Poll: How do you respond to the appearence of bias

[podonne]

Lets say you are two months into a meet at Random Major Track (RMT). As you handicap a race you see a single horse that exhibits RandomAttributeA. Over entiretly of the last meet, horses that exhibit this attribute won at a 12% rate, but from the records so far this meet, runners with this attribute are winning at a healthy 20% clip.

So do you A) favor this horse a little more because he exhibits an attribute that the track is obviously favoring, or B) find ill-favor with this horse because 20% so far means that the number is likely to regress to the mean (12%) so you can expect a probability of failures.

[raybo]

Lets say you are two months into a meet at Random Major Track (RMT). As you handicap a race you see a single horse that exhibits RandomAttributeA. Over entiretly of the last meet, horses that exhibit this attribute won at a 12% rate, but from the records so far this meet, runners with this attribute are winning at a healthy 20% clip.

So do you A) favor this horse a little more because he exhibits an attribute that the track is obviously favoring, or B) find ill-favor with this horse because 20% so far means that the number is likely to regress to the mean (12%) so you can expect a probability of failures.

Ride the horse 'til it quits. Anything that you can do that the public isn't, that is working, should be capitalized on.

[kenwoodallpromos]

Take advantage of the advantage.

[singunner]

What if RandomAttributeA is that the winning horse's name is 5 syllables? It depends entirely on the nature of the suspected bias, and said bias should be something that has historical precedent to back it up. If I found that Beyer Speed Figs were more profitable in April for the last 3 years, I wouldn't be betting like a mad man come April Fool's.

"Bias" to me just sounds like "theory". You theorize that a certain aspect of a track is giving better results to a certain group of horses. That's not saying that acting on a seemingly sound theory is a bad idea. We argue about relativity all the time, but it works so often that we don't hesitate to use it.

[46zilzal]

"Bias" to me just sounds like "theory". You theorize that a certain aspect of a track is giving better results to a certain group of horses. That's not saying that acting on a seemingly sound theory is a bad idea. We argue about relativity all the time, but it works so often that we don't hesitate to use it.
I used to belive that too until I began playing the East Coast tracks a lot, where there are more chances of rapidly changing weather patterns. The day Bellamy Road won the Wood was a classical example: everything that should be first at the first call did not stop.

It is not common, but it happens over and over again the same way: tracks get earlier at times way out of proportion to any other logic but bias.

[singunner]

That's what I meant by "historical precedent". If you have experience with the bias you think you're observing, there's no problem with acting on it.

[46zilzal]

That's what I meant by "historical precedent". If you have experience with the bias you think you're observing, there's no problem with acting on it.
you keep believing it does not exist, by my guest....... but every week or so energy distributions of winners tell me otherwise....OVER AND OVER AND OVER.

Funny, science tenants are based upon historical observations, just like this.

[singunner]

Somewhere along the way, you came to think I was arguing with you...

I only take offense with your improper use of the venerable colon and confusing "classical" with "classic", offenses that occur with such frequency in the general public that I scarcely recognize them anymore.

[46zilzal]

Somewhere along the way, you came to think I was arguing with you...

I only take offense with your improper use of the venerable colon and confusing "classical" with "classic", offenses that occur with such frequency in the general public that I scarcely recognize them anymore.

classical is used as an adjective not a noun in the context above.......
The adjective classical has 2 meanings:

Meaning #1: characteristic of the classical artistic and literary traditions

Meaning #2: adhering to established standards and principles THE LATTER WAS THE CONTEXT used.

or from Webster
of or relating to a form or system considered of first significance in earlier times <classical Mendelian genetics>

[singunner]

Would you care to give the link? Here's mine, with a million definitions, but I don't see yours:

Classical (http://dictionary.reference.com/browse/classical)

You wouldn't get in a fight with someone from UFC, why would you argue with someone who actually proofs their posts from time to time? You don't use proper capitalization or grammar, why are you defending your word choice? And the Webster quote you included backs up my claim, not yours. All I've got is your specious "Meaning #2:" with a single space after the colon.

Quintessential would have made sense and sounded better.

I think this is why you thought I was arguing with you to start.

[Valuist]

Saying that a bias is a "differentiation from the mean" is incorrect. A neutral track that produced a lot of similar numbered winners would be a differentiation from the mean. The mean, or expected results, become meaningless when a surface is truly biased because it is no longer a level playing field, (literally).

The guys I've meet who play this game for a living are ALL keen students of track bias. I have yet to hear of a full time professional horse bettor who doesn't believe in biases.

[Lefty]

Ride the bias till it throws you.

[46zilzal]

Would you care to give the link? Here's mine, with a million definitions, but I don't see yours:


just goes to show how DIFFERENT people use the language DIFFERENTLY when both are acceptable.

Don't be what Wayne Dyer describe as those people who MUST have things one way: MUSTERBATORS. I run into lots of them all the time. Like the colon there?

[bigmack]

Don't be what Wayne Dyer describe as those people who MUST have things one way: MUSTERBATORS. I run into lots of them all the time. Like the colon there?
Why does the fact that you're quoting Wayne Dyer explain so much about the voodoo that be you?

[podonne]

Man, you guys love arguing about language, reminds me of many boring days in college english.

So I think I am intrepreting this as that a bias is believable if it represents a sustained deviation from the entire population, rather than from itself. So a track is known to produce more front-running winners when compared to all other tracks, that's a bettable bias. A track has more front-runners so far this year that it did last year, that is not a usable bias (and should be bet against).

Am I close?

[Valuist]

Man, you guys love arguing about language, reminds me of many boring days in college english.

So I think I am intrepreting this as that a bias is believable if it represents a sustained deviation from the entire population, rather than from itself. So a track is known to produce more front-running winners when compared to all other tracks, that's a bettable bias. A track has more front-runners so far this year that it did last year, that is not a usable bias (and should be bet against).

Am I close?

Forget yearly numbers. Nowadays, when a bias comes up, it usually doesn't last an entire meet. Its important to differentiate the surface among individual days. Lets say you have a track that had a strong dead rail the first two weeks, after which point the track maintenance crew corrects it, and over-corrects it. The next two weeks, the rail is a paved highway. If you were to look at overall results, you may conclude the track had been neutral, when in fact it had been biased virtually every day; it just changed. Watch the races; where are the winners and runner-ups coming from? Were the results expected or head scratchers? If speed on the rail wins the first 4 races but it was all favorites, it may be wrong to assume there's a bias. I think the best situation is when you have a track that periodically gets biased; not every day. Tracks like the old Kee had such reputations that even casual players knew about the rail bias. Another factor to consider is wind. At a track like Arlington, mile races are run out of a chute. Its basically 5 furlongs due east, 1 furlong around the turn, and about 2 furlongs heading home (due west). On days with strong winds from the east, speed horses are going to struggle, regardless of the surface. Just a few things to keep in mind.

[podonne]

Forget yearly numbers. Nowadays, when a bias comes up, it usually doesn't last an entire meet. Its important to differentiate the surface among individual days. Lets say you have a track that had a strong dead rail the first two weeks, after which point the track maintenance crew corrects it, and over-corrects it. The next two weeks, the rail is a paved highway. If you were to look at overall results, you may conclude the track had been neutral, when in fact it had been biased virtually every day; it just changed. Watch the races; where are the winners and runner-ups coming from? Were the results expected or head scratchers? If speed on the rail wins the first 4 races but it was all favorites, it may be wrong to assume there's a bias. I think the best situation is when you have a track that periodically gets biased; not every day. Tracks like the old Kee had such reputations that even casual players knew about the rail bias. Another factor to consider is wind. At a track like Arlington, mile races are run out of a chute. Its basically 5 furlongs due east, 1 furlong around the turn, and about 2 furlongs heading home (due west). On days with strong winds from the east, speed horses are going to struggle, regardless of the surface. Just a few things to keep in mind.

So it sounds like you can help determine a true bias by relating the expected probability of winning for each horse (final odds) with the actual probability (the results), based on a subset of horses exhibiting a certain characteristic.

Also, I agree averaging can hide biases, but there are ways of testing for that situation stastically.