what the hell happened at the Big A

[46zilzal]

a MULTI-YEAR bias disappeared like a escaped convict overnight from Friday to Saturday. I have tracked this course EVERY CARD for 4 years now and it not only disappeared, but it changed POLARITY.

Ken. You know all about track maintenance, do you think they dug it up or brought in a lot of soil?

input appreciated

[Tom]

What changed?

I only palyed on erace yesterday - the first, and caught nice front end winner at 7-1. Are closers now winning?

[the little guy]

Personally I think the supposed speed bias on the inner has been basically a fantasy for the past ten years, however, this weekend has been ridiculous. It is almost impossible to pass.

[Tee]

To me it looks like a nice balance as to where the winners were coming from for the past 3 days.

Maybe it's just me :)

[the little guy]

I think if you dissect this weekend's races you will see it was a great advantage to being on the lead.

[shanta]

Personally I think the supposed speed bias on the inner has been basically a fantasy for the past ten years, however, this weekend has been ridiculous. It is almost impossible to pass.

Agree 100%.

Believe it or not the thread starter is complaining because "early" was NOT good on Saturday. Lookin at result charts thats kinda hard to grasp ha? lol

his definition of early is energy based rather than "positional" so thats the rub.

you want to see an early biased track in EVERY sense of the word??
PENN NATIONAL

Richie

[kenwoodallpromos]

Track kept its small inside post bias; total 1 workout Friday; very slow works earlier, sealed track works Saturday; Saturday results say "sealed" for the first several races, reflected in several races.
Since I claim NY to have among the best mainenance crews, even for off-tracks, and they try to make it safe and fair, I say they were working hard to keep the track raceable and had no choice but to super-seal it. The best guess of whoever was writing the track condition during the day must have assumed the track would lose it seal (speed) during the day. But it did not, even though "sealed" is not listed in early races.
Sand- I am not aware of how dumping extra sand on the track during wet weather would be easy to do while racing every day, or how a deeper track would speed it up like crazy. Steam rolling the track would do that!

[46zilzal]

T
Sand- I am not aware of how dumping extra sand on the track during wet weather would be easy to do while racing every day, or how a deeper track would speed it up like crazy. Steam rolling the track would do that!

That's what I am saying: it slowed tremendously

[46zilzal]

Believe it or not the thread starter is complaining because "early" was NOT good on Saturday.

his definition of early is energy based rather than "positional" so thats the rub.


I continue to be amazed at the clairvoyants of the world. Since when does asking an opinion qualify as complaining? One learns to adapt parimutually, I was just wondering at the cause that sprinters with multiple, significantly minus e/l's are winning for the FIRST TIME IN FOUR YEARS. Wouldn't that send up a red flag to you and your records? This was an observation noted by no less than four other people who use the same records as myself and reported independently. It is not an insignificant shift.

[nomadpat]

46, why didn't you ask that question originally? How in the world are people supposed to know that you were wondering "I was just wondering at the cause that sprinters with multiple, significantly minus e/l's are winning for the FIRST TIME IN FOUR YEARS." if you didn't state that? :confused:

[toetoe]

Zilz,

That's a great take on it, and I'm glad to get the heads-up. I didn't see much as my computer crashed about a hundred times. I was lucky to get in my losing pick-four entry. Did I mention I hated the Atreo Persaud horse?

One thing I see, which may or may not be parimutuelly useful, is that on some days, pressers take their best crack but, unlike "average" days, have little shot, whereas Silky Sullivan types have at least a prayer, like that John Morrison horse in the nightcap.

[the little guy]

You mean the John Morrison horse in the nightcap that Jan Rushton said got a brief freshening after breaking its maiden in November?


November 2003.

[toetoe]

Le Petit Guy,

Let's not Rushton judgement. :D I pray you're as fond of Suzy Walder (Waldman?) as I am. The sports screamers hate her, but as the greatest hetero fan of musical theatre, I gotta love her, if I may Garagiolize. I gotta believe, Tony, that it could be much, much worse. How? Two words --- Joanne Jones. :eek:

[shanta]

46, why didn't you ask that question originally? How in the world are people supposed to know that you were wondering "I was just wondering at the cause that sprinters with multiple, significantly minus e/l's are winning for the FIRST TIME IN FOUR YEARS." if you didn't state that? :confused:

you know only IF you have magical powers :lol: :rolleyes:

Richie

[cj]

Just so I can understand, what would a 20 mile per hour wind either up or down the backstretch do to these calculations? How is this adjusted for with energy or median?

For example, same horse might run this, and actually be the exact same performance:

+ Wind :45 1:10
0 Wind :46 1:10
- Wind :47 1:10

This is very common in New York. The energy numbers would be wildly different if there in no pace variant, even if the performance was very similar. How do you fix it?

[46zilzal]

adapt during the card to what is, rather than what should be

[Valuist]

Was it a speed bias or a rail bias? Somebody mentioned the closer in the last race; he was inside for most of the race and only came out late. I do think some horses got leads w/comfortable paces. After scratches, Jet Prospector became lone speed. All Hail Stormy also went kind of slow, and he still almost got passed by a horse closing on the rail.

[cj]

adapt during the card to what is, rather than what should be

I don't have a clue what you are trying to say, but I'm kind of slow.

If the hypothetical horse below showed up with those PP lines, how do you interpret them? The %E or %M ratings would be very different, yet could reflect the essentially exact same performance. Does Sartin use a separate pace variant to handle it? If not, how does this problem get fixed?

[PaceAdvantage]

46, do you think you could post some real time selections one day, to demonstrate how you adapt during the card to what is, rather than what should be? I think this would be very enlightening, and I for one would appreciate the effort.

[keilan]

46 – you don’t have to provide anyone with selections in real time, you have rights here. Redboard all you like, heck I myself hit a 87-1 just yesterday (the old grey in the mud angle). Tri paid 128k to the other two grey steeds. I wouldn’t call it a “gimme” but I’d have to say it was the ONLY wager if you’ve been following along. :lol: :lol: :lol:

[GaryG]

46 – you don’t have to provide anyone with selections in real time, you have rights here. Redboard all you like, heck I myself hit a 87-1 just yesterday (the old grey in the mud angle). Tri paid 128k to the other two grey steeds. I wouldn’t call it a “gimme” but I’d have to say it was the ONLY wager if you’ve been following along.Now THAT is a great post...ROTFLMAO! :lol: :lol: :lol:

[46zilzal]

I don't have a clue what you are trying to say, but I'm kind of slow.

If the hypothetical horse below showed up with those PP lines, how do you interpret them? The %E or %M ratings would be very different, yet could reflect the essentially exact same performance. Does Sartin use a separate pace variant to handle it? If not, how does this problem get fixed?
To go into a race without ANY pre-conceived idea how the track is playing and only responding to what bias, IF ANY, one sees TODAY.

I now have no idea what you are talking about. THOSE pp lines?????? When % median varies they DO NOT represent the same performances..

You will have to help me understand the question before I can answer it.

[46zilzal]

46 – you don’t have to provide anyone with selections in real time, you have rights here. Redboard all you like, heck I myself hit a 87-1 just yesterday (the old grey in the mud angle). Tri paid 128k to the other two grey steeds. I wouldn’t call it a “gimme” but I’d have to say it was the ONLY wager if you’ve been following along. :
My ONLY adversary in the parimutuel NO ONE else. I have never had the need to post selections, not now, not ever.

[46zilzal]

When there are certain characteristics of a track which various people have noted for years and years, then overnight those chararcteristics CHANGE dramatically, one wonders why. Having a numerical way of understanding how a bias presents itself helps a lot in adapting when it changes. As I mentioned, I was playing on line with two other people who use the same software as I do and they noticed EXACTLY the same thing and commented on it independently of my input. It would be one thing if I noticed it myself, but these other players are very experienced and very adept at the nuances of the program.

I was simply interested to know what factors might account for it. Plain and simple

[PaceAdvantage]

Hey 46, why is it that folks are more often than not confused by what you say, in that, they have no clue what you are talking about? Does this communication style of yours come naturally, or are you putting us on?

[46zilzal]

not putting anyone on

[traynor]

46zilzal wrote: <When there are certain characteristics of a track which various people have noted for years and years, then overnight those chararcteristics CHANGE dramatically, one wonders why.>

It's called randomness, an affliction that is prevalent in the real world. Models are averages, which are some up, some down, some in the middle. Models are not predictive--they represent historical trends, not future constraints. The very best a model can do is suggest a trend.

Without going all philosophical, you might be interested in searching Tversky's Law of Small Numbers and others on the fallacies in decision making that suppose short range results will replicate long range models. That is not meant to be insulting or demeaning in any way; the paper is a literate explanation of common fallacies that negatively influence decision making (a euphemism for "making losing bets").
Good Luck

[46zilzal]

I understand all that. I embrace randomness and quote that all the time. I just wonder if there was a logical explanation that's all.

[twindouble]

I understand all that. I embrace randomness and quote that all the time. I just wonder if there was a logical explanation that's all.



It's called randomness, an affliction that is prevalent in the real world. Models are averages, which are some up, some down, some in the middle. Models are not predictive--they represent historical trends, not future constraints. The very best a model can do is suggest a trend. quote; traynor.

I like to think in terms of prevailing conditions in horse racing not models,trends or randomness. That means your opening your mind to a new race, not relying on the above mentioned models or randomness.

T.D.

[kenwoodallpromos]

More soil; unusual maintenance; sealing; rain; freezing; high winds; Deeper harrowing of part of the track.
Those are the main (human and weather or a combination) conditions that would make a track change dramatically overnight.
As I said, sealing and heavy rain is what I believe caused what you and the others found. A couple of years ago I found the most profound daily changes were at the Il tracks.

[cj]

OK, let me try to be crystal clear and see if I can't get an answer, though somehow I doubt that it is coming anytime soon.

If Horse Zilzal runs a race on a track with a constant speed and no wind, let's assume he run in :22, :45, 1:10

His TE would be 60.00 + 57.39 + 52.80 = 170.19.

His Percent Median would be (60.00 + 57.39) / 170.19 = 68.98

Now, let's assume this horse had run the same day, the same exact ability, but into a strong headwind on the backstrectch. Nothing has changed except the wind. He now runs in :23, :46, 1:10

His TE would be 57.39 + 57.39 + 55.00 = 169.78

His Percent Median would be (57.39 + 57.39) / 169.78 = 67.60

Finally, let's then assume this horse had run the same day, the exact same ability, but with a strong tailwind down the backstretch. Nothing has changed except the wind. He now runs :21, :44, 1:10

His TE would be 62.86 + 57.39 + 50.77 = 171.02

His Percent Median would be (62.86 + 57.39) / 171.02 = 70.31

This is why I ask if they do a separate pace variant, which I've since been told by others they do not. If absolutely nothing had changed but the wind direction, you have three vastly different %M ratings, 68.98, 67.60, and 70.31. So I ask, how can this be reliable if they don't account for something as common as wind?

[GameTheory]

OK, let me try to be crystal clear and see if I can't get an answer, though somehow I doubt that it is coming anytime soon.

If Horse Zilzal runs a race on a track with a constant speed and no wind, let's assume he run in :22, :45, 1:10

His TE would be 60.00 + 57.39 + 52.80 = 170.19.

His Percent Median would be (60.00 + 57.39) / 170.19 = 68.98

Now, let's assume this horse had run the same day, the same exact ability, but into a strong headwind on the backstrectch. Nothing has changed except the wind. He now runs in :23, :46, 1:10

His TE would be 57.39 + 57.39 + 55.00 = 169.78

His Percent Median would be (57.39 + 57.39) / 169.78 = 67.60

Finally, let's then assume this horse had run the same day, the exact same ability, but with a strong tailwind down the backstretch. Nothing has changed except the wind. He now runs :21, :44, 1:10

His TE would be 62.86 + 57.39 + 50.77 = 171.02

His Percent Median would be (62.86 + 57.39) / 171.02 = 70.31

This is why I ask if they do a separate pace variant, which I've since been told by others they do not. If absolutely nothing had changed but the wind direction, you have three vastly different %M ratings, 68.98, 67.60, and 70.31. So I ask, how can this be reliable if they don't account for something as common as wind?
But is that really how the wind affects a horse? Will his final time really remain unchanged in all three scenarios?

[cj]

But is that really how the wind affects a horse? Will his final time really remain unchanged in all three scenarios?

Exactly, no, but I think you are smart enough to get the gyst of my argument. It is close enough to get the point across.

[GameTheory]

Exactly, no, but I think you are smart enough to get the gyst of my argument. It is close enough to get the point across.I get the argument, but doesn't its validity depend on the assumption that the wind doesn't affect the horse's "true" performance? With an additional assumption on top of that that if the performance does in fact remain essentially the same except for a different energy distribution caused by the wind that such change in energy doesn't matter and should be corrected for?

The second assumption seems somewhat reasonable given the first, but is the first valid?

[cj]

I think it is valid. At a track like Belmont, the horses run nearly the same exact distance both into and with the wind. All tracks aren't like this, most are slanted towards more run on the backstretch, so you wouldn't get exact most times. Speed figures tend to remain the same regardless of wind, so I'd say it is valid. I was just trying to keep it simple.

[46zilzal]

short answer is yes, a variant is applied to the raw data

[cj]

I know a final time variant is applied, but what about a separate fractional variant. That is the real question, I've been told no, maybe you know differently.

[46zilzal]

I know a final time variant is applied, but what about a separate fractional variant. That is the real question, I've been told no, maybe you know differently.
a separte one to my understanding

[twindouble]

I get the argument, but doesn't its validity depend on the assumption that the wind doesn't affect the horse's "true" performance? With an additional assumption on top of that that if the performance does in fact remain essentially the same except for a different energy distribution caused by the wind that such change in energy doesn't matter and should be corrected for?

The second assumption seems somewhat reasonable given the first, but is the first valid?

My assumption is, if your praying the wind will get your horse home, you bet the wrong one.


T.D.

[tahoesid]

Seems to me that is why the game is so hard. Things are always changing in the short term, even if they seem to be the same in the long term. That's why there is no way to figure out 100% of the playable races and you have to make do with a certain percentage of winners. It seems to zag when you are zigging and zigging when you are zagging. Always in a reactive mode.

[classhandicapper]

I think the biggest thing to take from these discussions of wind is that it effects the fractions and final time. More importantly, it effects them unequally not only from day to day, but from race to race because the wind is not constant in direction or intensity.

This is just one complication among many in trying to get an accurate appraisal of performance numerically. There are others, like changes in the run up, track maintenance during the day, changes of moisture in the surface, track biases, etc...

People usually combine a pace estimate with a final time estimate using a estimate of how the pace impacted the final time etc... There's obviously a lot of room for error in there.

Yet people bet off these numbers as if small differences have something to do with reality. Some even project big moves forward or backwards off small fluctuations in patterns of these figures.

The numbers (whatever set you use) are indespensible, but sometimes I think it's important to think clearly about level of accuracy we are working with.

It makes it a lot easier to bet horses that looks a little slower when the price is a lot higher, because for all you know, you may actually be betting the faster horse.

IMO, this is also why comparative and class style handicapping still should play an important role. It gives you another view of the horse's ability when you look at the quality of the competition, the trip relative to other horses in the race and how well he did vs. them. It's a view that bypasses all the challenges of making numbers when done well. Of course it presents other challenges, but IMO two views is better than one.

[twindouble]

Seems to me that is why the game is so hard. Things are always changing in the short term, even if they seem to be the same in the long term. That's why there is no way to figure out 100% of the playable races and you have to make do with a certain percentage of winners. It seems to zag when you are zigging and zigging when you are zagging. Always in a reactive mode.

In my opinion these guys are stretching the envolope with their caculations, to the point where the graft they create is suposed to lay over the race to predict the out come. Good luck!

The best way to determine if the track condition has effected the race is to look at the horses Past Performances. If the horse has no PP's and wins the race, your betting the wrong race because they are more than likely young horses that are learning to race. (risky bets) and nothing to compare with. When there is past performances to look at, if a horse run faster fractions than his best, you can be assured the track has a speed bias, by the same token horses that run below par time wise, that's the clue to look for. On a fair track, the pace is the determining factor time wise, those split times will tell the story along with the final time, when compared to past performances.

Aways keep in mind, young horses on the improve will crank out better times and it's good to take note when it's done on an off or deep track. You may be looking at something special. I still use the DRF speed and track variant but I don't lean on it, like any other factor.



Good luck,

T.D.

[traynor]

46zilzal wrote: <I understand all that. I embrace randomness and quote that all the time. I just wonder if there was a logical explanation that's all.>

Change is in the best interest of the track. People bet more when there is a chance for randomness, rather the (more reasonable) predictability. Kusyshyn did a big study of it as a consultant for track management. How do they flip a track? With a tractor in normal maintenance. Set the blades to dig a bit deeper, everything slows, and vice versa. "NYRA hates biases and predictability" is a reasonable assumption, based on manipulations of track surface.

Surface differences seem a more reasonable explanation than more esoteric ones (wind velocity, etc.).
Good Luck

[46zilzal]

It could be that rare situation where head comes up 25 times in a row. Poisson distribution or something akin to that

[cj]

I've done some research, and they don't use a pace variant, just a fraction of the final variant. This is worse than using no pace variant at all in my opinion.

[classhandicapper]

I've done some research, and they don't use a pace variant, just a fraction of the final variant. This is worse than using no pace variant at all in my opinion.

I think those guys do some brilliant things but I could never understand neglecting making accurate track variants for all calls etc..... It's one of the reasons I never put a lot of energy into that area. I was pretty sure that even when they had the right horse, it was often for the wrong reason.

[46zilzal]

I've done some research, and they don't use a pace variant, just a fraction of the final variant. This is worse than using no pace variant at all in my opinion.
who is they?

[46zilzal]

variant is applied to intermediate times as well, as it should be since it occured the same day

[cj]

The Sartin programs. Applying the same variant to fractions recorded over different parts of the track is asking for trouble. Wind is just one thing that can make the variants vastly different over different portions of the surface.

[46zilzal]

they are not the same

[cj]

they are not the same

I think you are mistaken, but we can just agree to disagree. Not important really.

[46zilzal]

I think you are mistaken, but we can just agree to disagree. Not important really.
well one of our members has the ORIGINAL code of the program and he explained how it was applied, so I have to disagree with you.

Years ago when I did my own pace numbers I saw that same problem with applying a variant to intermediates and I just made TWO variants.

[cj]

By all means, explain then. Maybe I'm getting bad info.

[46zilzal]

the programs use a proportional variant. That's all I am allowed to tell you as it is copyrighted and MY NAME is not on it.

[cj]

Yes that is what I was told. Which means, for example, if the variant for the day was 12 fast for 6f, you would assign 4 fast to the 1/4, 8 fast to the half, and 12 fast to the finish. This is not the same as making a separate variant. It is worse than not making one at all in my opinion.

One definite advantage of making your own program, you know exactly what goes into it, and its strenghts and weaknesses, which all man made numbers have.

[46zilzal]

NOT strictly PROPORTIONAL. There are some kinks there that are not in the public domain

[cj]

There is but one variant in the file. Regardless of how it is proportioned, it is done the same for races of the same distance, regardless of the wind. Which brings us back about 35 posts ago.

Its best to just let it go, unless you are saying I'm wrong about the variant. I've seen the files, nearly certain I am not.

[46zilzal]

there are MANY variations on the basic idea that handle it differently, your having see the files or not. I am currently testing yet another variation on the program and have about 6 different ones here.

FILES are one thing, readouts are another thing altogether

[PaceAdvantage]

So 46, for now (exclude the programs you are currently 'testing'), we can safely say that there is but ONE variant (a final time variant) and it is applied in some way or proportion to the first two fractions. This is the way the Sartinistas used to advocate dealing with the variant....

BTW, are they still using the DRF variant (with the exception of the program you are currently testing).

[GameTheory]

Yes that is what I was told. Which means, for example, if the variant for the day was 12 fast for 6f, you would assign 4 fast to the 1/4, 8 fast to the half, and 12 fast to the finish. This is not the same as making a separate variant. It is worse than not making one at all in my opinion.

One definite advantage of making your own program, you know exactly what goes into it, and its strenghts and weaknesses, which all man made numbers have.I use a proportional variant (I'm unsophisticated), but by time, not distance. So the adjusted pace times remain in the same proportion as the original times.

Example:

Raw times:

22.0 45.0 72.0

Adjusted final time after applying our single variant:

74.0

Adjusted pace times:

22.0 / 72.0 = .3055 (original ratio)
45.0 / 72.0 = .625

.3055 * 74.0 = 22.61 (maintaining the original ratio to the new final time)
.625 * 74.0 = 46.25

Adjusted times:

22.61 46.25 74.0



Now, in this case it is pretty close to the same result as if you did it by distance, but in more extreme scenarios and longer races the difference is more pronounced. You feel this is worse than no variant at all? This way I preserve the original energy distribution. We must remember that variants are pure fantasy anyway -- the horse ran the times he ran, not the adjusted ones...

[classhandicapper]

"You feel this is worse than no variant at all? "

It's hard to imagine that it's worse than no variant at all, but for anyone that has made seperate track variants for every pace call at some point in their horseplay (I have), there are some very obvious problems.

Wind is one very obvious issue, but there are others.

The runnup for some distances is not the same every day or even for every race. That has a bigger impact on the fractions than the final time proportionally, but it effects eveything. (2/5 at 2F is a lot while 2/5 at 6F or more is less significant).

There are no guarantees that the track surface speed is uniform. Track maintenance could make the 1st turn of routes faster or slower than other parts of the track. The chute could be faster or slower than the more used sections. All that means that various fractions are not impacted in some neat proportion to the final time variant.

By the way, these (and others) are all good arguments against putting too much faith in speed and pace figures to begin with, but we are forced to work with what we have. So we might as well try to be as accurate as we can get.

[46zilzal]

So 46, for now (exclude the programs you are currently 'testing'), we can safely say that there is but ONE variant (a final time variant) and it is applied in some way or proportion to the first two fractions. This is the way the Sartinistas used to advocate dealing with the variant....

BTW, are they still using the DRF variant (with the exception of the program you are currently testing).
yes to the first, no to the second

[cj]

Now, in this case it is pretty close to the same result as if you did it by distance, but in more extreme scenarios and longer races the difference is more pronounced. You feel this is worse than no variant at all? This way I preserve the original energy distribution. We must remember that variants are pure fantasy anyway -- the horse ran the times he ran, not the adjusted ones...

I'll keep it simple to not bore too many. Let's say a track is measured as 3/5s fast for 6f races. The exact numbers aren't important, but a proportional scheme might say at the 1/4 it was 1/5 fast, at the 1/2 it was 2/5 fast, and at the finish 3/5 fast. This is the same as saying each fraction was 1/5 fast, and uniform around the track. I have no problem admitting that most days, this is fine. But on other days, it can cause big problems.

Let's say we are at Belmont, and that very same 6f race variant is 3/5s fast. Due to wind, shade, or some other weather condition, it often happens that this 3/5s fast variant is not evenly distributed around the track. Another factor can be chutes, which often times don't get the same maintenance as the rest of the track and play faster or slower.

In the above, perhaps there is a 20 mile an hour head wind on the backstretch. In this case, instead of the backstretch playing 1/5 fast, it might really be 1/5 slow. The turn might then still be 1/5 fast, but now the stretch is 3/5 fast. This is why I say in conditions like this, no variant would be better than a proportional one for the internal fractions. Maybe that is a bit extreme, neither one is very good.

Compare this to the proportional.
Proportional at first call, 1/5 fast. Separate at first call, 1/5 slow.
Proportional at second call, 2/5 fast. Separate at second call, 0.
Proportional at third call, 3/5 fast. Separate at third call, 3/5 fast.

2/5s is a lot at the pace calls. The spreads can only get worse the farther away from a zero track you get. Imagine what this can do to energy or median ratings mentioned earlier.

[classhandicapper]

CJ,

It should be obvious I agree with you completely.

I'll just add one thing because I don't think most people don't appreciate this enough.

Inaccurate figures are probably the #1 source of longshot selections that win. That doesn't mean the figures selected the most likely winner or even an actual overlay. It just means that among reasonably similar contenders, they all have a decent chance. So if a bogus figure incorrectly points to 4th best horse as the most likely winner, it's possible to have the right horse for the wrong reason and have actually bet an underlay.

[Tom]

I agree with everyone! ;)


The purpose of making a variant is NOT to get it right in terms of how fast a track played - it is to predict a winner. I have used GT's method - used it for years, actually, back in my Sartin days. I am familiar with what 46 is using, and I use CJ's stuff today - with HTR to boot. I get different horses with each method, and each one seems to work well enough to make money, but they work differently on the same race. CJ may well some winners 46 will never see, and 46 will get winners CJ will throw out.
I find using CJ's figs and HTR togather is a good thing because I see the same race through different windows. I sometimes see stand outs wtih CJ numbers that look so-so in HTR numbers, and I see HTR stand outs that have noting special with CJ figs. how do I know which to bet on? Easy. The higher odds.
And I have recently bet horses based on good Sartin-type readingS the neither CJ nor HTR showed as anything special.

[cj]

No doubt about what you say Tom. We all know, at least I hope, that figures aren't the be all and end all of handicapping many races. But I still want the best ones I can have, otherwise I'd still be using BRIS or the TSN 50 centers.

[classhandicapper]

Tom and CJ,

I know what you guys are saying about looking for value, but I don't think I've had especially good results over the years when a horse had very good figures using one method and weaker figures with another by betting on the higher priced one with the higher figures.

In fact, that's the point I've been trying to make.

When one method is pointing to a high priced horse and the others aren't, more often than not it's because the one pointing to the higher priced horse is in error. Yes, some of them will win because some of any group will win. But they may not be overlays.

I've had my best results when I've been able to dig much deeper into the prior races and numbers and determine who actually had it right. Usually, it's the one that is picking the shorter priced horse (in which case I will pass), but occasionally it's not. That's where I think the real overlays are. That smaller group.

I've come to the conclusion that the wagering pools are amazingly efficient. More so than I thought at one time. The net of all the figures and methods used out there tends to weed out all the errors and methodology issues of individual figures very well and set the odds close to the probabilities most of the time. I think the mistakes are rarer than people think.

[GaryG]

In some cases figures will isolate a standout play. More often, at least for me, they point out the contenders. Then the handicapping begins, evaluating form and trainer intent. Nothing better than a top rated horse with positive trainer moves and peaking form. :ThmbUp:

[46zilzal]

whatever "hiccup" occured last weekend has reversed itself and the speedballs are back doing their thing with GUSTO on the Aqueuduct inner, just like the last 4 years.

[cj]

Most others seem to think speed did just fine last weekend.

[46zilzal]

P-O-S-I-T-I-O-N-A-L-L-Y maybe (since that is racing EVERYWHERE) but not energy D-I-S-T-R-I-B-U-T-I-O-N wise

[cj]

How would we know, you haven't posted the numbers so we can see what you are talking about.

[46zilzal]

How would we know, you haven't posted the numbers so we can see what you are talking about.

Well let's keep it that way.

[cj]

Then why bring it up in the first place?

[46zilzal]

Then why bring it up in the first place?
Independent of my observations, I wanted to see if anyone else noticed it

[kenwoodallpromos]

So yes, others noticed.
"Ken. You know all about track maintenance, do you think they dug it up or brought in a lot of soil?"
Since Aq is back to form, Adding to the cushion was not as likely done as sealing and the effect of rain and reactions to it.

[Tee]

Independent of my observations, I wanted to see if anyone else noticed it
Since your observations are coming from a energy distribution view, wouldn't anyone else need to be doing the same to notice these same observations? Ala the 2 other people who use the same software as u do and noticed EXACTLY the same thing and commented on it independently of your input

[46zilzal]

Since your observations are coming from a energy distribution view, wouldn't anyone else need to be doing the same to notice these same observations? Ala the 2 other people who use the same software as u do and noticed EXACTLY the same thing and commented on it independently of your input
yes, I was contacted by others who noticed the same hiccup but wondered if it has crossed over the more tradition observations as well

[46zilzal]

So yes, others noticed.
"Ken. You know all about track maintenance, do you think they dug it up or brought in a lot of soil?"
Since Aq is back to form, Adding to the cushion was not as likely done as sealing and the effect of rain and reactions to it.
yes, but such a dramatic change.....FIRST time in over 200+ race cards? One would look for something a bit more definitive....Then it could be as I said: that rare period where 26 heads came up in a row...

[shanta]

How would we know, you haven't posted the numbers so we can see what you are talking about.


here is the energy dist graph of today's 1st race at Aqu the # 5 is the winner the right of center line is early the left of center is late

the 5 has the 2nd biggest early "stick" so regardless of positional style ( visual) he is number 2 early on the readouts

btw I did NOT handicap or bet this race. I just entered the highest speed rating in last 3 races to illustrate.

hope this helps
Rich

[classhandicapper]

P-O-S-I-T-I-O-N-A-L-L-Y maybe (since that is racing EVERYWHERE) but not energy D-I-S-T-R-I-B-U-T-I-O-N wise

If I understand you, what you are saying is that horses near the front were winning and running well, but they weren't winning with the same energy distribution as recently.

In other words, I assume you mean the paces of the top finishers were probably slower (or something like that).

Assuming I understand you and you have accurate enough pace variants to even make that determination, what if the bias existed but many races just happened to develop with slower paces because of the specific horses within the races?

Then energy distribution would have little to do with anything. The horses near the lead would run slower paces and come home faster than usual producing different energy distributions among winners and the bias "may" still exist. This is one of the complications of bias determination to begin with. Pace matters and complicates the bias analysis.

[kenwoodallpromos]

You could email the track and ask them if they have anything more definitive than listing on the charts "sealed" and "rainy". They would know if they dumped a bunch of soil or anything else on the dirt. Ask them also the harrow depth that day.
Or wait for 200 more race days for it to happem again!

[46zilzal]

Then energy distribution would have little to do with anything. The horses near the lead would run slower paces and come home faster than usual producing different energy distributions among winners and the bias "may" still exist. This is one of the complications of bias determination to begin with. Pace matters and complicates the bias analysis.
Don't think so

[classhandicapper]

Don't think so

You would need to elaborate a bit, but if horses run faster earlier they come home slower and vice versa. The exact relationship might be debatable, but the general truth is not. So a series of either slow or fast paces could very well change the average energy distribution of the winners, while being unrelated to whether or not there was a bias favoring speed.

ex.

IF you could run the exact same horses twice on the same day (all else being equal) and the front runner went 22 45 once and 23 46 the next time and won both races I assume you would produce different ratings. However the track bias would be identical because it is independent of how the horses choose to run on it. Do that 9 times and the bias still doesn't change. It just gets more difficult to measure because pace complicates the analysis (as it always does to some degree)

[PaceAdvantage]

Elaboration isn't 46zilzal's forte.

[GameTheory]

Grunt.

[46zilzal]

Energy distribution is INDEPENDENT of velocity: that is what the program ENERGY was all about. One looks for distribution mismatches that favor a particular animal.

these two arrays, percent wise, represent the DISTRIBUTION of first an early/presser
35.88 34.16 30.91
versus a sustained/presser
33.98 34.16, 31.86 which you can see is more evenly distrubuted. Now if the total energy (velocity of those three fractions) is substantiatly bigger, one or the other, then that one has a better chance to prevail provided that style is dominant for said track and distance and is closer to what NORMALLY wins at this track and distance. At most tracks, on most days, sprints go to early, early/pressers, or pressers, while most routes go to pressers, sustained/pressers, or sustained runners. In any particular contest however, almost ANY distribution COULD win provided that the total velocity is superior. Horses with the wrong style for any track and particular distance usually have to have velocity superiority to overcome the "style fit" for that distance and race course. Knowing each race track's FINGERPRINT (style and distribution that wins the majority of races at a particular distance) relly helps when shippers come in from a course that might have favored a different style. One distance track combination that is really unique is Woodbine's 7 furlong standard which is like NONE other I have ever charted. Makes for easy eliminations for shippers if they cannot make the "fit."

A turfer, running as a LATE participant might distribute this way 32.71 33.19 34.08, getting MORE velocity (percentage wise) the further into the contest one goes. But another horse with an earlier distribution, provided it had velocity superiority, COULD project as the winner. On the lawn this is particularly true at 8 furlongs, but rarely at longer turf distances.

When one then adds velocity to these distributions, compares them to one another along with their relative rates of deceleration, one can figure out if there is a mis-match in favor of one or more which would suggest a wager. By the recording WHICH of these styles is NORMALLY dominant at a course and distance, then noting what is winning TODAY at this course and distance allows the observer to notice if there is an observable bias.

The earlier screen shot showed the early/late balance which is a quick way to look for that "fingerprint" for any particular distance/race course. One looks for the earliest one that is still within the range of normalcy for the distance. If two were dead even in all other categories, the earlier one would be the favored bet.

[kenwoodallpromos]

"a series of either slow or fast paces could very well change the average energy distribution of the winners, while being unrelated to whether or not there was a bias favoring speed."
While I believe that it is true that a horse loses energy relatively quickly at top or near top speed (there is racing that goes for many miles), the more "cushioning" that absorbs any of the horse's stride causes energy loss. The reason they call it a "cushion" is because the base is so hard. There may be some kind of formula of speed vs. energy loss but you will not hear it from me! Again the example is a sealed or maybe wet fast track- the pace matters but so does the firmness as it determines a certain absorbsion of energy- otherwise horses' speed and all races are very predictable since pace is very often able to be predicted, and the only thing left to consider greatly is the horse's cardio-vascular/lung capacity.
At this point in time cardio-vasculular and lung efficiently is a very minor variable because muscular efficiency and stamina is very much more important, as evidenced by your comments about pace, and by the banning of "milkshakes" to prevent the muscles from getting tired by lactic action.

[Tom]

yes, but such a dramatic change.....FIRST time in over 200+ race cards? One would look for something a bit more definitive....Then it could be as I said: that rare period where 26 heads came up in a row...

Might the cange in the TM speed ratings/variants have anyting to do with this?
Have they started yet? Aren't your pacelines equalized through these?

[46zilzal]

Might the cange in the TM speed ratings/variants have anything to do with this?
Have they started yet? Aren't your pacelines equalized through these?
wouldn't change the distribution only the relative velocities and there is a PATCH for said problem which has worked well so far

[Tom]

46: whatever "hiccup" occured last weekend has reversed itself and the speedballs are back doing their thing with GUSTO on the Aqueuduct inner, just like the last 4 years.


CJ: Most others seem to think speed did just fine last weekend.


46: P-O-S-I-T-I-O-N-A-L-L-Y maybe (since that is racing EVERYWHERE) but not energy D-I-S-T-R-I-B-U-T-I-O-N wise


CJ: How would we know, you haven't posted the numbers so we can see what you are talking about.

46: Well let's keep it that way.

CJ: Then why bring it up in the first place?


46: Independent of my observations, I wanted to see if anyone else noticed it




:lol: :lol: :lol: Hey ABBOT!

You guys made my night.....

[cj]

Grunt.

Did you see my example earlier in the thread about using a proportional variant for the pace calls?

[classhandicapper]

OK, I understand what you are saying about the energy distribution profile of the typical winner. I find that kind of analysis very interesting. However, I still think there are pace/bias issues as I was describing that complicate these things and that can lead to false conclusions. I'm not saying I have the solution. For me it's a subjective process.


Energy distribution is INDEPENDENT of velocity: that is what the program ENERGY was all about. One looks for distribution mismatches that favor a particular animal.

these two arrays, percent wise, represent the DISTRIBUTION of first an early/presser
35.88 34.16 30.91
versus a sustained/presser
33.98 34.16, 31.86 which you can see is more evenly distrubuted. Now if the total energy (velocity of those three fractions) is substantiatly bigger, one or the other, then that one has a better chance to prevail provided that style is dominant for said track and distance and is closer to what NORMALLY wins at this track and distance. At most tracks, on most days, sprints go to early, early/pressers, or pressers, while most routes go to pressers, sustained/pressers, or sustained runners. In any particular contest however, almost ANY distribution COULD win provided that the total velocity is superior. Horses with the wrong style for any track and particular distance usually have to have velocity superiority to overcome the "style fit" for that distance and race course. Knowing each race track's FINGERPRINT (style and distribution that wins the majority of races at a particular distance) relly helps when shippers come in from a course that might have favored a different style. One distance track combination that is really unique is Woodbine's 7 furlong standard which is like NONE other I have ever charted. Makes for easy eliminations for shippers if they cannot make the "fit."

A turfer, running as a LATE participant might distribute this way 32.71 33.19 34.08, getting MORE velocity (percentage wise) the further into the contest one goes. But another horse with an earlier distribution, provided it had velocity superiority, COULD project as the winner. On the lawn this is particularly true at 8 furlongs, but rarely at longer turf distances.

When one then adds velocity to these distributions, compares them to one another along with their relative rates of deceleration, one can figure out if there is a mis-match in favor of one or more which would suggest a wager. By the recording WHICH of these styles is NORMALLY dominant at a course and distance, then noting what is winning TODAY at this course and distance allows the observer to notice if there is an observable bias.

The earlier screen shot showed the early/late balance which is a quick way to look for that "fingerprint" for any particular distance/race course. One looks for the earliest one that is still within the range of normalcy for the distance. If two were dead even in all other categories, the earlier one would be the favored bet.

[classhandicapper]

If what you are saying is that the relationship between pace and final time varies from day to day and track to track because of cushion and other factors, I wholeheartedly agree. Not only do I agree, I think the best jockeys and trainers eventually figure these tendencies out and alter the average fractions at the track (through small changes in aggression and training) trying to maximize their chances of winning - which in turn tends to offset the advantage some styles would have had (at least to some degree).


"a series of either slow or fast paces could very well change the average energy distribution of the winners, while being unrelated to whether or not there was a bias favoring speed."
While I believe that it is true that a horse loses energy relatively quickly at top or near top speed (there is racing that goes for many miles), the more "cushioning" that absorbs any of the horse's stride causes energy loss. The reason they call it a "cushion" is because the base is so hard. There may be some kind of formula of speed vs. energy loss but you will not hear it from me! Again the example is a sealed or maybe wet fast track- the pace matters but so does the firmness as it determines a certain absorbsion of energy- otherwise horses' speed and all races are very predictable since pace is very often able to be predicted, and the only thing left to consider greatly is the horse's cardio-vascular/lung capacity.
At this point in time cardio-vasculular and lung efficiently is a very minor variable because muscular efficiency and stamina is very much more important, as evidenced by your comments about pace, and by the banning of "milkshakes" to prevent the muscles from getting tired by lactic action.

[Tom]

Energy distribution is INDEPENDENT of velocity: that is what the program ENERGY was all about. One looks for distribution mismatches that favor a particular animal.....



46 - that was one of the best descriptions of energy distribution I have ever read. Thanks for summarizing it like that.

[GameTheory]

Did you see my example earlier in the thread about using a proportional variant for the pace calls?Yes, thank you. I was grunting only in order to outdo 46 for the shortest most unelaborative reply in the thread. I'm competitive like that...

[cj]

Yes, thank you. I was grunting only in order to outdo 46 for the shortest most unelaborative reply in the thread. I'm competitive like that...

I got the grunt thing, was just wondering what you thought. I know you are a sharp guy, that is why I ask.

[GameTheory]

I got the grunt thing, was just wondering what you thought. I know you are a sharp guy, that is why I ask.I don't have much comment, since I'm not much of a pace guy. Times and variants really aren't very important to me, so I'm just curious, but don't have a lot to say about it.

You're probably right, but as others have said maybe other methods are also nearly right at different times, meaning each method will help lead to different winners at different times. My method preserves the original energy distribution as it actually occured. Maybe that is worth something. Your method probably works better for comparing pace ratings from different races. No reason one couldn't do it both ways...