Synthetic/Dirt figure comparision

[cj]

I mentioned this in the Derby Prep Figures thread, so I'll post it here. Because of pace differences, synthetic and turf races are almost always slower than dirt races in classier races and faster in cheap races. I don't like to directly compare different surfaces, but for a rough idea when a horse has never tried a different surface, I use the below chart.


Poly Dirt
110 120
105 113
100 107
95 100
90 93
85 87
80 80
75 73
70 67
65 60
60 53
55 47
50 40



Formula is 80 + (110 - Beyer)/30*40

[ny0707ny]

Last year I guessed the polytracks were slower times than dirt. Most people said the times were slower. I guess it is due to what the tracks are maded up of. They run more like turf with slower times. I never really looked into it though. It is nice to see I was right. The problem is though that you have so many different polytrack surfaces.

[cj]

I am not talking about the raw times, but the adjusted times, or speed figures. Raw times can be very, very fast on synthetics as we have seen at Santa Anita and Keeneland.

[Capper Al]

Thanks again CJ.

[ny0707ny]

Thanks :jump:

Have we seen any triple digit beyers on poly for the Derby preps? I think some in sprints but I don't think any on routes. I think the crop is very weak this year so I would be surprised if we did.

[Kelso]

Raw times can be very, very fast on synthetics as we have seen at Santa Anita and Keeneland.


CJ,
Do you think the 2007 stroll-in-the-park Blue Grass Stakes was just an aberation, perhaps due to it being the first big race on rubber?

[bobphilo]

CJ,
Do you think the 2007 stroll-in-the-park Blue Grass Stakes was just an aberation, perhaps due to it being the first big race on rubber?

Kelso, while we wait for CJ's answer, hope you don't mind me chimming in. I remember discusing the race with CJ last year, though we were mostly concerned with the effect of the pace on ground loss.

1) There was a perception by the riders (a self fullfilling prophacy in my view ) that it was impossible to win on the front end at Keenland so it had more to do with strategy than track effect. The fast fractions they flew home in showed it wasn't a slow tiring track.
2) Teuflesberg was the lone speed and tried to get away with as slow a pace as possible. Again, a strategy decision.

Combine the 2 and you get Teuflesberg going out super slow and no one willing to challenge him for fear of the dreaded bias against early speed. If the jocks had ridden that way on any dirt track the result would have been the same - provided the horses had the same affinity for that surface.

Whether one believes there was a bias or not, what happened was due to the jockey's perceptions and how it affected their strategy.

CJ's view may or may not differ.

Bob

[Kelso]

1) There was a perception by the riders (a self fullfilling prophacy in my view ) that it was impossible to win on the front end at Keenland so it had more to do with strategy than track effect. The fast fractions they flew home in showed it wasn't a slow tiring track.
2) Teuflesberg was the lone speed and tried to get away with as slow a pace as possible. Again, a strategy decision.

Combine the 2 and you get Teuflesberg going out super slow and no one willing to challenge him for fear of the dreaded bias against early speed.
Bob

Bob,
Thank you for chiming in. I worded my question very poorly ... my reference to "first big race" was actually directed at how the jocks/trainers had doped out tactics for the new surface, not at the track's actual characteristics.

I agree with your view that the race could have been run much faster (particularly by T'berg, whose first six furlongs - as I recall - each took more than 12 seconds) if the riders knew what CJs figures in this thread suggest.

[bobphilo]

Bob,
Thank you for chiming in. I worded my question very poorly ... my reference to "first big race" was actually directed at how the jocks/trainers had doped out tactics for the new surface, not at the track's actual characteristics.

I agree with your view that the race could have been run much faster (particularly by T'berg, whose first six furlongs - as I recall - each took more than 12 seconds) if the riders knew what CJs figures in this thread suggest.

You worded your question very well, Kelso. I understood that you were asking what part the riders/trainers perceptions dictated their strategy and the slow time. I felt that was precisely what directly determined how the race was run and only added that the reference to the surface to say that that was not the reason for the slow time.
T'bergs/s rider wanted to slow the pace on the front end since he felt that would favor the frontrunner bt leaving him with more to hold off the closer's late challenges. The others didn't press him since their perception was that the track was biased against early speed and that guided their strategy. i think I read you correctly.

Jockey's are not so much interested in running a fast time as they are in running a time that's just faster than their competitors, and will often ride tactical races to do so. Of course, the perceptions they base these tactics on are not always correct.

Bob

[cj]

I think you guys have it right. The early pace is a very small factor in routes on synthetics, just like on turf. None of the jockeys of the main contenders were worried about Teufelsberg on the front end no matter how slow he went. This is what leads to the bunched finishes on the stuff. While the Blue Grass was 9f, they only really ran 2 to 3f so horses that would have been drubbed on dirt stayed much closer at the end.

[classhandicapper]

I compiled this kind of data for myself for high quality turf/dirt racing, but have no info on Poly. Thanks.

[classhandicapper]

CJ,

We've discussed much of this in the past.

Like you, I believe this phenomenon is related to pace, but I also think there's a minor problem with the way all figure makers make their numbers.

According to the standard speed figure chart, a 1/5 of a second (and/or a length) is more significant at a shorter distance than at longer distance. That principle is demonstarted by comparing runners to world class times.

If a runner is 1 second slower than the best 100 yard dash runner, he is a slow sprinter.

If a runner is second slower than the best miler, he is a world class runner.

1 second is very significant in a sprint, but not as important at a mile and longer.

Almost everyone also understands that if I ran against a world class miler and we just jogged for 3/4 of a mile, he wouldn't beat me by as much as he would with a fast pace. That's the phenomenon we are talking about on turf and Poly.

What I believe many people don't quite grasp is that even though the race was at a mile, the effective distance was actually shorter given that we only ran hard for a 1/4 mile. That means that each 1/5 second (and or length) was more significant than in a faster paced mile.

That leads me to this conclusion about my example:

If the effective distance was shorter than a mile, we shouldn't be using the Time Charts and Beaten Length Charts for a mile. We should be using one for a shorter distance where each 1/5 of a second and length is more significant. That would take the compression out of the figures and help reflect the actual difference in ability.

The same principle is applicable to turf and poly etc...

The problem of course is that paces vary wildly on turf and poly. So there is no easy way to convert from the actual distance to the effective distance in order to know how significant a 1/5 of a second/length actually was in any given race.

In addition, even though the paces are slower in many turf and poly races, they aren't so slow that the horse's aren't using up at least some level of energy early in the race.

It's an extremely complex problem that may not have a perfect solution because you could easily argue that the turf/poly figures are correct but the dirt figures are wrong because the paces are often too fast. In any event, IMO one of the problems is the way figures are calculated.

[cj]

I agree, and basically I'm not really converting, just using a different speed chart for the different surface. I'm just doing it after, rather than before, making the speed figures. Both will result in the same outcome, or at least in very close proximity.

[classhandicapper]

I agree, and basically I'm not really converting, just using a different speed chart for the different surface. I'm just doing it after, rather than before, making the speed figures. Both will result in the same outcome, or at least in very close proximity.

Understood and agreed.

You know me. I just like to talk about these issues. :lol:

[bobphilo]

Understood and agreed.

You know me. I just like to talk about these issues. :lol:

Class, I too like talking about these things.

I recall discussing this very issue with CJ after last years Blue Grass – it must still be in the archives. Beyer also had noted that a similar situation existed with the figures for grass races where the real running doesn’t begin until the final fractions. Beyer deals with this by using the beaten length/points formula for dirt sprints in turf routes. I had suggested that that, perhaps, since they only ran the last 3/8’s in the BG that the formula for a much shorter distance be used instead of the 9 furlong one. CJ thought the idea had merit too.

However, I realize that since early pace can vary on grass and poly it would be cumbersome to have a different formula that changes with each pace scenario. It’s probably better to have a set formula for the typical pace for each surface, which is what Beyer does for grass and what it seems CJ is doing for grass and synthetic.



Bob

[Achilles]

Class,

I honestly believe that your explanation of why you need to use a beaten length chart for a shorter distance (on turf and probably fake dirt) is more easily understood than Beyer's explanation of the same adjustment in Beyer on Speed......and I really respect Beyer.

Nice job...............Jack

[Niko]

CJ,

We've discussed much of this in the past.

Like you, I believe this phenomenon is related to pace, but I also think there's a minor problem with the way all figure makers make their numbers.

According to the standard speed figure chart, a 1/5 of a second (and/or a length) is more significant at a shorter distance than at longer distance. That principle is demonstarted by comparing runners to world class times.

If a runner is 1 second slower than the best 100 yard dash runner, he is a slow sprinter.

If a runner is second slower than the best miler, he is a world class runner.

1 second is very significant in a sprint, but not as important at a mile and longer.

Almost everyone also understands that if I ran against a world class miler and we just jogged for 3/4 of a mile, he wouldn't beat me by as much as he would with a fast pace. That's the phenomenon we are talking about on turf and Poly.

What I believe many people don't quite grasp is that even though the race was at a mile, the effective distance was actually shorter given that we only ran hard for a 1/4 mile. That means that each 1/5 second (and or length) was more significant than in a faster paced mile.

That leads me to this conclusion about my example:

If the effective distance was shorter than a mile, we shouldn't be using the Time Charts and Beaten Length Charts for a mile. We should be using one for a shorter distance where each 1/5 of a second and length is more significant. That would take the compression out of the figures and help reflect the actual difference in ability.

The same principle is applicable to turf and poly etc...

The problem of course is that paces vary wildly on turf and poly. So there is no easy way to convert from the actual distance to the effective distance in order to know how significant a 1/5 of a second/length actually was in any given race.

In addition, even though the paces are slower in many turf and poly races, they aren't so slow that the horse's aren't using up at least some level of energy early in the race.

It's an extremely complex problem that may not have a perfect solution because you could easily argue that the turf/poly figures are correct but the dirt figures are wrong because the paces are often too fast. In any event, IMO one of the problems is the way figures are calculated.

thoughtful post...

I'm not a figure maker but couldn't some of the above be addressed by looking at the pace of the race and maybe pars. If 80 is the par and they ran 60 you may want to use 1/5 of a second. If the pace was 80 you may want to use your standard forumla. One reason why what CJ calls a negative pace race (basically early pace slower than the speed of the race by a couple lengths) can be thrown out because the numbers are deceiving. Or maybe even by looking at the percentage of energy to determine which scale to use. I think a lot of people do this in a sense by eyeballing the pace of the race but maybe it can be done mechanically too? I'm sure someone can run an algorythm or try it with regression analysis? Maybe someone already has the answer?

I always like to say..there's no stupid questions-only stupid people. Sometimes I'll raise my hand....

[rgustafson]

I mentioned this in the Derby Prep Figures thread, so I'll post it here. Because of pace differences, synthetic and turf races are almost always slower than dirt races in classier races and faster in cheap races. I don't like to directly compare different surfaces, but for a rough idea when a horse has never tried a different surface, I use the below chart.


Poly Dirt
110 120
105 113
100 107
95 100
90 93
85 87
80 80
75 73
70 67
65 60
60 53
55 47
50 40



Formula is 80 + (110 - Beyer)/30*40

CJ,
Very interesting comment in the Cristblog today by Steve T giving out some statistics that speak to exactly what you have listed here.

[classhandicapper]

thoughtful post...

I'm not a figure maker but couldn't some of the above be addressed by looking at the pace of the race and maybe pars. If 80 is the par and they ran 60 you may want to use 1/5 of a second. If the pace was 80 you may want to use your standard forumla. One reason why what CJ calls a negative pace race (basically early pace slower than the speed of the race by a couple lengths) can be thrown out because the numbers are deceiving. Or maybe even by looking at the percentage of energy to determine which scale to use. I think a lot of people do this in a sense by eyeballing the pace of the race but maybe it can be done mechanically too? I'm sure someone can run an algorythm or try it with regression analysis? Maybe someone already has the answer?

I always like to say..there's no stupid questions-only stupid people. Sometimes I'll raise my hand....

I think you have the right idea, but it would get complex.

I tend to just look at the pace figures and final time relative to PAR and get a feel for what I think happened without getting too caught up in calculations.

[Tom]

OK, so we use the BL figures for 6.5 furlongs in turf routes.....is the some advantage to using a different BL figure for the pace call? If the pace is the slower part of the race, would it be better to have the significance less per length? Instead of using 1 lenght is the basis of the rating, maybe use 2, or 1.5, or something along those lines, so that a horse running 112.4 and one running 113.1 are about equal?

[cj]

OK, so we use the BL figures for 6.5 furlongs in turf routes.....

There are a lot of problems with doing that as well, which is why I don't do it. When the pace is fast, it penalizes the horses that ran close up and backed up much more than it should for one.

[Murph]

There are a lot of problems with doing that as well, which is why I don't do it. When the pace is fast, it penalizes the horses that ran close up and backed up much more than it should for one.Interesting points made here on this subject. Class explains the situation very nicely but we still don't have a proper way to make this into a bet. Is anyone using a published method to deal with turf races only?

CJ, would you explain the logic of the formula you used for comparing the surfaces? [Formula is 80 + (110 - Beyer)/30*40]

I would like to see this thread continue until we can toss some brass tacks out to stick some figures on. Thanks for sharing your thoughts on this for everyone here.

Murph

[cj]

Interesting points made here on this subject. Class explains the situation very nicely but we still don't have a proper way to make this into a bet. Is anyone using a published method to deal with turf races only?

CJ, would you explain the logic of the formula you used for comparing the surfaces? [Formula is 80 + (110 - Beyer)/30*40]

I would like to see this thread continue until we can toss some brass tacks out to stick some figures on. Thanks for sharing your thoughts on this for everyone here.

Murph

The logic is just that the scale is tighter because the pace tends to be much slower. Therefore, good horses run slower overall, and bad horses run faster.

The formula was just based on the high end of the Beyer scale. Dirt figures of around 120 are near the very top on dirt, while 110 is near the top on synthetics and turf. So, that part was pretty easy. The hard part is finding the point where the figures are even. 80 was a good starting point and seems to work OK, but to be honest, 75 might be better which would change the 80 to 75 and the 30/40 to 35/45.

[trigger]

In regards to pace, I have always wrestled with why the trainer doesn't figure out for each horse what fractional times will result in that horse's best overall time and then tell the jockey to run him accordingly regardless of race pace.
Is it because the jockeys don't possess the internal clock necessary to run a horse this way.
Or, perhaps, the horse wants to run in a preferred position relative to the rest of the herd regardless of time and the jockey can't do anything about it without screwing up the horse's chances? :confused:

[RonTiller]

In regards to pace, I have always wrestled with why the trainer doesn't figure out for each horse what fractional times will result in that horse's best overall time and then tell the jockey to run him accordingly regardless of race pace.
I know trainers who do exactly what you just described. After handicapping the race and estimating what kind of 1/4 and 1/2 times they believe the race will be run in, and who will be racing where, they give jockey instructions on where to place there horse relative to the pack, the front runner or relative to another horse. This placement is meant to have their horse run their most advantageous fractions and final times. I'm told this is the best one can do, as nobody can just get on a horse and run a 23.2 (rather than a 23.6) quarter at will.

If the jockey follows the trainer's instructions and loses, they'll live to ride the horse again. If they disobey and win, they don't ride again. This is of course in stark contrast to the instructions "ride him as you see fit."

Ron Tiller
HDW

[classhandicapper]

Interesting points made here on this subject. Class explains the situation very nicely but we still don't have a proper way to make this into a bet. Is anyone using a published method to deal with turf races only?

Murph

Murph,

I consider myself an expert at understanding issues without finding a way to turn them into profitable bets. :lol:

Seriously, I think if you are armed with a decent set of pace and final time figures and understand the issue as I have presented it, it will allow you recognize "suspect final time figures" when comparing horses that switch surfaces. It will also allow you to recognize when a very slow pace allowed several horse to finish close together without it being a true indication of their relative ability. After that, you are on your own. ;)

I still tinker with formulas from time to time. If I ever come up with one that I think solves all the problems well enough to use, I'll post it.

[classhandicapper]

OK, so we use the BL figures for 6.5 furlongs in turf routes.....is the some advantage to using a different BL figure for the pace call? If the pace is the slower part of the race, would it be better to have the significance less per length? Instead of using 1 lenght is the basis of the rating, maybe use 2, or 1.5, or something along those lines, so that a horse running 112.4 and one running 113.1 are about equal?

Tom,

The problem is more than just beaten lengths. The actual time charts have to change too. The problem is that they can't be changed in some standard way because depending the pace, the adjustment would have to vary. It would be very difficult to understand all the relationships between pace and final time and the significance of a 1/5th second.

There is a reason my name is classhandicapper. ;)

The more I explored pace and final time, the more I realized I was working with approximations that were a tool and not an answer.

[cj]

Tom,

The problem is more than just beaten lengths. The actual time charts have to change too. The problem is that they can't be changed in some standard way because depending the pace, the adjustment would have to vary. It would be very difficult to understand all the relationships between pace and final time and the significance of a 1/5th second.

There is a reason my name is classhandicapper. ;)

The more I explored pace and final time, the more I realized I was working with approximations that were a tool and not an answer.

You have it pretty much right in my opinion. One other thing that I think you realize as well is that even adjusting the chart by the pace would be fraught with errors. Imagine a 6 horse field that goes in a very slow pace time. You use the "new" chart, but two of the also rans are cheap horses that actually couldn't run any faster. Their figures are now inflated. I'm not sure there is a numerical solution. It is just something you have to be aware of when you analyze a running line.

[46zilzal]

CJ,
Do you think the 2007 stroll-in-the-park Blue Grass Stakes was just an aberation, perhaps due to it being the first big race on rubber?
An article in this week's Racing Post hits the nail on the head as regards Polytrack surfaces.
Let surfaces bed in before lumping on at Great Leighs, by Richard Young.
"The first new race course to open in England in the last 80 years has been delayed by all sorts of difficulties (Great Leighs). The reaction of the jockeys was noticeably similar (good) to those who took part in the trials at the current Polytrack venues in that most felt the surface was excellent but one that will be riding on the slow side until it has taken time to bed in. Therefore kickback at the early fixtures is likely to be more than a couple of months time but it's worth remembering that few horses seemed disadvantaged by the levels of kickback in the early days at Lingfield, Wolverhampton, Kempton or Dundalk.

It would be unwise to go in with preconceived ideas when it comes to new racecourses but given the nature of the track, this will be another Polytrack venue that favors speed over stamina......it is best to keep an open mind until the surfaces beds down."

[cj]

46,

How are the tracks going to settle in when they keep changing them?

As for the original formula I posted in this thread, a customer pointed out it was wrong. Shows what happens when I try to use my memory! The actual formula is this:

=80+(Beyer-80)*(40/30)

[jonnielu]

Tom,

The problem is more than just beaten lengths. The actual time charts have to change too. The problem is that they can't be changed in some standard way because depending the pace, the adjustment would have to vary. It would be very difficult to understand all the relationships between pace and final time and the significance of a 1/5th second.

There is a reason my name is classhandicapper. ;)

The more I explored pace and final time, the more I realized I was working with approximations that were a tool and not an answer.

Has anybody ever just left time out for awhile to see if it really makes much difference?

jdl

[46zilzal]

46,

How are the tracks going to settle in when they keep changing them?


Only reporting the point of view of an arena where these tracks have a longer history.

[cj]

I forgot, you just post links and screenshots.

[46zilzal]

I forgot, you just post links and screenshots.
Yes, factually substantiated things.

[cj]

Thanks for being our own personal Google.

[classhandicapper]

Has anybody ever just left time out for awhile to see if it really makes much difference?

jdl

I have.

I started out as a speed handicapper. Then I incorporated pace. Along the way I discovered more and more of the complexities of making figures, seeing large disagreements between expert figure makers etc... and realized I was often dealing with bad information. So I stopped making figures and handicapped using class as my primary basis for measuring performances and ability (more sophisticated stuff than in the basic books of long ago).

My results actually improved very sharply. That's when I really broke out and started to win with some degree of consistency somewhere betwen 10-15 years ago (after more mixed results and small net losses for years before that).

I think that was primarily because many poeple already had access to figures and were using them, the old classing techniques weren't very sophisticated relative to what I was doing, and class was routinely attacked by poplular authors and figure handicappers.

Somewhere along the line, I started realizing that some of the weaknesses of class handicapping were the strengths of figure handicapping and vice versa. The figures also got more sophisticated. Now I'm what I would call a hybrid handicapper. I measure horses and performances using class and trip, but use figures as tools where my classing techniques are difficult to apply and where they help clarify results.

[Valupix]

Love ’em or hate ‘em, synthetic surfaces are here to stay.

I can only advise folks to adjust or end up driving yourselves nuts. They are never leaving this game.

[cj]

Love ’em or hate ‘em, synthetic surfaces are here to stay.

I can only advise folks to adjust or end up driving yourselves nuts. They are never leaving this game.



You could be right, but you could be wrong, and they'll never be installed enough places that you have to play them.

[46zilzal]

The installation and maintenance costs would bankrupt most tracks.

[Kelso]

class was routinely attacked by poplular authors and figure handicappers

CH,
While not "attacks," I've read several times on this board that a major weakness of class 'capping is that large differences in purse values ... differences that are now expanding due to slots/racinos ... make comparing horses' experiences much more tentative. Do you know of anywhere, in print or on the web, that the differences have been systymatically analyzed and reliably reduced to something such as a formula or chart?

Thank you.

[classhandicapper]

CH,
While not "attacks," I've read several times on this board that a major weakness of class 'capping is that large differences in purse values ... differences that are now expanding due to slots/racinos ... make comparing horses' experiences much more tentative. Do you know of anywhere, in print or on the web, that the differences have been systymatically analyzed and reliably reduced to something such as a formula or chart?

Thank you.

Kelso,

Purses are a problem. I rarely look at purses other than when comparing horses at the same curcuit.

If purses get larger at a specific track because of slots/racinos, IMO that track will eventually attract better horses to take a shot at that extra money. So the class structure will get more efficient over time. But at any given time it's a problem. A bigger problem is statebreds. Statebred purses are almost always out of line with the quality of the horses and generally remain that way. That's true even on the same circuit.

I think a better way to evaluate the quality of classes is to look at the average winning speed figure for the class. Beyer speed figure PARS are available for many classes at most circuits at DRF Simulcast Daily (which is a good source of info of all types).

If you want to get more sophisticated, you should keep the winning speed figure (and pace if you have the info) for each race on the circuits you play in a spreadsheet or in a database. That will give you the ability to get some class PAR averages that aren't kept by the DRF. It also gives you the ability to break them out by age, season (winter at AQU vs. summer at Saratoga for the same class example), and other categories to get a more refined feel for how various groups of horses stack up and improve over time with age.

That is always my starting point.

It also helps to keep track of shippers to your circuit that perform well or are beaten badly at a short price and any class moves on your circuit that are confusing to you.

Info like that can verify some of the PAR information or give you other clues about quality.

After that, you just have to get used to recognizing weak, average and strong fields within each class based on the actual horses in the race and how many fit ones there were.

After that, you start comparing their trips within the same race.

I hope that helps.

[Kelso]

I hope that helps.

It did ... very much. Thank you, CH. :ThmbUp:

[Overlay]

For dirt or turf races, data I'd seen showed a smooth impact-value progression in ranking a field (1/2/3/Front Half/Bottom Half) based on taking an average of the speed figures (Beyers) from those races out of a horse's last three starts that were run on the same type of surface as today's race or (if none of the horse's last three races were run on today's surface) the average from all of the last three races. (If the horse had started less than three times, the average was based on its races on the same surface as today, or (if none were on the same surface as today) on all of its one or two previous races.) Does anyone know whether a similar orderly progression applies either to synthetic figures as given, or to synthetic figures converted to a dirt equivalent?

[Grits]

Love ’em or hate ‘em, synthetic surfaces are here to stay.

I can only advise folks to adjust or end up driving yourselves nuts. They are never leaving this game.



Obviously, you agree with this. Let's all just "Get Over It."

http://www.ntra.com/content.aspx?type=feature&id=31989

[46zilzal]

The synthetics require a complete re-thinking of track maintenance.

[Valupix]

You could be right, but you could be wrong, and they'll never be installed enough places that you have to play them.

Sounds like you have your own solution. Pass over them when when you see them and quickly move on to what you like.

[Valupix]

Obviously, you agree with this. Let's all just "Get Over It."

http://www.ntra.com/content.aspx?type=feature&id=31989

Pretty much Grits.

If you can't get over it, forget about them and move on to what it is you can handle.

[cj]

Sounds like you have your own solution. Pass over them when when you see them and quickly move on to what you like.

How did you infer that from what I posted? I play them all the time. But, if someone doesn't like betting them, they don't have to do so.

You stated they are never going away, and that is certainly not the fact you pretend it to be.

[Valupix]

My apologies. A solution for those folks that continue to whine and cry about these surfaces.

Anything that has 3 positives to every 1 negative will always be far to good a thing to ever be scrapped. There is absolutely zero chance these surfaces will ever leave this game. Fight that fact all you want, but your wasting energy.

Personally I don't care one way or the other. I handicap them the same as I do any other track or race and my results are the same. Keep them or ditch them, it make no difference to me. But they ain't going away.

[Grits]

Pretty much Grits.

If you can't get over it, forget about them and move on to what it is you can handle.

It will be interesting to see how Keeneland responds upon the end of their meet, if their handle does finish down.

Trying to handicap their races has been difficult--to say the very least.

I wish tracks had looked into the alternative of using poly on their training tracks, as Keeneland did for some time before putting poly on the main. Though, of course, this was not an option for all racetracks.

[Valupix]

Grits: I live in Kentucky and I have not been allowed to wager on Keeneland or Turfway in months via internet or phone.

Not sure where you are based, but consider this. What if Del Mar decided to prohibit California on-line bettors from wagering on their signal. Or Saratoga to lock out the New York based gamblers, what effect would it have on their handle?

Most handle results I see from the vast majority of tracks these days have only a small amount of their handle from on track wagering. Most is either from intertrack wagering or on-line wagering. Cut out the on-line wagering that are based in the heart of your market and you will see significant handle decreases.

[asH]

It will be interesting to see how Keeneland responds upon the end of their meet, if their handle does finish down.

Trying to handicap their races has been difficult--to say the very least.

I wish tracks had looked into the alternative of using poly on their training tracks, as Keeneland did for some time before putting poly on the main. Though, of course, this was not an option for all racetracks.

check TVG Interview with Kee Pres, seems hes aware of the frustration, which is mainly why I posted it...says he will offer more information (?).....well anyway it may prove a bit easier for the figure cappers, seein that the biggest % of cappers use figures of some type or another... and I was just getting use to these 4K pk 4's.

I myself believe this anomaly is a result of the combination of good horses from different tracks, and less that of Polly alone. These environments (huge prices being the measure) happen during big card days such Derby or BC, and at Kee and Saratoga meets when the fields are less watered down... Most trainers/jocks/owners use Kee,Sar to strut their stuff

[andicap]

Grits: I live in Kentucky and I have not been allowed to wager on Keeneland or Turfway in months via internet or phone.

Not sure where you are based, but consider this. What if Del Mar decided to prohibit California on-line bettors from wagering on their signal. Or Saratoga to lock out the New York based gamblers, what effect would it have on their handle?

Most handle results I see from the vast majority of tracks these days have only a small amount of their handle from on track wagering. Most is either from intertrack wagering or on-line wagering. Cut out the on-line wagering that are based in the heart of your market and you will see significant handle decreases.

Would you happen to be a certain HTR user who has done well in tournaments? I remember that handle from the old Netcapper board and wondering if it's the same person -- he was from Kentucky as well.

[Valupix]

Would you happen to be a certain HTR user who has done well in tournaments? I remember that handle from the old Netcapper board and wondering if it's the same person -- he was from Kentucky as well.

The same.

[highnote]

46,

How are the tracks going to settle in when they keep changing them?

As for the original formula I posted in this thread, a customer pointed out it was wrong. Shows what happens when I try to use my memory! The actual formula is this:

=80+(Beyer-80)*(40/30)

CJ

What is the significance of 40/30?

You could have just written "*1.33". However because you found 40/30 to be important, perhaps you included the two numbers in your formulation for clarity and as a reminder?

Thanks.

John

[cj]

CJ

What is the significance of 40/30?

You could have just written "*1.33". However because you found 40/30 to be important, perhaps you included the two numbers in your formulation for clarity and as a reminder?

Thanks.

John

Yes, just a reminder, but 1.33 is rounded and I'm a bit anal that way anyway.

I toyed with different things, that is where 40 and 30 came from. Basically, 80 is even on both surfaces, and the 40/30 comes from a 120 on dirt being equal to 110 on other surfaces, and a 40 on dirt equalling 50 on others.

[andicap]

Just a word folks about Valupix -- I have "known" him on various boards and followed his performances in tournaments over the years and I can tell you he is a very, very sharp handicapper, always worth listening to.

I won't out his name -- so I'll just say...

Welcome to PA and I hope you stay around -- even through the b.s.
:ThmbUp:

[PaceAdvantage]

Welcome to PA and I hope you stay around -- even through the b.s.
:ThmbUp:And if you notice any b.s. that I fail to pick up, please, please let me know, and I will gladly bring the rake ....

[Semipro]

Yes, factually substantiated things.So if it's in print it's a substantiated fact.

[46zilzal]

So if it's in print it's a substantiated fact.
No, when it stands the test of independent observation.

Peer review.

[thelyingthief]

Speed figures provide far too small an edge: why do people even bother with them anymore?

Much as I admire the mathematician who informs me Pi is some utter number on the 13 millionth iteration, for my purposes, 3.14159 gets it done. Similarly, refining a number largely composed of incidentals (like who bumped whom at the quarter pole) hardly gives it any greater predictive power than its more coarse first, seems to me.

tlt

[Tom]

Speed figures provide far too small an edge: why do people even bother with them anymore?



Because your premise in incorrect. They might not prvide you with a good edge, but you can only speak for yourself. And lumping all figures under one umbrella is far to braod to really have any meaning.

[asH]

Because your premise in incorrect. They might not prvide you with a good edge, but you can only speak for yourself. And lumping all figures under one umbrella is far to braod to really have any meaning.

Tom,
I believe that was his point. Speed numbers, on their own, can only augment (if you believe) a cappers decision. A good capper as yourself wouldnt bet on the horse with the highest number, without scrutiny.

[Tom]

There you go...speed figures are bad because....then insert your own idea on how to use them. I let other factors augment my numbers.

[asH]

There you go...speed figures are bad because....then insert your own idea on how to use them. I let other factors augment my numbers.

I didnt say bad... for me not logical
its like stuffing a closet of suits into a single suitcase, describing the beauty of a rose with a word. I did explore them in my handicapping youth, tried to incorporate my knowledge in a number, spent a bit of computer time on it. but for me I could never be satisfied how a single number described the 'world'.
. did you have Desert Key?

[proximity]

question about this thread and the converter box on cj's site:

wouldn't it be a step better to obtain median or average beyer figures for each individual synthetic surface and turf course and then have the chart be based on deviations from these track specific numbers as opposed to using 80 for all situations?

for example, suppose that before polytrack that the average beyer at turfway was say 60 and after polytrack was installed the quality of racing stayed basically the same. nothing really changed, but now we're changing the average to 53. and if the average quality of racing was above 80 at santa anita, we'd be upgrading all of those horses. it seems to me that in each individual situation that half of the figures should be upgraded and half downgraded. does this make sense?

also, are there any database gurus out there that could produce a list of average beyer by track and surface? i still have a database of henry kuck ratings, although it's only broken down by track.

[cj]

question about this thread and the converter box on cj's site:

wouldn't it be a step better to obtain median or average beyer figures for each individual synthetic surface and turf course and then have the chart be based on deviations from these track specific numbers as opposed to using 80 for all situations?

for example, suppose that before polytrack that the average beyer at turfway was say 60 and after polytrack was installed the quality of racing stayed basically the same. nothing really changed, but now we're changing the average to 53. and if the average quality of racing was above 80 at santa anita, we'd be upgrading all of those horses. it seems to me that in each individual situation that half of the figures should be upgraded and half downgraded. does this make sense?

also, are there any database gurus out there that could produce a list of average beyer by track and surface? i still have a database of henry kuck ratings, although it's only broken down by track.

It just doesn't really work that way. Cheaper horses run faster than usual, and better horses run slower. It has been that way on turf for forever as well. Using the "track average" produces pretty poor results.

In the end, the adjustment only really helps if you think a horse is going to be as good on one surface as he is on the other. Those instances are pretty rare though.

[proximity]

. Cheaper horses run faster than usual, and better horses run slower.


i can appreciate that the gap in the raw poly/turf numbers can be tighter, but cheaper and better relative to what? i'd say the other horses/ races on the card that went into making the variant, yes? and if the (dirt equivalent) pars for all of these races were above or below 80 then we should increase or decrease all of these figures accordingly? i'm not trying to be difficult and can also appreciate that real world experiences should take precedence over theory, but i am struggling to understand this.

[fmolf]

It just doesn't really work that way. Cheaper horses run faster than usual, and better horses run slower. It has been that way on turf for forever as well. Using the "track average" produces pretty poor results.

In the end, the adjustment only really helps if you think a horse is going to be as good on one surface as he is on the other. Those instances are pretty rare though.
are you working form any pars at all when it comes to the poly numbers...wouldn't it be easier to compile par times for tracks with poly as if they were dirt races and then compare and come up with a multiplier or change over ratio?....i have been shying away from the poly tracks.... i find this very interesting....

[cj]

are you working form any pars at all when it comes to the poly numbers...wouldn't it be easier to compile par times for tracks with poly as if they were dirt races and then compare and come up with a multiplier or change over ratio?....i have been shying away from the poly tracks.... i find this very interesting....

This is exactly what I have been doing, and this is where the 80 "switch point" came from in my formulas. Pars that were below 80 were generally inflated with the move to synthetics, while pars that were below 80 were deflated. It didn't matter what track I studied, it was very consistent.

[proximity]

are you working form any pars at all when it comes to the poly numbers...wouldn't it be easier to compile par times for tracks with poly as if they were dirt races and then compare and come up with a multiplier or change over ratio?....i have been shying away from the poly tracks.... i find this very interesting....

it definitely is interesting.

if penn national switched over to polytrack tomorrow, how would you adjust the figures? would the par figures drop for 95 % of the races? what about belmont park?

[fmolf]

This is exactly what I have been doing, and this is where the 80 "switch point" came from in my formulas. Pars that were below 80 were generally inflated with the move to synthetics, while pars that were below 80 were deflated. It didn't matter what track I studied, it was very consistent.
then an 80 is an 80 no matter the surface and adjustments are made on the scale in almost equal proportions going up and down?...seems almost to neat and tidy...but your work is very commendable.... :ThmbUp:

[cj]

it definitely is interesting.

if penn national switched over to polytrack tomorrow, how would you adjust the figures? would the par figures drop for 95 % of the races? what about belmont park?

For mine, almost 100% sure the Penn pars would be higher since polytrack seems to inflate cheaper horses.

Belmont, the upper tier would be lower, the NY breds definitely higher. :)

[proximity]

For mine, almost 100% sure the Penn pars would be higher since polytrack seems to inflate cheaper horses.

Belmont, the upper tier would be lower, the NY breds definitely higher. :)

ok, so we're kinda coming at this from opposite directions. you'd change the class pars, use the normal beyer speed chart, and then adjust the resulting figures whereas i'd keep the class pars constant and just change the speed chart (increasing the value of 1/5 second).

in the winning horseplayer beyer talked about comparing all of his figures to the par figures at the end of each month to account for their tendency to decline as time goes by. perhaps when/if switches to artificial surfaces occur he could do something similar to ultimately determine what the speed chart for that particular track should be? he could start out just using his normal speed chart, then compare the figures of each class of horse to the median at that track, and a weighted average of such deviations compared to what they would normally be would tell him how much to adjust his speed charts for future use?

[cj]

ok, so we're kinda coming at this from opposite directions. you'd change the class pars, use the normal beyer speed chart, and then adjust the resulting figures whereas i'd keep the class pars constant and just change the speed chart (increasing the value of 1/5 second).


Yep, exactly the same thing in reality.

On another note, Beyer adjusts his turf figures by using a different beaten lengths chart. While this does help a little, it only partially offsets the figure making differrence. It changes the figures for non-winners, but the biggest problem is the winners!