Liverpool on verge of Premiership springboard

At long last, a Liverpool manager with a clear protocol for developing a systematic pattern of (attacking) play for his team. A trio of excellent signings added to a brilliant defence and a world-class striker. If only it had happened when Gerrard was 25. LFC supporters can finally herald, also, a manager who understands what fast, tricky wingers do on a football pitch. Tiki-taka might take some time but, equally, it might not.

Liverpool +19pts 14/1 e/w
Luis Suarez TGS 16/1 e/w

The Value Myth #1

The largely unchallenged notion of ‘value’ in horse racing betting houses a complex problem that I believe has received very little worthwhile attention. The origin of the term as it is currently understood in horse racing betting, so far as I can tell, is unclear but it almost certainly derives its modern context from similarities to the use of the concept in the game of poker and the game of Texas No Limit Hold’em in particular.

In short, ‘value’ as a strategy in horse racing betting has all the hallmarks of a myth. Here I am only concerned with turf horse racing for the simple reason that racing conditions on turf are hugely variable, whereas on the all weather surface there is a far more certain degree of regularity and repetition of racing conditions.

In racing betting the most common phrases regarding the word ‘value’ include: ‘At the prices [horse A] is value’; ‘Given that [horse A] is 4/1 and [horse B] is 14/1, [horse B] represents the best value’; and ‘[horse A] has a great chance but I’m looking for some value’. There are others, of course. What these examples refer to is the idea, roughly, that in any given race a horse’s price offered by the bookmakers may be ‘too big’. This means that in the subjective view of the person placing the bet, the horse has a much better winning (percentage) chance than the odds on offer imply. In one of the above phrases, for example, the person backing the 14/1 shot clearly believes that horse B would win the race in question let’s say around once in every seven attempts, giving the ‘true odds’ or ‘truer odds’ of the horse as 7/1. By backing the horse at 14/1 therefore, the person placing the bet believes he or she has ‘value’ or ‘the value’ because the horse will win the race – in theory at least – more times than the odds imply, giving the bettor ‘value’ or, put another way, a much better bet than should actually be available.

In Texas No Limit Hold’em the same idea is dominant. From the vast array of possible hands one will suffice as an example. Two players, with equal chips, with starting hands (Player 1) 6h7h v (Player 2) AsAd [67 v AA]. Pocket aces is the best pre-flop starting hand in No Limit Hold’em, they will win in this situation roughly 77 times per hundred played (77%). On the flop for these two players comes 2h 4h Js. Now player 1’s hand is favoured more than it was, as one more heart makes a flush, but AA is still dominant and expected to see out and win this hand roughly 60% of the time. Player 1 has increased his chances from roughly 22% to 40%. Both players bet to stay in the hand. The turn then brings a ‘blank’, 2c – a card that for each player is highly unlikely to improve their opponent’s hand. Player 2’s original hand started with maximum strength pre-flop, became more tricky after the flop, but has strengthened again as his opponent fails to catch the card that would make his hand. Player 2 is now an 82%-18% favourite. As such, player 2 makes another bet from his strong position. It is now up to player 1 to decide if he wants to pay to stay in the hand and try to catch a heart on the river that would make his flush and win him the pot. What determines what player 1 will do depends on the concrete value of the bet he is being offered. In scenario (A) player 2 bets 40,000 and the pot becomes 160,000. Player 1 knows his chances of winning the hand are roughly 20% or 5/1. He is being offered pot odds of 4/1 by his opponent (40k to win 160k) and so folds. In scenario (B), player 2 bets 20,000 and the pot becomes 140,000. This time player 1 is being offered 7/1 about a hand he is 5/1 to win. This is value, he calls the bet (20k) and the fifth and final card is turned over. Win or lose, player 1 was offered bigger odds of winning than the actual odds implied and so can be said to be making a ‘value call’. In other words if he repeats this bet at those odds hundreds of times, he will come out in front, winning more than he loses on this particular situation. In the long run, he will have found value, obtaining more money by winning events that were priced incorrectly. If Player 2 bet 30,000 making the pot 150,000, the odds for calling and seeing the final card would be 5/1 – exactly the odds of player 1 winning the hand over time. In the long run therefore, no advantage would be gained and no value can be observed.

In theory, this is exactly what the person placing the horse racing bet at odds of 14/1 is doing: they are looking to consistently make bets at odds higher and in some cases much higher than the ‘real’ or ‘true’ odds of winning the horse holds in their opinion.

The last bit is crucial. ‘In their opinion’; and this is where it gets interesting. You can carry out your own poker experiments regarding the percentage probabilities and decision making of various hands here: http://www.pokerlistings.com/online-poker-odds-calculator. Several inescapable observations emerge when considering poker hands as a means of assessing value:

-    the game conditions are clearly defined and operate unilaterally everywhere. One table, between 2 and 10 players per table, 52 cards only, the same 52 cards only, the same statistical probabilities per hand definition (e.g. with AA v KK the pre-flop odds always favour AA 8 times out of 10 (80%)). This never changes;

- there are any number of subjective decisions and psychological variables at play in a game of poker but none of these affect the actual statistical certainty of a particular situation, such as the example given earlier. The cards will fall and hands play out in exact accordance with their statistical values over a long period of time (long enough to sustain significant variance). The values are known and calculable in theory;

-    the application of statistical probability models is applied and developed alongside the use of inanimate objects – cards.

Now consider the depth of difference between these conditions and the vast (sometimes incalculable) array of variables that feature in assessing races in the sport of horseracing. They are too numerous for an exhaustive list but include:

-    the ‘game conditions’—in this case everything experienced by the horse in the course of their preparation and day of race—is incalculable in percentage terms. There is no race that can be said with any certainty to represent a clear statistical model on known information in any way similar to a poker hand;

-    a horse’s ‘chance’ can change radically at any point in time in the lead-up to a race but a poker card always affects a given hand in the same way, strengthening and weakening that hand accordingly. Of course, poker opponents cannot know for sure each other’s hand strength: that introduces the fundamentals of skill and experience that give the game its bottomless appeal. Racing has any number of indicators: ‘betting money’ may indicate a horse has a very strong chance (it often means nothing too); a horse may have a physical problem or be unwell, neither of which can be known (mostly) and neither of which can be adjusted in terms of probability; a horse may alter its physical state pre-race thus violently skewing any pre-held estimates of ‘chance’—a horse sweating and reacting nervously or even just looking less than bright in its coat could all indicate a depression in the level of their upcoming run: how can that be factored into a sustainable price model?

-    In poker the cards stay the same but the opponents change. In racing everything is in constant flux: stable form/health, opponents, ground, preparation, injury detection, weather, travelling, pace a race is run at—essentially every variable possible is in play precisely because racing deals with animate objects with wildly fluctuating game conditions over time. Virtually nothing is repeatable with the exact same conditions.

Perhaps one of the strangest yet most consistently used terms in racing betting is ‘[horse A] should be shorter’. We can see from the earlier example how this is a widely perceived expectation of ‘value’ if indeed finding a horse that ‘should be shorter’. In the context that underscores this discussion it is a flawed and confused concept, however. In analysing a race the search of all available criteria has the aim of finding the likeliest winner of the race: the horse with the strongest credentials. Precisely because the game conditions in turf horse racing are almost never repeated it is not possible to apply a rationale that is based on the repeatable sequence of events occurring limitless times in the future, as is the case with poker. To say a horse with odds of 14/1 ‘should be 7/1’ is to imply that if the race as it is run the first time is repeated as an event under the exact same conditions (same opponents, ground, track bias, wind direction and strength, horse’s health and physical condition, stable health etc) then the horse would win far more frequently than the original odds suggested. This is unknowable. There can be, of course, a certain amount of estimation. This is the case in poker where one has to estimate an opponent’s hand. In poker, this estimation forms part of the overall decision but that decision is still reliant on a series of statistically predictable and repeatable events where analysis can reveal if certain decisions are profitable in the long term. In racing a given race is never repeated making projected future outcomes invalid and irrelevant. “If the race had been run on fast ground (instead of soft) then the outcome would have been different” is a common refrain but its place as an analytical footnote, as one can see, is just that: it bears no relevance unless the same horses replicate the same race conditions (preparation, preliminaries, field size, draw [flat], etc) with the sole exception that the ground is officially described as ‘good to firm’ and not ‘soft’.

In racing there are distinctive characteristics of the sport that demarcate it from poker in betting terms. The most obvious is physical disintegration and repair. Rise and fall. A racehorse is rarely assured to run a race befitting the ability attributed to it. Exceed expectations and disappoint them. This is in part due to their often fragile physical confirmation but also because they are asked to repeat a performance usually on different courses, different ground, with different weight, against opponents of different ability and so on. It is not yet possible to know or gauge the exact physical exertions of a racehorse during a race and one measure of that has yet to be implemented as a standard, namely the weight of racehorses before and after racing. This would at least give some insight into the level of physical entropy experienced by the racehorse and the extent to which the racehorse has recovered from those exertions. Visual clues are on offer, as are audio soundbites from the trainer, but neither allow us to know for sure particularly as underlying physical problems are rarely disclosed to the general public. In poker the flop, turn and river cards offer increasing and reliable information about concrete theoretical and statistical outcomes. Percentages are therefore calculable; in turf horse racing, they are not.

A second example is the extent to which bettors are ‘forced’ to bet. The blind and ante structures inherent in poker are ‘forced’ bets shared equally among all participants to induce action and prevent persistent non-participation, giving poker a peculiar and fascinating strategic mode of play through which bet sizing and value betting become imperative. With horse racing, there is no such ‘forced’ inducement. One of the more interesting aspects of hearing some horse racing bettors talk about ‘value’ is that those who insist on it as a viable betting model happen to bet the most frequently. Somehow, somewhere, in all those races, there is ‘value’; and there is always ‘value’ because looking hard enough and structuring a view ‘just so’ means ‘value’ is indeed ‘present’ and obtainable. The far simpler view is that such talk is nothing other than a metaphysical abstraction: a layer of continual justification for betting that wraps around a series of events that take place irrespective of it. It seems a very strange state of affairs to talk more about this incalculable abstraction than the actual likeliest winner of a race, than what is most likely to happen given all the available information. The man-made formulation of a series of prices, comprised and altered/adjusted by strangers, has a total lack of impact on the end result. Curiously, it is almost never that a race is analysed for retrospective ‘value’. Only the very top-end races appear to feature any kind of look back at the prices that were in operation for the race; the vast majority of time the talk of ‘value’ disappears as soon as the race has passed, most likely because it was largely irrelevant in the first place: the key was finding the likeliest winner and backing it. For at this point it becomes obvious that the likeliest/eventual winner will always return at odds covering the full spectrum of prices.

An illustrative example is the 2012 renewal of the Oaks at Epsom. The race looked a very open affair as is often the case. The early favourite was Maybe, unbeaten as a 2yo, unable to win as a 3yo: a fairly common scenario and no doubt she was ‘value’ when ‘x’ price ‘if’ she returned to her 2yo form. The actual favourite was The Fugue, presumably ‘value’ given she was heavily backed before the off after cruising home in her trial race. At a much bigger price but with a profile among the best in the race was Shirocco Star: the likeliest winner for some was returned at 16/1 and denied by a neck. Presumably ‘value’ when an even bigger price because she was a big price. ‘Value’, you see, covers everything. And then there was Was. A super-expensive yearling purchase, owned by the richest and housed in the most powerful stable of all. Unremarkable form mattered not a jot. Very late on, she was backed from 50/1 to 20/1 and won, despite being mentioned barely at all in the multitude of betting previews. She hasn’t won a race since. Pick the ‘value’ out of that.

2013 Champion Hurdle

During the build-up to the 2012 Champion Hurdle I was at pains to stress that Grandouet was capable of running to a rating of 170+ and therefore beating Hurricane Fly. A physical problem prior to the Kingwell not only denied the opportunity to assess that conclusion it led me to an incorrect assessment of the runners that did line-up. The perils of reassessment. Rock On Ruby, unable to beat a clearly declining Binocular when last seen in December, delivered a run stamped with prominence, power, rhythm and class; I was there and watching, he won easily. In winning he acquired a rating of 170.

Both his win and subsequent transferral into the full care of new trainer Harry Fry leave the Champion 2m hurdle division in a small state of disarray. It is not uncommon for certain national hunt divisions to become unnaturally weak either in reality or potential. This weakness can materialise as a result of the emergence of a superstar performer in isolation, a simple lack of exceptional talent, or both.

That Rock on Ruby is not currently favourite despite being reigning champion and beating Hurricane Fly in the process is a measure of the exceptional nature of Hurricane Fly’s 2010/11 season that culminated in his thrilling win over Peddler’s Cross. It is quite some measure of Hurricane Fly’s impact that season that he both retains favouritism despite his defeat in March and that for that defeat expectation was still very high and his starting price correspondingly low despite coming off a far from ideal preparation. Put simply, Hurricane Fly was breathtaking for all of his winning season. Since then, the star has dimmed, despite winning two of his three starts last term. He will be a 9yo in March with the same imponderables as last season concerning his physical well-being. His price is understandable but he represents a declining force and as such he is not a project of much interest for the forthcoming season.

Rock on Ruby was beaten on his next start at Aintree over 20f as Oscar Whisky, relatively fresh from his somewhat predictable no-show in the World Hurdle, excelled at the trip and track he was made for. Rock on Ruby is a little hard to gauge: it’s hard to think of him as a ‘great’, as a repeat winner. He beat the steely Overturn as Binocular and Hurricane Fly declined in behind, leaving the overall shape of the race a little questionable. That in itself matters less than the challenge he will face, all being well, for a second time in March. Returning in the same health, form and physical condition is a challenge all too many succumb to but it is also the likelihood of more potent opposition that dampens enthusiasm about our current champion. If that opposition fails to materialise fully, his powers will need reassessment. One conclusion I cannot agree with is that Rock On Ruby was allowed to ‘get away’ from Binocular and Hurricane Fly who sat too far back and were given too much to do. That may soothe backers of those horses but they simply weren’t good enough to give chase; they never travelled as well as they did in their pomp because they couldn’t call on the same physical powers and ability.

On a steep upward spiral, Grandouet oozed Champion class before his injury. The appetite to back him is revived on viewing either of his last three runs: obviously the most recent stands out, cantering all over the eventual Champion Hurdle second in the International before sweeping past him and forging clear after the last. Hurricane Fly missed Champion Hurdles due to injury and so now has Grandouet. He will be six in March, having grown and filled out further a frame that still had more to come before his injury. The enforced time off could be the making of him: as things stand he has the physical attributes and form in the book to be the biggest player of all in March and is trained by the man who can.

All of which rightly overshadows his stablemate Spirit Son. Prolonged absence, as in his case, is never appealing. As such we know very little about him and he awaits assessment once he makes the track once again.

All eyes will be on the young guns from the Supreme Novices, too. Cinders and Ashes is a sharp technician over a hurdle but even he felt the pressure in March as he tangled with two flights, one of them the last (like three of the front four) before surging on giving Darlan too much distance to overhaul. Darlan deserves additional credit for his run and subsequent win at Aintree, coming as they did following a spectacularly crashing fall in the Betfair Hurdle when moving ominously well in the latter stages. It will be interesting to see how both handle the step up into open company.

Rhythm continues to be central..

Spain's Olympic team, in particular, were often shorn of it. 41 shots and no goals, many from close range, underlines both their own profligacy but also just how sensational the Senior team have been in their last three competitions. Brazil continue to flow through London 2012, meanwhile.

Out Now is a great example of a horse having everything in its profile without running a race. Backed from 16/1 into 4/1, he completely missed out the first fence and from there on was never in a rhythm and never in the race. He did manage to stay prominent for a while, but had nothing more to give a fair way out fading to 11th: such runs are often symptomatic of an underlying physical problem and nothing hinted at that before the race. The winner and second were up top the whole way, jumping with, yes, a lovely rhythm as the second attempted to make all but was tracked the whole way and passed by Bob Lingo, who I wouldn't have put forward as the likeliest winner at any stage until the race was off and they were past halfway. The winner was a big price, like Out Now, and this race is another fine example of one that had nothing to do with price and everything to do with analysis, with congratulations on that front to those that found him.