Faint Yet Pursuing -
Signals from the Market (Part 1)
|
|
Faint Yet Pursuing - Signals from the Market (Part 1) |
|
|
Article 12_004_1
18th January, 2012 |
The Market. The Pundits' favourite tool and the centre of everything in racing if you believe them. It is all-knowing and the Signals it sends out as the prices adapt as race time approaches are High Energy and have to be heeded. The prices will be pored over days, weeks, even months, before a bigger race. In the morning of the race when hardly anyone has had a bet even for the lowliest all-weather contest. It's usefulness is never questioned as it consulted endlessly through to the start of the event.
The Market analysis is tremendously simple apparently if you watch the coverage and listen to the equally simple punditising. If the price shortens then this is "..a Good Thing." (flat stress across that phrase and no emphasis on the 'Good', this one is a different format, 1066 and all that) and some mythical group of people, collectively known as 'They' are having their way with the rest of us. 'They' are hugely powerful eminences that the rest of us cannot hope to compete with and they control Destiny in a direct way. If the price lengthens then this is "..a Bad Thing.." and either 'They' have told the others to back off or another group of 'They' is having us over by waiting for tomorrow. This is definitely NOT a monotheistic religion and you have to pay homage to all the 'Theys'.
But, time to snap out of this 'Faith' based nonsense and see it for what it is. Treehugging twaddle. B2yoR has found in other areas that doing something physical can help to break the spell when the Pundit Priests are trying to hypnotise the rest of us. Let us take an example from Football and because of the reach of that sport a bit of cliched punditry that nearly everyone in the country will have been exposed to and is repeated so often it is never questioned. Even though it is wholly wrong and you could have a go at characterising it as believing in malicious sprites, like the Industrial Revolution and all that followed never happened. Which we shall and term it 'Woodwork Hugging'.
The highlights package of goals and notable incidents from the day's football is on the TV and you can be sure you will hear this format from the commentator or voiceover merchant from nearly every game. Here we go with the appeal to some voodoo action in it's first form - "...oh, and [insert attacking team name] denied by the post there...". Or they can turn it around in the "..oh, and the crossbar saves [insert defending team name] there .." and so on, forever. One thing to be very clear about here before going forward, if the ball hit any part of the frame of the goal and bounced away then it was an off target attempt and as much not a goal as the ball being retrieved from Row Z, upper tier, or the street outside.
While apparently a slightly flippant example you do have to do something active to break out of this belief in external influences and take some control back. Here we are after the match and the pundit interviewing the haunted looking manager of the losing team. "Well, you were a little unlucky there..", the pundit begins, "just needed the woodwork to have been a bit kinder with those two shots...". The manager responds with the ".. we just need a few of the balls to bounce for us at the moment, things go against you at the wrong end of the table..." gambit of blaming anything other than him and his team. Balls, guys, and not footballs. You lost here and are bottom of the league because of how you perform. You spend too much time in your own half defending and too little in the opposition half attacking. Because you are mostly chasing the ball being controlled by the other team in the wrong areas of the pitch the Normal Distribution of Chance Events will de detrimental to you. But, this is not 'Luck' and you can change the balance.
B2yoR has found it useful to say out loud "No they haven't" in response to the woodwork 'saving' anyone format and "No it didn't" in response to the spirit-of-the-crossbar approach of the goal's frame actively denying an attacker. It proves very useful in snapping out of the grip of the pundit's watch swinging like a pendulum before you. Keeps you thinking clearly about what is really going on and how much control you have.
We are going to extend this activity to deal with pundits and the Market. Imagine that the worrying ATR betting chimera Chapman Boyce has just told you that 'They' are in the process of prosecuting a huge GAMBLE because a horse has been clipped from 5/1 to 4/1 in the 4:30 at Wolverhampton's Friday evening meeting. Well before the 'Disco Party' crowd have made it to the course and that price change probably took £20 of bets to occur. Altogether now, "No they are not...". Next onto the plain frightening Channel 4 betting chimera Tarnya Mcririck (which one is working the other with their foot?). As they tell us that one horse is "..drifting like a barge...." you should be saying, as loudly as you dare in a room with other people present, "So bloody WHAT...".
First steps, and all that, but what about trying something else to get a handle on what is going on here. Something is not quite right because the research and analysis is being done by the person who does not believe in the Accepted Fantasy Proposition. The 'Establishment' figures are there promoting the importance of the Market and it's simplistic Shorten/Good and Drift/Bad approach. But, you will die waiting for the racing insiders to provide any Proof for the propositions they peddle (Whips effective, Market importance, etc.) so in the end you have to have a go yourself even though not best qualified. You also have to accept, in spite of the anti Faith stance overall, that you cannot kill the Bogeyman, whether he is called 'Frank' or whatever else.
So, the rest of this article, and the second part, give details of a brief study done by B2yoR on a range of horses in the 2010-11 2yo seasons which have been Supported in the Market (noted price shortenings) and those that have Drifted (noted price lengthening). The study covers around 4,700 individual runs during the two seasons roughly split 60%-40% between Supports and Drifts. The input data amount to around 20% of the total runs over the two years. This first part gives details of how the information was analysed and considers how well the Supports did overall and then split by trainer. The second article considers how well the Drifts performed to complete the picture.
Without giving the results away the study, unsurprisingly, showed that analysing the Market Supports and Drifts is a more complex activity than the Pundits would have it. But, 'complex' tends to be taken as a negative word indicating something that is hard to deal with and will make your brain hurt. In this case the positive view should be taken that complex means that gaining more understanding of what the Market moves are saying is actually revealing a beautiful and subtle structure which is far more interesting and diverse to contemplate then enclosing yourself in the 'Faith' box of the Pundit. As ever, the question niggling away is why some of the pundits or those around them do not do this sort of simple study to improve their knowledge. They could then talk more widely and usefully in the endless hours they have to preach to the masses on the subject. With some bafflement it is onward to the next section and an overview of the raw data and what was done with it.
The results for the 2yo races each season have recently had notes added which highlight horses which have had a notable change of price in the Market from the opening price to the final reported Starting Price (SP). Those which have shortened to the final SP are prefixed in the Results notes with 'Supp'. Those that have lengthened have 'Drift' if front of them. The opening price for each horse is recorded in the notes after the prefix. The final SP is given in the dedicated SP cell.
One point to be clear on is that these notes do not cover every change of price that may have occurred. There is some selection made as the results are entered which principally depends upon the size of change and, to a lesser extent, the trainer involved. Despite this selection the number of recorded changes over the two seasons came to 4,660 with 2,841 of those 'Supports' & 1,819 'Drifts'. There were a total of over 21,000 individual runs over the two seasons so the percentage with noted changes is around 22%.
The hope is that the sample size is big enough to produce worthwhile results and that the selection process has not added any, too problematic, bias to the outcomes. A full survey could be done relative easily by someone with access to one of the commercial databases. You would hope that organisations like the 'Racing Post', Attheraces & Racing UK would carry out such work regularly anyway given the amount of time they spend telling the racing audience how to think about the Market.
Another point to note is that the prices used are those reported generally and are based on the returns provided by those people paid to report on the market from the racecourse betting rings along with whatever input the betting shops have these days. Clearly many of the betting rings have very low volumes of turnover and it may take very little money to change a price notably. The amount of money per 1% of change will be variable at different meetings for example. But, these figures are those that are reported as 'Gospel' by the media sources listed so are valid in that sense of checking the correctness of the views they promote. With greater levels of turnover on the Betting Exchanges a similar study using their figures would be useful. B2yoR presumes that this has already been done by somebody but would welcome any guidance from those with more knowledge in that area.
Lots of caveats to start off with but, unlike the Pundits, declaring the problems and limitations at the start and not wanting to copy their approach of misleading everybody with their lack of checking and idleness.
A relative simple process to analyse the data and only one real item to note in the first step. Each individual run with a noted price change was processed to produce a database with the 4,660 runs in. Each database record containing extra details about the run as well as the opening price and the final SP. The main point to note is that the Price Change from opening figure to final was converted into the difference in the percentage of the probability that the prices represented. The change being a positive figure for both Supports & Drifts since it is the total change required because the labelling already defines the direction.
If the previous paragraph was confusing when it mentioned probabilities and turning SPs into percentages then you have some Homework to do to be betting sensibly, perhaps even safely. If you think a change in price from 50/1 on opening to 10/1 final is much bigger, and therefore more noteworthy, than a horse going from 2/1 to 5/4 then you are beginning from an unsafe grounding. Like someone given the job of launching large flares when they have never read the manual. You might get away with it but you would not want to be stood too close to them. Lots of ways for this thing to go wrong. [A quick check for a suitable link at this point to direct people to which explains this fully came up blank. So if anyone reading this needs that advice then drop B2yoR an e-mail to ask for it to be added to the Glossary on the site.]
To give a bit of feel for the changes involved the largest change for a Support was 21.6% which was for a Harry Findlay owned 2yo that won at Beverley. Which will have the 'Dave Nevisons' among those reading this drooling at the thought of how they are going to be proved right and B2yoR wrong. Of the 2,841 Supports noted around 6% were changes of 10% or more and just under 40% were between 5% and 10%. On the Drifts side the largest change was 19% and only just over 2% were more than 10%. Just over 25% were above 5%.
This set of data about individual runs was then grouped in up various ways to be analysed. The figures calculated for each grouping being the numbers of Wins and Runs (and so the Strike Rate), the Profit or Loss backing all runs in the grouping, the Return on Investment (ROI) along with checking averages for the Starting Prices and the SP Change. The Groupings done were :-
With that background the next three sections pull out some examples of the results that came out of the work. The Conclusions section then tries to pull together some of the results and compare them to two possible extreme views. The Pundit view of complete correlation between every nuance of the Market and the actions of the connections of the particular horse. The alternate view that no-one knows much and there is little worthwhile control and the changes are close to random.
The following table summarises the results attained by all of the runs designated 'Supports' in the study. The data comes from the years 2010 & 2011 and the first line in the table covers every run for the numbers of Wins, Runs and Strike Rate. The final two columns then show what the 'Return on Investment' (ROI) would have been for someone having a level stake £1.00 bet on every run. Negatives figures are losses and are colour coded in Red. For example an ROI of -0.02 means a loss of 2 pence on every pound bet, or 2%. Conversely a positive figure shows an overall profit and +0.33 would mean a profit of 33 pence on every £1.00 bet (invested), or 33%.
There are two columns for ROI returns and they relate to the two extremes of the prices available on each horse, ignoring occasional drifts before the price shortened to the returned SP. The first column is based on the 'Final SP' which is the figure that will be recorded in most publicly available published results. The second column is the ROI based on the Opening price for the horse which it shortened in from overall. This will be a longer price so a better profit or a smaller loss. The conclusion section below considers where a reasonable estimate of what price a committed follower of Supports might achieve. They clearly will not get the Opening price in most cases because they will be reacting to a price shortening. Equally, they can presumably do better than the final SP overall.
| Group | Wins | Runs | Strike Rate |
ROI (Final SP) |
ROI (Open SP) |
| 2010-11 | 613 | 2,810 | 21.8% | -0.02 | +0.33 |
| 2010 | 324 | 1,407 | 23.0% | +0.01 | +0.37 |
| 2011 | 289 | 1,403 | 20.6% | -0.04 | +0.30 |
| FTOs | 122 | 655 | 18.6% | +0.01 | +0.39 |
| STOs | 171 | 1,359 | 12.6% | +0.06 | +0.52 |
| SPL4 | 442 | 1,451 | 30.5% | -0.09 | +0.15 |
| SPG4 | 193 | 752 | 25.7% | +0.07 | +0.46 |
| SCL4 | 200 | 1,383 | 14.5% | 0.00 | +0.34 |
| SCG4 | 413 | 1,447 | 28.5% | -0.05 | +0.31 |
| SC Top 10 | 3 | 10 | 30.0% | -0.44 | -0.11 |
| SC >= 10% | 57 | 161 | 35.4% | -0.19 | +0.19 |
The lines in the table below the first line then show how the results vary if you split the overall data in two by some factor. Or, if you take a particular sample from within the overall group. The first two lines show the overall data for 2010-11 split down by individual years. This gives an indication of how consistent the data was in the two years. In this case the answer is almost worryingly consistent.
The rows labelled FTOs & STOs show the results for the just the debut runs in the survey (i.e. First Time Out or FTO), or the second runs (STO). These are sub-sets or samples from the complete data and a third figure for all runs later than STO would complete the set.
The rows labelled SPL4 & SPG4 split the overall data in two depending upon the Final SP. All SPs up to, and including 4/1, make up the 'SPL4' group. All returned SPs above 4/1 are in the SPG4 group.
The rows labelled SCL4 & SCG4 split the overall data in two depending upon the percentage change of the SP from opening price to final SP. Smaller changes up to, and including 4% of relative probability, are in the SCL4 group. All price changes of 4% and above are in the SCG4 group. Note, that these 'change' groups can include SPs of any length since they are about relative changes.
The final two lines are really for reference and to get a feel for whether there might be any difference in results if you only assessed the very largest changes in price. The SC Top 10 line shows the returns for the ten largest shortening of prices. To give a feel for the SP ranges then all ten started at 7/2 or less with the largest opening SP at 20/1 although all the others were in single figures. The lines labelled Supp >=10% shows the returns for every 'Supported' run in the survey which showed a 10% change in price, or higher.
The 'Conclusions' section below examines the results shown in the table to try to assess what the data might be telling us and how the survey could be improved.
The previous section gave an overview of the data collected at the top level without trying to split it down by any 'Agency' that might be causing the price shortening. If we presume that the Market is taking it's lead in these cases from some agent behind the scenes who has set this 'GAMBLE' up then how might we split it down by the agents? As a first step the data was split down by Trainers to see whether examples showed up of trainers with notable good, or bad, records with horses shortening in the Market. A trainer might a primary 'Agent' but presumably would have to have some input, or be able to be influenced, even if the driving forces were owners, or betting syndicates, or whoever in the background. The Trainer also makes a good place to designate the 'Agent' in a simple survey unless you have access to betting pattern data in the way that the BHA does. Grouping at the trainer level also produces reasonable sample sizes for around the largest 50 stables as another advantage.
The full details for all trainers will be available in a particular form during the 2012 season so the numbers are not going to be published here. Looking through the figures you find what you might expect which is a distribution of returns through from very good to rather poor and therefore larger losses. This need not have been the case but the unusually consistent overall returns for Supports noted in the previous section might well indicate a Normal Distribution (of Market knowledge of trainers' runners) working through to visibility given a large enough sample size. If you take the unworldly 'Pundit' view of every shortening in the Market being really meaningful then the overall results would look different. Nearly every trainer would be in profit on their supported runners, the overall ROI would be ridiculously high, and there would be only a narrow variation in individual trainers between very good returns and good ones.
Let us take three examples of trainers whose records seem to indicate a real link between Market shortening and expressed performance. Unlike the Pundit, try to improve our knowledge of what the Market is really telling us, or when it can be ignored. For reference, 80 of the 205 trainers represented in the survey made a profit or broke even.
The first example is a real surprise to B2yoR because Clive Brittain has long been one of those trainers held in below average regard. He could get winners with his very best 2yos but his overall returns were well below average given the type of 2yos his Arab backers provided him with. You could pick out several areas where he just failed to make the best of what he had. Then, despite being closer to eighty now than he is seventy (he was born in 1934) his record picked up and the last three seasons have produced above average results overall. Can he really have changed? All those visions of him spending hours forcing Radetsky (the horse, he isn't that old) backwards to get him to go onto the gallops in the Newmarket qualifiers for the World, inter-species, Pig-Headedness competition. Or moaning away like an old-timer because he can no longer do all his gallops work in the dark and have his horses washed down and back in their boxes before anyone else has left their bases.
But, look at his record with supported horses in 2010-11 and it is entirely possible to believe that stable confidence actually means something and leaks into the Market's domain. A solid number of representatives and an excellent Strike Rate at 36%. Look at the horses that won or placed compared to those that were unplaced and all but two of the largest changes in price are among the winners and placers. All but one of the winners opened in the Market at single figures but still made the noted shortenings in price while the unplaced horses mostly opened in double figures. An overall ROI over 100% on final SP figures and a profit in both 2010 & 2011 and a better record in the latter. Look at his record with a similar number of Drifters and it is completely different. Seems the old dog really has learnt something in more than 60 years at the Game.
William Haggas has shown hints over the years of having a much better than average handle on the real quality of his 2yos before they run. He is the only trainer, in B2yoR's memory, who has ever used an OR rating figure to characterise a juvenile in the annually published "2yos of [insert year]", for example. He said in interview about an older horse in a Group race in 2011 that ".. he only gets Johnny [Murtagh] over to ride if he thinks the horse is really competing to win.." & he seems to judge that split well enough. Haggas spent time with Mark Prescott while learning as an Assistant and perhaps he picked up some ideas there. Prescott priding himself on being a master of placing his horses and knowing exactly what OR figure they are ready to run to is a key part of that.
A solid profit overall and one in each of 2010 & 2011 but a less clear cut case overall than Brittain. Many more examples of supported horses and a loss on the horses with the largest SP changes and, mostly related, also a loss on horses with final SPs that get below 4/1. Of the 43 runs he had with noted Supports as few as six opened at double figures in the Market and none at more than 14/1. You could propose a theory that the Market actually over-reacts to the stable, and whatever agents surround it, and this drive some areas into losses. If you could identify a way to cut out the 'smoke' from the real targeted ones then really good returns would be possible. But, again, a subtle story needing work and far more interesting to pursue than the fantasy, 'Flash Harry', world of the pundit.
The real clincher with Haggas in believing some real link to the Market prices is his results with horses that drift in price. Fewer representatives overall, barely a winner from them in the two years and large losses however you group up the runs. Possibly this confirms the view that a number of the Supports are on reputation, or misinformation, or expecting too much of horses on debut (he is not good at tuning them for this run), or other reasons.
Rather than picking out another of the larger and more established stables for the final example perhaps interesting to look at a different level. There are useful facts to learn about Rod Millman's developing approach over the years from his results and Elaine Burke had a very good first season in 2011. Mrs Burke is the front for Karl Burke returning to training of course and he had a good record with juveniles before his enforced break for misdemeanours. The stable seemed to have a very good group of 2yos in 2011 and you would worry that the positive results owed a lot to that and might not be repeated. Mrs Burke is the daughter of trainer Alan Jarvis who has a dreadful record with his supported 2yos. You would probably not have to ponder too long about which of the trainers up close in her life had most influence in 2011.
To some extent a similar story to the Burke one with Keith Dalgleish who was the latest person with a Scottish passport (oh, they haven't ceded just yet, apparently) to take over at owner Gordon McDowall's Belstane Stables. He was in his first season with licence in 2011 and you wonder how much of the strong start owed to an unusual group of competitive 2yos. All four of his winners, from eight runners, were owner breds including one he took over from Mark Johnston, his main supporter in his days as a jockey. About 6-7 stones ago. So it just needs a few mares to produce a duff one together and he is in for a poor season. But perhaps his strong start and contacts will increase the numbers.
But, an interesting story to watch develop because he made a good job of handling the 2yos he had and got good returns overall. The small number of horses noted as Supports in the Market produced a better than 50% Strike Rate and solid to good profits in all grouping except when the final SPs got below 4/1. This final point turns up with a lot of profitable trainers and you start to wonder whether it is saying something fundamental about how the Market works. The biggest price changes married up with the best performances to aid the positive view. As with Haggas & Brittain the results with the horses which drifted in the Market add to the outlook that the Market says something useful in this case. No win from a similar number of Drifts to his Supports.
The previous section picked out a few trainers who showed profits following their supported juveniles and checking into the overall record suggested some real indicators were getting through to the Market. In this section the examples consider those were the Market signals are unclear, at best, and in some cases plain misleading or wrong.
The first example is really interesting because if you look at Tom Dascombe's overall record with 2yos noted as Supports and it is dreadful. Well above average losses overall and losses in all groupings. A longer priced FTO win helping to make that category a bit more respectable. The initial reaction was to make him the star example in this section and the losses mean that he is still a prime exhibit. Sit there listening to pundits telling you that Dascombe's associates ".. are backing this one.. " and reacting and you will find yourself having lots of bets and making big losses.
You could characterise this as the Market knowing little about the Dascombe horses. It is certainly too sensitive to input and information related to the Dascombe yard. At the anecdotal level you could understand why that is. Dascombe is a part salaried trainer backed by Michael Owen and Andrew Black. The second of those who made his fortune as part of the group that built up Betfair. Easy for an outsider to believe that all those millionaire footballer cronies and betting gurus, including scouse muckers from the Racing Post, must be up to something a lot of the time. Dascombe's personality means that he can encourage this sort of thinking just by being himself. He tends to be bullish and upbeat about his horses and uses the network social media fully to propagate those views. Which means the 'availability' of the information from the stable probably gives it a power which is often uncorrelated with actual performance.
So, at the headline level a trainer to avoid when it comes to Market moves because if there is anything worthwhile to be dug out of the signals it is very hard to identify from the background noise. And yet. Fiddle around with the raw data a little, sort it in different ways and on one particular way of looking at it most of the winners are in the top 20 examples and a 33% Strike Rate. You would still have to still have to work to make a profit because the Dascombe horses tend to open at short prices for the reasons above. While the Market looks clueless on Dascombe there is a faint signal in there if you know how to look. You would still have to sit through supporting some over-hyped disappointments like Juno The Muffinman, Kenny Powers & Electric Qatar. But, you might be able to usefully identify some real Market signals from the stable.
As a challenge to the Pundits perhaps they could do a bit of work to find out how to sift through the pile of meaningless Market moves for Dascombe to find the few really positive ones. Rather than sitting there telling the world that every one is a full-scale GAMBLE in action.
The second example is David Simcock and a younger trainer with a growing positive reputation. But look at his record with Supported horses and a poor record 2010-11 and above average losses in nearly every category and very hard to find any positives. Look at the horses with the largest changes in price and, unlike some trainers with poor records overall, there are no signs of better outcomes. Detailed study of those largest changes with the benefit of hindsight and it is hard to see what caused the changes. With one exception they really were limited ability horses who have achieved little since. In some instances it almost looked like hopefulness that an Arab owned horse trainer by Simcock in a weak race would be enough. But, certainly not any input from the Trainer feeding through to the Market.
Look at all his supported horses as a group and the size and types of move do not suggest a stable that is communicating with the Market in any direct way. Certainly not trying to set up the conditions for a Market move or getting involved in GAMBLES. Add in the shape of results with the Drifters he had noted and a picture of a set-up where the Market knows very little.
The final example is again a trainer who had his first season in 2011 and presents an interesting story to watch develop. Roger Varian had been Michael Jarvis' assistant for ten years before he took over the licence before the 2011 season with Mr Jarvis fighting long term illness and a battle he unfortunately lost during 2011. Look at the Jarvis record in 2010 and easy to believe in the small set of noted Supports indicating a real link to the Market. Not big moves but very good profits when those moves got above a cut-off percentage level which seems to be an indicator level for a range of trainers and is covered in the conclusions section. An especially good record with FTOs and he was always a trainer who showed clear patterns in how and when he started his better types out and they would be ready to win on debut.
With that background what would you expect from Varian? He is taking over an established stable and has retained most of the owners. Jarvis and his wife were still involved in the stable until Michael's death. Varian had been at the yard for years and would presumably carry on doing the same thing? But, look at his 2011 record and the total number of Supports is the same and the range of SP changes the same as Jarvis in 2010. But, a poor record overall and big losses in every category. Unlike Jarvis in the previous year the best runs by the Supports are in those below the cut-off price percentage change. Has something really changed or just a set of lesser horses in 2011? If the change is real is it with how the trainer approaches the job of training or with the communication channel to the Market now the personnel are subtly different?
What makes the results more baffling is that Varian's record with the juveniles noted as Drifts during 2011 was terrific. 'Retire abroad' stuff if you could count on them continuing. The Jarvis record was ok with Drifts and suggested a one-way message going to the Market in his day if there was one. You could propose a theory that the returns perhaps depend a lot on how the FTO runners go and Varian took some time to find his aim here. Something to watch for in 2011 to see whether this changes his record or not.
The first point to note with the analysis is that, on reflection, the selection of Supports and Drifts that have been recorded are too 'positive'. At the headline level in the Supports, for example, the sample records make up close to 14% of all of the 2yo runs over the two seasons. A large number of the other runs are not qualified as Supports because they are Drifters or the price did not change. Even so, the 14% figure is still a sampling, to some degree, within all Supported runs. But, the total number of winning runs recorded in the Supports section is 28% of all wins across the two years.
The selection process was more thorough in catching winning runs and those cases which produced the largest changes in price from the opening figure to the final SP. The following conclusions are likely :-
With those notable caveats documented the results still show some interesting features. The overall view is that in certain areas the Market can hold useful information when a price shortens. Reading this effectively requires real knowledge of how particular sets of horse's connections operate. Not only overall but when they are working within different situations, such as the opening price of a horse and how much the price changes in response to any support.
The overall figures for Supports show a profit of around 33% on Opening price but this reflects an unrealistic profit level for two reasons. Firstly, because of the points noted above. Secondly, because a backer trying to respond to Market Support to inform their betting are not going to get prices at the opening level overall. The final SPs show a Break Even figure betting on every selected supported run so would be expected to dip into losses with a full sample. However with a 33% range between opening and final SPs you could speculate that a committed follower could get close to breaking even. A fuller study would be needed to resolve this.
The figures which just analyse the largest changes in price actually show worse returns than the total figures. With hindsight this probably fits with the points above in that more of the losing runs have been captured because the vast majority of the larger price percentage changes will have been captured by the selection process.
The really interesting points come when comparing how trainers perform. The selection process should not show the same bias problems so the differences between trainers should be real results. A fuller study would not be expected to show much change in relative performance.
Rather than a flat response of profits for most trainers with Supported runs the results show a complete distribution of returns. From very high profits through to 100% losses. As covered to some level above you can start to categorise trainers into a small set of types. At one extreme those that appear to have no direct communication with the Market and little sign that anyone associated with their yards fills that gap in any meaningful way. The supported runs for these trainers will show losses in most groupings and the occasional wins just seem to be normal variation and not part of some directed pattern.
At the other end of the scale there are trainers who get solid profits on supported runs and other indicators seem to confirm this reflects real knowledge being input to affect the Market price. The associated indicators which seem to be solid links are better results with the more strongly supported runs and solid losses with runs that drift in the Market.
The first of those points introduces an interesting point which is to try to define what percentage change in price reflects a real input of insider knowledge. To make the point the Pundit sees all Market moves as significant when prices will move naturally as the Market develops and responds to inputs from every source. A horse may shorten, or drift, in the Market simply because it was introduced in the Market at the wrong opening price. The input of everybody, not insider GAMBLES, will then shift the price to the perceived 'Right' level with no directed insider influence. Remember, that all of the prices in this Study reflect the pre-race betting Market prices generally reported in the media. Not Ante-post or morning prices or the Exchanges, for example. Similar figures for the Exchanges would be useful and are presumably available 'somewhere'.
Looking through the data for various trainers B2yoR would suggest that a useful cut-off level to start classifying price changes would be somewhere between 4-5%. Sort many of the trainer's raw data by the SP Change percentage and you find the majority of the wins and better runs in those changes above this range. With some trainers, who have a poor overall record and losses, you can find an SP Change percentage level that the wins they get are grouped above. You could propose that changes below the cut-off point were often just normal shifts caused as the Market found it's final levels from the opening point.
With some trainers you could suggest that although they normally do not try to interact with the Market they occasionally do when they have a really strong feeling that they have a superior horse. A good example would be someone like John Quinn who shows losses overall. But, when he thinks he has a Listed+ quality horse the opening prices get very low and the horse gets supported further anyway. His record with a small number of these types 2010-11 then becomes profitable.
One of the things you would suggest out of this is that the Media should always report the Percentage Price change in their Market Move features. Not just the upper and lower price. To take this further they could pick out a small set of trainers and have an indicator on their Market Price display pages which showed when the price had changed by more than a qualifying figure. Do some useful research on-the-hoof and inform people as they do it.
This article is only 'Part 1' of the story because it only deals with runs by 2yo horses which were noted as shortening in the Market. The simplistic view in those cases is that the shortening represents a real link to the views of those associated with the horse and therefore a strong performance is expected. Hopefully this article will have made the point that reality is more complicated and subtle than that. There is some useful information hidden away in the Market shortenings but in a pretty low percentage of those moves and you need to know how to assess where the worthwhile pointers are.
The second part of the story comes in the follow up article as 'Part 2'. This deals with those horses which have been noted as lengthening in the Market and are termed 'Drifters' or Drifts for short. Again, there is a simplistic media view to compare the results of the study to. This is that the drifting is well correlated with connections of the horse knowing that it is not going to compete well in the race. A punter should avoid all drifters because they are not going to win. The performance by trainers with their Drifters compared to how those supported have performed over the same period should be instructive as well.
Taking the simplistic view we should probably expect that 'Part 2' will report poorer returns and almost certainly big losses for backing drifters. It ought to be easier for a Trainer, for example, to know that a horse will run badly because there are far more ways for it to 'not win' than be successful. To be simplistic again, there are nine places for a drifter to finish in a ten horse race where they will be unsuccessful and only one to win. Someone ought to have a harder job definitely knowing a horse will win than manage to finish in the range of spots that count as losing.
But, how about proposing an alternative view that many drifts in the Market are not caused by real inside information at all. Perhaps they are just indicators that no-one has a particularly positive view of the horse so do nothing to arrest the drift. The horses are also lengthening in price so unless the Market is well tuned perhaps they drift to long-term value prices and will manage small losses and possibly profits. How sure are those promoting the accuracy of Markets that it always knows when to put the brakes on when a drift starts? Given the way the media highlights drifts is there a feedback mechanism which pushes them out to longer prices which has nothing to do with the views of the horse's connections?
Only one way to find out which is to do some work and see what the data and results say. Time for you to have a think about what you are expecting to see in 'Part 2' so that you have a framework within which to review the results reported.