More About My Horse Race Calculator

[David Sklansky]

I'm not letting this subject go because I really think it could be very useful to those knowledgeable about racing and sellable to those who aren't.

It is important to realize that I wasn't suggesting that average times and SDs be a matter of simple calculation from past performance. Rather those parameters have to be expertly estimated for the specific circumstances.

If the horse is getting a better jockey, lower his time.

If it is getting a normally better trainer, lower his time but possibly increase his SD.

If he is moving to the turf for the first time raise his SD. Ditto if he might be involved in a speed duel. Or if you have little information to go by.

These are just a few examples of how the SHC should be used.

Another example is that it can help you decide whether a 40-1 shot is deserving of the third spot on a trifecta ticket.

The bottom line is that as long as you don't try to overuse the information the SHC puts out and make note of any consistent biases, you can't help but improve your results IF you know what you are doing.

[kdotsky]

This is very similar to the independent chip modeling calculator, especially used in sit and goes. Very easy to misuse, but very powerful if you know how to use it and what to take from it.

[David Sklansky]

Don't know what that is. Meanwhile I want to reiterate that The SHC should have the ability to input DEPENDENT events if one horse's relation to another is more clear than those horses relation to the field.

[Divad Yksnal]

Now that you have clarified things more about your gizmo I don't think you are crazy as I did when you posted the first time. I thought it was so silly that I assumed it was another one of your posts that aren't really about what they appear to be about. Perhaps like how easy it is to make a useless gadget appear useful if you market it properly.
[ QUOTE ]

It is important to realize that I wasn't suggesting that average times and SDs be a matter of simple calculation from past performance. Rather those parameters have to be expertly estimated for the specific circumstances.


[/ QUOTE ]

Ok, gotcha.

One thing that you seem to imply in the last thread, was that the exotic probables could be calculated with the information given. That is, knowing the probabilities the various horses will win. This goes to your dependance comments too. Which horse finishes second or third cannot be accurately extrapolated from win probabilities. (one of several problems with Ziemba) To illustrate, imagine a harness trot race where one horse is a near cinch, but breaks stride frequently. The second most likely winner starts from his outside and when the near cinch doesn't break, he is severely disadvantaged due to running styles. So he attains his wins/seconds/thirds mainly from those times the most likely winner breaks. So the exacta combo of the most likely winner onto the second most likely winner is not nearly as likely as the win probabilities would suggest. And this situation arises to a much less obvious degree in many races and if accurately assessed can be the source of very nice exotic prices.

You can solve the above by just having a different model for the place/show.


If you already addressed the above I either didn't read it or I don't understand what your technique is doing.

One thing that struck me immediately was that your gizmo could be helpful not just as a final time tool, but as a pace analysis tool. And I wouldn't be surprised if there was already software out there similar to what you proposed because the pace analysts are a precise bunch, using feet per second calculations, accounting for wind, path, humidity, temperature, weight, run up (the distance the horses run before the timer goes off), and even post race exertion (how much energy was expended just after the wire).

I might have more later since i do think the dependance of how horses run against each other isn't something that would be easy to do without your calculator. And it might help make pace analysis more accurate.

DY

[Divad Yksnal]

-I think your general technique would be more useful to help accurately assess the internal fractions of a race. To study pace, it is helpful to know when slow or fast fractions will aid or hinder certain runners. This is basic knowledge. The advanced application of it is quite involved, though, and software which sped this up would be sellable. This assumes it isn't already out there. There have been some very seriousminded pace practioners, some with strong math backgrounds.

-your technique would help with some other applications as well. I might get into that later.

-the concept of assigning probabilities to internal fractions is far from new. As a shortcut I used to assign probabilities to ranges, only getting interested when the race IMO shaped up in differently than the public would perceive. So in a crude sense this is similar to what you have proposed regarding final time. There wouldn't be any doubt my results would have been better with more sophisticated gizmos.

-to save time, which especially became necessary as multi tracking became available, I found that I could accurately project how a race would shape up *without even looking at time*. A method of assigning numerical values to running styles accomplished this and allowed more time for other more important factors. I could visualize how a race would run, seeing it in my mind's eye. with practice, I became very good at it, and when my certainty was high, I was usually right on. I would find this method difficult to translate into somehting your new advanced pace calculator could employ. but with effort it could be done.



-why gizmos, etc, lower EV in the long run? There was a time that crude speed figures such as Beyer's could beat the track. That changed a long time ago. In the beginning when I was a sohpisticated pace analyst, we enjoyed a tremendous edge. Both methods still allow profit, however, if they are excellently applied, but why not look for other less exploited avenues. Avenues that the general public and the influential handicappers/authors haven't touted?

-you can beat the track without ever assigning a time to your selections, never mind assigning probablities to those times and SDs. Yes, doing so should be possible, if you are indeed winning, but is it necessary?. IF i know a trainer wins in certain spot and it's one of those areas where the public doesn't have full access, I can assign him a probability of winning, and let the final time practioners scramble to figure out why they missed the bet. But even here one could work out estimates, but I don't think it would be worth it.

-the "poker objection" can't be fully applied to horse racing. Because in poker you MUST make a decision. If I am faced with a bet on the river, I can't just say, "pass". Whether I fold, raise, or call I could be making a mistake. There's no escape. In horseracing, I can just wait for a race where supreme accuracy isn't needed. But that isn't much of an objection. Having better accuracy can't really be a problem.

my post is boring crap David, so I hope someone is actually interested in this.

DY.