AND SPEAKING OF DICK MITCHELL: DISTANCE EQUALIZATION

[delayjf]

I'm reviewing one of Dick Mitchell books (Winning Thoroughbred Strategies). In it he claims that one could use non-linear regression analysis on the a horses rate of deceleration to acurately predict a horses times at different distances. At the time, he said it was not practical because fractional times had a lot of built in inaccuracies due the fractions being timed in 1/5 seconds.

Now that times are in 1/100 of a sec, any math gurus out there doing something like that.

[cj]

Even with times to the hundreth of a second, it would be impossible to accomplish with the exception of wire to wire winners. Beaten lengths are still just a guess at the pace calls, so 5ths, 100ths, what's the difference really?

[Rexdale You]

Del,,,
IMHO Simple arithmatic will tell you that a horse
moving at 22secs. per 1/4 covers 60ft.per sec.

A horse moving at 26secs. per 1/4 covers 50.76ft.per sec.

So you can see how much energy is lost or gained based
on speed. The problem now is will the horse be rated or
sent. All horses are not in a race to win,,,IMHO,,,

Rexyou :confused: :confused: :eek:

[Tom]

The only problem with projecting times from one disatance to another is the the horse has a say in the outcome. Horse do not understand math and so very often, the dummies don't run to the numbers we think they should.

[Hosshead]

This is very true. And another reason they often don't run to the numbers is when they hit the "wall" (their dist. limitation), they often don't decelerate, they STOP, and call a cab. And the distance from the start to the wall can be shortened by such things as weight and class of competition, not to mention off tracks, etc. A horse knows his limitations, - as long as we know what the horse knows, we're O.K.

[delayjf]

Even with times to the hundreth of a second, it would be impossible to accomplish with the exception of wire to wire winners. Beaten lengths are still just a guess at the pace calls, so 5ths, 100ths, what's the difference really?

All that is certainly true, the difference in 1/5 vs 1/100 is certainly a step in the right direction. Now you have at least eliminated the potential errors at each pace call, which could get as high as 1/19th of a second at each call.

Beaten lengths is another matter, we all know how inaccurate the BL can be at the 1/4 pole, at least for those several lengths behind the leaders. One solution might be to use the 1/2 mile pace call in routes, as the BL at that segment is probably more accurate due to the side angle view of the race.

I do know of one former Sartin guy who does incorporate non-lineair math in his pace calculations. Just wondered if anybody else does the same.

Being a Marine, just what kind of math are we talking here. Calculas, algerbra??

[cj]

Originally posted by delayjf
All that is certainly true, the difference in 1/5 vs 1/100 is certainly a step in the right direction. Now you have at least eliminated the potential errors at each pace call, which could get as high as 1/19th of a second at each call...


That can easily be cut in half, just add 1/10 of a second to each horses time. So a horse is listed as running a quarter in 23.1. We know that could mean 23.20 to 23.39, so assuming it means 23.20 invites the largest possible error. Assuming 23.30 cuts it in half. In the grand scheme of things I'm sure it makes little difference, but I do it anyway.

Craig

[delayjf]

Leave it to an Air Force to think of a better way.:D How's Italy, greet food I'll bet.

[lousycapperiii]

I don't know anything about Mr. Mitchell's method, however...

Using 10 feet per length? An 1/8 mile = 66 lengths... a horse's time that is 4.25 lengths back of the leader at the 1/4 mile in 23.12 seconds should be about 23.89 seconds at the finish, give or take a couple of 1/00ths of a second.

66 * 2 / 23.12 = 5.7093425

((66 * 2) - 4.25) / 23.12 = 5.525519

5.525519 / 5.7093425 = .967803

23.12 / .967803 = 23.889159 or 23.89 seconds

You could extrapolate a deceleration factor to get an approximate time for various distances at each track. Or you could use Rubin Boxer's equations which take into account weight carried and the distance from the rail. I believe it's still going to be a guesstimate.

-LC3

[Tom]

Ok, you time a quater in 22.06.
Great. What the hell good does it do you?
Was you horse 3 back, or 5 back? Depends on who called chart.
And if it is 3, howmuch is that really worth?

I use the figs in HTR, but when I am capping with the form and no software, I see a horse run 22.1 and antoher run 22.3, I don't figure much difference bewteen them.
What if the 22.1 was run on a fast day and the 22.3 on a slow day?
No you have the chartcallers ability and your abiity to make variants in the mix. And the Form's ability to transpose times correctly. So a fifth of a second ain't wroth the worry.
But then, what do I know? (That is a rhetorical question, all!)

:D

[cj]

Originally posted by lousycapperiii
I don't know anything about Mr. Mitchell's method, however...

Using 10 feet per length? ...

You are already in the whole, 10 feet is far longer than one length. I know many use it, but its been proven to be around 8 feet per length.

[cj]

Whole? Damn, I'm slipping these days...HOLE!

[lousycapperiii]

Originally posted by cjmilkowski
You are already in the whole, 10 feet is far longer than one length. I know many use it, but its been proven to be around 8 feet per length.

=============================

Mr. cjmilkowski,

Using your figure of 8 ft. per length...

82.5 * 2 = 165

165 / 23.12 = 7.1366782

165 - 4.25 = 160.75

160.75 / 23.12 = 6.9528546

6.9528546 / 7.1366782 = .9742424

23.12 / .9742424 = 23.73 seconds rather than 23.89

I believe 23.89 seconds would be closer to the final time of the horse... However, a stop watch may prove me wrong.

-LC3

[cj]

LC,

I have no idea what you are talking about, and don't want to know really. Measure most horses, its really that simple. Also read Charles Carroll's book, its quite definitive on the subject.

[JustMissed]

I Originally posted by cjmilkowski
LC,

I have no idea what you are talking about, and don't want to know really. Measure most horses, its really that simple. Also read Charles Carroll's book, its quite definitive on the subject.

Hey, did you change your signature from "I loath chalk" to "I am loath to bet chalk"?

JustMissed
;)

[Derek2U]

Lousy3 got a point, albeit mal formed. MAL = BAD ... its russian i think. anyways ... the issue being I think his math is ok but so
what anyways I agree with you. **

[Lance]

The updated version of Davidowitz's "Betting Thoroughbreds" reports on a study of horse length. The average is less than nine feet. I'm almost certain it is 8.75 according to this study.

[GameTheory]

Davidowitz's book says the U of Penn study claims 8.5 - 9.0 feet....

[Tom]

A length is exactly what the chart caller sees it as. Unfortunatelty, they differ from caller to caller, day to day, race to race, point of call to point of call.

The Empire State Building is exactly "X" centimeters hight, but measuring to the nearest foot is close enough. I know one thing for sure, it is higher than Madison Square Garden.

BTW, NYC friends, isn't MSG round? Shouldn't it be Madison Round Garden? Hey, I don't know, just curious.

[kenwoodallpromos]

I just read a webpage that says the average oblique length of a stud stallion is 161, mares is 149 (centimeters)(Sorry, I was out of high school before the new math!)(Does anyone know what horse's oblique length is? Thanks!) I think about .4 inches is a centimeter. I know it is not 5' or 33'. Maybe one of the #'s guys knows!!

[lousycapper4]

Originally posted by kenwoodallpromos
I just read a webpage that says the average oblique length of a stud stallion is 161, mares is 149 (centimeters)(Sorry, I was out of high school before the new math!)(Does anyone know what horse's oblique length is? Thanks!) I think about .4 inches is a centimeter. I know it is not 5' or 33'. Maybe one of the #'s guys knows!!

=============================

I believe oblique length refers to the body of a horse. 161 cm. would be slightly over 5 feet.

-LC4

[PaceAdvantage]

Some people just don't take hints or even overt signals very well, do they?

[cj]

You'd think he'd at least change his name and not make it so obvious...geesh!

[andicap]

Originally posted by Tom
A length is exactly what the chart caller sees it as. Unfortunatelty, they differ from caller to caller, day to day, race to race, point of call to point of call.

The Empire State Building is exactly "X" centimeters hight, but measuring to the nearest foot is close enough. I know one thing for sure, it is higher than Madison Square Garden.

BTW, NYC friends, isn't MSG round? Shouldn't it be Madison Round Garden? Hey, I don't know, just curious.

Tom, the building is not named after its shape. After all, this MSG is its third iteration. It's named after the original location -- it was built on Madison Square.

[raybo]

<I just read a webpage that says the average oblique length of a stud stallion is 161, mares is 149 (centimeters)>

Oblique means sloped or slanted, so I assume this is in reference to the horse's body in full stride. "161 cm" seems quite a bit short to me, huh (62.79")?