Contender Selection/Eliminations

[BJennet]

Originally Posted by J2EEDeveloper

In the first stage of race analysis, the color designations for top 3 are based on APV and CR. That immediately points out entries that may be glaringly deficient in this area (45%, 50%). Are there any generic guidelines for eliminations/contender designations at this point, using only the APV and CR?

Individual races vary, of course, but I am wondering if anyone has done any serious testing of the APV and CR designations. For example, how often does an entry with APV and or CPR 50% or less actually win? Or 60% or less? And how much confusion is introduced in the readouts by including such an entry as a contender?

Information, insights, suggestions, and ideas will be appreciated.
Thanks

Hi J2,

The late Guy Wadsworth, who wrote most of the late Sartin software told me that 90% of all winners were in the top five of APV, in all classes, and although he presented no validating charts or evidence, this seems accurate to me as a ballpark figure.

Just anecdotally, I've found APV to be most useful at the high end, above GR. II, and the low end, below 10k - not incidentally the areas in which the various pace numbers, including TE, tend to be less useful. It's not unusual to get a stakes race exacta by simply boxing the top three APVs. BTW, this is true for winners only sub-10k, not place and show necessarily.

The main problem with APV, as FTL's useful chart indicates, is that it's a very obvious and overbet factor, with subsequently no wager value under most circumstances. As far as its utility in eliminating 'false positive' TEs with cheap speed, it's not as accurate as simply eyeballing the PPs. In general TE 'false positives' are more likely to fall into the category of 'herd' horses - those that generally run well but prefer to lose. You can recognize them by the preponderance of yellow and green in their PP finish positions.

Best of luck.

Cheers,

B Jennet

[J2EEDeveloper]

Originally Posted by For The Lead

You are correct, but not specifically in the east. This is true in state bred races in "any" state. The way to handle this is, if you are able to determine that a goodly portion of the horses earnings come from state bred racing, you can lower the APV by 20% as that is approximately how much the purses are inflated in state bred races.

That is a fair approximation that will work in most cases. The question is, how do you determine the proportions of state-bred to open races when some or most may be in older races (early in the year) and not show up in the latest PP lines? A similar situation exists with f-claimers; some have same or near-same purses as open, but are generally considered about 20% "less in competition" (or claiming price, etc.).
Thanks

[J2EEDeveloper]

Originally Posted by For The Lead

Very interesting. This is much like the "surrender option" in blackjack, where losing 50% on a hand that has a long term expectation of losing more than 50% actually increases profits. (This assumes one "knows" when to exercise the option correctly )

Exactly, with the caveat that the Surrender option is used appropriately. Unless you are counting, and know the various probabilities involved on a hand-by-hand basis, Surrender is a poor choice.
Thanks

[For The Lead]

Originally Posted by J2EEDeveloper

That is a fair approximation that will work in most cases. The question is, how do you determine the proportions of state-bred to open races when some or most may be in older races (early in the year) and not show up in the latest PP lines? A similar situation exists with f-claimers; some have same or near-same purses as open, but are generally considered about 20% "less in competition" (or claiming price, etc.).
Thanks

I agree. It's a tough call. As I said "IF" you can make the determination then you can make some adjustment, otherwise, you're kind of forced to go with what you have.

[For The Lead]

Originally Posted by J2EEDeveloper

Exactly, with the caveat that the Surrender option is used appropriately. Unless you are counting, and know the various probabilities involved on a hand-by-hand basis, Surrender is a poor choice.
Thanks

That's what i said.

[J2EEDeveloper]

Originally Posted by BJennet

Hi J2,

The late Guy Wadsworth, who wrote most of the late Sartin software told me that 90% of all winners were in the top five of APV, in all classes, and although he presented no validating charts or evidence, this seems accurate to me as a ballpark figure.

Just anecdotally, I've found APV to be most useful at the high end, above GR. II, and the low end, below 10k - not incidentally the areas in which the various pace numbers, including TE, tend to be less useful. It's not unusual to get a stakes race exacta by simply boxing the top three APVs. BTW, this is true for winners only sub-10k, not place and show necessarily.

The main problem with APV, as FTL's useful chart indicates, is that it's a very obvious and overbet factor, with subsequently no wager value under most circumstances. As far as its utility in eliminating 'false positive' TEs with cheap speed, it's not as accurate as simply eyeballing the PPs. In general TE 'false positives' are more likely to fall into the category of 'herd' horses - those that generally run well but prefer to lose. You can recognize them by the preponderance of yellow and green in their PP finish positions.

Best of luck.

Cheers,

B Jennet

I agree in general, with an added note. A drop in class to win is fairly common, especially at minor tracks. Despite a good race that makes it seem the entry is in top form, when it returns to its "normal" class level (where it is relatively safe from hostile claims because it is overvalued) it tends to run more poorly than its numbers from the lower race might indicate.

Despite its many failings, the APV can still be used as an indicator--even if what is indicated is a the necessity of taking a closer look at the PPs.

[J2EEDeveloper]

Originally Posted by For The Lead

That's what i said.

I understand that, but other readers may not. They may think that all you need to know are the relative (generic) probabilities that the dealer's hole card will be such-and-such, based on full deck calculations. Like most other things in life, it isn't quite that easy to know when to use the Surrender option.
Thanks

[BJennet]

Originally Posted by J2EEDeveloper

I understand that, but other readers may not. They may think that all you need to know are the relative (generic) probabilities that the dealer's hole card will be such-and-such, based on full deck calculations. Like most other things in life, it isn't quite that easy to know when to use the Surrender option.
Thanks

J2,

Not to let this thread drift too much, but just wanted to add a couple of comments for recreational blackjack players. I played the game professionally during an earlier period of my life and continue to derive some of my income from it. As both J2 and FTL mention, unless you're counting cards, you should never take either the surrender or insurance options, which are extremely valuable under the right conditions. However, if you are counting, it's a relative no-brainer to make either of these decisions - certainly compared with almost any decision that needs to be made in horserace handicapping, a much more difficult endeavor.

Cheers,

B Jennet

[BJennet]

Originally Posted by J2EEDeveloper

I agree in general, with an added note. A drop in class to win is fairly common, especially at minor tracks. Despite a good race that makes it seem the entry is in top form, when it returns to its "normal" class level (where it is relatively safe from hostile claims because it is overvalued) it tends to run more poorly than its numbers from the lower race might indicate.

Despite its many failings, the APV can still be used as an indicator--even if what is indicated is a the necessity of taking a closer look at the PPs.

J2,

I use it in generally the same way.

B.