PDA

View Full Version : Robot suggestion


BW SoCal
4th May 2009, 05:27.30 PM
Ken,

I like that on the jockey trainer page we can see how hot or cold the trainer is (the last 30 days). However, is there a way to see how hot or cold the jockey is? Perhaps by using the same 30 days or 10 and 30 days for jockeys? The 10 days makes sense to me since jockeys ride a lot more often than trainers have horses in races (although being able to catch a trainer early in a streak is not so bad).

Thanks,

Bernardo

km
5th May 2009, 12:37.02 AM
thanks for the post Bernardo

30day Trainer stat is part of our regular racefile structure, but not the Jockey for the same period. We would have to convince HDW to stick that stat in there for us - not an easy task right now.

But there is something you can do now. as the Robot and Export can quickly test or query jockey stats for any period- pretty quickly with the HTR2 Robot and more efficiently later with Robot II

Jockeys have a lot more starts than trainers for the most part, and are harder to pin down on hot and cold streaks without separating the data. If you run the Robot you might look at a normal overall win rate, but a terrible record with Turf or Fillies, for instance. The Learn All report is fascinating to review with individual jockeys and trainers. I used to run them and print in the newsletter - looking for strength and weakness with the HTR data - and they all have some.

BW SoCal
5th May 2009, 01:28.18 AM
Thanks for the reply Ken.

On Robot 2 will we look under the connections for this and more information in the future?

thanks,

Bernardo

tomcat
5th May 2009, 08:40.57 AM
Field size IMHO would be a nice added feature, but by the number of horses exactly might be misleading.
Maybe <8 or >9 might be enough.

DanG
5th May 2009, 09:36.55 AM
I’m not trying to steer anyone off research, because we all can reach different conclusions with similar data.

For what its worth; I put in Jky record in last 15 days (actually I tried many time periods) and to be honest it was very little help.

My brother did an exhaustive study by graphing riders and trainers and it did not yield any profits whatsoever.

Back to the riders; It was a classic regress to the mean scenario because most of them have such a narrow client base. Once a hot streak takes place / often that stock has run out of their conditions and can struggle over the next time frame.

As Ken said; all riders / trainers have their strengths and many Jock’s have lost their edge after a spill, agent change and the terrible personal problems that follow a profession where eating disorders and risking your life are part of your job description.

Unfortunately the way I looked at rider’s streaks it led to as many false positives as it did useful information.

Bob
5th May 2009, 09:51.40 AM
Dan,

Did anything at all regarding the Jocks have any trends?

For example: Jocks on Turf horses, Jocks on closers vs front runners? Post Position Jocks?

Regards,
Bob G

DanG
5th May 2009, 11:24.07 AM
Dan,

Did anything at all regarding the Jocks have any trends?

For example: Jocks on Turf horses, Jocks on closers vs front runners? Post Position Jocks?

Regards,
Bob G
Absolutely Bob; (Other then the excellent Jky ratings, TJ% and record with horse)

1. Jkch: Also huge with specific trainers.
2. Trk & Course
3. Post Position =1 and or extreme outside & RS
4. Age / Layzone & Class: FTS and some riders are just used to race horses back into shape and not pop off the layoff.
5. Race Shape (Pre-race) I use a pretty wild method of measuring probable pace, but the QT3 is a great gauge of what the connections see pre-race. I feel it’s a slightly misunderstood factor that just because a race doesn’t line up early the way we thought so it’s not a valid factor. Many riders / trainers have great variance because they change there strategy because of the "probable" pace scenario. Ramon Dominquez is a classic example of a rider when given a certain pre-race pace set-up…he is actually a flat bet profit or not even close.
6. Range of the K of course is huge: The tote is an obvious one, but the Kline and K are tremendous indicators pre-race indicators. Some riders just fall off the face of the earth below a certain range.
7. Top rider double back after a poor race: This is rider specific, but for some top riders it’s a lay-up.


There are many others, but I ALWAYS attach the HTR DS (dist / surf) to each factor unless the rider has a small sample, and then just surface. Mixing turf, dirt & Artf really is apples, oranges and kumquats.

Also; you really need to attach win expectation to make sense of studying individuals. The tote can be used and is the ultimate value indicator, but I also use what we know pre-race such as the K. If a rider is averaging X fin position, but his rank of K says he should be finishing in Y; it tells an entirely different story then just straight stats.

99.9% of riders have such mixed stock that it’s much more difficult to make sense of their numbers then a trainer. The trainer has such control over placement, intent etc…while the rider is basically left at the mercy of his clients.

Ultimately its no different then any line of work however; just like in a restaurant there are people more skilled for meats, salads, desserts and then there are the rare few who can command the entire line and those are the ‘Top chefs…or in our sport what super agent Ron Anderson calls…Garrett Gomez.

Bob
5th May 2009, 11:56.35 AM
Dan,

Thanks! As always with your in-depth, thought-provoking posts...now I want to play with RII rather than set up marketing for my restaurants for the month of June and July! :(

:D,
Bob G

DanG
5th May 2009, 01:01.49 PM
now I want to play with RII rather than set up marketing for my restaurants for the month of June and July! :(

:D,
Bob G
Lol Bob! :)
As my family would admit that's why I volunteered to bake at 4:00 am. Had to keep those Calder / GP and Hialeah afternoons free! :cool:

km
5th May 2009, 06:32.09 PM
Tomcat wrote =
Field size IMHO would be a nice added feature, but by the number of horses exactly might be misleading. Maybe <8 or >9 might be enough.

Already in there as a Range Filter in both Robots Tom, you can set the Field Size to any setting you want and the LEARN MORE report has a complete breakdown when you run a test.

tomcat
7th May 2009, 04:15.06 AM
Y'know, after I wrote that, it was in front of me in Maxvel....shishh!
I wish you quit getting ahead of us.:)