Quote:
Originally Posted by Dave Schwartz
I perused the system you guys are talking about here. Actually very interesting approach.
Please know that I am not saying it has a snowball's chance in the very deep south of winning, but it is interesting.
It is interesting enough to discuss a little. It would, logically, put the average handicapper onto some exciting races and the occasional big day.
The author is putting forth the concept of "segmentation." Essentially, he is suggesting that the handicapper divide the fields into groups for a reason: to separate the most likely winners from each other (i.e. into different groups).
The next step is to look for the "right group" by applying some rules that (indirectly) have to do with how competitive the group itself is. Doesn't really matter how you do this, but occasionally you will wind up on some lopsided groups, which just scream to fit the system.
See, I read (and often buy) a lot of... what I call, "idea systems," ... just to see if the author has stumbled across something buried in their that has value.
So, allow me to offer a different approach that is actually based upon some of the same principles but makes a little more sense.
This will NOT be a winning system, but you can have a good time with it and (probably do a lot better).
Step 1: Divide the field into 2 groups based upon your belief of the odds in the race.
Group A: All horses that figure to be below 7/2. I don't suggest that you use just morning line for this but you can if you wish.
Group B: Everyone else.
Step 2: Handicap each group like they are in completely different races. The goal is to pick the best horse in each group.
This puts you on two horses in each race, one low-priced, and one higher-priced.
Summary
You will cash a lot of tickets with the Group A horses that will keep you alive and then cash the occasional Group B horse.
There are a lot of bells-and-whistles one could add to this, like:
A. Building a model of what wins at this track in Group B.
B. Instead of splitting groups based upon predicted price, put front-runners in Group A and the rest in Group B. (Group B will always be "the rest.")
C. Get creative with the groups. See how many ways YOU can come up with group segmentation.
D. If this is too simple for you, try creating multiple sets of Groups in each race. You could have Group A be early horses, class horses (what ever that means to you), recent high speed ratings, etc.
Then you could model which GROUP is doing the best.
This approach actually has a lot of potential for winning if you add a few bells-and-whistles as discussed.
Finally, personally, I am trying to do more for the don't-want-to-do-much-work handicapper. I think I will spend some time expanding on this just for fun.
If an average player (i.e. not a winning player) takes any "segmentation" approach, he will open himself to having some pretty good races simply because sometimes he will find himself looking past the obvious horses in the race and playing the best-of-the-rest.
Hope this helps.
Dave
|
I think it's a good idea. I touched upon this recently, in another thread, when I said that your best chance of showing a profit basically comes down to handicapping the horses that appear to have less of a chance to win. What I mean is, if you're betting multi-race exotics, like the Pick 4, sure, you have to include some favorites or "obvious" contenders. But if you think you can make a profit betting to win, you're going to have to come up with winners that pay well. Which means that you're going to have to get good at picking the winners that don't look like obvious contenders. The first thing to do is find races where the obvious contenders don't look that good. In other words, there isn't a strong favorite.
Of course, this kind of fits into the old ABC rating of contenders that Steve Crist, Barry Meadow, and others have talked about, the ranking of contenders to use in exotic wagers.
The bottom line is, if you're not good at picking the winners that aren't that obvious, you're pretty much screwed and you should not be betting horses.
