Energy distribution, again....

RanchWest

Okay, I can't find the thread on energy distribution.

I do know that I completely misunderstood the question. So, if I may, I will reframe the question...

If you use pace handicapping ala Sartin, Brohammer, et. al., do you use energy distribution and, if so, how?

All of what follows is primarily for dirt racing.

First, I use the seven Sartin factors, which are for the most part a combination of ability and distribution, to rank the horses in a race and arrive at a composite factor ranking.

I seldom use the pace line distribution of an individual horse. What I do is use the %Median of all of the lines of a race and build a model range based only on those races. I then see the number of races a horse is outside the range. If outside the range too often, the horse probably doesn't fit the race. I am especially leery of horses that are too often low %M.

I also use %M to determine how the field as a whole figures to distribute. If it is a low %M field, some of the horses will still have to go to the front and very often those horses will tire and the race becomes chaos. If it is not a low %M field, I assume a "normal" race and move on to other factors. Uncommonly, some races have very high %M and those can also devolve into chaos.

Many people look for an increased distribution (more early) in the most recent start if it is within a recent period of time, maybe 21 days. These horses often improve their endurance and fire well next out. .I probably should watch these more, but admittedly I don't.

I have not found a benefit to studying the distribution of the pace line. Maybe I should look at it more.

Dave Schwartz

↪RanchWest

I'm not following your process very well and I'd really like to.

What I do is use the %Median of all of the lines of a race and build a model range based only on those races.

So, if there are 10 horses, each with 10 pacelines, you through all 100 into a put and... WHAT?

Do you average all of those 100 races, ignoring surface & distance? (I'm thinking not.)

More, please.

RanchWest

Yes, I go through the "100" races and take the median of the %M's regardless of surface and distance. My logic is that that is the pace the horse is accustomed to running. You might disagree with the logic of using all values, but you are entitled to your opinion. I have not gone through it in any other way from what I have described above, so I am not claiming it is absolutely the best way.

Dave Schwartz

↪RanchWest

"Median" meaning the single value dead center in the sorted list?

I actually like the idea - so much so that it will wind up in the Studio.

Might want a little filtering. (Perhaps tossing wrong surface, way wrong distance.)

Dave Schwartz

BTW, THIS is why I love a forum.

I asked a simple question but apparently didn't make it quite clear enough. Yet, a wonderful idea comes out of the DISCUSSION that ensued.

RanchWest

Yes, the middle of the sorted list. I figure for the most part the extremes will end up being mostly irrelevant.

I don't remember exactly how I figured the range, but it is always a set value above and a set value below the median value. So, I am not depending on an exact value anyway. Just looking for horses that have been running extreme energy distributions for the current field.

My unscientific observation is that at Saratoga a horse that has been a late distribution has very little chance on the dirt.

RanchWest

As I described in the chaos thread, I also take the median %M of each horse and assign a value from 1 to 6 as being E to L. I am fully cognizant that this EPS is not to be a replacement for the positional EPS running style commonly employed, both should be employed in different ways.

It is this 1 (early) to 6 (late) system that I use to determine how early or late the field is, using the average of the 1 to 6 system. I tried to arrive at values for 1 to 6 that more or less evenly split up all horses, at least that was the goal. So, when a field is over 3, it is late. Under 3 is early. When the field is late, that's when chaos might ensue. With a lot of late horses in the field, they still have to have some on the lead and those horses often shift their distribution and that's why they back up. At least, that is my thought.

In this 1 to 6 system, I also calculate the variance and standard deviation from the mean, but I have not done enough study to know how to use those values. I suspect they have significance.

RanchWest

%Median, deceleration and VDC (velocity to deceleration) can also tell a couple of things.

1) the inability to handle a longer distance

2) a form cycle fall off of form

RanchWest

i just remembered. I do use VDC as one of many factors in selecting a pace line, giving limited preference to the line with the best VDC.

Dave Schwartz

i just remembered. I do use VDC as one of many factors in selecting a pace line, giving limited preference to the line with the best VDC. — RanchWest

How is that calculated?

RanchWest

I use Total Energy divided by deceleration. Does that tell you enough or do you need more? I don't think that is technically the Sartin way.

Dave Schwartz

I use Total Energy divided by deceleration. — RanchWest

What are the calculations for those 2 variables?

Total Energy
Decelaration

Mark

Why not just contact Ted Craven since VDC is apart of the RDSS software?

Dave Schwartz

Yes, Ted & I are always sharing ideas. LOL

Mark

↪Dave Schwartz

He does have New Pace and CSR in RDSS right?

RanchWest

The following is how I define the calculations. May not be "perfect Sartin".

deceleration is
( EP/f3 ) * 100

I believe EP is f1 + f2.

TEDC (which I use as VDC) is:
100 * ( f1 + f2 + f3 ) / deceleration

I think VDC is:
100 * ( AP + f3 ) / deceleration

I make no adjustments to the raw numbers for these because the adjustments would wash out.

RanchWest

↪Dave Schwartz He does have New Pace and CSR in RDSS right? — Mark

Ted told me publicly how to compute CSR. I use the BRIS speed ratings. I also use the same methodolgy on the BRiS pace figures.for 2f, 4f, 6f, and LT.

RanchWest

I also have a number I call Pressure that I equate to Quirin speed points on steroids. It suggests which horses will go to the lead. I then arrive at a mean number for that. So, here is what I have for the field:

Variance 0.611111
Std. Dev. 0.781736
Mean 1.888889 EP
Median Pressure 63
Top 3 Q 21
Giles Speed Points 3 (3+ 1C, 2 2C, 1 3C, 0 4C)
Hi Q % Tot Q 25.00

The first 3 numbers are for the EPS running style based on %M as explained earlier in this thread. These are actual numbers for the 8th at TDN today. With the mean EPS at 1.88, my anticipation was that the pace was not going to be unusual though a bit on the early side, suggesting some of the front end might collapse. The top 3 Q of 21 suggested the same. I had #9 as a pace Presser though positionally he was an S. His current class was highest in the field and pressure was 63,exactly at average. His Scott PCR was 3rd behind the coupled entry that went off at odds on, .3/1. So, the #9 did indeed stalk from 3rd and closed up the stretch. Interestingly, the uncoupled entry mate, #6, finished 2nd at 67/1, triggering an exacta of $163.40 on $1. The jock on #6 was the best in the field and the trainer of the two exacta horses was the best in the field. Of course, hindsight is 20/20, but just goes to show that most races do have working puzzle pieces.

Dave Schwartz

He does have New Pace and CSR in RDSS right? — Mark

Yes, he liked what he saw and appropriated it from me.

I have missed your point, though.

Dave Schwartz

deceleration is
( EP/f3 ) * 100

I believe EP is f1 + f2.

TEDC (which I use as VDC) is:
100 * ( f1 + f2 + f3 ) / deceleration

I think VDC is:
100 * ( AP + f3 ) / deceleration — RanchWest

Ah, this is very helpful.
So, this is not physics-based but rather a logical relationship between the numbers.

I generally use some form of a known physical science formula but really struggling with the velocities for fractions that use the Stretch Call (in any way) since it is a distance without a time.

It's easy when you are using pace numbers because you all the variables are known:

Distance
Lengths Behind
Core Ratings

RanchWest

I generally use some form of a known physical science formula but really struggling with the velocities for fractions that use the Stretch Call (in any way) since it is a distance without a time.

It's easy when you are using pace numbers because you all the variables are known: — Dave Schwartz

I don't understand. What is not known?

Dave Schwartz

I don't have a valid physics-based formula to use.

For example, my energy calculation is similar to energy produced by a wave.

It is a a physical science formula that is not known (by me).
Note: I am not smart enough to create my own from scratch, so I look for something in the science world that looks close and try to use it - without modification if possible.

Dave Schwartz

@RanchWest

Here's what I was looking for:

Dave Schwartz

A big THANK YOU to @RanchWest for your idea about the Median Energy.

I'm implementing via some reports in the Pace Handicapping module.

F1% & EP% for races where the horses ran within 1 length of the winner (or won by any margin).
Same but with Similar Surf & Distance within 1/2 furlong of today.

In addition, these 2 metrics (F1% & EP%) will show for the race in each pace line.
(This one takes some doing but will make it relatively easy to test your theories.

Ranch, I believe that your approach may have tremendous value when combined with the prediction of which call will be the most important in today's race. (Determined by the ES5 approach which considers the number of ES5+ horses and applies the two rules I've mentioned before.

Energy distribution, again....

Please register to see more

Categories