Speed rating inaccuracies

[jasperson]

There has been a lot discussion on this forum about inaccuracies in speed ratings and who has the best. In my postmorten examination of the 5 tracks I played saturday I didn't find over one race where I thought an inaccurate speed rating effected the outcome. I was using bris speed rating. Question how often do you think inaccurate speed ratings effected the race?

[cj]

An inaccurate speed figure never changes the outcome of a future race. What it might do is lead one to make bad bets, or miss some good ones.

[Dave Schwartz]

I think you have mis-stated the problem. Speed ratings do not determine the outcome of races.

We actually do not know what determines the outcome of races. If we do, then we have the Grail.

[Robert Goren]

I remember Beyer complaining when he first developed his ratings and they were not available publicly that trainers did not always try with their horse because they did not know that it had just ran a very good rating in their last race.

[cj]

I remember Beyer complaining when he first developed his ratings and they were not available publicly that trainers did not always try with their horse because they did not know that it had just ran a very good rating in their last race.

I would say a large percentage of top horsemen now use some type of figure to aid in the placing of their horses.

[thaskalos]

There has been a lot discussion on this forum about inaccuracies in speed ratings and who has the best. In my postmorten examination of the 5 tracks I played saturday I didn't find over one race where I thought an inaccurate speed rating effected the outcome. I was using bris speed rating. Question how often do you think inaccurate speed ratings effected the race?
You say that you were using the BRIS speed ratings.

How is it possible to identify the inaccurate speed ratings, if you only use one set of them?

IMO - in order to really observe the inaccuracies in the figure-making process - you would have to put several different sets of figures side-by-side...and then you would see that different figures sometimes favor different horses in the same race.

Unfortunately...figure-making is not an exact science...and it will never be...

[classhandicapper]

I think it's more than an accuracy issue. It's that even when they are accurate they may not reflect the quality of the horses.

[Robert Goren]

Some days the times are so all over the place that it impossible to come with a good number. It doesn't matter how different rating services you use, the numbers from that day are useless. The guys who make the numbers know which days those are, but you as a subscriber don't.

[aaron]

Making numbers can be extremely frustrating. At the current Belmont meet-for instance,you are dealing with probably 4-5 dirt races and 4-5 turf races on 2 different turf courses each day. With the different distances on the dirt and the 2 different turf courses usually with 2 turf sprints thrown in.it is extremely difficult to just make an opinion based on numbers.

[cj]

Making numbers can be extremely frustrating. At the current Belmont meet-for instance,you are dealing with probably 4-5 dirt races and 4-5 turf races on 2 different turf courses each day. With the different distances on the dirt and the 2 different turf courses usually with 2 turf sprints thrown in.it is extremely difficult to just make an opinion based on numbers.

Belmont is the toughest course for sure. Don't forget the varying rail settings on the turf too!

[jasperson]

Some days the times are so all over the place that it impossible to come with a good number. It doesn't matter how different rating services you use, the numbers from that day are useless. The guys who make the numbers know which days those are, but you as a subscriber don't. That is the point I am trying to make. Nobody has absolutely accurate speed rating or else Andy Beyers would not have to adjust his. We can deal with inaccuracies by using other handicapping factors.

[thaskalos]

That is the point I am trying to make. Nobody has absolutely accurate speed rating or else Andy Beyers would not have to adjust his. We can deal with inaccuracies by using other handicapping factors.
IMO...more mistakes are made in the INTERPRETATION of the speed figures, than in their construction.

Successful speed handicapping involves alot more than just blindly betting on the horse with the highest speed figures. We also have to analyze the circumstances under which these figures were earned.

[therussmeister]

Making numbers can be extremely frustrating. At the current Belmont meet-for instance,you are dealing with probably 4-5 dirt races and 4-5 turf races on 2 different turf courses each day. With the different distances on the dirt and the 2 different turf courses usually with 2 turf sprints thrown in.it is extremely difficult to just make an opinion based on numbers.
I'm deliriously happy when there are five dirt races. There are almost never more than four, unless some turf races are rained out. That is one of the main reasons I don't play New York, I don't feel comfortable with my figures.

[dansan]

speed figure's are overrated

[Greyfox]

An inaccurate speed figure never changes the outcome of a future race. What it might do is lead one to make bad bets, or miss some good ones.

cj Yours is the second post. :ThmbUp: NAILED THE ANSWER.
End of discussion.

Psst. An accurate speed figure doesn't either.

[Fastracehorse]

That is the point I am trying to make. Nobody has absolutely accurate speed rating or else Andy Beyers would not have to adjust his. We can deal with inaccuracies by using other handicapping factors.

.....for many situations; but you can measure horses's accurately in the past in others. You are right when other factors come into play. It should be mentioned that speed figs can predict the future too, by projection, which leads to overlays.

And overlays being one of the benefits of speed figs in conjuction w/ the 'genetic predispostion' of the game.

fffastt

[jasperson]

A little ancient history. After returning to Texas after a stint in Cal where I had been making my speed ratings and track variants from the charts in the newspapers. In 1988 when bris came on line I started using them. At that time they were giving both their speed rating and Andy Beyer's. I was also making my own again out of the newspaper charts. I tracked Lad using all three sometimes my horse won sometimes it was bris and sometimes it was Andy and sometimes nobody had the winner. Almost always we had the same top three horses so there didn't seem much difference between us. I chose to use bris because I was tired of the work of making my own. Present day my race track crony uses equibase and again there isn't much difference when we compare speed ratings. I don't know who has the most accurate speed ratings, but I am comfortable using bris because speed rating is not the only factor that I use in selecting my bet. If there are inaccuracies in bris's speed ratings I don't sweat them.

[gm10]

There might a small % of outcomes affected by it (e.g. trainer placing the horse in the wrong race because of the figures he has for it), but I'm not sure that's what you meant.

[GameTheory]

I'm still wondering what the the original question (the one in the poll) is supposed to mean?

[rufus999]

The basic geometry and somewhat limited trig and calc of the Beyer speed method produces a high measure of accuracy. However, using reliable models to predict unreliable behavior is an exercise in futility.


rufus:9::9::9:

[rastajenk]

I voted for <10, because, as others have stated, the answer is zero.

[jasperson]

I'm still wondering what the the original question (the one in the poll) is supposed to mean?
How did it effect you in picking a winner. I should have been clearer. Sorry

[mistergee]

the figure makers may be very good, the horses however are inconsistent and are more often than not seperated by tenths of a second

[therussmeister]

Because I feel there is not much difference in win percentage using various speed figures, I think it is most important to use the least popular figures, so when the figures you use gives you a winner the other figures didn't have you get the best price. That is why I make my own figures using a methodology I developed myself designed to give me a different answer than Beyer's frequently enough.

When I get a winner that didn't figure using Beyer's figures I almost always get at least 5 - 1. The flip side is sometimes my figures really, really hates that 3 - 5 favorite that runs off by 10 lengths, and I wind up feeling like an idiot, but since I never give out my selections it's my little secret. ;)

[pondman]

The basic geometry and somewhat limited trig and calc of the Beyer speed method produces a high measure of accuracy. However, using reliable models to predict unreliable behavior is an exercise in futility.


rufus:9::9::9:

Any time you utilize gimmicky math to produce an average, par, or variant, you'll run into problems. It's similiar to predicting human behavior by analyzing a bell curve. If you can put enough of the knowns together, you might have a high enough win % to make a profit.

Hopefully, over time a person gains enough experience with the unknowns to recognize the outliers, the hot horses going off at a good price.

[TheGhostOfOscarB]

Until you have made figures for a few years, you can't possibly understand how tricky it is.

I made my own numbers for 20 years in the 70's- 90's in NY.

The big problem is dealing with shippers, its not possible to make figs for 10 tracks at once, unless you have hired help.

As for published figs, the numbers in Brisnet are decent enough to use, and the sheets are , well, good, but very pricey, though since they haven't changed the price in decades, it could be said they are cheaper than ever.