Bris prime power by gender..

[johnhenry81]

FYI, I manually have kept records on the #1 bris prime power selection separate for males and females (albeit a limited sample size of 620) I show males with an advantage over females 35% to 30%, which contradicts large batch data having the post time favorite fems at a higher rate than the boys. Does anyone have a larger sample for prime power?

jh

[DeltaLover]

Originally Posted by johnhenry81

FYI, I manually have kept records on the #1 bris prime power selection separate for males and females (albeit a limited sample size of 620) I show males with an advantage over females 35% to 30%, which contradicts large batch data having the post time favorite fems at a higher rate than the boys. Does anyone have a larger sample for prime power?

jh

Here you go:

...................................| Matches| Win%| ROI| IV|
----------------------------------------------------------------
top_pp_female............| .....6458| 32.50| 0.88| 2.67|
top_pp_male.............. | .....8432| 30.94| 0.85| 2.55|
top_prime_power........| ...14890| 31.62| 0.86| 2.60|

[acorn54]

deltalover is typical nowadays of serious horseplayers, we all have and some of us are still constantly datamining which imo is trying to find a needle in a haystack, as to a profitable kernel.
i said before, just a few short years ago horseplayers with database capabilities on their computers had a substantial advantage, but nowadays, with the drf subscriptions to the drf formulator available, to the public as well as various database software available, it is passe.

[TrifectaMike]

Originally Posted by acorn54

deltalover is typical nowadays of serious horseplayers, we all have and some of us are still constantly datamining which imo is trying to find a needle in a haystack, as to a profitable kernel.
i said before, just a few short years ago horseplayers with database capabilities on their computers had a substantial advantage, but nowadays, with the drf subscriptions to the drf formulator available, to the public as well as various database software available, it is passe.

More importantly, DL has a methodology in place, which allows him to assess how well (or not) the public treats the data derived factors as expressed by the odds.

Mike

[acorn54]

Originally Posted by TrifectaMike

More importantly, DL has a methodology in place, which allows him to assess how well (or not) the public treats the data derived factors as expressed by the odds.

Mike

you are giving a lesson in how to digest the data. yes it is true, i have found it useful to use the data to see how the betting public at large uses the data.
basically horse wagering has become a game of who can outfox who at the betting windows.

[johnhenry81]

Thanks DeltaLover, interesting that the fems are more predictable....

jh

[DeltaLover]

What I find interesting about handicapping factors, is how they are performing in conjunction with other basic factors. A very common break-down I ma using when analyzing the handicapping factors in per- race base (meaning customizing them as much as possible based in the specifics of the race)..

Take a look in this shot:






Here I am analyzing two factors:

Top Briss Prime Power + Female

And X (a factor that came up last week, while I was handicapping a specific race).

Note that I break down the stats per favoritism (favorite, less than natural odds, natural odds and more than natural odds) and keep track of WIN – PLACE -SHOW betting, reporting percentage, ROI, PNL (profit and loss based on $100 bets and expected value for a $100 betting unit..

Note that although for TOP_BPP+FEM never shows any profitability, factor X looks a little promising, if we limit our bets on win and natural odds... I find this type of macro-handicapping to reveal very interesting meta-handicapping situations and tendencies, having the potential for some profitability..

[DeltaLover]

As you can see in the example for factor X, its strike rating is similar to the global favourite when limiting our sample to all the horses who left the gate at natural odds..

Of course the sample is small (only 22 matches with 36% winners out of them) and the pessimist is going to say that sooner or later, this stellar performance will regress to the mean and the profits will eventually be evaporated...

In the vast majority of the cases, the pessimist is going to be proven correct, albeit there might be cases where the pattern will continue for enough time to show a good profit...

The challenge for the bettor who happens to be aware of these kind of anomalies, is to realize their degeneration timing and jump to something new...

[Appy]

"The challenge for the bettor who happens to be aware of these kind of anomalies, is to realize their degeneration timing and jump to something new..."

In my neck of the woods that's called "sharpening the needle".

[raybo]

Originally Posted by Appy

"The challenge for the bettor who happens to be aware of these kind of anomalies, is to realize their degeneration timing and jump to something new..."

In my neck of the woods that's called "sharpening the needle".

Timing is everything, just ask any serious market investor!

That is why my program doesn't just have one selection method, it has 10 static methods, plus 1 method that is variable and can be set to produce another 7 methods that are already available, and can be customized to produce an infinite number of other selection methods. Continuous track testing of these methods (the 10 static and the 1 variable (using whatever method is currently set)), with all 11 sets of data going to separate record keeping sheets (all 11 at once), allows us to see things happening negatively to some methods and positively to other methods, as things are changing. It is the ability to decide, early enough, to switch methods that allows some of us to perform better than others.