Putting Beyers to work

[DigitalDownsJoe]

Ok so we know that the horse with the best figures is generally overbet and therefore you dont generally get much value there. So how do I find overlays??What patterns should I be looking for? I think a major step i need to take is figuring out what the figures are going to mean for this horse, with this jockey, on this surface, at this distance, with this many days of rest, with these types of workouts. There is so much info, but what do you all do to put it together, and make it work for you. I wanted if there were other outside factors which might also lead you to believe the horses figures will be lower or higher for this particular race??? Thanks :)

[pondman]

I wanted if there were other outside factors which might also lead you to believe the horses figures will be lower or higher for this particular race??? Thanks :)

I'd look at the thread "Average Beyer Improvement of winner". I think the author is in the right ballpark, but I don't agree with the speed philosophy much.

When a jockey has an inexperienced maiden runner and gets an early lead, they'll usually relax the horse, and if nothing runs at them, the jockey will hand ride the horse and not ask anything more from it. And so there are times when you'll see a dramatic improvement in it's follow up race. You'll see this in lower claimers on the West Coast. with a repeat win. I think it's sandbagging-- but if the horse wins anyway there isn't any problem.

[cj]

Buyers? Really?

[turninforhome10]

Gotta know how the number is earned. On the lead, come from behind or so on. Find 4 or 5 horses with the same beyer from different tracks and watch the races.

[OTM Al]

Buyers? Really?

They are awarded by the stewarts don't you know.....

[turninforhome10]

They are awarded by the stewarts don't you know.....
When you get the stewarts award, you must then get the key to the quarter pole.

[DigitalDownsJoe]

"beyers" my bad...I say that(REALLY??) to my kids all the time... I felt probably a little like they do when I say it to them lol OK now that we got that outta the way, care to make a response?? :)



Buyers? Really?

[Overlay]

Incorporate multiple fundamental (but non-redundant) factors in addition to speed (such as class, condition, and pace) in your handicapping, rather than just speed (or any other single factor) alone.

Instead of focusing exclusively on using the factors to isolate the one most probable winner through a process of elimination (which will generally lead to underlays, as you noted), compare all the horses in a field to each other by using surface- and distance-specific probabilities associated with measures of those factors (particularly measures that rank the field from top-to-bottom, rather than applying to just one specific horse) to develop a composite probability for each horse.

That probability can then be used as a basis of comparison with the actual odds or payoffs of horses or combinations to find overlays.

[dansan]

I never look at beyer numbers

[098poi]

Putting buyers to work

I thought this was a political thread.:faint:

[Steve 'StatMan']

Buyers? Really?

Yeah, I opened the thread thinking I was still in Off-Topic.

No worries DigitalDownsJoe!

[pondman]

Instead of focusing exclusively on using the factors to isolate the one most probable winner through a process of elimination (which will generally lead to underlays, as you noted), compare all the horses in a field to each other by using surface- and distance-specific probabilities associated with measures of those factors (particularly measures that rank the field from top-to-bottom, rather than applying to just one specific horse) to develop a composite probability for each horse.

That probability can then be used as a basis of comparison with the actual odds or payoffs of horses or combinations to find overlays.

Your are taking a real number under 1, normalizing it, scaling it somehow, into a ranking of 1 to 9(assuming this is a 9 horse field?)

Why would you want to do this?

The beyer is already a normalize number which includes surface, distance, and class built into it's product.

I'm not a beyers person. I don't even look at a horses performance much. I prefer the jockey fell off the horse during its last race.

But...Just think this is sending a person on a goose chase.

[Steve 'StatMan']

The beyer is already a normalize number which includes surface, distance, and class built into it's product.


Not supposed to be. Class isn't supposed to be in it, and while different adjustments are made for variants based on the distant of the race, and the numbers are "supposedly" an equivelent measure of performance at various distance and surfaces - if a horse performs good numbes at one distance but performs differently at another distance, they end up with differing figures.

Differnces can happen in the form cycle (go up or down depending how many races after a layoff or initial career starts. Different surfaces, better to use different figures, or project, ignore, use ofther information if the horse hasn't race at the surface or distance (breeding, trainer angles). Pace situations may be different - a horse may run better with an easy lead if he gets it, and not run back to a fig 'earned' with an easy lead.

Trip notes - some, but not all things that happen in a race get into the small comment at the right of the PPS, as well as the results chart.

Some races and figures, therefore, can give us a clue, some can be misleading, and sometimes it is best to either look farther back in the PPS, ignore them, project a difference or a range for today, etc.

I wish was far better at knowing when a horse was going to deprove in todays race. But these are key things to think about.

[Overlay]

Your are taking a real number under 1, normalizing it, scaling it somehow, into a ranking of 1 to 9(assuming this is a 9 horse field?)

Why would you want to do this?

The beyer is already a normalize number which includes surface, distance, and class built into it's product.

I'm not a beyers person. I don't even look at a horses performance much. I prefer the jockey fell off the horse during the race.

But...Just think this is sending a person on a goose chase.

I'm talking about rating various factors on a common scale, and combining them according to the principles of probability to produce a reliable composite value that reflects the relative strength of each factor with respect to different distances and surfaces, and that produces projected winning probabilities that allow wagering based on a comparison with actual odds.

With regard to your comment concerning how Beyers are developed, why, then, do Beyers exhibit varying predictive power with respect to races run at different distances and on different surfaces? Also, reliance on any one isolated handicapping factor (no matter how powerful or multi-faceted the factor may be) will sooner or later lead to loss of pari-mutuel value, even if the predictive ability of the factor may remain unchanged. That's why consideration of multiple factors is necessary.

[pondman]

Also, reliance on any one isolated handicapping factor (no matter how powerful or multi-faceted the factor may be) will sooner or later lead to loss of pari-mutuel value, even if the predictive ability of the factor may remain unchanged.

Only a few factors can be quantified, "time" being the current, major focus of the crowd. Most of the other factors can not be distributed at all. If I'm correct (which I know I am), and I say don't look at time at all because there are better methods, what other factors do you stat people include?

I think this is what the OP is asking...

I believe a horses performance will improve at certain (not in all) purse levels after a maiden win. I believe a horses performance will improve at certain (not in all) purse levels after being shipped...

Both of these statements are true, but complex and beyond contemporary math.

[Overlay]

Only a few factors can be quantified, "time" being the current, major focus of the crowd. Most of the other factors can not be distributed at all. If I'm correct (which I know I am), and I say don't look at time at all because there are better methods, what other factors do you stat people include?

I think this is what the OP is asking...

I believe a horses performance will improve at certain (not in all) purse levels after a maiden win. I believe a horses performance will improve at certain (not in all) purse levels after being shipped...

Both of these statements are true, but complex and beyond contemporary math.
I agree that time is a factor that deals by nature in numbers, and thus lends itself to quantitative measurement. But I think that other factors (such as the ones you mentioned), are also quantifiable through the use of tools such as impact values or multiple regression, and can be given proper numerical weight by themselves in the overall handicapping picture (as Quirin did in Winning at the Races, for example), without being considered only in light of the effect that they will have on a horse's time or speed figure today.

[cj]

Ok so we know that the horse with the best figures is generally overbet and therefore you dont generally get much value there. So how do I find overlays??What patterns should I be looking for? I think a major step i need to take is figuring out what the figures are going to mean for this horse, with this jockey, on this surface, at this distance, with this many days of rest, with these types of workouts. There is so much info, but what do you all do to put it together, and make it work for you. I wanted if there were other outside factors which might also lead you to believe the horses figures will be lower or higher for this particular race??? Thanks :)

Here are some I use:

Surface switches: If the horse is running well on a surface and switches, I'll assume his figures will regress. If the horse is one that should have some talent but is running poorly, I'll expect an improvement.

Stretch outs: If the horse is a winning front runner, I'll usually expect a decline. I also don't expect deep closers to improve going an extra turn.

[DigitalDownsJoe]

So generally horses who are middle runners to stalkers with some life in stretch will do generally better going an extra turn..? I would love to see some hard data about that..Horses going an extra turn and running style correlated into final results..Im sure it differs from track to track..