some image

Having fun, writing about the stuff I like

Mind Games in Horse Racing

Oscar Foulkes October 25, 2015 Horse Racing No comments
Arezzo-wins-Listed-Settlers-TrophyHorse racing is pitched as “the intelligent bet” because there are so many dimensions that can be critically examined. I’m going to illustrate this by means of a race that was run yesterday at Kenilworth.

This does come with an advance apology. Much of what follows may be somewhat technical (or, ‘higher grade’, as some may say), but I’ll try to keep it as accessible as possible. It’s not for nothing that I’ve used “mind games” in the title of this post!

On Monday last week, the field was published for the Woolavington Handicap, to be run over 2400m at Kenilworth:

Zante 4yo filly 61kg
Deputy Ryder 4yo filly 58kg
Desert Swirl 5yo colt 58kg
Friendly Tibbs 5yo gelding 57.5kg
Ovidio 5yo gelding 55kg
Arezzo 4yo colt 52.5kg
Parachute Man 4yo gelding 52kg
Forte de Ouro 4yo colt 52kg
A Time To Kill 5yo gelding 52kg
Private Doowaley 6yo colt 52kg

A handicap is a race in which the horses carry weights – assigned by the handicapper – that will theoretically result in a dead heat. In other words, the weight is used to equalise the relative merit of the runners. However, the South African merit rating system, as used by the handicappers, is something of a blunt instrument, in that horses often end up with ratings so high that they are rendered uncompetitive at the weights they then have to carry.

On this basis, I immediately wrote off Zante and Deputy Ryder, both of which are good fillies, but nowhere near good enough to carry 61kg and 58kg against a decent field of colts.

With the exception of Ovidio, many of these horses met a month earlier, in the Settlers Trophy. I’ll go into the rationale shortly, but immediately upon seeing the Woolavington field, I sent Jono Snaith this message: “I think Ovidio is a big runner if he gets the extra 400m (assuming his 4.5 length beating of A Time to Kill is a fair reflection).” Like Arezzo, Ovidio is trained by Justin Snaith. My interest in all this comes about as a result of my tiny shareholding in Arezzo.

Jono was confident that he would get the distance. My doubts lay in the first two generations of Ovidio’s pedigree, which is loaded with horses who were best up to about a mile. However, his grandam was by Kenmare, a solid 2400m influence. Further stamina can be found in his third generation, but it needed to leapfrog the more immediate ancestors.

As you can see, all of this analysis is multi-dimensional, bringing together not just handicapping, but also genealogy.

Before we get to the actual races that make up this analysis, a bit of background on handicapping. Weight has a drag effect, which is measurable – Newton’s Second Law in action. The conversion of weight to lengths beaten over different distances is not widely publicised. The work of Phil Bull, the founder of Timeform, is key to this, but given that he did his work without the benefit of computers, the conversion rates are in user-friendly round numbers, rather than pesky decimals.

Establishing the data for ‘lengths beaten’ is a minefield of its own, because of the different ways of calculating it. For the sake of this explanation, let’s just accept the data as recorded officially.

Timeform do not publish their pounds-per-length conversion. This is what the British Horseracing Authority (BHA) uses:
5 furlongs: 3lbs per length
6 furlongs: 2.5lbs per length
7f-8f: 2lbs per length
9-10f: 1.75lbs per length
11-13f: 1.5lbs per length
14f: 1.25lbs per length
15f+: 1lb per length
(one furlong is approximately 200m)

Locally, Charles Faull uses these conversions:
1000-1400m: 3lbs per length
1500-1900m: 2lbs per length
2000-2300m: 1.5lbs per length
2400m+: 1lb per length

If we use 2lbs per length at 1600m as the mid-point, then a mathematical suggestion is:
1000m: 3,20lbs
1200m: 2,67lbs
1400m: 2,29lbs
1600m: 2,00lbs
1800m: 1,78lbs
2000m: 1,60lbs
2200m: 1,45lbs
2400m: 1,33lbs

For an apparently scientific view on the same stuff (yielding a conversion of 2.26 pounds per length for a mile run at a pace that results in a time of 98 seconds), click here.

For the rest of the exercise, I’m going to use 1.4lbs per length for 2400m, and 1.75lbs per length for 1800m, which is a kind-of consensus between BHA and an Excel assisted extrapolation from 2lbs per length over 1600m.

Still with me? OK, let’s get into some form study, by looking at the last meeting beween Ovidio and A Time To Kill:

Kenilworth 1800m on 29 August 2015
Pos Horse Wgt Beaten Relative Merit
1st Ovidio 56.5kg 0 113
4th A Time to Kill 54kg 4.5 lengths 100

To calculate relative merit, multiply the distance beaten by the pounds-per-length conversion, and then add/subtract the difference in weights (expressed in lbs). Relative merit can be equated to ‘merit rating’, but I haven’t used that terminology here because I’m working with a small group of horses, not the entire racing population.

The form of this race suggests that Ovidio is 13lbs better than A Time To Kill. My message to Jono Snaith was prompted by the difference in weights in the Woolavington, in which Ovidio carried just 3kg (6.6lbs) more than A Time To Kill, making him the better handicapped of the two.

At the time, I thought that A Time To Kill had improved over the additional distance of the Settlers Trophy (he goes into this carrying the notional 100 rating from the race against Ovidio):

Durbanville 2400m on 26 September
Pos Horse Wgt Beaten Relative Merit
1st Arezzo 52kg 0 102
2nd A Time To Kill 52kg 1.25 100
3rd Friendly Tibbs 58kg 1.75 113
4th Forte de Ouro 52kg 2.15 99
5th Parachute Man 52.5kg 3.65 98
6th Paddy O’Reilly 64.5kg 4.4 123
7th Desert Swirl 58.5kg 6.4 109

To keep things simpler, I have rounded the decimals when calculating relative merit.

Right, so what did they do when they all raced against each other yesterday (I’ve carried Friendly Tibbs’ relative merit through, using him as the link horse here):

Kenilworth 2400m on 24 October
Pos Horse Wgt Beaten Relative Merit
1st Ovidio 55kg 0 110
2nd Desert Swirl 58kg 1.25 115
3rd Friendly Tibbs 57.5kg 2 113
4th Arezzo 52.5kg 2.2 101
5th Parachute Man 52kg 4.45 97
6th A Time To Kill 52kg 5.7 95
7th Zante 61kg 6.95 113

The suggestion is that A Time To Kill ran a bit below best yesterday, as well as Ovidio (marginally), but the result more-or-less stacks up with what the horses had done against each other previously, adjusted for weight differences. Perhaps Ovidio isn’t quite as good over 2400m as he is over 1800m. Time will tell.

Arezzo’s jockey thought that he was disadvantaged by a slowish pace down the back straight, as well as losing a shoe, but he finished more-or-less where expected, relative to Friendly Tibbs.

Anyone looking at the result of this race without taking the weights into account would deduce that Zante did not do anything significant. She was, after all, unplaced, finishing nearly seven lengths behind the winner. Under handicap conditions she’s likely to struggle until her official merit rating is reduced. However, the J&B Reserve Stayers, on J&B Met day, is a conditions race in which she’ll carry 55.5kg. Under that weight she’ll be much more competitive.

I find that stayers’ races (e.g. 2400m+) are fairly reliable from a handicapping perspective, because the horses are generally mature, and are therefore not prone to the rapid improvement one sees in young horses.

The analysis I’ve laid out above may look complex, but the point is that analysis/logic is possible. And, even if you aren’t gambling, it adds an entirely different dimension to a day at the races.

I’ve left weight-for-age out of the above analysis, which takes into account the improvement a horse makes as it matures, expressed as a difference in weights carried between horses of different ages. This varies according to distance run and time of year. I’ll do a separate post about this soon.

Add your comment