Creating speed and pace figures?

[twindouble]

I hope you number guys don't mind me asking a question or two on this subject because you know I'm skeptical about betting just numbers and I have no clue how many factors go into creating them. Plus it's been a hot topic here more than once that I know off. If it's top secrete I'll understand.

I stated in another thread that with some tracks moving to Polly material I thought the Beyer figures would go by the wayside, primarly because it would screw up what ever method he uses to factor in track variants. So my first question is, how important is track variants to your numbers? Plus have you adjusted to the new material and times?


Thanks,

T.D.

[cj]

Polytrack has not eliminated the need for daily track variants. The surface still changes speeds from day to day, and even within the same day on occasion.

A person who makes figures isn't betting just numbers, any more than a trainer handicapper is betting just trainers.

[twindouble]

Originally Posted by cj

Polytrack has not eliminated the need for daily track variants. The surface still changes speeds from day to day, and even within the same day on occasion.

A person who makes figures isn't betting just numbers, any more than a trainer handicapper is betting just trainers.

It's answers like this that flustrate me. I didn't infer the guys that make numbers just bet numbers, I asked what other factors go into them. Nowhere did I say the pollytrack eliminated daily variants, I asked how you and others have adjusted to them.


T.D.

[sjk]

TD,


I'm not sure what you mean by other factors. Everyone who makes numbers probably does it a bit differently so I don't know if they would all agree with this.

It is all about comparing times of horses running on the day in question with projections of how fast they ought to run based on (same surface) past races in an effort to determine how fast the track was.

How you handle outliers and race-to-race differences play into the final figure but the main thing you have to work with are today's times and horse's past performances.

I have treated Polytrack as dirt with good results.

[twindouble]

Originally Posted by sjk

TD,


I'm not sure what you mean by other factors. Everyone who makes numbers probably does it a bit differently so I don't know if they would all agree with this.

It is all about comparing times of horses running on the day in question with projections of how fast they ought to run based on (same surface) past races in an effort to determine how fast the track was.

How you handle outliers and race-to-race differences play into the final figure but the main thing you have to work with are today's times and horse's past performances.

I have treated Polytrack as dirt with good results.

Thanks, you at least answered my question on the Polly. So your saying I'm wrong thinking the Polly effected the variants and it's no different than a dirt track creating numbers.

By other factors I mean, bias, loss of ground, trouble in the race, quality of horses that ran that day or any other day when it comes to variants and so on. Is it all just raw numbers?

T.D.

[sjk]

If you had the point of view that polytrack was not interchangeable with dirt then making numbers would be much more troublesome at the beginning of a meet where all of the horses were suddenly changing surfaces. Many races would be unplayable or would be played based on very limited data.

As I said above I treat poly as dirt which allows me to play the poly meets from the beginning. Of course if I am wrong and this is mixing apples and oranges I will pay a price in time. Last winter at TP seemed to go well so I hope I am on the right track.

As to the other factors you mentioned, the quality of horses that ran on a given day is reflected in the time data. I make a small bias adjustment in my program but it is separate from the figure calculation. I don't have any way of measuring ground loss or trouble so I ignore those items.

[twindouble]

Thanks, sjk. I have to run, I'll continue tonight.

[cj]

Originally Posted by twindouble

[/b]

Thanks, you at least answered my question on the Polly. So your saying I'm wrong thinking the Polly effected the variants and it's no different than a dirt track creating numbers.

By other factors I mean, bias, loss of ground, trouble in the race, quality of horses that ran that day or any other day when it comes to variants and so on. Is it all just raw numbers?

T.D.

Dirt and Turf are two totally different surfaces, yet you make figures pretty much the same way. You look at how fast the horses ran according to the clock, then you look at how fast you think they should have run. The difference is a projected "variant" for that race. You do every race on the card, and try to determine an overall daily variant, or see if there is maybe a split somewhere in the day. Why would polytrack be any different?

I personally don't really use bias, except as a possible factor in projecting how fast the winner should have run. I definitely do not use ground loss. The only people I know that do are Thorograph and The Sheets. Personally I think that provides more distortion than it does better figures, but that is a personal choice. Many agree, many others disagree. The quality of the horses running that day is of course a huge part of my projection of how fast the horses should have run.

I factor in the pace of the race a bunch as well. It often clearly indicates why the vast majority of variants that don't seem to fit with the day occur. Let's say there are five sprints run on the day, and they look like this: 10 fast, 10 fast, 5 slow, 9 fast, 11 fast. The 5 slow seems to stick out and looks crazy. Most times, I will look and see the horses in that race ran extremely fast or extremely slow to the pace call. This is especially useful in turf races.

I'm sorry I didn't answer your question better the first time. I was obviously mistaken thinking you understood at the very least the basics of making a speed figure. Since you obviously don't, it is probably wise that you don't use them. Instead of giving me a bunch of shit, at least give the benefit of the doubt that I tried to answer. Perhaps your questions could have been phrased better?

[RXB]

If anything, Polytrack will enhance the performance of speed figures because it removes some of the early pace bias.

[BIG49010]

I am making Poly figs for the first time, and was wondering how close I am on the first 3 days. I had Wed. and Thur. the same, with Friday being a full second faster. It was hard to tell, but I thought it was raining last night, and my assumption was the track got faster.

Thanks for comments.

[sjk]

I am guessing you are talking about TP since you mention night racing.
I have all three days between 24 and 26 fast. New material must be making things faster than last year.

DATE	TRACK	 VAR
9/6/2006	TP	-25.0415010916118
9/7/2006	TP	-23.6846652013384
9/8/2006	TP	-25.7855028831288

[twindouble]

I'm sorry I didn't answer your question better the first time. I was obviously mistaken thinking you understood at the very least the basics of making a speed figure. Since you obviously don't, it is probably wise that you don't use them. Instead of giving me a bunch of shit, at least give the benefit of the doubt that I tried to answer. Perhaps your questions could have been phrased better?

cj; Thanks for responding in the manor I knew you are capable of but if you think that was a bunch of shit, you don't know me. Yes your right, I could use an editor, just concider that next time. hand shake.


T.D.

[twindouble]

Originally Posted by cj

Dirt and Turf are two totally different surfaces, yet you make figures pretty much the same way. You look at how fast the horses ran according to the clock, then you look at how fast you think they should have run. The difference is a projected "variant" for that race. You do every race on the card, and try to determine an overall daily variant, or see if there is maybe a split somewhere in the day. Why would polytrack be any different?

I personally don't really use bias, except as a possible factor in projecting how fast the winner should have run. I definitely do not use ground loss. The only people I know that do are Thorograph and The Sheets. Personally I think that provides more distortion than it does better figures, but that is a personal choice. Many agree, many others disagree. The quality of the horses running that day is of course a huge part of my projection of how fast the horses should have run.

I factor in the pace of the race a bunch as well. It often clearly indicates why the vast majority of variants that don't seem to fit with the day occur. Let's say there are five sprints run on the day, and they look like this: 10 fast, 10 fast, 5 slow, 9 fast, 11 fast. The 5 slow seems to stick out and looks crazy. Most times, I will look and see the horses in that race ran extremely fast or extremely slow to the pace call. This is especially useful in turf races.

I'm sorry didn't answer your question better the first time. I was obviously mistaken thinking you understood at the very least the basics of making a speed figure. Since you obviously don't, it is probably wise that you don't use them. Instead of giving me a bunch of shit, at least give the benefit of the doubt that I tried to answer. Perhaps your questions could have been phrased better?

What came after "I'm sorry" kind of got whipped away with the rest of the paragraph on second thought.

To suggest I don't understand the "basics" by inquiring how these figures are made with you knowing I've been playing he horses for almost 5 decades now is a hit below the belt. Perhaps you could have phrased that whole paragraph better.

The first three paragraphs covered what I was looking for, I also could have been given the benefit of doubt. Works both ways.

I made a joke of the whole deal but believe me, when it comes to many posts that I think are off the wall, I could have thrown out more shit than this board could handle. I chouse not to do that for obvious reasons.

Anyway, do I dare inquire any further? Yes I will, it can get rather booring here now and then.


T.D.

[twindouble]

Originally Posted by RXB

If anything, Polytrack will enhance the performance of speed figures because it removes some of the early pace bias.

RXB, why would you conceder early speed a "bias". A bias to me has more to do with the track condition, inside, outside bias favoring speed or closers.

Are you saying the Polly tracks favor closers, speed tires on that material?


T.D.

[RXB]

Frontrunners dominate on most dirt courses. Closers are more dangerous on grass. If two horses are running the same speed figures, the horse with the higher early pace rating is the more likely dirt winner. On grass, it's the horse with the superior closing fraction.

Poly, from what I've studied so far, doesn't really promote any style of runner. It's less favourable to early speed than dirt, but more favourable than grass. Pace matters from an internal dynamics standpoint in each individual race but speed figures by themselves seem to be more important on Poly than on the other surfaces.