Synthetics

[smalltimer]

TGJB,
You can look at that a couple of different ways.
If you have a group of 10 horses, with good dirt speed, but with limited Poly ability, I\'ll take the proven Poly pedigree most of the time.  Why?  Because when the majority of the field can\'t stand up on Poly, and my debut runner is out of a high producing bloodlines, sometimes they don\'t have to be that good to take the purse.  If he catches them on a regular dirt surface, he\'s apt to get it handed to him, but put him/her on their in-bred preferred surface and it sure brings \'em together.
Just a comment.

[BitPlayer]

TGJB -

So, when you are making a set of figures for a race like the Blue Grass (in which many runners are moving from dirt to synthetic) or the Arkansas Derby (in which many runners are moving from synthetic to dirt), do you use all the horses, or rely primarily on those with experience over the surface?

[TGJB]

There is no set answer on that-- I look at the race, the surrounding races, and do what makes the most sense overall. I\'ve attached the BG, which I got about the same as Andy (don\'t know about Len).

[fkach]

TGJB,

When you eventually get a large enough sample of horses moving back and forth between synth and dirt, it might make sense to chart the figures by the quality of horses.

1. Horses that typically run between -5 and 0
2. Horses that typically run between 1 and 5
3. Horses that typically run between 6 and 10
4. Horses that typically run between 11 and 16

etc....

In addition to trying to understand whether the figures are accurate on average, it might be interesting to see if there is any flattening at the extremes relative to dirt.

Thanks.

PS: I would be willing to pay for some specific studies of your database - assuming the price was reasonable and info I need is available. You might want to consider that as another source of income because I\'d be willing to bet you\'d get other customers.

[BitPlayer]

TGJB -

I notice that there seem to be minor inconsistencies between your ground loss info for the Blue Grass and the Trakus data posted on the Keeneland site.  For example, the Trakus data indicates that Monba ran more than 2 lengths further than Cowboy Cal.  Since he also beat him by a neck, I\'d expect the difference in their figs to be at least one TG point and yet I see that the difference is only 0.75 points.

I assume that you would have to pay Trakus something for the right to incorporate their data in your figs.  What has been your experience with respect to the reliability of the Trakus data in the places where it is available?

I apologize in advance if this has already come up and I don\'t remember it.

[TGJB]

Trakus has approached us, and if they ever get going at a lot of tracks I\'ll take a close look at their data. No real opinion at this point.

[miff]

Bit,

About two years ago a few of us looked at TG ground loss(checking replays) for about 100 races in NY.Honestly did not find one instance where we saw a major discrepancy from the TG track guy.For a non computer type trakus set up, TG seems to have ground nailed pretty good in NY.Of course, a good computer program trumps the naked eye in any endeavor.

Now I am not a plant, but the same exercise showed \"fig\" differences for the Rags. Since they do not show ground on the product, it was tough to tell if we were dealing with track speed differences or ground loss differences when comparing between the TG and Rag figs.

Mike

[TGJB]

Miff-- the way to tell would be if there were discrepencies within a race.

[miff]

JB,

The differnces we found were not in beaten lengths as on both sets the relationship was close. The figs were different scale to scale which means that either variant or ground was not agreed.

Mike

[TGJB]

If the relationships within the races are close, the difference has to be in how we did the race as a whole, not ground loss.

[miff]

Makes sense, it\'s track speed then.

Mike

[BitPlayer]

TGJB –

Despite your protestations about learning nothing, you have (or have access to) more knowledge than your customers.  Even if you can only guess which horses will jump up when moving from synthetic to dirt, one issue on which you might shed some light (perhaps in the context of your Derby/Oaks seminar) is how big the jump-ups of synthetic-to-dirt horses have been.

Without having seen the figs for the SA Derby, I\'m guessing that Colonel John still needs a big forward move (say, just for example, 4 points) to justify his status as probable second favorite.  It would be interesting to know what percentage of the synthetic-to-dirt jump-ups have been 4 points or larger.

In general, I\'m surprised you haven\'t done more with quantifying the sizes of new tops.  Tops seem like a huge part of TG handicapping.  Even if ThoroPattern suggests that a new top could be forthcoming, the handicapper needs to guess at the probable size of the top in order to know whether it would make the horse a contender.  From reading the ROTW, it\'s obvious that you have developed a feel for what size jumps are just too big to expect.  For less experienced users, it would be useful to have some statistics (perhaps grouped by age, number of starts, or existing performance level) about what percentage of new tops are more than 4 points, more than 3 points, etc.

On the flip side, for horses who have just run new tops, it would be interesting to sort those horses by size of the jump-up, and ask how they perform in their next outing.  What percentage pair the new top, run an off race, etc.?

[TGJB]

On the narrower question, the sample size is way to small to do a study, but Heatseeker/Zenyatta and a couple of others going cushion/dirt have jumped about 3-4 points. It\'s more of an issue with the older horses, since spring 3yos jump forward pretty regularly.

We do studies something like the ones you mention for the Derby seminar every year, and will again this year.

Everything else aside (like our small staff and one programmer having to spend time on several other projects in the works that will actually produce revenue), the number of variables involved in the type of studies you mention (age, number of races, surface, trainer, spacing, etc.) is very high. We did what we could to give the inexperienced user something to work with.

[BitPlayer]

TGJB –

Thanks for the info on synthetic-to-dirt jump-ups.

Regarding my other thoughts, I\'m sympathetic to your need to run a business.  I\'ll share what I have regarding jump-up sizes in hopes that some others find it informative.  I took a look at all the tops run on or before the first Saturday in May by 3yos who ran in a Triple Crown race in 2004 through 2007.  I chose that sample because that\'s what I had access to through the TG archive.  It\'s obviously biased because tops by horses who fell by the wayside during the spring were excluded.  For a race to qualify as a top, the horse had to have run in at least two prior races.

Including all tops (n = 130), the jump-up sizes were:

1.25 to 2 – 32.3%
2.25 to 3 – 32.3%
3.25 to 4 – 20.0%
4.25 and up – 15.4%

Surprisingly to me, including only horses who had run at least 5 prior races (n = 69) did not seem to make much of a difference.  The jump-up sizes were:

1.25 to 2 – 30.4%
2.25 to 3 – 34.8%
3.25 to 4 – 24.6%
4.25 and up – 10.2%

Including only horses with a prior top of 5 or better (n = 57) did reduce the percentage of big jump-ups.  The jump-up sizes were:

1.25 to 2 – 42.1%
2.25 to 3 – 31.6%
3.25 to 4 – 17.5%
4.25 and up – 8.8%

[TGJB]

Jimbo-- here\'s a post of mine from a few days ago (that was completely misunderstood by one person who thought I was talking about handicapping synthetic tracks, which I was not). It touches on your question only in that I don\'t know the answers at this point. The only thing I would add is that there have now also been a couple coming from cushion that jumped up on poly, although there is also the Pletcher-at-Keeneland issue to deal with in that regard.

This thing is a major pain in the butt. The good news is that with the keeneland handle drop and the articles that have been written recently, Santa Anita may go back to dirt, and others may gradually follow.

I don\'t know the answer to the compaction question, but I do know that the figure relationships within races hold up on all surfaces.

On your Pyro comment-- close to right about him, wrong about the seminar overall.


TGJB Wrote:
-------------------------------------------------------
> So I\'ve done a bunch of figures over the last few
> weeks for horses going from synthetics to dirt,
> and here\'s what I know-- nothing.
>
> The horses have mostly been coming from California
> (that and Turfway are the only ones running over
> synth over the winter). Several of the ones
> switching to dirt have jumped way up. Some have
> not.
>
> The ones going from Cal synth or turf to Kee synth
> or turf have run right to their figures-- if there
> has been a jump-up I\'ve missed it. That\'s one
> reason I don\'t think it\'s the California figures
> in general-- the other is the horses that have
> gone the other way over the last few years, some
> of which I\'ve been responsible for (Even The
> Score, Super Frolic, Student Couincil, Pinata)
> have run to their numbers in California. Also,
> lots of Calder horses have run much faster once
> they got to California.
>
> All of which means that what we have on our hands
> is a guessing game, as Beyer points out in a
> column in Wednesday\'s DRF. A lot of people here
> have posted a lot of theories, and for all I know
> one or more could be right. But I\'m pretty
> convinced that a) the figures themselves are
> right, and b) that by itself is of limited
> usefulness in dealing with the surface switches.