PDA

View Full Version : HTR Maturity Adjustment...


DanG
6th February 2007, 11:35.40 AM
Disclaimer…

[This is not tested in the traditional sense. It’s based on something I do and has proven effective under live fire for myself at least. Proceed with caution if you alter HTR ratings with this unscientific research]

It’s that time of year when the 3yo and even 4yo suddenly blossom and grow up seemingly overnight. HTR’s Pedigree rating is powerful at pointing out those with the genetics capable of improving and those that will hit the wall as distance / competition increases.

Ken writes often that “age” is a much underrated handicapping factor. Let’s look at a horse’s age vs. a human to get a feel for what were actually betting on.

Average life span:

Human (United States) = 77.2yo / Thoroughbred horse 24yo
(BTW: The human life expectancy is rising dramatically lately. I know HTR has added several years to mine! :) )

In terms of the human scale each year of a horses life = 3.2 years. There is a flaw with this metric…A horse can reproduce and 1 ½ years while a human can realistically reproduce at 13yo. (Note: I picked 13 because a couple years ago at my stepdaughter’s softball party one of the 13yo showed up pregnant. If THAT doesn’t give a parent the chills. :eek: )

So, considering a horse “matures” at a faster rate (approximately twice as fast) lets adjust the scale to each year of a horses life 2X or = 6.4 years.

This produces the following rough equivalency table…

Horse’s actual age: Human equivalent:
01 06
02 13
03 19
04 26
05 32
06 38
07 45
08 51
09 58
10 64
11 70
12 77


I admit these are major league ballpark estimations, but the spirit of the scale is relevant.

So, when we wager on a 2yo at Saratoga or Del Mar we are wagering on a 10yo athlete, give or take.

We then ask an animal to start competing in our triple crown at approximately 15yo. An entirely separate issue but I would LOVE to see the Triple Crown pushed back one more year and abolish 2yo racing. Won’t happen, but there is no law against dreaming.

What to do with this data?...Well, beyond the realization of just how young these animals are as they rapidly develop, we can approximate a “maturity adjustment” to use with various HTR ratings.

(Note: The cream of the crop improve at a much greater rate than the population as a whole. By the cream of the crop I mean the major HTR pedigree horses and the random freaks we see from time to time.)

HTR has several ratings that are on the same scale. (PER, PAC, EPR, SOR, PSCN and Class etc…)

I tried (Again, this is an approximation) to include a similar population in each age group. (I.e.…Sex, range of EPR, Mid – Major tracks only.)

Let’s take the average PSCN produced by each age group:

Horses Age: Avg. Winning PSCN:
02 90.7
03 93.8
04 95.5
05 96
06 96
07 95.7
08 95
09 94.3
10 92.6
11 92
12 91


As we can see the average horse levels off in ability at approximately 5-6 years old. (Or the equivalent of a 35 - 40yo human)

The majority of growth occurs between 2 and 3 with another significant spurt at 4.

This time of year we have many new 3 & 4yo coming off layoffs and we are basing some ratings off their past immature form. If we approximate the “average” expected growth of a normal horse, we can reasonably say for each 150 days a 4 yo. is laid up that a 1 pt maturity point can be rewarded. (Within these scaled HTR ratings, Cramer #’s etc require a different scale.)

“Each 150 days off = 1 point.”


In addition, if the animal has a monster pedigree, connections and / or raw ability this “maturity adjustment” can more than double.

The obvious drawback being, you are “rewarding” animals who have been laid up for various reasons. The usual lack of conditioning and reason for absence rules still apply. However, these “grow-up” plays as there called can produce some nice payoffs.

A very powerful angle is when a grass horse shows a significant jump in their 2yo grass races as compared to their dirt form. When as a 3yo they are debuted in a dirt / artificial race and exceed their prior “best” grass ratings they are very live when they switch back to the lawn. The reverse in surfaces is also true although not quite as dramatic in my experience.

(This obviously becomes far more powerful when HTR’s pedigree, workout and PPX information show all systems are go.)

Nothing written in stone here. Just something I find useful and I hope you do as well. “Lurkers” beware, this does apply to you…this technique needs HTR’s ratings to be most effective! :D

Zaf
6th February 2007, 03:08.52 PM
Very interesting thought provoking post Dan. Excellent !

Donnie
6th February 2007, 03:39.17 PM
Dan-
You have actually have raised another question in my mind....

How are new horseplayers being "trained" today? The age chart you listed above is one of the first things I learned years ago. We call them "babies" but in their life cycle, at age 2, they trully are babies! Sometimes I wonder if we should be racing even the 3 yr olds. As you stated, the average horse tends to peak between 4 and 5 and by age 6 they are on the downside of their racing career. Since you and Jim love baseball, I guess you could liken it to a ball player. He shows promise at the age of 15-18. Drafted into the leagues at 19-21. But by the time they hit their 30's they are usually on the downside of their careers, save for the freaks on the steriods, milkshakes or other enhancers. Do new players look at the age of the horse?

Kinda like the conditions of the race....somedays I think I am the only person in the joint actually reading them. A lot of horses can be tossed simply by holding them up to the conditions. Never thought to hold the template of age up to the other rankings found in HTR. Very interesting concept! Thanks for throwing that out on the board!

njcurveball
6th February 2007, 03:58.17 PM
Drafted into the leagues at 19-21. But by the time they hit their 30's they are usually on the downside of their careers, save for the freaks on the steriods, milkshakes or other inhancers.

Two words BRIEN TAYLOR! Worst draft pick in the history of baseball!

Even worse was that his rookie card which I think was a minor league or special one was going for $50 or more at one time and there were more buyers than sellers!

http://en.wikipedia.org/wiki/Brien_Taylor

Good point Donnie and nice post Dan!

DanG
6th February 2007, 04:00.00 PM
Thanks Zaf & Donnie…

Your right Donnie and it’s probably another discussion for a new thread concerning the changes in modern racing.

We all read the historical great thoroughbreds and how durable they were. We will never again see a “Citation” go 19 / 20 as a 3yo, which followed an 8 / 9 at two!

As you say, the “old” line of declining physical performance has changed somewhat in baseball. If the same “enhancers” are used on the youngest thoroughbreds then it stands to reason their career may be shortened.

As you said; not looked at by most and one of Ken’s recurring mantras is…”It’s the extreme stats that matter”

mr. posh
6th February 2007, 05:05.25 PM
Donnie And njcurveball If interested inbaseball cards, contact me at promosinc@msn.com.
Have good fifties card collection for sale.

Gambler
6th February 2007, 05:27.30 PM
Great post Dan. You really gave me something to think on. You too Donnie. I never really gave conditions much of a look. Maybe thats because Ive never taken the time to read on them. Hopefully I can use both of these in the future.

DanG
6th February 2007, 06:10.55 PM
Sidebar:

Obviously if the theory is valid in “grow up” plays…then the reverse is also true to a degree after a horse crosses the 5 to 6yo mark. An 8yo (John Henry aside) will show inevitable wear and tear. The further you go back in their form cycle the more severe the negative adjustment.

tbrown
6th February 2007, 06:18.45 PM
Very interesting read.
Thanks.

Tomcat talked about the grow up play at a seminar - Keenland? - a couple of years ago.

km
6th February 2007, 07:21.50 PM
Good thread, i have tested the age factor quite a bit in relation to the PER cycle.

= most horses run their lifetime top at age 4 or 5 (if they make it that far in their racing career)

= almost every thoroughbred, except the few that are injured early, will improve it's figures during its 3yr old season, or after its first 4 initial starts. In other words, a lightly raced horse is going to get better, often dramatically, so using past speed figures for comparison is not a profitable method of predicting outcomes with developing horses. (But it is a great way to find the chalk).

= the higher the PED, the more likely the horse will improve dramatically, the higher the PED the greater the likelyhood of improvement over time if there are changes such as dist-surf-class transitions.

= Once young horses have reached the point where they can handle regular 5.0f or longer workouts, it is a clear sign of coming improvement. The ability to drill successive 5.0f in good time = 1.00 or so (but often 1.01 or 1.02 depending on the track) is key indicator of a new PER top coming up and horse maturing. Unfortunately, many circuits are lax regarding reportage allow the works to go unreported and so the barn is the only source of that "inside information".

MVM
6th February 2007, 10:32.46 PM
A couple of things I do that probably seem strange:

1) For "established" sire ratings (4 crops to race or more) I use a formula that relates the FT rating to the PED rating for horses up to July of their 4 yo year.

2) The PED and SOR are very intertwined. For each range of PED ratings, there is a distribution of SOR ratings in which they are most effective.

Just a suggestion, but do some research relating the TRN, SOR, PER, PED and FT ratings.

I found some great stuff, but it is a bit abstract, not certain that it would be of interest.

DanG
6th February 2007, 10:50.01 PM
A couple of things I do that probably seem strange:

1) For "established" sire ratings (4 crops to race or more) I use a formula that relates the FT rating to the PED rating for horses up to July of their 4 yo year.

2) The PED and SOR are very intertwined. For each range of PED ratings, there is a distribution of SOR ratings in which they are most effective.

Just a suggestion, but do some research relating the TRN, SOR, PER, PED and FT ratings.

I found some great stuff, but it is a bit abstract, not certain that it would be of interest.
MVP:

Knowing your intellect you have discovered this long before I.

There is a fascinating interplay between the ratings you described that I have only recently have understood.

As you said, its not easily explained / documented but there are powerful implications that this is a cutting edge approach. A method completely lost on the public and that is more than we can ask for! :)

BTW: When you say “established sire ratings” are you basing that on foals produced, or outside data you import?

MVM
6th February 2007, 11:57.45 PM
Established Sire ratings>

After a few years the PED ratings (I meant PED not sire) become less volatile.

When a sire has few crops to race, the rating is very sensitive (take a look at what happened to Street Cry's D/R PED rating after the BC Juvenille).

This sensitivity is not necessarily a bad thing. For instance, HTR PED rating for Unusual Heat T/S was up in the 900s very early on, before everyone and their brother picked up on it.

As far as the interplay of rankings, there are nearly an infinite number of permutations.

I rarely if ever use the rankings of the ratings, rather I use the charactersitics of the distribution of that rating within a race, so for me there are something close to a billion possible scenarios to test.

I'm not a masochist, so I am not about to attempt that kind of research manually.

About 5 years ago I wrote a generic data miner application for work, and I used that to examine the HTR output.

Basically, the program defines a race object based on one or more sets of ratings, and then relates every other rating to all races of that "type".

I'm not a masochist, so I am not about to attempt that kind of research manually.

About 5 years ago I wrote a generic data miner application for work. The specific purpose was to relate medical conditions and their treatment regimens to outcomes. But the base logic is used for marketing research, or anything else where you "know what your looking for but don't know where it is". I used that (modified) program to examine the HTR output.

I know there are a lot of smart handicapper's out there, so I couldn't imagine that I would intuitively come up with some angle to research that had been somehow overlooked all these years, so I figured the best way to go about things was to let the program combine every single piece of data in the manner described above and crunch the numbers.

There are a million non-sensical "angles" that emerge from this process, but many are worthy of a second look, and utilize ratings in combinations that I never would have thought to combine.

BTW, a strong negative TV/IV angle that applies to a large number of entrants is almost as valuable as a positive one, and far more likely to go forward. My handicapping (prior to HTR)was based on exclusion, rather than the positive factor handicapping that makes up a major part of the $ and $$ logic, so HTR has been the perfect ying to my yang.

Ritz
7th February 2007, 12:07.35 AM
A stand alone caliber of physical maturity dependent on date of birth within the isolated ability of a horse amongst their peers is not an easy task. Age, within the realm of T-Breds, can be a nebulous thing. I'll consider the age of a contender in correlation with their running mates when the prime of their athletic life is in full blossom and those of the competition are under/over developed.

DanG
7th February 2007, 11:36.50 AM
Great answer!

Thanks MVP; as I said, there are some serious intellects in the HTR family!


BTW, a strong negative TV/IV angle that applies to a large number of entrants is almost as valuable as a positive one, and far more likely to go forward. My handicapping (prior to HTR)was based on exclusion, rather than the positive factor handicapping that makes up a major part of the $ and $$ logic, so HTR has been the perfect ying to my yang.
This is a very important statement.

Incredibly under valued by the masses. Virtually all handicappers are looking for positives in contenders vs. severe negatives. It’s a great approach that produces real value.

DanG
7th February 2007, 11:44.24 AM
A stand alone caliber of physical maturity dependent on date of birth within the isolated ability of a horse amongst their peers is not an easy task. Age, within the realm of T-Breds, can be a nebulous thing. I'll consider the age of a contender in correlation with their running mates when the prime of their athletic life is in full blossom and those of the competition are under/over developed.
Very true Ritz;

It’s not something you can easily universally apply. It almost becomes a “feel” thing with the more early 3yo seasons you experience. Ken I’ve found is excellent at anticipating the explosions in improving young horses. Mike Dee is also very adept at it.

Horse racing is one of the rare games that the player actually improves at the more mature they become and that’s a beautiful thing! :)