Saturday Beyers

[cj]

Vosburgh, Girolamo, 105
Flower Bowl, Ave, 97
Beldame, Life at Ten, 100
Turf Classic, Winchester, 100
JCGC, Haynesfield, 107
Cotillion, Havre de Grace, 104
Norfolk, Jaycito, 87
Lady's Secret, Zenyatta, 100
Yellow Ribbon, Hibaayeb, 92
Goodwood, Richard's Kid, 105

[PhantomOnTour]

A 107 for Haynesfield? That seems high.
I got the JCGC in 110-104-102 Quirin style which is far below the Gr1 par. Further off the par than Beyer has it....hmmmm.
Races 3 and 8 were both similar in their deviance from par (didn't use the off turf race #5 for 2yr olds) but the JCGC was much slower compared to par; esp on final time.

[cj]

A 107 for Haynesfield? That seems high.
I got the JCGC in 110-104-102 Quirin style which is far below the Gr1 par. Further off the par than Beyer has it....hmmmm.
Races 3 and 8 were both similar in their deviance from par (didn't use the off turf race #5 for 2yr olds) but the JCGC was much slower compared to par; esp on final time.

10f is tough at Belmont because of the weird starting position.

[gm10]

Vosburgh, Girolamo, 105
Flower Bowl, Ave, 97
Beldame, Life at Ten, 100
Turf Classic, Winchester, 100
JCGC, Haynesfield, 107
Cotillion, Havre de Grace, 104
Norfolk, Jaycito, 87
Lady's Secret, Zenyatta, 100
Yellow Ribbon, Hibaayeb, 92
Goodwood, Richard's Kid, 105

Haynesfield 107 and Zenyatta 100?

That is one of the funniest things I've read in a while.

[Charlie D]

Haynesfield 107 and Zenyatta 100?

That is one of the funniest things I've read in a while.


Not really, when you consider Haynesfield was not beating tomato cans.

[the little guy]

Haynesfield 107 and Zenyatta 100?

That is one of the funniest things I've read in a while.


It's hard to argue with you. Given who finished second in each race, how could he not have run 15 to 20 points better.

It's just a racing oddity I guess.

[gm10]

It's hard to argue with you. Given who finished second in each race, how could he not have run 15 to 20 points better.

It's just a racing oddity I guess.

You know I suspect that you are making these numbers for Beyer to earn some pocket money, that's how bad they are.

[cj]

You know I suspect that you are making these numbers for Beyer to earn some pocket money, that's how bad they are.

So, you think Switch is suddenly better than Blame, or at least ran faster yesterday?

[the little guy]

So, you think Switch is suddenly better than Blame, or at least ran faster yesterday?


Wait, he thought they numbers were wrong the other way?

Come on....even he's not that clueless.

[gm10]

So, you think Switch is suddenly better than Blame, or at least ran faster yesterday?


I don't think that Haynesfield ran 7 points quicker than Zenyatta.

Blind Luck ran a 100 in the Alabama. Blind Luck was beaten a few months earlier by Switch. I'd make Switch 101 and Zenyatta 105 yesterday.

I'd make Haynesfield about 104. Blame ran a bad race. Maybe he doesn't get 10F, maybe he needs a target to run that. Who knows, but he was certainly uninspiring.

[cj]

I don't think that Haynesfield ran 7 points quicker than Zenyatta.

Blind Luck ran a 100 in the Alabama. Blind Luck was beaten a few months earlier by Switch. I'd make Switch 101 and Zenyatta 105 yesterday (don't forget the extra 3lbs) she was carrying.

I'd make Haynesfield about 104. Blame ran a bad race. Maybe he doesn't get 10F, maybe he needs a target to run that. Who knows, but he was certainly uninspiring.

You can't use weight when talking about Beyers, because he doesn't use it.

How could you get four points between Switch and Zenyatta? That isn't the Beyer scale. If Zenyatta gets a 105, Switch gets a 104...period. If you give Haynesfield a 104, Blame gets a 99.

Now, maybe with your insights you could write Beyer and have him adjust his methodology, but until then his charts are pretty set as the basis of all his numbers.

[Charlie D]

I use with and without weight in the calcs and Haynesfield comes back with the better fig.


Attach a FWIW to the above.

[gm10]

You can't use weight when talking about Beyers, because he doesn't use it.

How could you get four points between Switch and Zenyatta? That isn't the Beyer scale. If Zenyatta gets a 105, Switch gets a 104...period. If you give Haynesfield a 104, Blame gets a 99.

Now, maybe with your insights you could write Beyer and have him adjust his methodology, but until then his charts are pretty set as the basis of all his numbers.

They aren't my insights, they are the accumulated knowledge of racing horses for 4 centuries. They have been proven empirically and theoretically.

Anyway, by all means let's stick to the scale and give Switch a 104. She ran a cracker and has taken the measure of a filly rated 100 before, so why not.

[the little guy]

You can't use weight when talking about Beyers, because he doesn't use it.

How could you get four points between Switch and Zenyatta? That isn't the Beyer scale. If Zenyatta gets a 105, Switch gets a 104...period. If you give Haynesfield a 104, Blame gets a 99.

Now, maybe with your insights you could write Beyer and have him adjust his methodology, but until then his charts are pretty set as the basis of all his numbers.


gm10 knows better.

[gm10]

So, you think Switch is suddenly better than Blame, or at least ran faster yesterday?

How do you rate both winners, btw?

[cj]

They aren't my insights, they are the accumulated knowledge of racing horses for 4 centuries. They have been proven empirically and theoretically.

Anyway, by all means let's stick to the scale and give Switch a 104. She ran a cracker and has taken the measure of a filly rated 100 before, so why not.

The thread is about Beyers, so I think it is fair to stick to the Beyer methodology. Maybe you can start a thread about the GM10 ratings for the Lady's Secret and JCGC and explain how your ratings are made. Until then, it is really kind of pointless to talk about them.

[cj]

How do you rate both winners, btw?

I do my figures two days after the races, always.

[the little guy]

If Zenyatta had run in yesterday's Jockey Club Gold Cup, and everything else had been the same, she would not have won. It's really as simple as that. Whether or not she is a better horse than Haynesfield is a different discussion.

[gm10]

The thread is about Beyers, so I think it is fair to stick to the Beyer methodology. Maybe you can start a thread about the GM10 ratings for the Lady's Secret and JCGC and explain how your ratings are made. Until then, it is really kind of pointless to talk about them.

OK, well, with all due respect to the BSF which I genuinely find excellent under most circumstances, I don't see how Haynesfield can be rated 7 points ahead of Zenyatta. That just does not make any sense to me.

I'm not going to give the details of how I calculate mine, but (in general) they take pace shapes into account. If a winner closed strongly into a slow pace, then the last part of the race will count for more than the early or middle part. Just using final times is too rigid an approach imo.

It's like the old discussion in statistics between linear regression and non-parametric regression. When you have a set of data, do you say 'well, I've got this great linear regression package that I learned in uni, and the data is just going to have to fit a line', or do you say 'OK, data, I'm totally open about this, data, you tell me whether you follow a linear model or not'? Linear regression will work in many circumstances, but non-parametric regression will almost always work. It's a bit more work, you need to be more open-minded, but it is worth the effort.

It's the same with speed figures. Do you just stick to final times or are you willing to let the pace, surface, course, going determine what the speed figure actually was.

Granted, it is technically not a speed figure anymore, but who are we kidding, which useful speed figures are purely speed-based.

[cj]

Granted, it is technically not a speed figure anymore, but who are we kidding, which useful speed figures are purely speed-based.

This is just further proof people try to make Beyer figures out to be more than even Beyer himself does.

Once you start "projecting" better numbers because of slower paces, you are guessing how fast the race might have been. It works sometimes, other times it doesn't. Both methods have ups and downs.

Personally, I do numbers differently according to pace and surface, but I won't pretend there aren't problems with that method either.

[Charlie D]

Mmm. i'd not be so sure TLG.

Similar Early fractions and she was a few lengths closer to the leader than Blame at the mile in Lady Secret (which was slightly faster)

Zen has run run 10f in around 1.21.00


So at a guess you would have to say she would overcome Haynesfield unless Haynesfield found more under pressure from Zen. May have been very close actually if the latter did happen to take place.


Thats my 2 cents worth anyway.

[gm10]

This is just further proof people try to make Beyer figures out to be more than even Beyer himself does.

Once you start "projecting" better numbers because of slower paces, you are guessing how fast the race might have been. It works sometimes, other times it doesn't. Both methods have ups and downs.

Personally, I do numbers differently according to pace and surface, but I won't pretend there aren't problems with that method either.

I'm not guessing, I'm letting the data tell me how the race was won today and how it was historically won over C&D.

In the case of Haynesfield, keeping this principle in mind, I fail to appreciate his effort. He had dead easy fractions, no pressure, a neutral surface, was not running particularly fast at any stage of the race and yet he finished in one of the slower times of the race's history. 107 just doesn't sound right.

[cj]

I'm not guessing, I'm letting the data tell me how the race was won today and how it was historically won over C&D.



What I am saying about guessing is that you can project all you want that a faster pace could have led to a faster time, but while it may be right, it may be wrong, and it certainly is wrong for some horses in every race.

The final time of the JCGC compared to history means nothing to me. I'd rather compare it to the other races yesterday.

[gm10]

What I am saying about guessing is that you can project all you want that a faster pace could have led to a faster time, but while it may be right, it may be wrong, and it certainly is wrong for some horses in every race.

The final time of the JCGC compared to history means nothing to me. I'd rather compare it to the other races yesterday.

I'm not projecting anything.

Let me know once you've done Haynesfield's number, I'd like to see a justification of that 107. Maybe your projection method can provide the answer.

[cj]

I'm not projecting anything.

Let me know once you've done Haynesfield's number, I'd like to see a justification of that 107. Maybe your projection method can provide the answer.

Well, again, you are getting away from a pure speed figure then...final time and variant. If you are using the pace scenario you can't really compare it to Beyer, simple as that. It sounds like to me you are projecting what horses could have done if the pace wasn't slow.

I haven't offered an opinion on the JCGC...I haven't even looked.

[Charlie D]

The figs are only relevant when you compare them to others and in this case comparing the BSF awarded Haynesfield, he ran better than Zen according to BSF methodology yesterday.

If you want to compare Haynsefield to other JCGC winners then you need thier BSFs, but little point in that as they are no longer around. :)

[gm10]

Well, again, you are getting away from a pure speed figure then...final time and variant. If you are using the pace scenario you can't really compare it to Beyer, simple as that. It sounds like to me you are projecting what horses could have done if the pace wasn't slow.

No that's not what I'm doing at all.

A pure speed figure is not 'pure speed' if it based on final time and variant. Speed is distance/time. If you adjust time (up to several seconds) first, it isn't speed anymore.

I haven't offered an opinion on the JCGC...I haven't even looked.

I know, I am just interested in what you think of the 107 once you have.

[Charlie D]

which useful speed figures are purely speed-based.

Mine and they help me find Arc winners too.

[cj]

No that's not what I'm doing at all.

A pure speed figure is not 'pure speed' if it based on final time and variant. Speed is distance/time. If you adjust time (up to several seconds) first, it isn't speed anymore.



I know, I am just interested in what you think of the 107 once you have.

In the US, final time adjusted by variant is what people refer to as a speed figure. Since I'm in the US, that is what I'll go with for now.

A 107 is low historically for the race, by a lot.

[Charlie D]

I once read this by Charles Mingus


"Overcomplication is commonplace"


I thought, mmm, this applies to many, many thoroughbred handicappers i think. :)

[gm10]

In the US, final time adjusted by variant is what people refer to as a speed figure. Since I'm in the US, that is what I'll go with for now.

A 107 is low historically for the race, by a lot.

Fine, it is like 'processed cheese', I suppose. Yes it's cheese, but it's not, really.

[illinoisbred]

I'm curious about the Beyer figure for Redding Colliery-the winner of the Hawthorne Gold Cup,run over a very slow tiring surface. I make my own Quirin-style figures and this looks like the biggest I've assigned all year here-116.

[Charlie D]

Today i read this by thelittleguy

The silliness never ends.



I had to agree with him.

[senortout]

Damn....7 to 1 on a horse I love! I been waiting all day for odds like these...and I go into my adw betting screen to punch him hard, thinking how unbelievable 7 to 1, 7 to 1, 7 to 1, I never get 7 to 1 on ANY horse I've doped out this well.

Long story short, perhaps you've guessed by now, I bet the odds instead of the post position which was 6......

There'll come another day.

senortout

[gm10]

I'm curious about the Beyer figure for Redding Colliery-the winner of the Hawthorne Gold Cup,run over a very slow tiring surface. I make my own Quirin-style figures and this looks like the biggest I've assigned all year here-116.

I have an identical number for him as I have for Haynesfield. No BSF published yet.

[Charlie D]

Damn....7 to 1 on a horse I love! I been waiting all day for odds like these...and I go into my adw betting screen to punch him hard, thinking how unbelievable 7 to 1, 7 to 1, 7 to 1, I never get 7 to 1 on ANY horse I' doped out this well.

Long story short, perhaps you've guessed by now, I bet the odds instead of the post position which was 6......

There'll come another day.

senortout


We like to think we are, but we are not perfect senortout. :)

[bisket]

i think the beyers for both of these races are about right where they should be. the mistake you don't want to make is to compare a beyer that was earned on turf or poly track to a beyer earned on dirt. a beyer earned on poly will always be lower than one earned on dirt. the problem with beyers on poly is in the variant. i do think if you are comparing say rich's kid's performance with zenyatta's the beyers are accurate. i wouldn't dream of comparing zenyatta's and rich's performance with haynesfield's race using any speed figures for that matter!!! looking at handicapper's comments on these races shows who is stupid enough to do this. i would suggest anyone reading them to disregard them if they plan to make any money in the upcoming cup races.

i think poly track has really highlighted the problem's with speed figures when using them to judge the speed of races between different surfaces. there are very few horses that can run equally as well on both turf and dirt. although, its not as difficult for horses to transition from poly to dirt and vice versa. so its now become more important to be able to judge the speed of the races. i see both the andy's have found this to much of a challenge for them :p . maybe someday they'll get over their frustration? or find another job.....

[cj]

i think poly track has really highlighted the problem's with speed figures when using them to judge the speed of races between different surfaces.

I don't know many (any?) good handicappers that do this.

[WinterTriangle]

Vosburgh, Girolamo, 105
Flower Bowl, Ave, 97
Beldame, Life at Ten, 100
Turf Classic, Winchester, 100
JCGC, Haynesfield, 107
Cotillion, Havre de Grace, 104
Norfolk, Jaycito, 87
Lady's Secret, Zenyatta, 100
Yellow Ribbon, Hibaayeb, 92
Goodwood, Richard's Kid, 105

Interesting to compare BSFs to Performance Figs, so for fun, I did.

Haynesfield ran a sizzling -88 and Richard's Kid a sizzling -82.

I say sizzling because they are very high PFs.

Zen, Gio Ponti, Big Brown's etc, usually avg in the -75 area

[Steve R]

Interesting to compare BSFs to Performance Figs, so for fun, I did.

Haynesfield ran a sizzling -88 and Richard's Kid a sizzling -82.

I say sizzling because they are very high PFs.

Zen, Gio Ponti, Big Brown's etc, usually avg in the -75 area
The Haynesfield -88 correlates roughly with BSF 114 which is actually the JCGC par figure. Zenyatta earned a -60 in the Lady's Secret which correlates with BSF 104. That's about a 6 1/2 or 7-length differential according to Beyer. But here's the kicker. BRIS gave Haynesfield a huge 116, roughly equivalent to a BSF 118, and Zenyatta a 101, equivalent to a BSF 101. IOW, the BRIS people would argue that the differential between the two is much greater than indicated by BSFs or Performance Figures.

[gm10]

The Haynesfield -88 correlates roughly with BSF 114 which is actually the JCGC par figure. Zenyatta earned a -60 in the Lady's Secret which correlates with BSF 104. That's about a 6 1/2 or 7-length differential according to Beyer. But here's the kicker. BRIS gave Haynesfield a huge 116, roughly equivalent to a BSF 118, and Zenyatta a 101, equivalent to a BSF 101. IOW, the BRIS people would argue that the differential between the two is much greater than indicated by BSFs or Performance Figures.

I rated Zenyatta's performance higher (=better) than Haynesfield's.

This will obviously invite a lot of ridicule, so I'll add straight away that it's based on race dynamics not just final times. If the race is won by a late closer, how fast did he/she actually close, and off which fractions set by the leaders? If the horse is won by the pace setter, which fractions did he go, and how well did he finish the race given the those early fractions.

To extrapolate such figures to another race is of course another matter, but in this case of 10F at Churchill, it probably bodes well for Zenyatta.

[Charlie D]

I rated Zenyatta's performance higher (=better) than Haynesfield's.

This will obviously invite a lot of ridicule, so I'll add straight away that it's based on race dynamics not just final times. If the race is won by a late closer, how fast did he/she actually close, and off which fractions set by the leaders? If the horse is won by the pace setter, which fractions did he go, and how well did he finish the race given the those early fractions.

To extrapolate such figures to another race is of course another matter, but in this case of 10F at Churchill, it probably bodes well for Zenyatta.


Explanation please.

[gm10]

Explanation please.

Think of a final time-based speed figure S as

S = 1/3 * early + 1/3 * middle + 1/3 late (I)

where early = middle = late = S.

The quantities on both sides of (I) are the same, but the right-hand side allows for flexibility. Which is necessary - the 1/3 doesn't apply to every track, surface, distance, pace scenario. A turf route on firm ground would be more like 1/4 1/4 1/2.

Of course you need to estimate the replacements for 1/3 1/3 1/3 and have some numbers for early/middle/late.

[delayjf]

The Haynesfield -88 correlates roughly with BSF 114 which is actually the JCGC par figure. Zenyatta earned a -60 in the Lady's Secret which correlates with BSF 104. That's about a 6 1/2 or 7-length differential according to Beyer

personally, I'm always a bit dubious when comparing SF of horses coming from different distances when the distance difference is greater than 1/8th of a mile.

I don't dispute the figures, but I don't think Zenyatta's SF represents her ability - 1 1/16 is not her best distance and from my perspective, she was only running hard for the last 1/16.

I do my figures two days after the races, always.
How many hours a day does it take for you to calculate your figures??

[bks]

cj: I'm not familiar with the details of Beyer's approach, but I thought it involved comparing the final times of different races at the same distance at the same track on a given day [if that's clear].

From memory, the Lady's Secret was the only race at 1 1/16 that day, and it went in 142.9 or so.

The Goodwood was the only race at 1 1/8 that day, and it went about 6.2 slower, at approx 149.1.

Zenyatta's final time was faster than Richard's Kid's when broken down to average speed (she averaged just under 6.06 per 1/16th during the race) while Richard's Kid averaged slightly higher (just over 6.06).

So how is his figure 5 points higher? She went faster and finished the race at a rate that, it seems clear, would have put her under the wire at 1/8 in less time. Zenyatta, as we know ran the final 1/16 in less than six seconds. I take it that's irrelevant in Beyer's calculations?

Thx.

[cj]

cj: I'm not familiar with the details of Beyer's approach, but I thought it involved comparing the final times of different races at the same distance at the same track on a given day [if that's clear].

From memory, the Lady's Secret was the only race at 1 1/16 that day, and it went in 142.9 or so.

The Goodwood was the only race at 1 1/8 that day, and it went about 6.2 slower, at approx 149.1.

Zenyatta's final time was faster than Richard's Kid's when broken down to average speed (she averaged just under 6.06 per 1/16th during the race) while Richard's Kid averaged slightly higher (just over 6.06).

So how is his figure 5 points higher? She went faster and finished the race at a rate that, it seems clear, would have put her under the wire at 1/8 in less time. Zenyatta, as we know ran the final 1/16 in less than six seconds. I take it that's irrelevant in Beyer's calculations?

Thx.

All that matters is final time. Of course horses of equal ability at a shorter distance should run a faster average speed than those at a longer distance.

Zenyatta's final time of 1:42.97 would be about a 107 on the Beyer chart. Richard's Kid's final time of 1:49.18 would be a 108 on the Beyer chart.

There are some adjustments to the Beyer chart, fairly new, for synthetic races. The exact adjustments have not been published to my knowledge, but I've figured them out. Beyer used a variant of about fast 8 points, so the adjusted for variant figures would have been 99 for Zenyatta and 100 for Richard's Kid. After the adjustment for synthetics, they move to 103 for Z, and 104 for RK.

Now, why they eventually changed to 101 and 105, I have no idea. You'd have to ask Beyer or whoever does the SoCal figures. There clearly is not a 4 point difference on the final time, somewhere another "adjustment" was made.

[bks]

thanks, cj, very helpful.

[Valuist]

[QUOTE=cj]You can't use weight when talking about Beyers, because he doesn't use it.

How could you get four points between Switch and Zenyatta? That isn't the Beyer scale. If Zenyatta gets a 105, Switch gets a 104...period. If you give Haynesfield a 104, Blame gets a 99.

Now, maybe with your insights you could write Beyer and have him adjust his methodology, but until then his charts are pretty set as the basis of all his numbers

I think he thinks the transitive property somehow relates to speed handicapping. If A beats B and B beats C, A must be better than C? Maybe in Algebra class....not at the track.

[gm10]

[QUOTE=cj]You can't use weight when talking about Beyers, because he doesn't use it.

How could you get four points between Switch and Zenyatta? That isn't the Beyer scale. If Zenyatta gets a 105, Switch gets a 104...period. If you give Haynesfield a 104, Blame gets a 99.

Now, maybe with your insights you could write Beyer and have him adjust his methodology, but until then his charts are pretty set as the basis of all his numbers

I think he thinks the transitive property somehow relates to speed handicapping. If A beats B and B beats C, A must be better than C? Maybe in Algebra class....not at the track.

No, I don't believe in the 'transitive property' at the track. I think that weight matters. Zenyatta beats St Trinians by a nose but carries 9lbs more. Who in their right mind is going to say that St Trinians deserves an identical speed figure? Any method that rates them equally is ultimately flawed imo. The same principle applies to Zenyatta/Switch (to the lesser extent of 3lbs). That's all I was saying.

[rastajenk]

Then you're talking about a rating that might be called a power rating, or a performance rating, or some other name. But when horses are a nose apart, then I'd give them the same speed rating. And I am of a right mind; at least, I was the last time I checked.

How many times does this have to be explained? If weight matters to you, that's fine. But we're talking about stopping the timer and figuring with beaten lengths; that's a speed rating, which, if I recall correctly was the point of the original post in this thread.

[OTM Al]

If weight matters so much then you need the weights of the horses themselves, which you can't get. The weight loss caused by Lasix is going to throw any such attempt into great question as well. This more than overcomes some five or ten pound jockey weight difference. I believe it is weight that reduced the performace of Rachel this year more than anything. She was carrying quite a bit more this year than last due to maturation. Saw her up close both this and last year at Saratoga in the paddock and the difference was marked. If you really think 9 lbs can make such a difference, just throw a couple hundred on the animal's frame and tell me it won't be a factor.

Such small amounts of difference to an animal of the size of a horse is minimal at best. Zenyatta is a very big horse and therefore should well be able to carry a proportionally larger amount of weight than smaller rivals all else equal. She's never been properly handicapped compared to handicap conditions in England where you know well that high weights still very often win.

[gm10]

Then you're talking about a rating that might be called a power rating, or a performance rating, or some other name. But when horses are a nose apart, then I'd give them the same speed rating. And I am of a right mind; at least, I was the last time I checked.

How many times does this have to be explained? If weight matters to you, that's fine. But we're talking about stopping the timer and figuring with beaten lengths; that's a speed rating, which, if I recall correctly was the point of the original post in this thread.

Well I call it speed rating, Racing Post call their weight-adjusted Topspeed speed ratings. It's as valid as labeling the going-adjusted BSF as a speed rating. They are all speed-based, none of them are purely speed ratings.

A performance rating is something completely different, it is certainly nothing like that. A power rating, well I don't know the definition of that.

Anyway, yes the thread is about BSF, and I disagreed with them, nothing else.

[gm10]

If weight matters so much then you need the weights of the horses themselves, which you can't get. The weight loss caused by Lasix is going to throw any such attempt into great question as well. This more than overcomes some five or ten pound jockey weight difference. I believe it is weight that reduced the performace of Rachel this year more than anything. She was carrying quite a bit more this year than last due to maturation. Saw her up close both this and last year at Saratoga in the paddock and the difference was marked. If you really think 9 lbs can make such a difference, just throw a couple hundred on the animal's frame and tell me it won't be a factor.

Such small amounts of difference to an animal of the size of a horse is minimal at best. Zenyatta is a very big horse and therefore should well be able to carry a proportionally larger amount of weight than smaller rivals all else equal. She's never been properly handicapped compared to handicap conditions in England where you know well that high weights still very often win.

Yes, the official BHB view on the effect of weight is an OVER-estimate, but most authors would still agree that the used weight-to-lengths conversion is about 75% correct.

You are definitely correct that the mass of the horse makes a big difference, and that we don't have that data (they do in Hong Kong actually). However, once a horse has established a level of ability, and we know that it is carrying extra weight, then we can still estimate the effect of that extra weight. There are quite a few authors who have written down the mathematical equation of the effect, so it easy to have a reasonable estimate of it.

Furthermore, as a recent study highlighted, the effect of DEAD weight (=jockey) plays (relatively speaking) a much bigger role than the actual body mass of the horse. This is confirmed by observation imo. Horses who carry less weight are better at getting out the gate, choosing position, going into a gap etc.

I think the Zenyatta/St Trinians example is a perfect example. It just took Zenyatta a few strides longer to get going, and it showed at the finish. Try running up the stairs with a bottle of coke in a rucksack on your back and then without. There is a difference.

[Tom]

[QUOTE=cj]You can't use weight when talking about Beyers, because he doesn't use it.

How could you get four points between Switch and Zenyatta? That isn't the Beyer scale. If Zenyatta gets a 105, Switch gets a 104...period. If you give Haynesfield a 104, Blame gets a 99.

Now, maybe with your insights you could write Beyer and have him adjust his methodology, but until then his charts are pretty set as the basis of all his numbers

I think he thinks the transitive property somehow relates to speed handicapping. If A beats B and B beats C, A must be better than C? Maybe in Algebra class....not at the track.

It's not unusual to see a Beyer broken out from the rest of the card. They were are two different distances, and I have seen lots of cases where that happens. He also makes periodic adjustments when his quality control tells him that some tracks or distances are our of wack. Save a Winners Book and then look back at the same track in 9 months - you are apt to see some differences as further information is analyzed.

[RXB]

[QUOTE=Valuist]

No, I don't believe in the 'transitive property' at the track. I think that weight matters. Zenyatta beats St Trinians by a nose but carries 9lbs more. Who in their right mind is going to say that St Trinians deserves an identical speed figure? Any method that rates them equally is ultimately flawed imo. The same principle applies to Zenyatta/Switch (to the lesser extent of 3lbs). That's all I was saying.

For the record, the margin was 1/2 length between Z and St T, not a nose.

There are all kinds of things that are not taken into account besides just weight, several of which I think impact finish positions and final times more than weight does. Pace, bias, trip, etc. can all have a significant effect on a horse's adjusted final-time figure. Beyer has never claimed that his number is anything other than an adjusted speed figure.