Making Turf Variants/Figures

[highnote]

If you have thoughts on making turf variants to be used for making speed figures I'd appreciate your thoughts.

I put together a set of Turf par times for Saratoga and have started to try to make turf variants there. It is hard to know how good the figures they produce are going to be because most days there are only 2 races run over each of the inner and outer courses. I expect that the variants will vary widely.

[cj]

If you have thoughts on making turf variants to be used for making speed figures I'd appreciate your thoughts.

I put together a set of Turf par times for Saratoga and have started to try to make turf variants there. It is hard to know how good the figures they produce are going to be because most days there are only 2 races run over each of the inner and outer courses. I expect that the variants will vary widely.

I've said this before, might want to do a search for specifics and I wrote about it for HANA. If I were starting over I'd use a significantly different value on turf for time than I do dirt. Not only is a beaten length much more significant on turf, so are smaller differences in time among races.

[highnote]

I've said this before, might want to do a search for specifics and I wrote about it for HANA. If I were starting over I'd use a significantly different value on turf for time than I do dirt. Not only is a beaten length much more significant on turf, so are smaller differences in time among races.

Thanks. I'll check the hana archives and search on pa.

David Edelmann had some interesting thoughts on making turf figures, but he didn't talk about making turf variants.

[cj]

Thanks. I'll check the hana archives and search on pa.

David Edelmann had some interesting thoughts on making turf figures, but he didn't talk about making turf variants.


The key for turf in my opinion is having an accurate read on several horses on the card. With moving rails, different courses, and sometimes super slow paces, it is a challenge but it can be done.

I do think pace has to be a part of variant making in some way. Final times will leave you scratching your head without it.

[TonyMLake]

If you have thoughts on making turf variants to be used for making speed figures I'd appreciate your thoughts.

I put together a set of Turf par times for Saratoga and have started to try to make turf variants there. It is hard to know how good the figures they produce are going to be because most days there are only 2 races run over each of the inner and outer courses. I expect that the variants will vary widely.


Why not start with the published Tomlinson/mudder numbers? They're not horrific and they're probably the most available set of numbers.

Somehow combining them with your other speed numbers sounds like a decent start to me.

[cj]

Why not start with the published Tomlinson/mudder numbers? They're not horrific and they're probably the most available set of numbers.

Somehow combining them with your other speed numbers sounds like a decent start to me.

Aren't Tomlinson numbers for determining how a horse would handle a new surface and/or distance? How are they going to help assess track speed?

[highnote]

The key for turf in my opinion is having an accurate read on several horses on the card. With moving rails, different courses, and sometimes super slow paces, it is a challenge but it can be done.

Nick Mordin made projected U.S. turf variants by using every horse that raced on a given turf course on a given day if the horse finished relatively close to the winner.

I do think pace has to be a part of variant making in some way. Final times will leave you scratching your head without it.

I agree. James Quinn had a some good ideas and I may try to incorporate some of his techniques. Although, Beyer had some good criticism of some of the flaws in his method.

Cary Fotias had the best turf figures I've used, but his method was proprietary.

So at this point I'm making my turf variants the same as my dirt variants. I haven't made turf figures, yet, but I imagine that the key to using them will be analyzing pace.

Also, I may use Edelmann's ideas for quantifying beaten lengths.

Currently, I'm calculating final times of runner-ups using Charles Carroll's method of using 8 feet for a beaten length and then measuring the Time Per Length, (i.e., the time it takes a horse to run one length.) Then I add up the Time Per Length for the runner-ups' beaten lengths to calculate their final times.

I find that my dirt figures are correlated with Beyer, but different enough to add pari-mutuel value. I also changed the scale of my figs to be analogous with Beyer so that I can more easily compare them.

[TonyMLake]

I meant to say the mudder/turfer/Tomlinson ratings but here's a decent read:
http://www1.drf.com/misc/tomlinson.html

[highnote]

I've said this before, might want to do a search for specifics and I wrote about it for HANA. If I were starting over I'd use a significantly different value on turf for time than I do dirt. Not only is a beaten length much more significant on turf, so are smaller differences in time among races.


In 5 1/2 furlong sprints at Saratoga the beaten lengths don't seem much more significant than on dirt -- if at all. But as the races get longer I'm sure the sprint to the wire is faster on turf than on dirt. So it makes sense that the beaten lengths get more significant.

[cj]

In 5 1/2 furlong sprints at Saratoga the beaten lengths don't seem much more significant than on dirt -- if at all. But as the races get longer I'm sure the sprint to the wire is faster on turf than on dirt. So it makes sense that the beaten lengths get more significant.

Yes, covered most of this (I think) here and on HANA. Sprints don't vary as much, but there is still a difference between turf and dirt IMO.

Here is the link to the HANA issue:

http://horseplayersassociation.org/jan15issue.pdf

[classhandicapper]

One of the appeals of using "closing time" to me is that any variance in the speed of the turf course won't have as big an impact on 1 or 2 furlongs as it will on an entire race.

If the track 8/5ths fast at 8F, that's approximately 2/5ths fast at 2F.

If on another day it's 4/5ths fast, that means a lot when you are comparing final times to the other day (4/5ths), but it's only approximately a 1/5th difference in the final 1/4.

Naturally, it's even less if you are looking at just the last 1/8th.

Looking at just the straightaway times also mitigates the issues of rail settings. So even if your track variant is only ballpark, you should be able to make some excellent closing time figures.

[JohnGalt1]

Since most tracks card few turf races, turf variants are unreliable on a daily basis so I don't use a variant at all.


EXCEPT, I use variants for Gulfstream races, because they move the rails to preserve what turf course :) they have. The average DRF variant for 2014 routes was 19, so I deduct 1 tick per point over. A high variant could mean the rails were out causing a slower final time.

A 1:37.2 with a 30 variant I rate as 1:35.1. Any variant lower than 19 I use the actual time of the race.

Another exception--- When a horse only has, or his only good races, are on good or yielding turf, I factor in the variant like Gulfstream.

Some examples of 2014 route turf variants, CD 11, Aqu 23, AP 15, AP about 20.

This works for me. I hope it helps.

But I always prefer rating of firm and NOT using any variant.

[AndyC]

I have found whole race or final time speed figures on turf to be entirely useless.