Hugh Taylor Selection Angles -
18th March, 2013
Regular readers of the B2yoR website will perhaps remember some issues higlighted in the 'Pointless Punditry' articles. Of particular relevance here are the way humans think covered in the 'Blink' book and the inability of most 'experts' to be able to predict the future to any worthwhile level in 'Future Babble'. Another book - 'Thinking, Fast and Slow' (TFS) by Daniel Kahneman covers both issues, and a wide range of others, in considering how the way we think can hinder us in getting the best result in many situations.
A brief recap of the path to this point. In 'Blink' author Malcolm Gladwell introduces how the human thinking process is split between a fast acting, intuitive, process of an inner 'you' and the slower and more concious thinking we undertake. In TFS the author covers this in huge detail and terms the intuitive, snap judgements as 'System 1' & the slow, hard work thinking as 'System 2'. A major point to note is that we are lazy to a large extent and we will let System 1 do our thinking for us if it appears to come up with a sensible story. Which it is very good at doing but it is prone to the wide variety biases which hamper our ability to get to the correct solution in many instances.
Further, the TFS book is very good at considering the problems caused by thinking and biases and it can help us in understanding & fixing the 'worthless Pundits' issue. Consider this quote from the 'Illusion of Validity' chapter.
"..We knew as a general fact that our predictions were little better than random guesses,
but we continued to feel and act as if each of our specific predictions was valid.....I coined the term for our experience the illusion of validity."
Which, as an aside, helps to explain why racing Pundits can spend decades presenting the same approaches and thinking without worrying about how worthwhile it is. Assisted by an audience that thinks the same way and wants confident Pundits regardless of the actual long-term 'cost'. Which is never totalled and considered.
On the way to suggesting some methods to break this way of taking decisions the TFS book highlights some examples of the problems it causes. For this document let us take a couple from the world of Financial Markets which, while not exact, have a lot of obvious similarities to gambling on horses. But, try suggesting to a person 'playing the stock market' that they are like a horse player and they will be appalled. They do not see themselves as gamblers but high powered experts applying their superior knowledge and skills. Let us see whether we can puncture that bubble.
The TFS book goes through a number of examples in research into trades, both buying & selling of shares, stockbrokers & individual investors made to see how much skill is being displayed. Conversely, to be able to identify how much is due to luck. Whether an individual trader wins or loses in a set period are just random fluctuations which could be entirely reversed in the next period if little skill is involved. Rather than working through those examples let us cut straight through to this summary :-
".. the year-to-year correlations [the measure of skill involved] between the outcomes of mutual funds
is very small, barely higher than zero, The successful funds in any give year were mostly lucky; they had a good roll of the dice.
There is general agreement among researchers that nearly all stock pickers, whether they know it or not - and few of them do -
are playing a game of chance. The subjective experience of traders is that they are making sensible educated guesses
in a situation of great uncertainty. In highly efficient markets, however, educated guesses are no more accurate than blind guesses."
A quote which makes betting on horses seem more realistic and 'honest' than playing the Stock Market. Your failures incur personal losses quickly and you have to think about how to do better and not spend your time lobbying for the salary bonus you think is your due, no matter how incompetent you actually are. But, the last sentence of the quote needs re-reading and careful thought in how it applies to your betting approach. Face up to the fact of whether you are a 'blind guesser' fooling themselves with an illusion of validity, or do you have a real 'edge'?
TFS's author gives an example where he was invited to talk to a group of investment advisors for a company with many wealthy clients. He asked for some data prior to his visit and was surprised when he was given 8 year's worth of information. This covered the trades individual advisors had made and the end of year results comparison of all advisors. He ran the correlation test on the data to see whether the 8 years showed any persistent patterns which could be assigned to the skill of any individual. His calculations showed that the advisors were involved in a 'dice rolling competition' and almost zero real skill was involved. The real surprise for the author was that when he presented these results in his talk to the board, senior managers and the advisors they did not seem to be shocked. By the following morning when he as leaving he felt that all those involved were colluding in ignoring the message and instead 'sweeping it under the carpet'. The Illusion of Validity at work and it being too hard to face up to the fact that they, and many other companies, were involved in a game of chance and not superior skill. When being driven back to airport by one of the advisors after the visit the author reports this response :-
" [the advisor] told me, with a trace of defensiveness, 'I have done very well for the firm and no one can
take that away from me'. I smiled and said nothing. But I thought 'Well, I took it away from you this morning,
If your success was due mostly to chance, how much credit are you entitled to take for it?' "
At which point let us consider how we can improve things and break out of the problems that these cognitive illusions and biases create. In TFS the next chapter is titled 'Intuiton vs Formulas' and covers what you would expect. Rather than using humans thinking they are applying skill when they clearly are not could we use a simple algorithm instead. Lay down a simple set of rules and just have people follow those and not start adding their own intuitive thinking, or even their full thinking. This approach is not popular with people at large and seems counter intuitive. It especially riles people with the illusion they are highly skilled professionals.
Studying the two approaches started in the field of medicine by looking at how well expert physicians did, in diagnosing illnesses and prescribing treatments, when compared to combining a small set of scores by a simple rule. These showed that the algorithmic approach did better. The following quote summarises the results from a large set of similar studies, including those in areas outside of medicine :-
"About 60% of the studies have shown significantly better accuracy for the algorithms.
The other comparisons scored a draw in accuracy, but a tie is tantamount to a win for statistical rules, which are normally
much less expensive to use than expert judgement. No exception has been convincingly documented."
The book provides a lot of other examples of illusions and biases which could be applied to betting on horses. These include a good section on how to solve your Risk Aversion and deal with losing runs if the responses you have to those limit you. But, let us now turn the thoughts about experts fooling themselves, people wanting confident experts, etc., and apply this to Racing & Pundits/Tipsters. How much are they fooling themsleves and us? How much are we fooling ourselves that our favourite 'winner picking' methods are proven and useful? Could a person be a better punter, and tipster or pundit, if they just used a small set of proven angles in a systematic way? Not just choosing what they like to use and what appears relevant in these cases and mixing it all together into a 'Black Art'.
B2yoR has tried applying this thinking to it's own betting process and it can be painful to realise how much of what is used could be characterised as 'unproven black arts'. Even starting a 2yo season with good intentions of systematic behaviour can get overtaken by the pressure of time in the second half of the season and means you end up falling back too much onto unproven & intuitive thinking. Which, as the TFS book explains at length is decorated with problems.
Note that there are Tipsters who work in this more systematic way using a small set of Angles. The most obvious would be Time- and Speed Figures users. B2yoR would be a strong supporter of the use of a, well produced, set of these numbers as the 'Top Level' Angle coupled with a simple algorithm to turn the top raters in any race into selections if they qualify through using the simple rule set. In theory the 'Race Trends' approach is systematic. but only to a certain level and will always rely a little too much on the future being like the past. Unable to anticipate new trends as they develop, adapt to odd circumstances etc. But, still a reasonable starting point to try to build an algorithmic tipping approach on.
B2yoR will be trying to implement the knowledge gained in 2013 and run a set of Systematic 'Angles' betting approaches along with the more 'artistic noir' or 'intuitive' approach. But even in the second case there will be a checklist of input that HAS to be assesed and not just those parts you feel you can weave into a believable story to satisfy your lazy 'System 2' approval.
While going through this process the thought came up about why not apply this 'Systematic Process with Angles' approach to another Pundit or Tipster. Look at the selections they make and try to classify them into a set on Angles and see how systematic they are. Hugh Taylor (HT), of the At The Races website, was chosen for several reasons. He puts up regular selections and also puts up a good desciption of why he has chosen it and what angles he uses. Unlike much Punditry which is empty of reasoning or just puts up the odd reason which will be different in most races. He has also proved his ability to make profits over a long term, at advised prices, so he must be doing a lot 'right'. There must be some real skill being displayed and not just a game of dice.
One question might be why look at HT selections at all if he is already profitable? There are lots of answer to this. Firstly, to document what he does and try to learn from his approach. But also to try to find areas where his methods are weak so that you could make the returns better by cutting out the selections he makes built on suspect foundations. It is also easier to criticise someone else than admit your own shortcomings and time wasted on unproven tools.
The rest of this document gives a description of the results so far of trying to classify HT's approach. It is termed an Interim Report because there is only 2 months data so far. A much larger sample will be needed to get to definitive conclusions. But, enough data has been collected to produce a draft classification of the Angles & Approaches HT uses and start to think about how profitable they are? How systematic is HT in his approach? and so on.
A couple of other reasons for producing an interim report. Publish it prior to the Flat Turf Season start as a hint to others to think about how they bet. Also to share a bit of insight and see whether it pans out in the longer term. As a tease to encourage you to read the rest of this report think on this. Over the 2 month period HT's selections made a big profit at his Recommended Prices (RP) and still made a small profit at SP. But, you could have made an even bigger return on your investment (ROI) than that at RP but still based entirely on following HT's selections. How? Read on.
Back to Top of Page
Section 3. gives an overview of how it is planned to investigate the data recorded about Hugh Taylor's (HT) selections. A primary aim is to characterize each selection by a small set of Angles which are assigned a particular code depending upon the angle. This should ideally be around 6 Top Level angle codes and 8 as the maximum. Having more than this range is an indicator that there is not enough similarity between the different selections to believe that they are worth analysing. Also, more angle codes mean that the sample number in each will be small until a large amount of data is collected.
2.1 Top Level Selection Angles
This section gives details about the top level Angles that have been assigned at this interim stage. There are 11 detailed in the table below which is too high a number. 'FLATNH' and 'TRNRFM' have both been assigned just once and should probably be dropped or the selections assigned to a different existing Angle code if possible. The '1STHCP' code has been used twice and these records could be included into the 'SOFOR' code if necessary. This would leave 8 Top Level codes although 'SWEL' is a special case of the 'GNG' code so could be combined to leave a workable set of 7 if none of the other codes can be reassigned.
|1STHCP||Horse in it's First Handicap Race. Typically on 4TO after three qualifying runs. The reasoning should highlight why the horse should be capable of performing above the level it has shown so far. Possible to fold these into the SOFOR Angle but left as separate item at this Interim stage.|
|BTOR||Better Than OR. A horse with recent form which HT believes proves it can perform to a level above it's current BHA rating (OR = Official Rating). For example, HT's opening sentence might say a horse is "well ahead of the handicapper..".|
|COFOR||Competitive Off OR. A horse that has looked "in form" in it's recent outings without appearing notably better than it's OR. The supporting reasoning whould highlight why the horse is a value price in the day's race set-up. Also used, on occasions, where it has been difficult to assign a single Top Level issue and the selection seems more a sum of a number of lesser positives.|
|CRSFM/RH||Course Form or Course related issues. Mostly used for horses with a good record at the course the day's race is at. RH indicates HT believes the horse would be better on a Right Hand course. Probable that in long term this category could be a catch-all for all course specific items.|
|FLATNH||A one-off use by HT on a day when the snow meant the only meeting was a 'Bumpers for Jumpers' affair on the AW. HT's angle was a horse with proven flat form taking on inexperienced NH types in a 'Flat' race. Will be used rarely & possible that a 'Miscellaneous' category is needed.|
|GNG||Going. Used where HT bases his selection on a horse who is proven on the surface they will run on in the race. They will usually have been running on a surface that is unsuitable for them in recent outings. Note that this change could be any combination of Turf, AW & going within either type. But, note that HT sees the Fibresand surface at Southwell as an important angle and therefore a separate Top Level category is used.|
|SHPBTR||Shaped Better Than the final Result. A horse who has run better in a previous race than the placing and summary in a normal Form Book would indicate. This categorisation rests on the large amount of Video Review that HT undertakes (every British race?). The "Eyecatchers" that HT highlights should all fit into this category. The Angles listed in the Eyecatcher table below are all angles which fit below the 'SHPBTR' Top Level category..|
|SOFOR||Scope For Improvement Past the Current OR. Horses who do not have proven recent form that they are competitive off the OR they will run off in the race. The two main sub-groups would be, firstly, horses with few career runs whom HT thinks can still develop further for some reason. The other would be horses who have won, or been competitive, off a notably higher OR in the past. These would include horses coming back from a long injury break and also those with long losing runs resulting in their OR declining.|
|SWEL/ST||Running at Southwell on the Fibresand Surface. Part of the 'Going' Angle but used so regularly by HT as to merit a separate Top Level category. The main sub-group would be those horses with proven Fibresand form returning to SWEL having run elsewhere recently. The smaller sub-group would be horses on their first run at SWEL where some statistic(s) suggests they have a better than average chance of being suited to Fibresand racing. These would be noted as 'SWELST' to split them from the 'proven' horses.|
|SWITCH||Switch to another Trainer, a 'Switcher'. Used where HT believes a move to a different trainer will produce an improvement over recent form. Note that this is not used solely to indicate the first run for the new trainer. It could be as many as 3-4 runs before the horse becomes a selection if it's form seems to be steadily improving or the trainer has now found a suitable race set-up.|
|TRNRFM||Trainer Form. For a trainer whose recent runners have been competing better than the stable average. Another one-off which perhaps needs to be put into a 'Miscellaneous' category and 'Trainer Form' is mainly used as a supporting reason below a Top Level category. On one occasion HT had two selections for the same trainer on one day and the second seemed to have little else to recommend it beyond a feeling the trainer was 'in form' [Does this exist?].|
Back to Top of Page
Below the Top Level Codes there is then expected to be a small set of items which HT highlights in his selection text which adds support to his reason for choosing this horse to tip. The table below details the wide range of positive reasons that HT has used in the first two month period. They have been placed into four groups for reasons that relate to the 'Horse' itself, the 'Race' conditions and set-up, the Trainer and Jockey plus an 'Other' category for those items not easy to fit into the other Groups.
The current set of 19 supporting reasons, plus the caveats listed in the next section, is much too large to be feasible to analyse fully. It also goes against the starting point of the analysis that the outcome should be a simple algorithm to implement which a small set of variables to score and then grade the final result. Many of the supporting reasons are rarely used and one approach would be to drop them entirely unless there is evidence that they add anything to the analysis. The obvious simplifying approach would be just to use the four Groups, or perhaps just three and drop the 'Other' set. The supporting reasons would then be aggregated at the Group level. At this interim stage it is acceptable to have a full list of supporting reasons, and caveats, so that the exclusion or aggregating of similar items can be identified.
|HORSE Related Reasons :-|
|RS||Run Style. The way a horse typically runs through a race will be an advantage. For example, a front-runner likely to get an 'Easy Lead' or a hold-up horse in a race which should have a fast pace.|
|PD||Pedigree. Where the horse's pedigree suggests it will be suited to the conditions of the race. For example, 'Sires for Going' or a horse stepping up in distance with a pedigree which suggests this is a positive. Occasionally it will bring siblings in as part of the reasoning.|
|HD||Headgear. Where is horse is running in some form of Headgear which HT believes will improve it's expressed race performance.|
|CD||Course and/or Distance items. Could be removed and made part of the Top Level 'CRSFRM' category. Left in as a separate item at this level at present. Would include horses who had a good strike rate at a particular course, had won or placed in the equivalent race in a previous year, and so on.|
|TM||Time or Speed Figure. HT produces his own Time Figures for the AW courses. Used to indicate a horse who has run to a higher time figure recently than would be expected for a horse with it's OR level, or for the Class of race it has been running in.|
|FR||Form Franked. Used when other horses from a previous race the selected horse has have run in have gone on to win, or run to a competitive level, in their subsequent outings.|
|OD||OR Down. The horse will be running off a lower Official Rating than in recent outings.|
|RACE Related Reasons :-|
|DS||Race Distance. The distance of the day's race is a change from recent outings. This could be a either a decrease or increase in distance and still be a positive depending upon the selected horse's aptitude.|
|PC||Pace of the Race. How fast the subject race is likely to be runs. Depending upon the horse's Run Style this could mean a slower, or faster, than average pace.|
|LS||Lesser Opposition. The selected horse will be facing less able opponents than in previous outings. This could be because it is in a lower Class race or simply because the assembled field is mostly uncompetitive off their ORs for various reasons.|
|DW||Draw. The horse either has a very good draw in the day's race [for it's Run Style?] or has been hindered by wide draws in recent outings.|
|TRAINER/JOCKEY Related Reasons :-|
|TF||Trainer Form. Where HT believes the Trainer's recent runners have performed to a level above the usual level. Perhaps better when used to identify a trainer who has had a period of horses performing poorly and now feels they are running to the normal level. Best when the trainer has identified the reason for the poor runs, perhaps an illness with the horses or a change of feed used.|
|TO||Trainer Other. Catch-all for other reasons HT assigns to being solely down to the trainer's choice.|
|LO||Laid Out for this race. Used when HT believes the horse may have had it's training designed so that it's performance peaked on the day of the race. Examples so far indicate that HT does not contact the trainer to check the view is correct.|
|JK||Jockey. HT believes the jockey used will give the horse an advantage in the race. Sub groups include the use of Apprentices claiming weight allowances whom HT rates as better riders; the use of a highly rated professional jockey after amateurs or lower ranked professional have ridden recently; the use of a jockey thought to be adept at riding a certain course or delivering a certain type of Run Style which will suit the horse. Note that HT used to be a Jockey's Agent and clearly feels he can judge a jockey's abilities himself. [Interesting to compare this to John Whiteley's views on people being able to subjectively assess jockeys.]|
|JB||Jockey Booking. Where the booking of a particular jockey is deemed significant. For example, a high profile jockey riding for a smaller stable when he usually would not.|
|TJ||Trainer & Jockey combination. Occasionally used by HT when noting that a trainer has a notably high Strike Rate when engaging the use of a particular jockey. The implication being that the trainer would not 'trouble' the jockey unless the horse was thought to be competitive to win the race.|
|OTHER Reasons :-|
|MK||Market Support. Denotes that HT feels the horse has been the subject of Market Support in some previous runs. Usually couched in the "well backed" format. [How is significant support judged?]|
|SD||Something Different. Catch-all item for recording that some other positive is not just 'noteworthy' but better than that. For example, below the 'SHPBTR' Top Level category the strength of finish a horse produces, or perhaps the high confidence, probably unusually so, of connections gleaned from public sources.|
Back to Top of Page
As well as detailing positive items relating to a selection HT will often add some warnings or cautions if there are some perceived negatives with the race set-up. An example would be a horse where the distance, going & race conditions are ideal for a strong effort from the horse but an inexperienced jockey is riding. These warnings are termed as 'Caveats' for this study. The table below gives details of the codes used for the Caveats identified as being used by HT in his tipping pieces.
Note that some of Positive Reasons which add support to a Selection, detailed in section 2.2, can also be negatives for some horses. For example a distance change from 7 to 6 furlongs, or vice versa, might be seen as positive or negative depending upon the horse in question. Those items and codes are not repeated in the table below.
|HORSE Reasons :-|
|UF||Unproven Form. The horse has not shown any worthwhile form during it's career or a recent run which looked improved may have been mis-read in some way. Not just the opposite of 'Form Franked'. For example, HT might say that a 4TO horse in it's first handicap might actually be as "useless" as the first three outings have suggested. In another case HT might say that a horse returning from a long break or injury may simply not be capable of the same level of form anymore that it had previously shown.|
|OU||OR Up. A horse running off a higher Official Rating that it has not proved it can be competitive off. In theory could be seen as the opposite of the positive 'OD' item but 'OU' a clearer way of highlighting the caveat than negating a positive. Also, unlike a Distance Change for example, an increase in OR should not ever be seen as a 'positive' so 'OU' is an independent item from 'OD'.|
|BH||Behaviour. Highlights problems with the horse's general attitude which may prevent it from displaying it's full ability. For example, a horse that gives trouble loading into the stalls or makes slow starts would be noted by this item.|
|RACE Reasons :-|
|HI||Higher Class or In-Form Opposition. The horse is running in a better class race or against tougher opponents than in it's recent races.|
Back to Top of Page
On the Wednesday 'Form Factor' programme on 'At The Races' (ATR) HT will list a set of 'Eyecatchers' he has spotted in recent races. He will illustrate these by showing a video of the races and talking to the pictures. Part of HT's summary will cover what he had seen in the horse's run which he though was eyecatching. He will then go on to list any other positives or negatives about the horse in general. He may also give an indication of what sort of race set-up to look for when the horse next runs. The Categories listed here and the assignment of positives and caveats therefore depend on the notes taken of what HT said. Unlike the case with his 'Selections' where the text is produced by HT and published on the ATR website.
At the Top Level all of the Eyecatchers ought to fit under the 'SHPBTR' (Shape Better Than the Result) category. The classification here would then be identifying the useful sub-groups below that top level. For the most part that is the case but HT did list 6 x Eyecatchers on one program where the real angle was a switch to racing on Fibresand. With Southwell closed from mid-November HT felt many horses had missed their chance to show their true form through to the end of January. With Southwell due to re-open in early February HT took the chance to use the Eyecatcher slot to show videos of 6 horses who would be 'interesting' at Southwell. Videos were shown of the horses' recent Polytrack runs but mostly just to show they were still alive and not running too well. Not typical 'Eyecatcher' stuff. Therefore the 'SWEL' Top Level category has been used to keep those (non)-eyecatchers separate.
Long time readers of the B2yoR website will recognise that the majority of the eyecatcher groupings in the table have direct relation to items recorded in B2yoR Run Notes. 'BoPc' has long been used and the 'TvWL' or 'TvOk' formats used to note horses whose final position was at odds with the way they went through the earlier stages. The Run Notes also sometimes record when a horse was still in a prominent position late in the race but faded badly. For example the code '3rd-1f' would indicate a horse still 3rd at the Furlong Pole but finished much further back. The 'Travel Well but Fade' is a major part of HT's approach. Presumably there are only so many 'clues' that anyone can find by watching race re-runs and probably a good sign that there is a solid match between the B2yoR & HT output.
|BOPC||Best Of Pace. A horse who sets, or is involved in, too strong a pace in the early stages of a race but produces a better than average effort given that. Will usually do best of the prominent runners in terms of final placings. The winner and placed horses in the race will typically have been in the midfield or further back at halfway.|
|FNWL||Finished Well. A horse that produces an unusually strong finish in a race. Often the eyecatcher will pull clear of the other runners in the last one furlong.|
|GGLT/M||Got Going Late /Maiden. Usually denoting a horse that was held up in the midfield or rear of a race that was run at a slower pace. The horse will then struggle to make ground in the later race 'sprint' but will still be going forward, relative to the other runners, at the line. May indicate a horse that needs a longer trip or a stronger pace in the race. On occasions it denotes an inexperienced horse in a maiden race who realises what is required too late in the race and produces a solid late finish. In that case the GGLTM code will be used.|
|RANOK||Ran Ok. Used when it is not clear from the Video and the notes that the horse did anything notable in the race. The positive reasons added will often identify what may be the underlying cause of this horse being an eyecatcher. For example, a trainer switch and the horse has just 'run ok' but better than it had been for the previous trainer.|
|SWEL||Southwell Switch. Denotes the set of unusual 'Eyecatchers' which were put up just before the Southwell course re-opened for racing.|
|TVWL||Travel Well (& then Fade). Horses that are noted travelling well in early race but then fade to finish well back. Sub-groups include handicapped horses who need a shorter trip, faster going or need the run to regain full race fitness. It can also be used for inexperienced horses in maiden races.|
|UNLK||Unlucky in running. The classic eyecatcher group whereby a horse is stopped from producing it's full effort by being blocked by other runners. The horse making ground on a rail who gets cut off and the jockey has to stop riding, for instance. Intuitively, these ought to fail the test of being difficult to find in a normal Form Book. They will be widely known and may well be overbet on their next outing just because of being 'unlucky last time' without anyone trying to quantify how much they should be capable of. The unsubtle end of 'eyecatching'.|
Below the Top Level Categories listed in the table above for 'Eyecatchers' they can also then be supported by other other positive items or caveats can be noted. The same groupings are used below the Top Level for Eyecatchers as for the Selections so are not explicitly repeated in this section.
Back to Top of Page
The Introduction covered the background to the study and why Hugh Taylor's (HT) selections were used. Section 2 detailed how the initial sets of Angles and supporting reasons have been categorised. The purpose of approaching the data in this way is explained by considering the following summary drawn from ideas in the 'Thinking, Fast and Slow' (TFS) book.
Many studies have shown that a simple alogrithm to combine scores from a small number of factors will nearly always perform better than expert judgement, especially in medium to low validity environments. Amongst the reasons for this are that experts are overconfident and overvalue their own judgement when set against other sources of information. They also are prone to many biases which are part of the human decision making process and, further, tend to be inconsistent in applying their knowledge. Doctors coming to different decisions when faced with the same data but in different circumstances, for example.
The first aim of this study therefore is to identify the meaningful factors that HT is using and to be able to cut out those which are being used as a crutch to pad out the selection text. When HT did a 'Form Factor' special with James Willoughby the latter used the following description - "..when the reasoning given for the selection is for 'support' [of a hopeful pundit] rather than illumination of the audience". There are clearly too many top level Angles and items of positive & negative reasoning in the categories listed in Section 2. But, this is an interim stage of the study so the first aim is to capture all of the possible inputs. They can then be cut out explicitly based on an algortihmic decision. For example, the number of uses and the strike rate of selections which include a specific item.
It would be possible to use a statistical tool called 'Multiple Regression' to check all of the items in Section 2 along with their uses in combination. But, the low sample sizes make this an issue and the TFS book offers support for a simpler approach anyway. Consider these quote from the book :-
"..One can do just as well by selecting a set of scores that have some validity for predicting the outcome and adjusting the values to make them comparable (by using standard scores or ranks). A formula that combines these predictors with equal weights is likely to be just as accurate in predicting new cases as the multiple regression formula..."
"..it is possible to develop useful algorithms without any prior statistical research. Simple equal weight formulas based on existing statistics or on common sense are often very good predictors of significant outcomes."
"The important conclusion from this research is that an algorithm that is constructed on the back of an envelope is often good enough to compete with an optimally weighted formula, [i.e, after multiple regression applied] and certainly good enough to outdo expert judgement."
Having read those quotes you could probably put together your own summary of where this study should go. The B2yoR plan is to do more than turn an envelope over but not get involved in multiple regression. Use a simple Strike Rate, Return on Investment (ROI) and common sense method to produce a final simple algorithm. What this should be capable of doing is ranking HT's selections into a Good-Average-Bad or Confident-Middling-Hopeful range. Then avoiding the selections at the lower end and raising stakes for the Good/Confident end. At present HT assigns nearly every one of selections to be single points win. This desparately needs fixing because anyone reading the selection text regularly will know that HT displays a range of confidence level.
For example, if you compare the text and angles he produced for Master Of The Sea's 8/1 win compared to the text for Shearian's 2nd in a 3 runner race. The former read as being really confident and the positive reasons held together as a sensible 'story'. In the text for Shearian it was hard to see a positive reason at all and it all sounded 'hopeful' stuff that a horse would have wised up in a few days. Even if Shearian had managed to win then, in the long term, you would not feel the kind of 'supporting' arguments were of equivalent worth of those for Master Of The Sea.
It would also be useful to know what statistics, data sources and proven research HT uses for each of the Top Level Angles and the items below. The importance of each factor could then be partly assessed by HT's preferences and methods rather than just back-fitting the strike rates by working through the selections. Further, it would be useful to understand whether each factor he highlights as a positive or negative in each can assumed to be at the 'Average' level if it is not mentioned. Does HT have a list of factors for each possible selection that he goes through, for example. Or is he picking out factors to support a pre-existing story? No criticism here since he is proven profitable, but instead to improve on what he does and show others what is useful. How to be 'more like Hugh'.
An idealised outcome would be to have HT filling in a Database Form with the all of the small set of factors on it and setting them to -1, 0 & +1, or 0, 1 & 2, or whatever the factor score ranges are. This would be fed into the algortihm which would pop out how good a selection this was and what the staking level should be. To take it back further, to an extreme, if we could encapsulate precisely what HT is identifying correctly in coming to a selection we could apply that approach to every race and find our own qualifiers and bets. HT in AI.
If B2yoR was asked what he thought were the main methods and Angles HT used having done two months study what would the response be? Top Spot would go to Video Review. As well as being the main positive factor for a lot of selections it underlies a lot of positives items as well being used for the 'Eyecatchers' he identifies. Even in other Top level Angles like BTOR, & especially COFOR, the Video work often means that HT feels he has seen 'something different' about a horse or race as an Angle to take forward. The Form Book doesn't get much of a mention and reciting lengths beaten & pounds carried rarely used. Presumably, like many successful tipsters, he has got beyond that simplistic approach. He is more interested in 'Class Levels' of horses and how this can be inferred from Videos, Time-figures and so on. He then works this view against the set background of the Official Ratings to look for advantageous set-ups. Form franking is more interesting to him and trying to mine seams of strong form where runners from individual races compete above average in their subsequent outings.
Time figures and simple sectionals get regular mentions but B2yoR feels he could include this area more fully and drop some of the long list of lower level factor. After that you are into the long list of angles and stats that get irregular mentions and needs some weeding out. The 'Going' Angle for horses already proven on the surface, including the Southwell fibresand specialists, feels 'real' and would be added to Video work, Timefigures and Class Levels/OR interaction as a definite angle to keep. But, note that this does not include Going angles which rely on other statistics such as pedigrees and 'Sires for Going'. Other than that a possible to keep with 'Trainer Switches' but with it needing to be analysed more thoroughly. Hard to feel positive about the 'Jockey' angles that have occasionally popped up.
So, for argument's sake, a back-of-an-envolope summary for where the final, simple algortihm, System might go try this outline. Cut down the Top Level Angles to those based on the areas given a positive mention in the last two paragraphn and require a 2, 4 or 6 score for each based on the confidence in the top level factor text. Group a small set of lower level factors into 4 sets as 'Other TOp Level', 'Horse', 'Race' & 'Trainer/Jockey/Other. In each set the scoring might be 1 point for a positive reason and -1 for a caveat. Each of the four sets is totalled and the final figure added to the top level Angle number. B2yoR can already hear 'Black Art' tipsters and handicappers either laughing loudly or being audibly appaled by such simple nonsense. Probably in the same way that the Paediatric Specialists were when the decision the give medical intervention to help a newborn baby was, successfully reduced to rating 5 variables as 0, 1 or 2 and ensured a more consistent response by all specialists. Try reading 'The Checklist Manifesto'.
Not enough data has been collected yet to start settling on the final variables and algorithm but let us briefly consider a few Top Level Angles. BTOR (Better Than OR) has been assigned for 15 selections of which one was a non-runner. The other 14 have produced 4 winners (29%) and an ROI of +0.54. That relies on two wins for the same horse but still a solid return so far and worth persevering with. SOFOR (Scope Off OR) unlike BTOR is used for unproven horses who probably have not placed recently and are not 'in form' as the BTOR horses have usually proven. This has been assigned to 9 selections of which none have won and none have even made the first three. The average recommended price is longer than BTOR group as you would expect for less proven horses. But, the selection text for these can feel 'weak' when being read. A suspect area whereby more evidence is needed to see whether it is worthwhile waiting for the odd long SP winner to turn-up to 'save' this group from regular losses.
The COFOR (Competitive Off OR) feels like a suspect Top Level Angle since horses competing 'OK' in recent races doesn't feel like much of an Angle. It was initially used where reading the selection text made it difficult to assign a single overarching Angle. HT would typically start by saying a horse was running well then give some excuses for it's recent places or explain why today's set-up should see a better performance. Looking at the tables the Top Level Angle SHPBTR gets listed for the majority of COFORs so perhaps breaking up COFOR is the answer. But, the 20 selection assigned to that Angle have produced 5 winners (25% SR) and a profit with a +0.15 ROI. So clearly worth looking further at which low level factors appear in the five winners.
If you believe that Video Review is the core of what HT does then the result so far for the selections assigned as SHPBTR (Shaped Better than the Final Result) would not put you off looking further into this area. Perhaps reassigning some of the COFORs and seeing how that changed things. The 5 x SHPBTRs have produced 3 winners and a solid profit, plus two horses who ran well at double figure SPs with one of those winning next time out.
Looking at the lower level factors for patterns is more time consuming and the highlight point would be how few times some of them are used. MK (Market, well backed previously) has only been used 3 times as an example and gave 2 short priced losers and a 4/1 winner. Inconclusive but the over-riding feeling is that something used so sparingly feels like a 'crutch' to pad the text rather than something to take too seriously. B2yoR would perhaps need to declare a bias here.
At the end of this section it is worth highlighting some of the questions for HT which occur out of the work. B2yoR would encourage others to pose them to HT as part of those he receives as input to his 'Form Factor' programme. They are all very interesting in their own right. But it would also save HT having to answer the same 3 questions every other week, which are - "I can't get on at the recommended price, what are you going to do about it?"; "How far to should I chase the price down down for a selection before it becomes unprofitable?"; "How can I produce Aatishoo!, sorry, a tissue, for a race?".
A final point to end this section is to note that the questions above might be taken to suggest that HT might be slacking in some way. Worth reiterating that HT was chosen for this survey because he is profitable to follow and tries hard to educate is readers in his methods. The idea here is to be able to take an approach which is already well above the average level of tipping and punditry in Britain and see whether it can be improved still further. With many of the pundits that would not be a reasonable target because the starting point would be too low. There is not enought there to begin with to try to 'polish'.
One last thing, the answer to the question of how you could have done better, in ROI terms, than backing all of the selections at recommnded odds? Back every selection the next time it ran as well as, or instead of, on the day HT selected it for a race. Doing that would have produced 11 wins from 38 runs at a 29% Strike Rate and for an ROI of +0.76. A study in it's own right to follow that up. The 11 winners had all won or placed in the race HT selected them for which is mildly interesting. Perhaps a subsidiary System based on not much more than backing in-form horses where HT has seen something in them recently.
Back to Top of Page
On the 'Form Factor' programme on At The Races (ATR) Hugh Taylor (HT) puts up a number of 'Eyecatchers' each week. These he will illustrate by showing the horses running in the races which he thought they were eyecatching. HT will talk to the pictures shown and explain what he is seeing that makes him think the horse will be capable of running to a higher level at some point in the future. Other positive or negative factors, not directly related to the race video, will then be added. HT makes it very clear that the horses are not 'back this next time' recommendations. The eyecatchers' chances and suitability for making a selection in their next race would depend on the the circumstances of that race and the prices available.
HT would say that the main purpose of the 'Eyecatchers' is educational. Assisting the viewers to understand how he looks at a race in Video Review and what factors he looks for. At present, the results of following the Eyecatchers in their subsequent runs are not recorded nor put up on the ATR website. The record of HT's selections since 2009, at recommended prices, are on the website. Since this study is aimed at understanding HT's approach the decision was taken to record the Eyecatchers and also their results in their next two outings. The details of this work are given in table 5.3 & 5.4 below. The Eyecatchers were showing a loss on their next runs until an upturn in March. At the time of putting this report together backing the Eyecatchers on their next run would have produced a small profit. Backing them on their second runs after being highlighted would have made a smallish loss.
As with the Selections the decision was taken to try to Classify the Selections into a small set 'Types'. Again, the aim was to see whether some factors are more important, and profitable, than others. Usually, by definition, the Eyecatchers fit within the SHPBTR (Shape Better than Final Result) top level Angle. The task was therefore to assign the set of items below this Angle which characterised the Eyecatchers. These second level items are detailed in Section 2.4 and it is worth noting that many of these match those used by B2yoR. There are presumably only so many sets of clues can be picked up from Video Review.
It is worth highlighting that there is no 'Text' from HT to analyse. Instead handwritten notes have to be taken to summarise the commentary he adds to the demonstration videos. This could be a less accurate and checkable method than having HT authored text to work on.
Back to Top of Page
|Price||+ / -||Tot +/-||ROI||+ / -||Tot +/-||ROI||Course||Date|
|Master Of The Sea||DaviesNAT||DaviesST||WINC||17-Jan||1||2.75||4.5||2.75||2.75||4.50||4.50||NWBY||09-Feb||1||8|
|Bertie Blu Boy||WilliamsonL||EavesT||WTON||18-Jan||99||0||2.5||WTON||29-Jan||9||9|
|Rapid Heat Lad||HollinsheadR||TudhopeD||WTON||21-Jan||1||6.5||11||6.50||13.16||11.00||21.50||WTON||08-Feb||11||11|
|On The Hoof||EasterbyMW||HopkinsM7||KTNA||30-Jan||3||2.25||3.5||-1.00||15.66||-1.00||27.50||WTON||08-Feb||5||7|
|She's Some Girl||FaheyRA||MucHughB||LNGA||03-Feb||5||8||12||-1.00||15.66||-1.00||38.00|
|Master Of The Sea||DaviesNAT||DaviesST||NWBY||09-Feb||1||8||8||8.00||19.41||8.00||46.50|
|The Betchworth Kid||BellMLW||McCarthyJA||NWBY||09-Feb||11||50||100||-1.00||18.41||-1.00||45.50||NWBY||17-Mar||9||16|
|Day Of Destiny||GivenJG||SullivanJP||SWEL||12-Feb||2||2||3.5||-1.00||19.41||-1.00||47.50|
|Lady Of Burgundy||UsherMDI||NewnesL5||KTNA||14-Feb||7||20||22||-1.00||16.41||-1.00||44.50||KTNA||20-Feb||9||9|
|Comedy House *||MadgwickM||TateR7||LNGA||15-Feb||4||2.5||3.3||-1.00||14.41||-1.00||42.50|
|World Freight Girl||IvoryDK||QuinnJ||KTNA||28-Feb||9||14||20||-1.00||11.41||-1.00||47.50|
|Come On Blue Chip *||D'ArcyPW||MurphyMJM5||WTON||09-Mar||10||8||12||-2.00||10.91||-2.00||48.00|
|Trigger The Light||KingA||HutchinsonW||WWCK||10-Mar||4||5||5||-1.00||8.91||-1.00||46.00|
|Bach On Tow||GardnerS||FahyD||TAUN||11-Mar||4||7||12||-1.00||7.91||0.11||-1.00||45.00||0.62|
Back to Top of Page
|Master Of The Sea||WINC||17-Jan||1||2.75||4.5||BTOR||GG|
|Bertie Blu Boy||WTON||18-Jan||99||0||2.5||BTOR|
|On The Hoof||KTNA||30-Jan||3||2.25||3.5||BTOR||SH|
|Master Of The Sea||NWBY||09-Feb||1||8||8||BTOR||GG|
|Lady Of Burgundy||KTNA||14-Feb||7||20||22||COFOR||SH|
|World Freight Girl||KTNA||28-Feb||9||14||20||COFOR||SH|
|Trigger The Light||WWCK||10-Mar||4||5||5||CRSFM|
|Rapid Heat Lad||WTON||21-Jan||1||6.5||11||GNG|
|The Betchworth Kid||NWBY||09-Feb||11||50||100||GNG|
|Come On Blue Chip||WTON||09-Mar||10||8||12||GNG|
|She's Some Girl||LNGA||03-Feb||5||8||12||SOFOR|
|Bach On Tow||TAUN||11-Mar||4||7||12||SOFOR||SH|
|Day Of Destiny||SWEL||12-Feb||2||2||3.5||SWELST||CO|
Back to Top of Page
|+ / -||Tot +/-||ROI||Course||Date|
|Bertie Blu Boy||WilliamsonL||BakerG||WTON||29-Jan||9||9||-1.00||0.00||WTON||09-Feb||6||7|
|No Mean Trick||MidgleyPT||FentonM||WTON||04-Feb||5||3.5||-1.00||-7.00|
|Henry The Aviator||JohnstonM||FanningJ||LNGA||15-Feb||2||2||-1.00||-4.00|
|Lean On Pete||PearsO||WTON||01-Mar||3||2.5||-1.00||-12.00|
|EYE Date||= No Run Since|
|My Son Max||O'GormanPJ||27-Feb|
Back to Top of Page
|Bertie Blu Boy||16-Jan||WTON||29-Jan||9||9||BOPC|
|Henry The Aviator||06-Feb||LNGA||15-Feb||2||2||FNWL||BT|