Par times

[socantra]

In making par times, is the $10k claimer still considered to be the baseline?

socantra...

[JackS]

Socantra- Yes, 10k is primarily used because it reaches even the very cheapest tracks. 10K might even be the true value of NG Stks at a few of these bull rings.

[cj]

Socantra- Yes, 10k is primarily used because it reaches even the very cheapest tracks. 10K might even be the true value of NG Stks at a few of these bull rings.

The problem with using 10k as a baseline is that they are not even close to equal at many tracks.

[socantra]

That's why I was wondering, cj. I'm not trying to make par times. I was comparing various pars and trying to get a context to work with.

As I remember, the old 'instructions' for doing pars were to collect the finish times for older males on a fast dry track, and use the average of that as your $10k par. I know that's over-simplistic, and its been a while, but I was wondering if that were still the general guideline.

I'd have to agree that $10k is not very comparable at many tracks these days, if it ever was. Do you find any level that is comparable?

socantra...

[JackS]

CJ- True but you've got to start somewhere. I've used 16K for a few years but doubt that there are anymore accurate then 10K. Purse structure may be another way to go. 10K pars with a little effort could easily be adjusted to more closely fit if one were to produce a simple "standard" . If opperating a DB there should be no problem at all.

[cj]

I agree, you have to use something as a base, but then you need to see how horses do when shipping around and adjust your figures.

[thelyingthief]

I use the Cynthia Publishing pars, and they provide supplemental materials which, among other things, details at what corresponding class level a horse from any given track and class level can be expected to ship successfully to another. Thus an FE horse of the 10k class may ship to WO at the 10k level successfully, but an FE 25k claimer can not with success enter any higher than a 16k level at WO.

I myself do not use the information, prefering instead to compile such data on a case by case, meet by meet basis for the tracks I play. I can not attest to the accuracy of, nor the methods used to assign these values. Contact me if you have need of any specific track to track breakdown. I doubt if this would be met with hostility by Cynthia...

(these examples are purely fictional)

[J-bred]

I would think that if one were making par times, that rather than use just the claiming price, would a more accurate measure be claiming price + 60% of the purse? Because that is what the owner actually receives if the horse wins the race and is lost to the claim. Obviously a 10 claimer with a higher purse than another 10 claimer at a different track should attract better horses. :confused:

[betovernetcapper]

Decided to check out claiming prices and purses:

Pim claiming price $7500 value to winner $5700

Mnr claiming price $5000 value to winner $6728

The purses tend to reflect other elements like slot machines, then just quality of horseflesh.

[Speed Figure]

The way I make my class rating, say each horse ran a 80 speed figure. The class rating would be. PIM 87, MNR 84.

[Dave Schwartz]

Obviously a 10 claimer with a higher purse than another 10 claimer at a different track should attract better horses.

Not neccessarily true. Sometimes it is to attract any horses at all.

For example, take a place like Canterbury Downs, which is geographically situated a long ways from everywhere else. They need slightly higher purses to encourage trainers to stay.

The only true way to do it is as CJ says - to consider what levels can ship to which track. It is an almost insurmountable problem as one range of class levels may ship over neatly while others don't. And to complicate the issue even more, it can be seasonal.


Regards,
Dave Schwartz

[socantra]

All good points, but I am not trying to build a set of class pars. I'm looking at several sets of class pars and am curious as to what level represents the raw data. In traditional terms it was the 10k level. Is that still generally the case?

socantra...

[JackS]

Socantra- 10k Claimers are generally used because this class level is represented at almost every track in N.A. To use another level higher would restrict some comparisons.
If you prefered for some reason to use a different class level, try 12k. This class should cover 80-90% of all tracks. 15k-16k, maybe 70-80% etc, etc.
In honesty, I can't see an advantage in doing this but, who knows?

[John]

If I remember right Gordon Pine talked about pars on his web site.

Socantra,you should ask Pine your question, No one is better than him with pars.

John

[socantra]

.
In honesty, I can't see an advantage in doing this but, who knows?

Basically I am looking for the average time run at a given track overall, from the raw data rather than the construct of class pars.

For example, the Cynthia pars have a nice smooth progression of 4.5 furlong races at Arlington Park, running all the way from 10k maiden claimers to graded stakes races. Unless Arlington runs a whole lot more 4.5 furlong races than I think they do, I have a feeling most of those lines are bogus, and that I might be better off with the raw averages of 4.5 furlong races run.

I'm not sure if it would be of any value or not, but I would like to find out, and I'm reasonably sure that the par for 4.5 furlong graded stakes races is not of much value.

socantra...

[cj]

SoCantra,

I do something similar for pace. I don't care what the final time or class of race happens to be, I track the percentage of the final time the WINNER ran to the pace call. This is pretty consistent. I don't use the cheapest or best horses on the grounds though.

If my "par" for Belmont at 6f is 64%, and I see the race is run in 1:10, I know the pace time for the winner should be around 44 4/5 (70 * .64.)

You could easily figure out the average final times for all the distances at all the tracks if you have a large database.

[J-bred]

I used to use a similar "pace-ratio" method. The flaw in using the winner's times at each fraction is that the closers are running at the same velocity as the speedballs when they hit the first quarter marker; it merely takes them a little longer to accelerate to that rate of speed.

[cj]

I'd like to know how this is a "flaw?" Also, I don't use the first 1/4 at all.

[J-bred]

The flaw in all current rating methods is that they don't measure energy expenditure. They measure the time required to cover X distance, and some go one step further and extrapolate average velocity from this information. However, energy expenditure is not measured by time or average velocity but acceleration, as per the formula F(orce) = M(ass) x A(cceleration).

Take two efforts, where the horses ran the following fractions:
1st 1/4 next quarter
:23 :24
:24 :24

Which horse expended more energy in the "next quarter"? Modern pace methods will give each the same rating for running a :24 quarter but these were definately not equal efforts, because more acceleration was required by the 2nd runner to be able to maintain that velocity.

[cj]

We aren't talking about the same thing.

[J-bred]

Looks like I diverged some from CJ's question.

Take a race where the "leader" runs the first 1/4 in :22, and the "closer" is 1 second behind at the 1/4. Meaning the leader ran at an average rate of 60.0 fps and the closer ran at a rate of 57.4 fps. But the reality is that while the leader is indeed running at the rate assigned to it, the closer is running at the same rate of velocity. Typical pace figures assume that this closer ran that quarter at a rate of 57.4 fps when in reality it is travelling at 60.0 fps at the point where the first 1/4 ends and the next fraction begins. The only difference between the two is that it merely took a little longer for the closer to achieve the same velocity as the leader. This can cause a skewing when you are measuring the acceleration rate of the horses in the portion of the race after the first 1/4.

By the way, allow me the chance to compliment you CJ for all the hard work you do in making your figures, and putting them out there free of charge, which I consider to be at least as good or better than anyone else's ratings. :ThmbUp:

[Dave Schwartz]

J-Bred,

What you have described is an easy correction. It has been in our software for 15 years.


Regards,
Dave Schwartz

[Secretariat]

Why bother with Pars? Use the Projection method- using whatever set of speed figures you are comfortable with to derive your variants. For example, your estimate of the speed of the race should be discernable by looking at the estimated speeds of the horses in the race. See how far over or under the horses are daily and adjust accordingly.

[Dave Schwartz]

Sec,

Man, am I trying to get that to work well. Very difficult (for me anyway). At least in an automated fashion.

I would sure be open to any suggestions you might have, or a pointer to some articles.

Regards,
Dave Schwartz