Predicting Speed Rating from Past Ratings

[TrifectaMike]

Originally Posted by Magister Ludi

Perhaps there is a cycle of equine performance consistency. W3 could be the minimum performance consistency point.

Bingo!!!

Mike

[TrifectaMike]

Originally Posted by PhantomOnTour

You are?
You have started some threads that, on the surface, seem pretty deep and not what I'd expect from someone new to this game.

I used a tale to make a point.

Mike

[ReplayRandall]

Originally Posted by TrifectaMike

I used a tale to make a point.

Mike

And your "point" is..............what?

[Greyfox]

Originally Posted by TrifectaMike

I used a tale to make a point.

Mike

You think that you are being Socratic in your approach on these subjects, educating as you go - like a sage educating the unwashed and illiterate peons.
This message board has a pretty large enrolment.
Math whiz that you are, there among that roster several that can run rings around you when it comes to picking and playing the track.
Except for those who laud your contributions, I don't think that you appreciate that.

Once again, me thinks, you should consider reading Eric Berne's Games People Play.

http://www.amazon.ca/Games-People-Pl.../dp/0345410033

[JJMartin]

Personally I don't look back further than the last 2 races for pace line and speed figures, however the max figure of the last 3 seems to work pretty well also. Running style can be pivotal when comparing tracks. The amount of variables that have to be taken into consideration precludes the efficacy of any single factor by itself. There are just too many varying conditions per individual race to accomplish a consistent model of reliable predictability, that is to say one that produces a profit with regularity.

[Cratos]

Originally Posted by Magister Ludi

Perhaps there is a cycle of equine performance consistency. W3 could be the minimum performance consistency point.

This is not a circularity problem as you are suggesting. It is a straightforward question of assigning weights to "speed ratings" with respect to the horse's past history and I am saying no with the present data (Trakus excluded) because distance is unknown in the so-called "speed ratings."

And you being an engineer should know that speed by definition is the ratio of distance/time. DISTANCE for the horse is not provided anywhere in the current PPs.

However if you want to do an analysis of time and weigh the metrics you should start with "pace" which would be the independent variable or the predictor of the final time or internal fractions by calculating the slope of the horse's running curve.

[ReplayRandall]

Originally Posted by JJMartin

Personally I don't look back further than the last 2 races for pace line and speed figures, however the max figure of the last 3 seems to work pretty well also. Running style can be pivotal when comparing tracks. The amount of variables that have to be taken into consideration precludes the efficacy of any single factor by itself. There are just too many varying conditions per individual race to accomplish a consistent model of reliable predictability, that is to say one that produces a profit with regularity.

I agree with you JJ, but there has to be a expiration date applied to any running lines, due to loss of potency. Mine are 180 days, and yours?

[JJMartin]

Originally Posted by ReplayRandall

I agree with you JJ, but there has to be a expiration date applied to any running lines, due to loss of potency. Mine are 180 days, and yours?

I like under 40 days but not always possible.

[ReplayRandall]

Originally Posted by JJMartin

I like under 40 days but not always possible.

When it comes to replays only, 35 days is the cut-off.

[Tom]

So how does one use this to handicap a race?
Is there an example, or aren't we done yet?

[classhandicapper]

Originally Posted by Tom

So how does one use this to handicap a race?

IMO, you shouldn't. That's what I alluded to before.

In my study, I controlled for every variable I could think of that causes figure fluctuations. The results I got (from an elite stats guru) on how to weight each race were very sensible and very productive.

The problem is that 95% of all horses have recent races with a mixture of distances, surfaces, track conditions, extreme trips, trainer changes, large class moves etc... Some are lightly raced and clearly moving forward and some have clearly seen better days and are on the way down. So if you used a formula like that to weigh their races in order to predict the future, there's almost no way you are on the right path.

You'd have to have a clear understanding of what you should do in each of those exceptional situations (which is practically every horse). I know I don't have that understanding. Instead, most people tend to eliminate recent unrepresentative races and instead focus on races they believe are representative of the horse's current form under conditions as similar as possible to today.

All that said, IMO, the controlled study was worth the effort because I have a better feel for how to weigh races in a general sense when I do have multiple races to consider.

[Greyfox]

Originally Posted by Tom

So how does one use this to handicap a race?
Is there an example, or aren't we done yet?

An example? That ain't gonna happen with Theoretical Mike....I mean Trifecta Mike.

[cj]

Originally Posted by Greyfox

An example? That ain't gonna happen with Theoretical Mike....I mean Trifecta Mike.

Yep...I was a fool for posting in this thread, but deleted it now.

[thaskalos]

Originally Posted by TrifectaMike

Let me ask a question, which appears to have an obvious answer.

I am fairly new to horse racing and have heard about the importance of speed ratings to predicting winners.

So, I think I've come up with a good idea.

"I'll take a horse's last four races and assign each speed rating a weight"

Last speed rating weighted by w1
Second back speed rating by w2
Third back speed rating by w3
Fourth back speed rating by w4

The question; how to assign values to the weights?

Some here have probably done something similar. Anything goes.

Mike

I have been doing something similar for the better part of the last 20 years. What I have found is that these speed ratings are very class, distance and pace-sensitive. You cannot, in my opinion, assess them just in accordance to the way they appear on the PP page. If the most recent speed rating is at the wrong class, or from the wrong distance...then it's insignificant. If the last race portrayed peculiar pace characteristics...then that speed rating too might be insignificant.

The disparity that we see in these speed ratings cannot be attributed solely to the inconsistency of the horse; the chief causes for this disparity are horse placement and race dynamics. It is my contention that speed ratings have very poor predictive value when they are looked upon without the benefit of corresponding pace ratings. It is also my contention that the newcomer to our game is asking for trouble if he focuses on the speed ratings alone when he first takes up the game. Speed ratings belong in the arsenal of the EXPERIENCED player...because he alone understands their limitations. To the newcomer...all these speed ratings provide is a key-hole view of a great big world.

[Robert Fischer]

Personally, I enjoyed the thread(and perhaps there will be more shared?).

TrifectaMike at the very least, presented an example.