Jockey/Trainer win %

[raybo]

Here's a question for those of you who have database query capabilities. I have often wondered how viable it would be to add the jockey's win %(last 12 months) and the trainer's win %(last 12 months) and use that either as a stand alone grade for picking winners among true contenders or as a separator between closely graded true contenders. So, if any of you have done queries of this type or know of others who have checked out this particular scenario, your response would be greatly appreciated. Possible qualifiers for contenders could be top 5 Prime Power (Bris), or top 5 speed ratings within the past 3 or 4 races at or near the distance, or possibly top 5 post time odds horses.

[pmd62ndst]

I found that when comparing contenders that have similar speed ratings, here's how I prioritize trainer/jockey combinations:

1. High Trainer Win % / Low Jockey Win %
2. Low Trainer / Low Jockey
3. High Trainer/High Jockey
4. Low Trainer/High Jockey

Trainers are undervalued by the public while jockeys are overvalued. Mind you, this priority list is based on ROI and not Win %. High Trainer/High Jockey will give you the best Win Pct but it's a poor bet.

I'll try to get some numbers for you.

PMD

[pmd62ndst]

First, let's determine what a contender is. For this example, let's say a contender is the horse with the highest Beyer earned in it's last 3 starts and any horse within 5 Beyer points (1 length) of that mark within it's last 3 starts.

Querying the database gives us a good size sample of 93478 entries.

In this sample, Trainers with >= 20% YTD Win pct represent the top 1/5 of Trainers while <= 7% YTD Win pct represent the bottom fifth. Likewise, Jockeys with >=18% represent the top fifth while <=9% represent the bottom fifth.

Here are the findings:

1. Top Trainer + Bottom Jockey
Wins: 496/2314 (21.4%) ROI: - $209.50 (-4.5%)

2. Bottom Trainer + Bottom Jockey
Wins: 1045/7631 (13.7%) ROI: - $1582.00 (-10.4%)

3. Top Trainer + Top Jockey
Wins: 1843/6432 (28.7%) ROI: - $1830.00 (-14.2%)

4. Bottom Trainer + Top Jockey
Wins: 384/1830 (21.0%) ROI: -$602.80 (-16.5%)

On a side note, judging by the entry numbers, looks like top trainers like to team up with top jockeys and vice versa. Looks tough for a bottom Trainer to get a top jock.

This study is based on all thoroughbred races available from DRF from 2/1/2003 to yesterday.

PMD

[BIG RED]

How about changing it a bit. One year, with the DRF paper, I had almost every day of CAL. meets (exept FPLX ). Just to try something that was easy to find, I listed jockeys by highest wins in the race with trainers with highest win % . When they both cocurred, it came up positive, big. I'd be interested if you could do that. I'm sure, as usual, time will take its toll!

BIG RED

[raybo]

Thanks for your info PMD, it concurs pretty much with what I have thought all along, that a top 10 trainer is worth more, monetarily, than a top 10 jockey. However, since I play superfectas primarily, I'm not overly concerned with the ROI for winning horses, only that I get the winner. I adjust my play according to the expected lowest payout for any given superfecta wager. I currently select 41% winners with 1 selection but miss another decent percentage to my 2nd pick. I am trying to include more of those 2nd picks in my winner's selection using other criteria as further qualifiers or separators between my top 2 picks. I am wondering if, between 2, or possibly 3, contenders that I grade very nearly the same, would the one with the highest "trainer win% plus jockey win %" be a viable separator and steer me to the actual winner more often. Sometimes, of course, it is feasible and adviseable to put more than 1 horse on the win line, but the odds must support this play. There is a point of diminishing returns concerning the costs of superfecta tickets just as with any other type of wager. I have found that the most profitable ticket, in supers at least, is 1 selection for the win line. One can load up the other lines more cheaply with higher payouts often possible when low favored horses finish in the money. So, the key to higher profits, for the way I wager at least, is to hit more winners.

[shanta]

Originally posted by pmd62ndst
First, let's determine what a contender is. For this example, let's say a contender is the horse with the highest Beyer earned in it's last 3 starts and any horse within 5 Beyer points (1 length) of that mark within it's last 3 starts.

Querying the database gives us a good size sample of 93478 entries.

In this sample, Trainers with >= 20% YTD Win pct represent the top 1/5 of Trainers while <= 7% YTD Win pct represent the bottom fifth. Likewise, Jockeys with >=18% represent the top fifth while <=9% represent the bottom fifth.

Here are the findings:

1. Top Trainer + Bottom Jockey
Wins: 496/2314 (21.4%) ROI: - $209.50 (-4.5%)

2. Bottom Trainer + Bottom Jockey
Wins: 1045/7631 (13.7%) ROI: - $1582.00 (-10.4%)

3. Top Trainer + Top Jockey
Wins: 1843/6432 (28.7%) ROI: - $1830.00 (-14.2%)

4. Bottom Trainer + Top Jockey
Wins: 384/1830 (21.0%) ROI: -$602.80 (-16.5%)

On a side note, judging by the entry numbers, looks like top trainers like to team up with top jockeys and vice versa. Looks tough for a bottom Trainer to get a top jock.

This study is based on all thoroughbred races available from DRF from 2/1/2003 to yesterday.

PMD

thanx for sharing P.M.D. very enlightening info. you should post more often .you have a lot of good info that you have shared in your posts! you remind me of that old e.f. hutton commercial :"when you talk people listen"!
Richard Pizzicara

[sq764]

Being that I am a harness player, this may not apply to t-breds. But.. I think both the trainer and jock are overvalued. I think people undervalue the horse.

Sure, the great trainers can condition a horse better, can place them properly, have access to the better drugs, etc.. And the good jocks are typically smarter, can get more out of a horse, etc..

But the horse is the biggest part of the equation IMO. (Which is why when I do play t-breds, I only ever use Validator 2. It puts zero emphasis on trainer and jock and it performs wonderfully)

[hurrikane]

IMHO the trn and jk are both important. Can you beat the ponies without them. Sure. Can you beat the ponies with just them. Sure.

But the best IMO is to use both the horse and the connections.

[JackS]

We can't help but notice the jock/trainer stats in every race which often leads to public underlays. What about the lower ranked jocks who have rode the same horse to a win previously? The better trainers give us confidence that the horse might actually be fit enough to win since this is where trainer stats come from. A low ranked jock who has 50 races at this track with four wins and possibly two of those wins on the horse he's riding today, needs to be temporarily upgraded to top jockey status. Top jocks are consistantly hired to ride the better horses which continues to add to the mystique of "heck of a Jockey" and IMO little realization that the horse he's riding is also quite a bit above average for this race. I'm always looking for angles to beat anything that I think overbet.

[raybo]

<But the best IMO is to use both the horse and the connections>

This is my opinion also. I tend to believe that one could use the trainer/jockey stats as a confirmation for a top graded horse. When a horse grades best, at least with my program, he IS the best. The problem lies in his ability to put that "best" on the track today. Who knows the horse better than the trainer? And who gets instructions from the trainer? So, I will pursue this line of thought some more.
I would not attempt to use trainer/jockey stats alone as a selection method but combining them with a solid grading system seems to make sense.