On-Track Observation

[traynor]

Originally Posted by DeltaLover

Nice post..

I would add to:



that this 'difficulty' is magnified by the stochastic nature of the game which is pretty accurately reflected in the odds offered by the various pools...

Either we want to accept it or not, the take out in this game is so large, that it is quite possible to evaporate the edge of even the most astute horse bettor, as his competitors improve their abilities to handicap and play.

One of the best learning experiences of my life was realizing how sharp the bettors at Fraser Downs were(are) in relation to bettors elsewhere. The result was nearly doubling my ROI by the simple expedient of passing races in which my top choice was over 3/1. My ego can take it. My bank account loves it.

[traynor]

Originally Posted by Tom

Like it or not, they are horses - living creatures,not data points.

Of all things in handicapping, that is one of the most important to understand.

[traynor]

Originally Posted by thaskalos

And the process is so subtle, that most of us won't even know when that point is reached.

That is what makes it so interesting. No matter how good/accurate/profitable/whatever some approach may be at any given moment, it is only a matter of time (usually short) until others replicate it and drive the available profit down into the red.

The only way to survive is to stay ahead of the crowd. I like it like that.

[Poindexter]

Are there any services out there that do this type of analysis? In other words they are at the paddock of say Del Mar, they post notes on to their web site and a novice would then have access to it for a fee. Say $10 bucks a day. Similar to a clocker report. Seems like if someone was reallly good at observing these things and explaining them to the layperson, they could do well with such a service. Guys like me couldn't tell the difference between a horse who is dead lame and a horse who looks ready to win a grade 1. I would probably pay the $10 bucks a day and see if it helps me and continue on as long as it did help me.
If I start tossing all of my winners because of this service, I won't stay with them very long .

[traynor]

Originally Posted by Poindexter

Are there any services out there that do this type of analysis? In other words they are at the paddock of say Del Mar, they post notes on to their web site and a novice would then have access to it for a fee. Say $10 bucks a day. Similar to a clocker report. Seems like if someone was reallly good at observing these things and explaining them to the layperson, they could do well with such a service. Guys like me couldn't tell the difference between a horse who is dead lame and a horse who looks ready to win a grade 1. I would probably pay the $10 bucks a day and see if it helps me and continue on as long as it did help me.
If I start tossing all of my winners because of this service, I won't stay with them very long .

I think it is a lot like writing your own software (even relatively simple) as opposed to using someone else's. There is something added to the process when you realize that it is your money that will be wagered on the output/observations that is totally lacking in purchased services.

People (of widely varying skill levels--the caveat being the same as for software "if it so good, why does he/she need my money, rather than using it herself/himself?) have been marketing various visual analyses and trip notes for many years. I have never found one worth buying (other than to try it out). Whenever I try to take shortcuts and save time, it costs me money.

Don't cut yourself short. It isn't that difficult.

And that is not to say that there are no services out there that are worthwhile. It is just that I have never found any.

[eqitec]

I've experimented with systematic paddock observations and measurements somewhat according to the principles of The Body Language of Race Horses by Ainslie & Ledbetter. My efforts have included development of an iPad app for recording such observations at the paddock and immediate transmission of my observations via the track WiFi to my friends.

The image below shows a mocked up sample from the iPad app of a few horses from tomorrow's card at SAR.

Based on my experimentation, I can tell you that this is not easy to do. The time available for such observations is very limited, and one's view of the horses are badly obstructed most of the time, especially at SAR where the paddock layout and hordes of paddock insiders make it virtually impossible to see any given horse for more than a few seconds. DMR was a little easier due to the paddock layout. TAM & FL were the easiest of the other tracks where I've tested.

Consequently, I've had to restrict my observations to an exception-based system. Rather than trying to observe all horses for all body language features, I observe for only exceptionally good or bad body language features which standout in the few seconds when my view of the horses are unobstructed.

[traynor]

Originally Posted by eqitec

I've experimented with systematic paddock observations and measurements somewhat according to the principles of The Body Language of Race Horses by Ainslie & Ledbetter. My efforts have included development of an iPad app for recording such observations at the paddock and immediate transmission of my observations via the track WiFi to my friends.

The image below shows a mocked up sample from the iPad app of a few horses from tomorrow's card at SAR.

Based on my experimentation, I can tell you that this is not easy to do. The time available for such observations is very limited, and one's view of the horses are badly obstructed most of the time, especially at SAR where the paddock layout and hordes of paddock insiders make it virtually impossible to see any given horse for more than a few seconds. DMR was a little easier due to the paddock layout. TAM & FL were the easiest of the other tracks where I've tested.

Consequently, I've had to restrict my observations to an exception-based system. Rather than trying to observe all horses for all body language features, I observe for only exceptionally good or bad body language features which standout in the few seconds when my view of the horses are unobstructed.

The difficulty is identical to that experienced by those learning (and using) various coding schemas (for example, language coders in therapeutic, counselling, or conflict situations). Observations (other than simple yes/no, as in the case of bandages) are of phenomena that occurs on a range. Unless that range (and the various points on that range, or the precise yes/no point on each of those ranges) is explicit, repeatable, and exact, the observations go directly into the zone mentioned above regarding Korzybski: labeling something, then responding to the label rather than to the thing labelled.

For example, muscling is not a boolean factor, and "well-muscled" is an arbitrary, subjective evaluation that could be interpreted differently at different times. With Secretariat at one end of the spectrum and Eeyore at the other end, nearly all horses will fall somewhere in between.

[senortout]

You guys....you lost me at 'carpenter and nail'

[traynor]

Originally Posted by senortout

You guys....you lost me at 'carpenter and nail'

It is really simple. It only seems difficult or complex when people try to explain it.

Watch the opening of Boss (series). Watch the shot of the golfer. You don't need to know anything about golf, or how to swing a golf club--it is glaringly apparent (or should be) that it is a clumsy swing.

Observational handicapping is not much more complex than that. Same with trip handicapping.

[cashmachine]

Originally Posted by DeltaLover

Of course I can be wrong! The way to prove this (or any other similar) point is simple and this is what I am proposing...

We can easily keep a (public) data base with related, paddock inspection data and check their validity after we have gather enough of them.

The null hypothesis, is that this kind of a process will not have a significant impact to the bottom line..


Let's see if it holds true of false...

I am doing it several months and can report significant success with it. I watch videos of trackworks that happen couple days before the race; I grade horses on 8 dimensions. All grading happens strictly before actual race so in no way knowledge of race results can get into my grading. When I grade videos, I don't know odds, past history, jockey, trainer - I know absolutely nothing about the horse. Of course I remember some catchy names but I don't remember how those horses did in the past, at least not at a conscious level. After the race happens I enter results into my database and program automatically compute evaluation of my grades.

Some marks are absolutely useless; but 4 of them raise my ROI to about -0.1 (ROI of random bet is about -0.2). Again, I don't know odds of horses when I grade video. When I combine basic grades into combos, they become more powerful. One combo has ROI -0.01 (count is 198 cases out of 1860 observations). One combo has ROI +0.21 (count is 100 cases out of 1860 observations). Keep in mind that I compute ROI of place bets, so individual payouts are small and single large win cannot affect my ROI. I proved to myself beyond any reasonable doubt that video handicapping does add some value to handicapping process.

[traynor]

Originally Posted by cashmachine

I am doing it several months and can report significant success with it. I watch videos of trackworks that happen couple days before the race; I grade horses on 8 dimensions. All grading happens strictly before actual race so in no way knowledge of race results can get into my grading. When I grade videos, I don't know odds, past history, jockey, trainer - I know absolutely nothing about the horse. Of course I remember some catchy names but I don't remember how those horses did in the past, at least not at a conscious level. After the race happens I enter results into my database and program automatically compute evaluation of my grades.

Some marks are absolutely useless; but 4 of them raise my ROI to about -0.1 (ROI of random bet is about -0.2). Again, I don't know odds of horses when I grade video. When I combine basic grades into combos, they become more powerful. One combo has ROI -0.01 (count is 198 cases out of 1860 observations). One combo has ROI +0.21 (count is 100 cases out of 1860 observations). Keep in mind that I compute ROI of place bets, so individual payouts are small and single large win cannot affect my ROI. I proved to myself beyond any reasonable doubt that video handicapping does add some value to handicapping process.

A key point in your success (and the lack of success by others) may well be the process you use (bolded above). There is an almost overwheming desire in people to "be right." That desire corrupts observations to a greater extent than most are willing to admit. "Studying the past performances" and then looking at the horses in the paddock or warmup may VERY strongly bias the observer toward the favorite, a horse with a "big number," a horse with a top jockey/trainer, or some other critieria unrelated to physical observation. That bias very strongly affects what people see, and how they interpret what they see.

There is a long, involved explanation for all this that has to do with heuristics and ""preserving cognitive resources" that I will bypass.

The bottom line is that objective observation is mandatory. The process described above--because it does not seek validation for "decisions" already made--may be much more useful than other approaches.

[davew]

You need to see post parade and warm-up if you are going to bet any maiden races with first and second time starters. If you skip all of these races, I can see Deltas point.

[jk3521]

Speaking of observation, I observed Rotundo Jr. in the paddock yesterday. Unmistakeable w/ that bowtie ! Also all the horses in the Travers looked great in the paddock, but my eye was caught by #3, he looked fantastic ! Then I looked down on my Form and said no way! May be one to put in my Stable Mail for somewhere down the line.

[BIG49010]

RTN the last couple days have been broadcasting Saratoga in HD, it sure helps this topic.

[fmolf]

what i do is handicap the race narrow down the field then narrow it down further by paddock and post parade inspection.I then make my decision on who to bet based on these observations in conjunction with the tote odds.It is not very often at all that i will wager on a horse solely on appearances.For all the naysayers though any horseman will tell you a horse on his toes,with dappled coat ,a tail held outward off the rump and defined muscularity are all positive signs of a happy healthy racehorse.as an aside it is important to view horses as they enter and warm-up on the track as some, especially older horses do not exude any energy until closer to actual post time