 with 5Dimes, Marathon 5.0 with quite a few firms and after commission 5.0 - 5.05 is roughly the price at the exchanges
In reality if you are talking about a fair price then 5.0 is the price you should be recording at. Yet in reality by recording at Pinn prices you would be recording at 5.13, the very top of the market if you ignore GoBetGo as an unestablished firm.
In reality, we’re not talking about a fair price of 5.0. The reality is that no one should have an issue with recording Pinnacle’s price, which is a lot sharper than most books. The reality is that anyone offering 5.0 is not competitive. Pinnacle Sports is a reputable company, and there are no issues getting on and my thoughts would be that Pinnacle have a better handle on the true probability than the other books. And yes, I wouldn’t even mention GoBetGo - I'd never even heard of them before!  

Whether or not 5.13 should be the price actually recorded is another issue. The price recorded should be by an independent source such as Football Data, or a certain amount of time after the bet has been released. Especially in thinner markets, or at softer (longer) prices, the recorded price needs to reflect what most subscribers can expect to get. Hence my suggestion that the service actually record the price they themselves obtain 30 minutes or whatever after releasing the selections. Matt:
Don't get me wrong, recording to Pinnacle is certainly not a bad way to record. There's far, far worse ways. But to almost insinuate that you are in some way leading a bright new future in how services should record their bets is a little OTT to be honest.
I think, as services, we do ourselves no favours by obfuscating how our numbers are recorded.
There's plenty of services that record their bets very fairly and leave plenty of room for subscribers to beat stated figures. Suggesting otherwise (and insinuating that only you record fairly) just further perpetuates the idea that all/most tipster services are dodgy firms run by shady characters. You've no need to go down that road. Enthuse the positives of your own service by all means (and it is a good service) but slightly off to do it by insinuating other services are not playing it straight.
I am familiar with very few services, so I am not in a position to suggest or insinuate anything about most services or the people who run them. If subscribers are happy with the service, then the retention rate will be high. If they are not, then they may complain or leave unfavourable comments. All I am advocating is that the subjective nature of recording prices should be removed. Either use an independent resource for prices or show evidence that the prices claimed are achievable. It won't happen, but it's important to put the idea out there.
Appreciate you probably were not deliberately doing this, and I'm probably being too precious, but kind of annoys me when most folk seem to think tipster services are dodgy so when a fellow service operator perpetuates that myth it winds me up a bit.
No personal offence intended, keep up the good work with the blog and the service.
No offence taken. I appreciate that as a part-timer, I have more freedom to advocate for such changes, but ultimately it is the subscribers who matter, and if they are happy with the status quo, then so be it. 

I appreciate Matt's input on this topic, and as most of you know, because Football Elite deal mostly in the top leagues, price recording isn't a huge issue for him and as mentioned above, his official prices hold up well against his FTL numbers based on the independently recorded Pinnacle prices from Tuesday and Friday afternoons. Not quite the same longevity as Coral, but Football Elite has also been around for a while, which in itself speaks volumes. 

Granada, Andalusia

The Wisdom of the FTL crowd is in question after their last outing, losing for the second straight week, down 6.85 points and up for the season by just 2.4 points.

This week, there's a lot of interest in the Newcastle United v Leicester City game, with no less than 13 experts (easily a record) having an opinion - more than twice as many as on any other single EPL game this weekend. The verdict is a Home win for Newcastle, one that I agree with as I said in a previous post.

Two matches are No Bets with opinions evenly divided. Manchester City are a popular choice to beat Tottenham Hotspur, while a Sunderland win at Southampton is a surprising unanimous verdict.

There have been some comments or emails about the issue of price recording - Matt from Football Elite left one on the last post which I shall address later this weekend, time permitting. We have a pretty full slate of entries with only Skeeve, Daily 25 and Talkies Tips missing in action, which is customary, understandable and highly unusual, respectively.

For those of you acca-watchers out there, @ValueBankFooty has scaled back his activities this week, shooting for a Home treble comprised of PRESTON NORTH END, PETERBOROUGH UNITED and STEVENAGE.

The XX Draws started the weekend in typical fashion - a 90th minute goal breaking up the perfect draw in Granada, but the game ending Under 2.5. Maybe we got the bad luck out of the way early. The last time an XX Draw selection lost to a 90th minute Away winner on a Friday night in Spain, we went on to make 7.96 points over the next two days. To paraphrase the old advertisement that some of you may Englanders of a certain vintage may remember:
It's great frustration you get, placing your sure draw bet, on Granada
For those of you who missed the rather clever pun in the Post Subject, Andalusia - and a loser... I'll get my coat.

Wednesday, 15 October 2014

I Gotta Horse!

I'm a little surprised that my post on conditional bets and the issue of how to fairly record prices hasn't received more attention, but it received a comment from Greekvik which was in favour, and raised a good point about two basic 'demands':

I really like the idea of conditional bets as a fair and useful tool.
As punter we have basically two demands
a) At the time when we decide if we want to sign up or renew a subscription we want a benchmark to compare the performance from different services.
b) When we have decided to follow a service our interest change. Now we want an estimate of the value of a potential pick at the time we place the bet and with the best price that is available among the bookmakers where we are allowed to place a bet. In reality we have no more a direct interest in the recorded price from the service. Our direct interest is to know if a potential bet has value or not with the odds we can obtain. The focus has moved from the recorded price from the tipping service. The focus is now the available odds for a potential pick. To make a wise decision we want an estimate from the tipping service of the lowest odds where a potential pick still has value.

The concept with conditional bets is flexible both for the punters and for the tipping services. As punters we will not follow the same strategy. But if we are serious about our betting activity we will all focus on an estimate of the value of a potential pick.

For the tipping services the concept is too flexible. The service might have different strategies for the lowest odds that should be accepted.
A very valid point that subscribers and non-subscribers have different concerns. A non-subscriber is interested only in that the reported results are fair, i.e. what they themselves can reasonably expect to obtain were they to sign up.

It's probably not reasonable for someone living in North Korea to expect to be able to match reported returns from a UK based service, but someone in Little Piddlington should be able to. It's up to the individual to join or not, and the service's responsibility ends at recording fair prices. For me the only fully transparent way to do this is to either have an independent third-party take care of it, or show evidence that bets were matched at the recorded price after the selections were sent out.

Greekvik's second point is in line with my own thoughts on the subject. Punters these days are far more sophisticated. Prince Monolulu never had to tell his clients what price to back at. He sold them a tip, and that was it. These days, sending out a selection with no mention of price is not going to fly with most people.

Having said that, there is still a lot of vagueness around. Services offer selections with the prices subscribers should be able to obtain, but unless the service states what their true price is, it puts the subscriber in a spot if the price suggested isn't there.

For example - Back Newcastle United (v Leicester City) this weekend - 2.2 Paddy Power, Coral; 2.15 Sky Bet, Boylesports, BetVictor, 888Sports, Ladbrokes, et al.

But say that 2.15 has gone, and the best I can find is 2.1. Is that still value? Well it might be, or it might not be, but I am paying my money, so don't I have a right to know? Am I paying for a selection and a vague idea of price, and to just do as I'm told, or am I paying for the selection with an accompanying valuation, and trusted to make my own (informed) decision? Well, you're paying for what you signed up for, but what SHOULD you be paying for?

If the service included their estimate of Newcastle's 'true' price, the subscriber would be in a position to make an informed decision. Comments along the lines of 'Newcastle are 2.15 and there is some value at that price' don't help. What does 'some' value mean? 0.1%? 10%? My preferred approach is to give my 'true' price, the current Pinnacle Sports price and let the subscriber decide.

For example from the last Newsletter, these were all unofficial mentions involving newly promoted teams, but the unedited paragraph read:
Some honourable mentions of matches involving newly promoted teams. At 9.8%, West Ham United are very close to 10% value versus newly promoted Queens Park Rangers (1.77 / 1.943) and in France, Olympique Marseille are 12.4% value to win at newly promoted Caen (1.85 / 2.08). In Germany, I have Eintracht Frankfurt at 2.12 (Pinnacle 2.45) versus newly promoted Koln, and in Italy two newly promoted teams meet – Empoli v Palermo where the home side offers value at 2.41 over my 1.91. 
 
I think this works well. Prices fluctuate, but with the Cassini price there, subscribers can make their informed decisions, and as Greekvik says, not all punters will follow the same strategy.

Via email, one subscriber commented:
Want to add that what you are doing with the price recording and your recent blog post is what any professional service should be doing. As far as I know you are the only one. Much respect!
That comment was appreciated, and I don't think I'm the only one, or at least won't be for long. It's the next step in the evolution of betting services in my opinion. Educate and inform. Understand that different people have different thresholds, and don't fret over recording the very best prices. A service making a small profit or even a small loss at prices recorded independently will earn a better reputation than one self-recording bigger profits, but

Having used them as an example, I should probably reveal where I am in regard to Newcastle. I have them at 2.016 this weekend, so 2.116 would be 5% value, while 2.217 would be 10%. 2:1 if you want a Correct Score prediction!

I was about to post this when I saw Steve M has a well-timed piece on dodgy tipsters. On the topic of exposing the latter, Steve writes:
There are not many places to complain and make sure others hear about it. I am working on such a thing as we speak. A place that will keep track of every tipster service and people’s reviews of them (Good and bad).
It's human nature for people not to comment favourably, unless pressed perhaps, but an independent forum where potential subscribers can look for views would be a good idea.

Finally a mention that the official hit count for this blog has now surged through the 700,000 mark (although Blogger has recorded an even more exciting 930k):

Monday, 13 October 2014

FTL Update 13.Oct.14

With a lot of people taking the weekend off, there wasn't a huge amount of activity in the FTL.Abromo still has one match to go, but win or lose, his position won't change so here are the current standings.  First the "up by 10 or more elite":

Second and third traded places (joke intended Randolph), while Mountain Mouse squeaked (another joke, I am on fire!) into the 10+ group which lost Football Investor, actually the biggest mover in terms of places dropping six. The "up by less than 10" group is:
Football Elite climbed three places, but otherwise all very quiet. The "Down by less than 10" group has these members currently:
while the final five are:
The grand totals are:
Twenty entries currently in profit for the October monthly prize as Club Havana looks to win back-to-back prizes:
@ValueBankFooty is in good form this month, and while one big winner can change things in a hurry, the top five are the ones to beat right now.

October is rather a short month with the Internationals rather getting in the way, but it's unlikely FIFA will re-arrange the calendar to fit in with us, so we'll play the cards we're dealt.

November 1st weekend is not only the start of November (he says somewhat redundantly) but also the second group matches in the Erskine Cup.

Sunday, 12 October 2014

Obfuscation And Conditional Bets

The ongoing issue of the price at which a selection or tip should be recorded continues to be debated. I attempted to explain my thoughts on this on Twitter last night, but the limit of 140 characters means that this medium doesn't lend itself to considered thought, so here's another attempt to get some of my ideas and opinions across. The post is written with football betting in mind, and assumes a majority of service subscribers are UK based.

In a nutshell, the problem is that most tipsters want to record the highest price they can get away with, (the better the profits appear to be, the greater the probability of new subscribers signing up in the future), while everyone else wants the recorded price to accurately reflect the quality of that service.

While the price on a casino bet is fixed, prices on sports events are dynamic. Not only do prices change between the opening of a market and the close at kick-off, but at any one moment in time, there are broad differences between sportsbooks themselves. Manchester City play at home to Tottenham Hotspur next Saturday, and prices for the home team range at this moment (using Sky Betting's Oddschecker site) from at least 1.4 (Spreadex) to 1.5 (four firms). Oddschecker include just 22 books, the most notable omission being that of the punters' favourite Pinnacle Sports.

Services have two choices. They can either record their official service price themselves, or they can effectively outsource that process.

Self-Recording

If services choose the first option, then they need to take a few considerations into account. When does the price get recorded? What book or books are offering the price? Do most subscribers have access to place bets at those books? How solid is that price? And most importantly how many subscribers have a realistic expectation of obtaining that price?

The first consideration is not so important with top games such as the EPL. Obviously a service can't record a price from prior to sending out selections, and needs to wait a certain period of time before recording them.

In regard to what books are offering the price recorded, it would be somewhat disingenuous in the City v Spurs example to record 1.55 based on a price that was offered for five minutes by a recently opened sportsbook in Outer Mongolia. Most subscribers, and probably not the service themselves, would be unable to take such a bet.

In a liquid market like a Premier League game, it would be hard to take issue with a service sending out a City recommendation recording 1.5. It's currently generally available at bigger named firms (Coral, Ladbrokes, William Hill, Stan James) and although most serious investors will long ago have had their accounts with these firms either closed or limited, that's an issue for the individual to work through. If those four are offering 1.5, most serious bettors will be able to obtain at least that price. It would be unfair to the service to exclude these names, and a price that is available at more than one book is likely to be more stable than one stand-out price would be, e.g if Labrokes were offering 1.52.

Things get a little more complicated in a less liquid market such as the Conference. Here one service can move the markets, and apparently does if the TFA emails are anything to go by! Graeme mentions one service, Blogabet, (with which I am not familiar) saying:

Given Mark can only quote odds which are available at the time of release, he is pretty relaxed about odds movements as at least when he releases his bets, the odds are there. The fact they might only be there for 10 seconds is something that doesn’t sit comfortably with me but then again, that’s an issue for Mark and Blogabet.
While 10 seconds might be an exaggeration, recording a price almost immediately suggests that the service provider understands that it's a price that will not be attainable for the majority of subscribers, and that's what shouldn't sit comfortably with anyone. Essentially in these illiquid markets, the prices are too soft to support the money coming in. As Graeme eloquently phrased it:
“some bugger took 0.85pts off one at Pinny at 7pm on the dot”
“We all go for same bets at once in poorest liquidity league”.
It's apparently a big problem if prices are moving by that much. I'm not sure I fully understand waiting until a certain time to release selections anyway. If you let people know what price they should be looking for at a minimum, then why not send the selections out as soon as you have them? Some people might like a set time, but no time will be convenient for everyone, and if anything is going to trigger a plunge in prices with the books, it is dozens of people suddenly piling in on a soft price at 7:01pm on a Thursday evening or whenever.

One final thought for anyone favouring self-recording of prices. Use the prices that you yourself are able to obtain at some point after sending out selections, and ideally after more than 10 seconds has elapsed. It shouldn't be unreasonable to self-audit and show subscribers evidence when sending out the update or the following set of selections.

Out-Sourced Recording

Rather than self-record official service prices, another option is to take the price as recorded by an independent third party at a future time. It's the approach I choose to take because it takes away any concerns about timing or availability, and frankly I have the freedom to do this because this is a hobby for me, and official returns are less important to me than perhaps they are to others. The method I use is to take the prices from Pinnacle Sports as recorded by Football Data Co uk  It isn't perfect, but it's totally transparent. Pinnacle are (or were) widely available with their policy of using incoming money to refine their lines, rather than to close or limit accounts. They are competitive, although they can often be beaten, (for example they currently offer 1.456 on Manchester City).

The prices are taken on either a Friday afternoon (for weekend matches) and a Tuesday afternoon (for the midweek games). I am not aware of anyone providing 'official' closing (kick-off) prices, but 24 hours or so before most games is acceptable.

Conditional Bets

Apparently my Twitter attempt at explaining the logic behind 'conditional bets' apparently failed miserably, with Football Investor writing:
I really don't get the concept of conditional picks. I can see it alienating a lot of members wrong side
I'll have another go here. Basically it's a necessary protection for the service against a dramatic price movement after releasing selections. The key to remember here is that any service worth its salt is not sending out selections to be backed regardless of price. The name of the game is value, and a selection is only made because it represents value assuming a certain price can be obtained. It would be unfair to a service to expect them to send out selections early in the week, and hold them accountable against prices perhaps three or four days in the future, especially when the service states "do not make this bet if you can't get better odds than..."  There has to be some protection. This policy has been well received:
I value your service very highly, so highly in fact it is the only one I am subscribed to as a 'paid service'. I think you're very fair, overly so in your recording of prices and can totally except you cannot record prices of draws below 3.15 because of their lack of profitability long term.
So that's the concept. Not one person has suggested this is in any way unfair, (at least not once the logic is explained to them). Out-source the bet recording to an impartial third-party, and protect yourself against drops in price by having a minimum in place. To me, it's a win-win. The service can send out selections knowing that they will be recorded at or above a certain price. The subscriber knows what price to look for, and if they can beat that price (as is usually the case) then great, if they can't, then the selection was never a value bet anyway. How any member could get upset about such an idea seems odd to me. I suppose if a service simply sends out a selection with no mention of price, then later saying it was a no-bet won't sit well, but surely no one simply says "Back Dover Athletic" and leaves it at that. A selection is only a selection at a certain price. If the price moves wildly, then the value is on the other team. It's how value works.

As I said earlier, the official prices I am recorded against aren't hugely important to me. I know that they are beaten more often than not, but I'd rather have a P/L that reflects the worst case scenario than have an inflated one that just doesn't reflect reality.

The ultimate measure of a service is its retention rate. If subscribers return year after year, that's the true measure that you are doing something well. Every subscriber keeps their own P/L and I doubt that any two have the same returns after a full season. And it's fine with me that all (including my own) will be above my official recorded results.

Saturday, 11 October 2014

Liquidity Pooling

Fizzer555 had a comment on Betfair's Cross Market Liquidity Pooling:

My first thought was, how are they going to deal with different commission rates. I don't use the Betfair Asian Handicap markets very often because of a lack of liquidity and uncompetitive pricing - but the commission rate was only 0.75% and you could use it to get small bets matched near kick off at prices equivalent to the match odds market but at lower commission. Not sure when they changed it but just looked at some League One matches today and the commission rate is now 5%, so problem solved for Betfair and one less opportunity for us.
Again, I don't believe that Betfair help themselves here by offering so many different AH markets on matches. A look at the Portsmouth v Mansfield Town game currently in progress shows no less than 33 different AH markets ranging from Portsmouth -4.0 (have Portsmouth EVER beating anyone by more than four goals?) all the way to Mansfield Town -4.0. Most of those markets will have seen zero traded, and simply put people off. When you need to scroll down the page for 5 minutes before you see any signs of life, the tendency is to give up. Why they don't start with a sensible +/- AH market, and then add adjacent ones if the justification is there, I'm not sure. Actually, I am sure - it's simply easier to plonk 33 markets down and move on.

For the same reasons as Fizzer, I rarely use Betfair's AH markets myself, but I would if they were competitively priced. Unfortunately a 0.75% commission rate means nothing to anyone paying the Premium Charge so that was never an incentive.

Betfair's AH liquidity is surprisingly low. As Fizzer points out, their incentive low commission rate of 0.75% seems to have quietly gone away, so the excellent point about different commission rates has been resolved, and the chances of similarly targeted incentives in the future is reduced because of the 'Pooling'.

Had I read my "Premium e-mail Newsletter" from Betfair yesterday, I would read about this earlier. Here are some more comments and plans for the release:
That’s [the AH] a simple example, but in practice the range of bets we will show and match will be much wider. The first sport to see liquidity pooling will be Football and initially this will apply to four markets: Match Odds, Draw No Bet, Double Chance and the -0.5 to +0.5 lines in the Asian Handicap.
We hope to load the first matches using this new logic early in November, but how widely and
quickly we roll this out will depend on how testing goes between now and then. Of course we need to test not only that the logic works as expected, but that performance is good too and above all our customers are safeguarded.
There is also a proposed change to the Correct Score markets:
We currently offer all quoted scores from 0-0 to 3-3 in Football Correct Score (C/S) markets, with “Any Unquoted” listed as a selection to cover all possible scores where one or both teams scores four or more goals. In many situations, perhaps where one team is a strong favourite like Barcelona at home to Elche, “Any Unquoted” will be the favourite in that market. It’s effectively an inexact proxy for “Barcelona to win and score heavily”. 
We’ve had some customer feedback asking for “Any Unquoted” to be split into three selections, “Any Other Home”, “Any Other Away” and “Any Other Draw”. That would give customers more choice – those who wanted to back just Barcelona to score a hatful should get a slightly better price, while those who still want to bet on (or against) all three outcomes can dutch them.
We are going to trial running correct score markets with those three selections instead of “Any
Unquoted” later this year. With football markets offered every day at this time of year there will never be a point where we don’t already have a correct score market loaded on the site, so it won’t be possible to simply switch all markets over to the new selections. That means that we’ll start loading new markets with 19 selections as “Correct Score New”. Already loaded markets will remain as they are now, with 17 selections and called “Correct Score”. Once all correct score markets with “Any Unquoted” have been settled, all correct score markets will have the new template and we’ll rename them all simply “Correct Score” as now.

Show-Me State

Another winner for the road 'dogs in the MLB play-offs this morning, as the Kansas City Royals beat the Baltimore Orioles in the first game of the ALCS. A little history for you non baseball fans, but the last time the Royals made it to the post-season was in 1985, when they won the World Series (their first, and so far only win). Their opponents in the World Series that year were their fellow Missourians from the National League - the St Louis Cardinals, and they are alive and well this season too, opening tonight in the NLCS v the San Francisco Giants.

And if that isn't exciting enough, I have more. That 1985 World Series wasn't the first 'All Missouri' World Series. That was in 1944, when the World Series was between two teams from one city - St Louis - the Browns and the Cardinals, who both shared the same stadium. The Cardinals won 4-2, but the quality wasn't that high due to the war. In fact, the Browns had a one-armed player (Pete Gray) on their team! One final twist is that the Browns later moved to become the Baltimore Orioles (the Royals' ALCS opponents this year).  The Giants don't have a Missouri past or present, having moved from New York in 1958, but this is their fourth NLCS series versus the Cardinals with the Giants winning the two most recent in 2002 and 2012.

Some interesting news from Peter Webb's Bet Angel Blog in regard to Betfair's plans for a November roll-out of "Cross market liquidity pooling". Basically an unmatched bet will appear in other markets. The simplest example used by Betfair is this:

Perhaps the simplest example would be the home team in Match Odds, and the same team -0.5 Asian Handicap. Currently if you place a bet on the Match Odds today, and if it’s unmatched we will only advertise the bet in Match Odds. The same is true of Asian Handicap. With the new bet matcher, if your bet is unmatched in Match Odds we’ll also show it in the Asian Handicap -0.5 line for customers to match there. The reverse will also be true: if your Asian Handicap -0.5 bet is unmatched, we’ll make it available for other customers to match not just there but in the Match Odds market too. Note that your bet will still only be matched once, with whoever requests the bet first.”
In my opinion, Betfair offer far too many markets, most of which see little to no money traded, and do not help themselves by having superfluous markets such as Under / Over 0.5 goals. While I understand that it costs Betfair nothing to put these markets up, I can't help but think they would be better served by concentrating interest on the major markets.

Friday, 10 October 2014

Top Drawer Results

I came across an old (2009) Kevin Pullein piece from the Guardian yesterday while I should have been working. Nothing there that readers of this blog shouldn't already know, and the article shows its age, and the suggestion that the draw is rarely quoted sub 3.4 suggests the draw receives more attention today than may have previously been the case:
• Low-scoring teams are involved in more drawn matches
• Odds offered on draws often represent better value for money
Two factors, above all others, influence the result of a football match – the difference in ability between the two teams and whether they are likely to become involved in a high-scoring or low-scoring contest. The first has been discussed here before. The second is overlooked by some bookmakers and one consequence of their omission is that there is yet another category of games in which the draw is sometimes offered at bigger odds than it should be.
The fewer goals there are likely to be in a match the more likely it becomes that both teams will score the same number. A good way of illustrating this point is with some figures from Premier and Football League games played during the past 10 seasons. The average number of goals scored in those games was 2.55. Overall, 27% of games were drawn. However, low-scoring teams were involved in more stalemates than high-scoring teams.
Teams whose games averaged between 2.5 and 2.6 goals drew 27% of the time. Teams whose games averaged between 2.9 and 3.0 goals drew only 25% of the time. But teams whose games averaged between 2.1 and 2.2 goals drew as much as 31% of the time. You can see that across this range the incidence of draws varied by 6%, depending on whether teams tended to become involved in open or tight contests.
Bookmakers generally ignore the style of a team's play, taking into account only the strength of their play. Consequently, some bookmakers rarely quote a draw at shorter than 12–5 (3.4).
However, there are occasions when the odds should really be shorter – namely, when one or both of the teams has a history of low scores.
Fulham's games this season have averaged only 1.8 goals, less than any other team in the Premier League. It is no coincidence that Fulham have drawn a higher proportion of their games than any other team in the Premier League – 39%. Wycombe's games have averaged 2.0 goals, fewer than any other team in League Two. And they have drawn 39% of their games, the second highest proportion in League Two. When one or both teams have a tendency to become involved in low scores, the draw can sometimes represent value for money.
After a much discussed drought of draws in the top leagues last season, in particular the EPL, Serie A and the Bundeslayga, they are back (with a vengeance) this season - except in Ligue 1. There's always un, as they say in Paris.
It won't last, so don't try this at home, but if you had backed the draw in every EPL game (using Pinnacle's prices) this season, you would be +11.52 points after 70 bets.

If you had been a little smarter and backed only in matches where the draw probability was > 20%, you'd be up a whopping 32.34 points on the season from just 45 bets. Only one draw has come from the 25 longer priced matches - favourite-longshot bias at work, although like gravity, the effect is a little weaker at the outer limits where draw prices are found than it is close to the centre where the hot favourites are merrily orbiting away.

This return is significantly better than the XX Draws' return from this league it must be said, which is looking rather languid at just 3.32 points from 14 matches.  

While the EPL strike rate for draws of 32.9% is at least within the bounds of being reasonable, in the Bundesliga, it is not. Only 63 matches have been played, but the draw percentage is a ridiculous 38.1% (implied price 2.62). Unlike in the EPL, a large number of draws have come from games where you would least expect them. The 22 matches where the draw is priced at 4.0+ would have made 18.28 points, while a benchmark of 'sensible' matches would have made 16.76 points. The XX Draws in this league? Up 7.55 points on 11 selections. This is a league where the draw price is traditionally a lot longer than the other top leagues due to the higher goals per game, (the lowest this season is 3.34) but if that number stays close to 2.60 goals a game (lower than the EPL and La Liga) this season, the carried over draw prices will be rather generous until it starts coming down. 

As for Ligue 1, the draws after 90 games are at a ten tear low. The 10 year average is 30.2%. This season? 23.3%. The strange thing is that goals are not up on previous seasons, so   Backing all draws would have lost 11.02 points. Backing all sensible draws (sub 4.0) would have lost 24.61, so as in Germany, the unlikelier draws are hitting. 6 of the 15 priced over 4.0 came in, for a profit of 13.59 points. 

France has the additional challenge of a lot of too low prices. My XX Draws exclude any price under 3.15, and backing the 14 matches in this range would have lost 7.81 points. The XX Draws are -3.93 in this league from 11 selections.

A lot of ifs, would haves, could haves etc. above, because no one knew at the start of the season that such a set of results would occur and the fact is that the draws will regress to more realistic numbers as the season progresses. That the XX Draws are under-performing the "draws under 4.0 benchmark" is disappointing, but the currently high numbers in the EPL, the Bundesliga and La Liga will come down and likely Ligue 1's will come up.

La Liga isn't one of my favourite leagues, with the perennial overpowering duo of Real Madrid and Barcelona complicating things, but it's interesting that with these two removed, backing every draw in Spain would have made 10.89 points.

I dare say I shall revisit this topic later in the season, because I do like my draws.

stats counter
page counter