Building Par Times for Saratoga - Need Help

[jeebus1083]

I have a compilation of three seasons (2007 through 2009) worth of data at Saratoga for both Dirt, Turf, and Inner Turf saved on an Excel spreadsheet. However, I am having a hard time trying to build a table of par times that seem to "make sense". I referred to Dave Schwartz's tutorial on building your own pars at his website, and am still lost.

Since there isn't any $10k claimers at Saratoga, and the fact that many of the claiming events now have n2L, n3L, B, or n1y conditions, trying to determine proper "levels" is becoming a complete and utter headache.

I'm not interested in purchasing anyone's par times. I want to do this on my own to attempt to gain an understanding in the Spa's surfaces, and to see what goes into building your own pars.

Any help or advice would be appreciated.

[jeebus1083]

An example of what I'm dealing with:

Between 2007 and 2009, 18 open claiming races for males 3up were run at 6F on Fast/Good Dirt. The breakdown (Class - # of races):

Clm16000n1y - 2
Clm20000 - 2
Clm20000b - 1
Clm20000n2L - 2
Clm25000 - 3
Clm25000b - 2
Clm25000n2L - 2
Clm25000n3L - 1
Clm35000n2L - 2
Clm50000 - 1

Short of "faking" it, how can you honestly derive a "par" for any of these classes, especially with limited data?

[illinoisbred]

Try starting by looking at claiming or allowance levels you have large samplings. If you have any classes with say 20 races try throwing out the fastest 5 and the slowest 5 then average the remaining 10. I've always found that to be a pretty representative time. After doing your large samples,see if your smaller ones fit in any sensical manner.

[illinoisbred]

Wow,that is a small group.Actually,you just have 6 truly open races.

[Robert Goren]

Mixing data from different years is a no-no at most tracks. Things like run up locations can change.

[jeebus1083]

Allowance/Optional Claimers 3up Males on Fast/Good tracks at 6F:

n1x/n2L (with 1 non-OC n1x, 1 non-OC n2L, and 2 OC n1x lumped) - 4
n2x (2 OC50k n2x and 1 OC35k n3L lumped) - 3
n3x (OC75k) - 1
OC100000b - 2

Total allowances: 10

[jeebus1083]

Mixing data from different years is a no-no at most tracks. Things like run up locations can change.

Then how does one create pars? Out of one's :-X? Don't you have to average "something"?

:)

[Horseplayersbet.com]

Then how does one create pars? Out of one's :-X? Don't you have to average "something"?

:)
Averages work if you have a very large sample. Still, there are basic assumptions, for example 4,000 open claimers will run slower than 7,500 open claimers.
Once you've made them for one track, you can use the same principals for other tracks.
If you are out on a few classes, it will show up very quickly when those races are run, and changes can be made.
Today, one can go to Equibase and get all the charts for any track dating back to the early 90's.

[Dick Powell]

Good luck with your project. Luckily, you are not doing it back when Howard Battle was race secretary who had allowance conditions that have not been seen since.

[Robert Goren]

Then how does one create pars? Out of one's :-X? Don't you have to average "something"?

:) Making good dirt pars for a short meet which runs a lot of turf races may not be possible.

[illinoisbred]

I believe Tom Brohamer outlined a quick,easy way to construct par times from small samples. Basically,use your middle class (25,000) and work the lower ones up by adding 1/5's and your higher ones down by subtracting 1/5's. I've never resorted to that so I have no idea if its a useful technique.

[illinoisbred]

Have you been making figures for that circuit (Aqueduct, Belmont) or any of the major ship-in venues? You may have to rely on projecting figures for the dirt.

[jeebus1083]

I have not. I only have the times for Saratoga.

[PhantomOnTour]

As some have stated already, you will have to extrapolate. In your example I would toss everything with a n2l or n3l after it but I always use n1y or B designated races in my par calcs.

An n1y is usually non winners of a race this year or a certain number of months (usually 6) or sometimes non winners of a turf race this year or certain number of months. A Clm20n1y field isn't necessarily a 'lower class' than Clm20 field, it may just have some layoff horses who haven't had a chance to win one this year or run on the turf yet this year.

If you accept the previous paragraph you have a few races in the Clm20-25 range to work with. Maybe 6 or so? If their times cluster you may have your starting point. Look at all Alw and MSW races you can find as well. It's nice to know where these 2 classes fall in the Clm ranks. Alwn1x is usually equal to Clm20-30 so if you find that the Alw race times are near those Clm20-25 races you already have then that is conformation of those Clm races as a good starting point.

I admire your desire to make your own pars. I do the same thing every year. I hope you make your own figs as well, even if just for one meet. The experience is enlightening. Good luck and holler if you have any questions.

[illinoisbred]

I agree with Phantom -I break out the nw1y but have never really found any difference with the open races so I re-include them in the sample. Question-are you planning on making Quirin-style pace/speed figures,or Beyer-like figures?

[jeebus1083]

I agree with Phantom -I break out the nw1y but have never really found any difference with the open races so I re-include them in the sample. Question-are you planning on making Quirin-style pace/speed figures,or Beyer-like figures?

I was looking to do a Quirin-style series of pars. They are simply easier to make than a Beyer scale.

[Dave Schwartz]

Jeebus,

The problem you are having is well-stated: Since there are no $10k claimers at Saratoga, how do you manufacture pars for them?

The answer lies in having the class levels for other tracks in place so that you can compare them as you called it, Quirin-style.

The short way is to take a track that you know should have highly similar class levels (like BEL) and compare the "firm" class levels they have in common and assume that the difference in class levels is the same.


Let me give you a quick bit of help here:
http://www.horsestreet.com/BBSImages/SarClassLvls2010.png

Thus, you can assume that a $16k claimer is about a "102."

Hope this helps.

Dave