Formula designing Speed Rating

[wiretowire68]

Hi hoping for some help with a function because I am bad in math. Okay I have designed a speed par similar to Hong Kong/u.k. for U.S. based on Class and a couple of other wrinkles in terms of my own speed figure. I realize everyone has opinion etc. I have a simple formula but the problem I am having because I do not know how to place in the function of google sheets or excel. I have no problem coming up with a speed rating based on the winner. The problem I have is when the horse runs quicker than par. If horse runs 109.40 and 3 year par is 108.30 my speed with what I am doing is good compared to the class par. The problem is if the horse runs runs a 107.00 ie faster than par, the speed rating my formula has the speed dropping below as opposed to a higher speed figure. The formula is Class Par is 75+(Race Time-Par Time)*10 no problem if winner Speed Rating. Again, anyone know the formula if Race is better than par time and Speed Rating becomes Greater >. Help appreciated????

Chuck

[TexasDolly]

Hi hoping for some help with a function because I am bad in math. Okay I have designed a speed par similar to Hong Kong/u.k. for U.S. based on Class and a couple of other wrinkles in terms of my own speed figure. I realize everyone has opinion etc. I have a simple formula but the problem I am having because I do not know how to place in the function of google sheets or excel. I have no problem coming up with a speed rating based on the winner. The problem I have is when the horse runs quicker than par. If horse runs 109.40 and 3 year par is 108.30 my speed with what I am doing is good compared to the class par. The problem is if the horse runs runs a 107.00 ie faster than par, the speed rating my formula has the speed dropping below as opposed to a higher speed figure. The formula is Class Par is 75+(Race Time-Par Time)*10 no problem if winner Speed Rating. Again, anyone know the formula if Race is better than par time and Speed Rating becomes Greater >. Help appreciated????

Chuck

The terms need to be reversed in the Parenthesis. THEN, if the race time is the same as the par time your number will be 75. If the race time is lower than the par time your number will be higher than 75 and if the race time is higher than the par time your number will be smaller than 75.

TD

[wiretowire68]

i thought so..is there a way to program that without having to it every time i put a time

[ranchwest]

i thought so..is there a way to program that without having to it every time i put a time

Just save your sheet and open it each time you need to use it. That will retain any formulae that you have.

[wiretowire68]

Help and not sure if good idea or bad idea but other handicapper's speed figures who put forth extra variables when they do their figures particularly when it comes the adjustments to the speed using variant to adjust. My question or looking for advice when it comes to say a bonus to the speed figure calculated or a penalty to deduct within the trip. I understand players use post position as deduction(more in terms of routes then sprints) Although if you look at post position winning percentages relative to the winning post/rail(as in Bris)

Wondering does anyone have thoughts ideas on this...as I am considering as Dave Schwartz does whereby he has added winning running style to the equation. In terms of formula calculation would you give bonud point or two to a horse in the equation for a bias would you penalize a horse points forother factors affecting performance. I believe particularly if no workouts are published that this a negative/ especially on a lower percentage trailer.

Just interesed in ideas for this or is it itoo much depth

[wiretowire68]

I have also read some work on friction affecting speed, Whereby the more ground needed when hung out wide/trouble in the trip. The Friction that I have read about is regard to kick back which obviously creates tiredness or zaps energy. I know a gentleman wrote formula using force and friction etc affecting speed besides the many other millions of factors that cause a horse not to run to the speed or the form we see. Again anythoughts or ideas. Would also ask for some ideas on reading material besides the regular guys on speed. Thanks.

[ranchwest]

I like Handicapping Speed by Charles Carroll.

[Dave Schwartz]

Wondering does anyone have thoughts ideas on this...as I am considering as Dave Schwartz does whereby he has added winning running style to the equation.

wiretowire68,

Actually, I add running style (of the winner) to the PAR MAKING process.

From my annual "State of the Pars" document within the 2022 Par Times.

Par Times Process
A huge advancement was how the impact of the much-improved class levels allowed the pars to be so much easier to build. The process for that was:
1. Build the core distance pars using the Standard Class Levels for Older Male and weighting the races based upon Time decay and Running Style of the winner.

The big change here was that the Running Style weights. I’ve been using this for several years. It begins by asking the question, “At which call does the ultimate winner first challenge for the lead?”

When I began it was a 4-3-2-1 approach, meaning 4 points for a 1st call winner and 1 point for a finish line winner. Then it moved to 6-4-2-1 and now is at 9-4-2-1.

In other words, a winner who first challenges at the 1st call is 9 times more predictive than a winner that is not within 1 length prior to the last furlong.

I would not be surprised if the optimum turns out to be something like 12-4-1-1.

2. Determine the Track Time Pattern (i.e. developed by Dr. William Quirin back in 1979).

3. Build out the remaining distances by turning on all classes/ages/sexes, etc. and averaging them, then adjusting them to fit based upon the track pattern.

While this is the basic approach I’ve always used, the new improvements to the process are showing year-over-year to produce greater accuracy. Heck – we were even competitive at KEE last year!


Historically, KEE has been all over the place. It might be because of the 2 meets - which appear to be substantially different.


Since you're working so hard for this, if you'd like a short call to discuss what you're doing, just email my Customer Service VA (Kelsey) (Kelsey@HorseStreet.com) and she'll set you up with an appointment.


Dave

[wiretowire68]

Hi Dave, Well would love to work for you but with todays computers and ways of data mining/machine learning people do not need people anymore. I was more or less saying that I subscribe to your pars theory. I have noticed because I have taken the time to input all winning times from every race in my spreat sheets for Aqueduct and Santa Anita as well as track condition and Class of race and boys/girls and ages, will do the same for my own track(Woodbine when it starts and concentrate on New York/California.

I have noticed which I never really paid attention to in the past is the obvious but not talked about alot is the weather in the east and how it really affects running times for most classes and the subtle changes from some races say 1st race to the 4th race and when races are run back to back at the same distance.

There is though the odd random Sustained winner or Presser but for the most part it is the Early or the EP that win the majority of races and you have found that it is trending that way It appears that way.


Chuck

[Dave Schwartz]

Hi Dave, Well would love to work for you but with todays computers and ways of data mining/machine learning people do not need people anymore. I was more or less saying that I subscribe to your pars theory. I have noticed because I have taken the time to input all winning times from every race in my spreat sheets for Aqueduct and Santa Anita as well as track condition and Class of race and boys/girls and ages, will do the same for my own track(Woodbine when it starts and concentrate on New York/California.

I have noticed which I never really paid attention to in the past is the obvious but not talked about alot is the weather in the east and how it really affects running times for most classes and the subtle changes from some races say 1st race to the 4th race and when races are run back to back at the same distance.

There is though the odd random Sustained winner or Presser but for the most part it is the Early or the EP that win the majority of races and you have found that it is trending that way It appears that way.

Chuck

"Hi Dave, Well would love to work for you but "

I was not offering a POSITION. I was offering to HELP.

Early or the EP that win the majority of races and you have found that it is trending that way It appears that way.

Not actually what I am saying.

More like, "Only races where the winner was on or near the lead at the 1st call are useful in predicting par times."


"Late" winners win a fair proportion of races. The definition of late being "Not within 1 length at the 1st call."

The later the winner challenges for the first time, the slower the time of the race.

Now, the implication of that last part is critical.

Let's say that you've determined that "par" for this races is 103 and your stats indicate that in order to win, a horse must have been able to run (say) within 3 lengths of par, which pegs "100" as the minimum.

That premise will be way wrong if the eventual winner does not challenge at the 1st call!

If the eventual winner comes from a 2nd call challenger, the REQUIREMENT will be perhaps 2-4 lengths slower!

And later than that could easily be 10 LENGTHS SLOWER!


IOW, Par times (as typically used) really only apply to front runners!


Dave

[wiretowire68]

part was just a joke Dave as truthfully If I could do this type of research for a living. I would be a very, very happyperson. And I appreciate the sound advice going forward

Chuck

[wiretowire68]

In my own experience, again I think it depends on the trainer, connections but for me what I see from watching races and particularly when it comes to running styles that you can examine the pace all you want but I personally believe and I have seen it and bet them many times befoe a horse is actually ready, is a horse that has broke their maiden as an E horse whether it is 1st asking (horse was fit enough) or either pressing or sustained, that when that horse comes steps up to face winners and change class etc, that many times the horse next that trainers instructions are to get back to trying to win on the lead or unti they learn...and it is vice versa--thus again, if the plan is to stretch them out or to find out what kind of horse will this horse will be in terms of their earnings potential.

Chck

[wiretowire68]

Maybe in plain English this time. I believe pace. However, I understand why certain horses win or are consistent. I believe it is difficult for a trainer to hide his horse from the public considering the sophistication of handicapping today. I have noticed over the years, Pace is important, many winners come from the E and the EP's. If you were to look up the par times of all individual classes age etc. My point is simple.. As an example when you consider horses moving up in class or down from their beginnings as racehorses, trainers get them ready and you can interpret that whatever way you want. My belief is besides being bred to do this..it comes down to their fitness and their class. I have seen it in many many races, horses step up after a win.. run nowhere or hit the board. In the case of maidens, the higher percentages win on the lead, then you will see, the classic bounce scenarios because the trainer now instructs to teach the horse to come from off the pace or sustained effort and vice versa.

They want to know what they have in terms of their asset and can he/she earn and will we make up ourmoney spent and if they have the patience. So the experiments take place each race at the horseplayer's expense. My thought process is to look for the positives/negatives that are part of the process leading up to potential win. So in saying that what negative numbers I can use after a speed calculation is made or can I add a point to his score due to as you say..the trends of 1st Call, 2nd Call etc. There are factors such as kickback, post position, etc. (negatives in the current cycle vs. positive one in previous.)

So looking to add that to my final speed calculation.

Chuck

[steveb]

(Race Time-Par Time)


i have not read thread thoroughly, but that bit above taken from your 1st post is a big NO NO.

[wiretowire68]

and why is that?

[steveb]

and why is that?


because if something ran 1 second faster than par at 5 furlongs, that would be more impressive than something running 1 second faster at 10 furlongs.


or another way to say it, with your way every 1/10 seconds equals 1 point at any distance, and that is nonsensical.



your way will have them the same regardless of distance.


Proportion is EVERYTHING.


that's the vanilla explanation.


all the other stuff is extraneous to that.


your 57 and 10 could be anything and are basically meaningless.

[wiretowire68]

Ok Steve well as far as developing your own speed formulas here in North America what do you recommend. I am trying to come up with a with a neural prediction. In terms of Beyer numbers or other speed numbers that show if the number in the horses cycles in chart vs. the par time of the track(three year best) and current par vs upcoming race. I keep up daily data on certain tracks in terms of winning time/run style and class/age etc, The question is how to take the normal Speed equation convert it. Here the norm is obviously to subtract Par time from Horse time, which gives your +- seconds. Thus the guess of Beaten Lengths to a horse'e accelerating point to its deacceleration. i.e. Early Vs. Late.

CL

[Tom]

Why not just use the Beyer numbers. No way you going to make better fugures on your own. I love spaghetti, but bI buy it in a box, I don't roll my own.

[steveb]

Ok Steve well as far as developing your own speed formulas here in North America what do you recommend. I am trying to come up with a with a neural prediction. In terms of Beyer numbers or other speed numbers that show if the number in the horses cycles in chart vs. the par time of the track(three year best) and current par vs upcoming race. I keep up daily data on certain tracks in terms of winning time/run style and class/age etc, The question is how to take the normal Speed equation convert it. Here the norm is obviously to subtract Par time from Horse time, which gives your +- seconds. Thus the guess of Beaten Lengths to a horse'e accelerating point to its deacceleration. i.e. Early Vs. Late.

CL


the easy and plain vanilla formula could be...


(par_time/race_time-.9)*1000


your par in this case would = 100 in Beyer language.


so if your par was 68 for 6f and some nag runs 69 then....
(68/69-.9)*1000 = 85.5

if runs faster than par say 67.5....
(68/67.5-.9)*1000 = 107.4





then you would be on same scale as Beyer for winners(beaten runners is a little more complex).


of course your variants and breakouts are another topic, as is how your pars are determined.


anybody that subtracts the race time from the standard time(or par) though, is living in a fools paradise.
you can quote me on that.
i also don't think that is the standard way in your country.
i have been through all this before with another on this forum, so am not going through it all again.
you can find the previous posts if you care enough.


your early and late and all that type stuff is another topic.
you need to get your overall done properly, before you get into the deeper stuff like pace and sections.