I have been asked to explain the thinking behind a selection I gave on the Mathematician’s forum earlier in the week.
That selection (Fathom Five, which won) , my Lincoln selection (Smokey Oakey, discussed on an earlier post here, which placed) and yesterday’s selection (Bogside Theatre, which lost) all came from an attempt to move beyond the basic VDW method as described in considerable, though not complete, detail in the March 1981 article.
The essence of that method is to note the three most consistent from the first five or six in the forecast (five for non handicaps, six for handicaps) and the four in the field with the highest ability ratings (total win prize money divided by races won), and then consider whether “any of the three probables have good claims for support”, starting with the probable with “the highest ability rating”. (Quotes from March 1981 article.) And in an earlier letter VDW had further simplified things a bit by suggesting that those who found balancing the various factors difficult should limit themselves solely to those situations where the highest ability rated of the probables was the class horse in the race (the one with the highest ability rating in the field) – item 36 of “The Golden Years of Van der Wheil”.
We know that VDW regarded that method as an elementary one to which only a little was added in his booklet “Systematic Betting”, and that the totality of what he had set out in that booklet and all his contributions to the Sporting Chronicle Handicap Book represented little more than a fraction of what he knew (letter of February 1996 reproduced in Tony Peach’s “Systems in My Racing”, page 6).
VDW’s comments in the February 1996 letter would have come as little surprise to those who had studied his contributions from 1978 because from his very first discussion of a race (the 1978 Erin, won by Prominent King, reproduced as item 8 of “The Golden Years of Van der Wheil”) he showed that he stretched the boundaries of the method summarised above, eg:
1) in the Erin discussion VDW showed that he was happy to go beyond the three most consistent from the first five (the Erin was a non handicap) in the forecast, Prominent King being the fourth most consistent;
2) in other early letters he showed he was happy to go for horses not even among the first five/six of the forecast (initially hinted at in the letter reprinted as item 13 of “The Golden Years ….” With his reference to Strombolus, and confirmed in letters reprinted as items 15 and 30 in respect of Love from Verona and Son of Love);
3) in the March 1981 article he showed that he was flexible about taking a horse’s last three completed placings at face value, in respect of Gaye Chance;
4) in chapter 4 of “Systematic Betting” he intimated that he was prepared to accept as a bet, as distinct from merely being a selection, a horse not within the top four of the ability rating ranking (Park Express), though given the horse’s position in the supplementary, time-based ability rating VDW used with less exposed Flat horses, there was some ambiguity. But that ambiguity was removed in his discussion of the 1988 Mackeson, won by Pegwell Bay (reprinted in “Betting the VDW Way”).
So from day 1 it was clear that VDW did not stick rigidly to horses identified by the very simple numerical picture generated by marking off the three most consistent from the forecast and the four with the highest ability ratings, and much of the work for those who wish to understand the basic method fully is to try to establish VDW’s boundaries – eg which horses within the first five or six of the forecast but not with one of the three lowest consistency rates should be taken as consistent horses (eg Prominent King and, later, Righthand Man); which horses from much lower down in the forecast should be taken as consistent horses (eg Love from Verona and Son of Love), and how to try to reduce the consistent horses to three probables when, as often, there are more than three consistent horses in a race.
But VDW’s final contributions, in which he mentioned as bets horses such as Arthur’s Minstrel, hinted at more radical departures from the simple first numerical picture, and it is those hints I have been exploring over the Winter, encouraged by what I see as evidence that other VDWers, notably Lee from the old Gummy forum with his selection of Top Dirham (19/07/04), have also thought it worth pursuing this avenue.
My selections of Smokey Oakey, Fathom Five and Bogside Theatre are examples of work in hand to explore whether decent bets can be identified by a more sophisticated approach to consistent form than is needed for the 80+% bets found by a more rigid application of the basic method. An approach where the basic assumption is that it may be possible to deduce from a horse’s career record that he (or, with Bogside Theatre, she) can be viewed as consistent and being targeted at a given race, bearing in mind that VDW wrote that most of the time horses were not running to win the race they were contesting but to get them ready to win a future race. And once one believes one has understood what VDW meant by the phrase “the easy cracking of the handicap” one is inevitably drawn to explore what VDW described as “the deeper evaluation of form …” as trainers try to get their horses ready to win.
With the above in mind as the context, on to Fathom Five and Bogside Theatre (Smokey Oakey having been discussed in my post on the Lincoln);
Fathom Five ran in the 2.20 at Epsom on 21 April, and my pre race post on the Mathematician’s forum was simply:
“Fathom Five strikes me as interesting in this race.
His record suggests he is better in the first half of a season, all five wins having been in April, May or June. His best performance to date was back in June 2007 when he won a class 2 handicap at Newmarket off 92. It took him over a season to down to near that mark, and on his first run of the 2009 season he won the equivalent of today's race off 93. Six relatively undistinguished runs followed, and today he has his first run of the year off 91.”
Looking at the race in methodological detail:
The ability ratings for each runner generate the following ranking, highest to lowest:
Indian Trail
Holbeck Ghyll
Fathom Five
Fol Hollow
Hoh Hoh Hoh
Bertoliver
Ocean Blaze
Rocket Rob
Rievaulx World
The first six and equals in the Post’s forecast were: 11/4 Indian Trail, 9/2 Rocket Rob, 5/1 Fol Hollow, 6/1 Fathom Five, 7/1 Hoh Hoh Hoh, 10/1 Bertoliver, Ocean Blaze
From these seven horses, those with the three lowest consistency totals were:
08 Rocket Rob
14 Fol Hollow, Ocean Blaze
17 Bertoliver
so these were the automatic consistent horses.
We then have, from within the first six and equals in the forecast, two potential discretionary consistent horses, Fathom Five and Hoh Hoh Hoh (both 18), and from outside the first six and equals Holbeck Ghyll (13). On the rules VDW had for deciding which of such horses should be treated as consistent horses, Fathom Five is a qualifier and the other two are not, so we end up with five consistent horses.
With more than three, VDW tried to get down to just three probables by the device he illustrated but did not explain in his 1978 Erin evaluation. Applying that device to the five did not in fact help, as none was eliminated, so there were five probables. Taking them in ability ranking order and focusing on each horse’s last three runs:
Fathom Five: progressive profile with last race beaten 4.1l in a class 312. A probable with form.
Fol Hollow: a regressive profile in that, after a close 3rd in a class 106 on his penultimate run he ran poorly in a class 76, . Not a form horse and therefore not a probable with form.
Bertoliver: as per Fol Hollow (very poor performance lto after decent run in a higher class).
Ocean Blaze: as per Fol Hollow (weak performance lto after better run the time before in higher class).
Rocket Rob: progressive profile with good runs in races of successively higher class. A probable with form.
So just two probables with form, and Fathom Five higher on ability and with better form: the class/form horse.
But was Fathom Five a class/form horse strong enough to back? On the normal “winner in race” characteristics VDW required of class/form horses dropping in class (in FF’s case from a 312 to a 109), quite clearly no: he lacked what I term, having picked it up from Lee, ultra consistency. So normally that would be the end of the analysis: a clear class/form horse, well ahead of the only other probable with form on both ability and form (and with no obvious negatives on capability issues such as trip, going etc), but although the most likely winner from VDW’s presumptions, some way short of being an 80+% probable winner.
However, as he was high up the ability rankings (3rd in field, behind the top ranked but non form horse Indian Trail and Holbeck Ghyll, whose last run was over a year ago and a poor one, I explored his career record to see if analysis of that suggested he was being placed to win. And here the facts set out in my Mathematician forum post became clear, ie he seemed to be a first half of a season horse, all five wins having been in April, May or June. His best performance was in June 2007 when he won a class 2 handicap at Newmarket off 92. It took him over a season to down to near that mark, and on his first run of the 2009 season he won the equivalent of the 21 April race off 93. Six relatively undistinguished runs followed, as a result of which he contested the race on 21 April off 91. He had thus proven he could win a race of the class of that on 21 April off a slightly higher mark, he had shown by his last run in 2009 (beaten less than 5l in a class 312) that he seemed to have retained his ability, and because he had contested the race successfully in 2009 it was reasonable to assume that connections were trying to win it again this year.
In short, the career context suggested that one could reasonably excuse the lack of ultra consistency one normally requires, and regard what was otherwise a strong class/form horse as one worth backing.
Turning now to Bogside Theatre yesterday, the ability rating ranking was:
Mudawin
Track Record
Gordonsville
My Arch
Swiss Act
Bogside Theatre
Dazinski
Halla San
and the first six and equals in the Post forecast were:
10/3 Halla San, 7/2 Bogside Theatre, 5/1 Dazinski, 11/2 My Arch, 8/1 Gordonsville, 12/1 Swiss Act.
From these six those with the three lowest consistency totals were:
13 Halla San
14 Bogside Theatre, Dazinski, Swiss Act
15 Gordonsville
which were automatically consistent horses. In addition there were two potential discretionary horses to consider: Mudawin (6) and Track Record (16), but neither met VDW’s requirements.
With five consistent horses the Erin device is applied and, unlike with the Fathom Five race, in this case it eliminated Dazinski and Gordonsville (not the first time the device has eliminated the eventual winner), leaving three probables.
Taking the probables in ability order:
Swiss Act: a regressive profile with poorer runs in races of successively lower class. Not a probable with form.
Bogside Theatre: a progressive profile, albeit a poor one, with two poor runs and then a really good one last time out in the highest class of the last three. A probable with form.
Halla San: a regressive profile, with a so, so last time out run after a better one in higher class. Not a probable with form.
So Bogside Theatre emerged as the class/form horse, albeit with a poor last three race profile and only 6th overall in the ability rating ranking. Ordinarily a class/form horse a long way short of 80+% probable.
Probably I should have ended the analysis there, because she was too far down the ability rating ranking (unlike both Smokey Oakey and Fathom Five) but I explored her career and saw possibilities as set out in my post on the Mathematician’s forum which I copied to this blog yesterday.
I feel rather ambivalent about Bogside Theatre. I think it is clear from the betting on the race and the way she was ridden that the horse was there to win. Equally it is clear that she hadn’t got what it took to win, and after leading for about 85% of the race folded rather tamely. And apart from the progressive Dazinski, the horses who beat her were all higher in the ability ranking for the race.
There were two differences between Bogside Theatre on the one hand and Smokey Oakey and Fathom Five on the other. The latter two had won races of comparable class to the ones for which I was considering them and were in the top four of their respective ability rating rankings. BT hadn’t won a race of yesterday’s class and was only 6th in the ability rating rankings. Another negative I overlooked but may well have been relevant is that on her last run BT failed to win – albeit narrowly – a slightly more valuable but actually less strong race, so it was expecting a lot to see her winning the stronger race off 3lb higher in the official ratings. Normally one would expect a horse to improve for a run, but BT was fit for her first Flat run from hurdles races, so there was less basis for expecting improvement.
In sum, it is one thing to be able to be confident that a horse is in the race to win, but another to relax too far the basic VDW criteria, and with Bogside Theatre I probably relaxed them too far. For despite what is often written about VDW on various forums, his win prize money divided by number of races won ability measure is a good guide to the respective abilities of runners and one should not go below one of the top four without very careful thought.
