Beyer Figs Bogus?

[classhandicapper]

I don't think there's a right or wrong answer to whether you should be using figures or not.

On the downside, everyone is using them. So if you use them they will drag you towards the same line of thinking everyone else has even in races where there are figure maker errors/disputes or the hard to quantify impacts of the track, pace, race development etc.. make the figures less representative of ability.

Looking at who is beating who and how in a qualitative fashion without much regard to the times will put you in the minority and sometimes get better prices, but there are clearly times there isn't enough information to make those subjective judgments as well as the clock.

This has been an intellectual battle for me for decades. They are 2 different ways of measuring the same thing with different pluses and minuses.

1. Which is the better way to go?

2. Can you combine them more effectively than either as a stand alone?

3. Can you use one method on some races and the other on other races more effectively?

This is what the evidence suggests to me so far.

All the database studies I have been doing since 2014 (and ongoing) using good quality automated class ratings and publicly available speed figures suggest that in terms of accuracy they are very similar (speed figures less than 1% better). One would hope that a good class handicapper using experienced judgement could make up the gap over automated numbers. So they are more or less the same.

In terms of value, class is slightly better (a few % higher ROI).

If you combine them, you can increase win% over either as a stand alone, but you will lower the ROI.

Speed figures definitely help lot with lightly raced horses (maidens, ALW etc..) where the quality varies widely at the same class level and information is limited for a more subjective analysis.

That has been my experience as a gambler also.

It doesn't really matter all that much (though I prefer weighting subjective quality over speed for the better ROI). The thing is, analyzing fields qualitatively on your own is a LOT more time consuming and difficult than looking at speed figures someone else calculated. It can take many years to get good at classing in the same way it can take many years to make good figures, but good figures are available publicly.

[CincyHorseplayer]

Originally Posted by the little guy

Here's what successful players actually don't do....ignore relevant factors that can be at least reasonably evaluated, whatever they may be.

They also don't go on message boards and dismiss proven methods of assessing horse's relative talents. That's not to say that people don't have different ways of coming to conclusions, as there are many roads to success, but dismissing variables that are proven to be important is not one of those avenues.

Glad to see you turning some of these clay pigeons into orange dust. Common sense is deadly!

[thaskalos]

Originally Posted by classhandicapper

I don't think there's a right or wrong answer to whether you should be using figures or not.

On the downside, everyone is using them. So if you use them they will drag you towards the same line of thinking everyone else has even when there are figure maker errors or the hard to quantify impacts of the track, pace, race development etc.. make the figures less representative of ability.

Looking at who is beating who and how in a qualitative fashion without much regard to the times will put you in the minority and sometimes get better prices, but there are clearly times there isn't enough information to make those subjective judgments as well as the clock.

This has been an intellectual battle for me for decades.

1. Which is the better way to go?

2. Can you combine them more effectively than either as a stand alone?

3. Can you use one method on some races and the other on other races more effectively?

This is what the evidence suggests to me so far.

All the database studies I have been doing since 2014 (and ongoing) using good quality automated class ratings and publicly available speed figures suggest that in terms of accuracy they are very similar (speed less than 1% better). One would hope that a good class handicapper using experienced judgement could make up the gap. So they are more or less the same.

In terms of value, class is slightly better (a few % higher ROI).

If you combine them, you can increase win% over either as a stand alone, but you will lower the ROI.

Speed figures definitely help lot with lightly raced horses (maidens, ALW etc..) where the quality varies widely at the same class level and information is limited for a more subjective analysis.

I applaud you on your work ethic.

[AltonKelsey]

Originally Posted by Tom

It's always the figures, never the guy who has no clue how to use them.
Even when MANY other use them successfully.

Guess it would be too hard to admit you're not smart enough to understand them.

Why should anyone admit ineptitude , when self delusion is so easy?

[thaskalos]

IMO...the gambler should first of all seek to attain as much "understanding" as he could possibly gather about the game of his choice. Only by relying on this "understanding" can he subsequently be able to position himself favorably against the odds that he needs to overcome in order to profit in the endeavor. And, in horse racing, anyone who ignores the speed and pace aspect of this game can't help but leave his understanding of the game woefully incomplete.

We may not be able to profit solely by depending on "figure handicapping"...but we do ourselves a terrible disservice by going the directly opposite way...and denying the "truth" that these figures represent.

[Maximillion]

Imo almost anything you can say about this game can be disputed... someone can say one thing and the exact opposite may be true in a different set of circumstances.
However once a horse has enough starts under its belt... its pretty hard to argue with too slow,at least imo.

[JohnGalt1]

To demonstrate some inconsistencies and problems with speed figures, and how using the wrong ones or misusing them can lead to poor selections, I have looked at some races from recent cards.


I print Bris pps, so don't have access to Beyer figs.

I would like someone to post Beyer figs and Time Form figs for the following races.

As you will see, the races I'll use are comparable to each other--same tracks surfaces, and distances.

From the June 23 Canterbury Mystic lake Derby card---

Race 3 Burma Road 6/3/17 1 1/16 Turf CD and 5/20/17

1:42 lost by 3/4 lengths--DRF fig/var 91-11 Bris speed rating 86
1:41.3 lost by 2 lengths--DRF 91-9 Bris 91


Race 5 Mr. Jägermeister 5/19/18 6f dirt and 8/20/17

1:10.4 won 89/18 100
1:10.4 won 89/15 91


Race 7 Huntsville 6 f 6/2/18, 7/16/17. and 6/10/17


1:10.1 won 94-7 86
1:10.1 won 94-6 89
1:09.2 3 lengths back 95-11 92

Race 10 Trident Hero 6f 11/18/17, and 7/14/17


1:11.2 1/2 length back 85-18 90
1:11.0 4 lengths back 84/16 83


Race 10 Shoot Craps 6f Belterra 4/17/17, and 4/28/17


1:10.3 won 89-6 99
1:10.2 won 90-11 88


June 24 Santa Anita card.


Race 6 Be A Lady 6 1/2 dirt 4/29/18 and 1/28/18

1:17.0 4 1/4 back 79/16 84
1:17.3 Neck back 80-20 90


In these almost identical races the Bris speed figures vary wildly and could lead a handicapper into making a wrong decision on which horse to play.


Variances like these aren't the majority, thank goodness, but bad information even rarely can be deadly to bank rolls.


So could someone look up the Beyers and Time Form ratings for these races and post them please?

[classhandicapper]

To demonstrate some inconsistencies and problems with speed figures, and how using the wrong ones or misusing them can lead to poor selections, I have looked at some races from recent cards.

This is the issue I've been bringing up for decades.

At various times I've had access to Beyer, Timeform, Ragozin, Thorograph, and Logic Dictate figures. For many major stakes I've had 4 or 5 sets at the same time. Even if you normalize them for things like ground loss, weight, pace and other methodology differences, they often disagree, sometimes significantly. That means any odds line you may be contemplating may often not reflect reality. It will reflect the figures you just happen to be using - which could be wrong.

So how do you know if that 5-1 shot you are betting is an overlay or if you are making a bet off an inflated figure or if one of the other horses has a deflated figure?

You can't.

I tried solving that problem by looking at all the major figure differences, studying the race cards and times carefully, and trying to determine who had it right in that case. That's a lot of work and often leads to greater confusion instead better betting value. That's not a very good solution.

What I've learned to accept is that even though very smart people are making these figures, the nature of figure making is that they are at best subjective approximations of how fast a horse ran and at worst contain some significant errors. So it is probably best to not form strong opinions on the merits of a horse based on numbers a few points one way or the other. I think it's best to eliminate the obviously slower horses and use other methods for separating those that are "ballpark" fast enough.

[Maximillion]

Originally Posted by classhandicapper

This is the issue I've been bringing up for decades.

At various times I've had access to Beyer, Timeform, Ragozin, Thorograph, and Logic Dictate figures. For many major stakes I've had 4 or 5 sets at the same time. Even if you normalize them for things like ground loss, weight, pace and other methodology differences, they often disagree, sometimes significantly. That means any odds line you may be contemplating may often not reflect reality. It will reflect the figures you just happen to be using - which could be wrong.

So how do you know if that 5-1 shot you are betting is an overlay or if you are making a bet off an inflated figure or if one of the other horses has a deflated figure?

You can't.

I tried solving that problem by looking at all the major figure differences, studying the race cards and times carefully, and trying to determine who had it right in that case. That's a lot of work and often leads to greater confusion instead better betting value. That's not a very good solution.

What I've learned to accept is that even though very smart people are making these figures, the nature of figure making is that they are at best subjective approximations of how fast a horse ran and at worst contain some significant errors. So it is probably best to not form strong opinions on the merits of a horse based on numbers a few points one way or the other. I think it's best to eliminate the obviously slower horses and use other methods for separating those that are "ballpark" fast enough.

Are they that far apart?
Of the 2 I have used- Beyer and Timeform,I seldom see them that far apart,especially when you know that TF rewards early speed (as it should be)Is there a recent race of any level that you could point out that there seems to be a very large difference of opinion?

[classhandicapper]

Originally Posted by Maximillion

Are they that far apart?
Of the 2 I have used- Beyer and Timeform,I seldom see them that far apart,especially when you know that TF rewards early speed (as it should be)Is there a recent race of any level that you could point out that there seems to be a very large difference of opinion?

Not every horse or every race, but if you ranked the horses from fastest to slowest off their last race the rankings would be a bit different in many races and you would find some differences of several lengths (some even more than that).

[Some_One]

Originally Posted by JohnGalt1

To demonstrate some inconsistencies and problems with speed figures, and how using the wrong ones or misusing them can lead to poor selections, I have looked at some races from recent cards.


I print Bris pps, so don't have access to Beyer figs.

I would like someone to post Beyer figs and Time Form figs for the following races.

As you will see, the races I'll use are comparable to each other--same tracks surfaces, and distances.

From the June 23 Canterbury Mystic lake Derby card---

Race 3 Burma Road 6/3/17 1 1/16 Turf CD and 5/20/17

1:42 lost by 3/4 lengths--DRF fig/var 91-11 Bris speed rating 86
1:41.3 lost by 2 lengths--DRF 91-9 Bris 91


Race 5 Mr. Jägermeister 5/19/18 6f dirt and 8/20/17

1:10.4 won 89/18 100
1:10.4 won 89/15 91


Race 7 Huntsville 6 f 6/2/18, 7/16/17. and 6/10/17


1:10.1 won 94-7 86
1:10.1 won 94-6 89
1:09.2 3 lengths back 95-11 92

Race 10 Trident Hero 6f 11/18/17, and 7/14/17


1:11.2 1/2 length back 85-18 90
1:11.0 4 lengths back 84/16 83


Race 10 Shoot Craps 6f Belterra 4/17/17, and 4/28/17


1:10.3 won 89-6 99
1:10.2 won 90-11 88


June 24 Santa Anita card.


Race 6 Be A Lady 6 1/2 dirt 4/29/18 and 1/28/18

1:17.0 4 1/4 back 79/16 84
1:17.3 Neck back 80-20 90


In these almost identical races the Bris speed figures vary wildly and could lead a handicapper into making a wrong decision on which horse to play.


Variances like these aren't the majority, thank goodness, but bad information even rarely can be deadly to bank rolls.


So could someone look up the Beyers and Time Form ratings for these races and post them please?

You're unable to grasp the concept that tracks can change in speed?

[Some_One]

Originally Posted by Maximillion

Are they that far apart?
Of the 2 I have used- Beyer and Timeform,I seldom see them that far apart,especially when you know that TF rewards early speed (as it should be)Is there a recent race of any level that you could point out that there seems to be a very large difference of opinion?

I think there was a wide gap in the TC races because of the early pace Justify ran which are not considered in the Beyer figs

[Maximillion]

Originally Posted by Some_One

I think there was a wide gap in the TC races because of the early pace Justify ran which are not considered in the Beyer figs

Yes good point... this came to mind almost right after I posted,as the Belmont didnt really interest me this year and i didnt even handicap it.In the everyday races though I seldom see large differences.

[cj]

Originally Posted by JohnGalt1

To demonstrate some inconsistencies and problems with speed figures, and how using the wrong ones or misusing them can lead to poor selections, I have looked at some races from recent cards.


I print Bris pps, so don't have access to Beyer figs.

I would like someone to post Beyer figs and Time Form figs for the following races.

As you will see, the races I'll use are comparable to each other--same tracks surfaces, and distances.

From the June 23 Canterbury Mystic lake Derby card---

Race 3 Burma Road 6/3/17 1 1/16 Turf CD and 5/20/17

1:42 lost by 3/4 lengths--DRF fig/var 91-11 Bris speed rating 86
1:41.3 lost by 2 lengths--DRF 91-9 Bris 91


Race 5 Mr. Jägermeister 5/19/18 6f dirt and 8/20/17

1:10.4 won 89/18 100
1:10.4 won 89/15 91


Race 7 Huntsville 6 f 6/2/18, 7/16/17. and 6/10/17


1:10.1 won 94-7 86
1:10.1 won 94-6 89
1:09.2 3 lengths back 95-11 92

Race 10 Trident Hero 6f 11/18/17, and 7/14/17


1:11.2 1/2 length back 85-18 90
1:11.0 4 lengths back 84/16 83


Race 10 Shoot Craps 6f Belterra 4/17/17, and 4/28/17


1:10.3 won 89-6 99
1:10.2 won 90-11 88


June 24 Santa Anita card.


Race 6 Be A Lady 6 1/2 dirt 4/29/18 and 1/28/18

1:17.0 4 1/4 back 79/16 84
1:17.3 Neck back 80-20 90


In these almost identical races the Bris speed figures vary wildly and could lead a handicapper into making a wrong decision on which horse to play.


Variances like these aren't the majority, thank goodness, but bad information even rarely can be deadly to bank rolls.


So could someone look up the Beyers and Time Form ratings for these races and post them please?

Happy to post the TimeformUS numbers but I'm not sure where this is going. What makes a race comparable to another? Tracks change speed and horses are at different points in their form cycle. Anyway, here goes:

Burma Road 114, 114

Mr. Jägermeister 120, 98 (You realize the 98 was as a young 2yo, right?)

Puntsville 99, 113, 106 (Assuming the middle race is 9-16)

Trident Hero 88, 86 (Assuming the first race is 8-18, not 11-18)

Shoot Craps 93, 88 (I put the most recent first, you had dates reversed above)

Be A Lady 100, 89

[GMB@BP]

I never understood, until this past few months, how much the raw times are still being used in any meaningful way. Learn something new I suppose.