Beyer Pace

[Secretariat]

Was revisiting Beyer on Speed and looking at Beyer's Pace Chart on Pg .133 and his BL chart on Pg .135.

From a 45.0 time (106 Beyer) to a 48.0 time (23 Beyer) there is a 83 point difference. This covers 3 seconds or 15 fifths (0.2). So 83/15 = 5.53 points per fifth. But if one uses his basic approach of taking 1/(Time *5) which would be 1/(46 *5) one gets .004348 or 4.34 points per fifth. It is interesting that his pace times are mathematically constructed differently than his final time charts. He did not use the approach as he used in Picking Winners for the pace chart and didn't explain why a fifth of a second is 5.53 in his Beyer on Speed pace chart, and not 4.34 as he outlines the approach even earlier in the same book.

Second his Beaten Lengths for Pace. I noticed that between 46.0 and 46.6 (or 0.6 seconds), there is 16 point Beyer differential on the time chart. 16 points equates on his beaten length pace chart to 3.75 beaten lengths. Hence, 0.6/3.75 means a beaten length at 4F = 0.16 seconds.

Third, since .2/.16 is 80% this means a length is 80% of the Beyer points at a fifth of a second. So since a fifth is worth 5.53 points, then 5.53 *.8=4.42 per length. So a length is worth about 4.42 beyer points based on his chart. (although his length chart doesn't necessarily conform probably due to rounding issues)

A formula I've seen for beaten lengths for Beyer is 15/(race distance) * (beaten lengths), but it is not in alignment for the 4F distance.

My research using published par times shows a 45.80 at 4F for 6F race. Looking at Beyer's 4F pace times comparing with C10 claimers shows that a 45.80 at 4F to a 1:11 at 6F. A 1:11 is a 92 on the Beyer Chart. A 92 on the Beyer Pace chart is a 45.50. The par time of 45.80 is an 84 on the Beyer pace chart. So overall my conclusion is the overall chart is off by about 8 points on his scale (92-84).

Have seen little on comments online about the accuracy of his Pace Chart and am interested in what others have found.

[cj]

I was never able to make much sense out of it years ago. The ones in Steve Davidowitz's book made even less sense to me. I'll have to go back and re-read those sections again someday.

I think honestly, if I were starting over, I'd use the same value for time I use for beaten lengths at the finish for all the pace calls. In other words, if a .10 seconds is worth a point for a 6f race, .10 would be worth a point at 2f and 4f as well.

[thaskalos]

I think honestly, if I were starting over, I'd use the same value for time I use for beaten lengths at the finish for all the pace calls. In other words, if a .10 seconds is worth a point for a 6f race, .10 would be worth a point at 2f and 4f as well.
If you want your fractional pace ratings to be in alignment with one another, and with the final time rating...then you can't have the tenth of a second equal the same point measurement at every call of the race.

A quality sprinter may win wire-to-wire in 22:00 - 45:00 - 1:09.00

It's a lot easier for the average horse to run within 5 lengths of the 22:00, or even the 45:00...than it is for this average horse to approach the 1:09.0

That's why the point allowance for the length gets larger, as the distance gets smaller.

[EMD4ME]

Was revisiting Beyer on Speed and looking at Beyer's Pace Chart on Pg .133 and his BL chart on Pg .135.

From a 45.0 time (106 Beyer) to a 48.0 time (23 Beyer) there is a 83 point difference. This covers 3 seconds or 15 fifths (0.2). So 83/15 = 5.53 points per fifth. But if one uses his basic approach of taking 1/(Time *5) which would be 1/(46 *5) one gets .004348 or 4.34 points per fifth. It is interesting that his pace times are mathematically constructed differently than his final time charts. He did not use the approach as he used in Picking Winners for the pace chart and didn't explain why a fifth of a second is 5.53 in his Beyer on Speed pace chart, and not 4.34 as he outlines the approach even earlier in the same book.

Second his Beaten Lengths for Pace. I noticed that between 46.0 and 46.6 (or 0.6 seconds), there is 16 point Beyer differential on the time chart. 16 points equates on his beaten length pace chart to 3.75 beaten lengths. Hence, 0.6/3.75 means a beaten length at 4F = 0.16 seconds.

Third, since .2/.16 is 80% this means a length is 80% of the Beyer points at a fifth of a second. So since a fifth is worth 5.53 points, then 5.53 *.8=4.42 per length. So a length is worth about 4.42 beyer points based on his chart. (although his length chart doesn't necessarily conform probably due to rounding issues)

A formula I've seen for beaten lengths for Beyer is 15/(race distance) * (beaten lengths), but it is not in alignment for the 4F distance.

My research using published par times shows a 45.80 at 4F for 6F race. Looking at Beyer's 4F pace times comparing with C10 claimers shows that a 45.80 at 4F to a 1:11 at 6F. A 1:11 is a 92 on the Beyer Chart. A 92 on the Beyer Pace chart is a 45.50. The par time of 45.80 is an 84 on the Beyer pace chart. So overall my conclusion is the overall chart is off by about 8 points on his scale (92-84).

Have seen little on comments online about the accuracy of his Pace Chart and am interested in what others have found.

I've had the same thoughts. I remember reading if you're beaten 1 length at 5F, it's worth 3 points. 2.4 at 6F and 1.8 at 1M.

When double checking his numbers, I find them off more often than not.

Nevermind the times where his team makes a collasal error. Last year his team gave a horse at EMD a 58 (for winning and all runners behind were calculated off of a 58 winning beyer). I made so much money pounding against those horses as the true winning beyer of the race was a 38. About 4 weeks after the error, I guess they caught it and changed the number.

I was hoping they would not notice and leave it be forever.

Many disparities to say the least.

Also, they adjust for pace now a lot. I hate that. Let the number be based off of final times/adjusted for variants. I used to love seeing a great horse get a low number because they closed into a super slow pace with a fast last quarter. Now they adjust the figure up to reflect the slow pace.

YUCK!

Loved the secret advantage to knowing which figs were total BS and which were logically conceived. Now, times have changed.

[Cratos]

Was revisiting Beyer on Speed and looking at Beyer's Pace Chart on Pg .133 and his BL chart on Pg .135.

From a 45.0 time (106 Beyer) to a 48.0 time (23 Beyer) there is a 83 point difference. This covers 3 seconds or 15 fifths (0.2). So 83/15 = 5.53 points per fifth. But if one uses his basic approach of taking 1/(Time *5) which would be 1/(46 *5) one gets .004348 or 4.34 points per fifth. It is interesting that his pace times are mathematically constructed differently than his final time charts. He did not use the approach as he used in Picking Winners for the pace chart and didn't explain why a fifth of a second is 5.53 in his Beyer on Speed pace chart, and not 4.34 as he outlines the approach even earlier in the same book.

Second his Beaten Lengths for Pace. I noticed that between 46.0 and 46.6 (or 0.6 seconds), there is 16 point Beyer differential on the time chart. 16 points equates on his beaten length pace chart to 3.75 beaten lengths. Hence, 0.6/3.75 means a beaten length at 4F = 0.16 seconds.

Third, since .2/.16 is 80% this means a length is 80% of the Beyer points at a fifth of a second. So since a fifth is worth 5.53 points, then 5.53 *.8=4.42 per length. So a length is worth about 4.42 beyer points based on his chart. (although his length chart doesn't necessarily conform probably due to rounding issues)

A formula I've seen for beaten lengths for Beyer is 15/(race distance) * (beaten lengths), but it is not in alignment for the 4F distance.

My research using published par times shows a 45.80 at 4F for 6F race. Looking at Beyer's 4F pace times comparing with C10 claimers shows that a 45.80 at 4F to a 1:11 at 6F. A 1:11 is a 92 on the Beyer Chart. A 92 on the Beyer Pace chart is a 45.50. The par time of 45.80 is an 84 on the Beyer pace chart. So overall my conclusion is the overall chart is off by about 8 points on his scale (92-84).

Have seen little on comments online about the accuracy of his Pace Chart and am interested in what others have found.
Andy Beyer's thinking was correct, but his mathematical approach in defining his thought process was incorrect.

A horse race performance curve is nonlinear and with deceleration as the race becomes longer the value of the length cost the horse more in terms of work which is measured in joules.

With a little effort of math manipulation this can be converted into time

[cj]

If you want your fractional pace ratings to be in alignment with one another, and with the final time rating...then you can't have the tenth of a second equal the same point measurement at every call of the race.

A quality sprinter may win wire-to-wire in 22:00 - 45:00 - 1:09.00

It's a lot easier for the average horse to run within 5 lengths of the 22:00, or even the 45:00...than it is for this average horse to approach the 1:09.0

That's why the point allowance for the length gets larger, as the distance gets smaller.

I understand this. I probably wan't clear on what I was talking about though. The difference between a 22 and a 23 would still be the same. It is the horses not on the lead that not be penalized as much.

If currently we have this for 6f:

2f .10 = 3 points
4f .10 = 2 points
6f .10 = 1 point

and these ratings:

22 = 100
45 = 100
1:09 = 100

Another race with these times would get these ratings:

23 = 70
46 = 80
1:10 = 90

That part wouldn't change. But, horses not on the lead would be different, using the same beaten lengths adjustment as at the finish. So a horse that is .20 behind at each call in each race would look like this:

Using the original value of a beaten length:

Race A: 94, 96, 98
Race B: 64, 76, 88

I think using the version I recommended is a better match for what we see visually:

Race A: 98, 98, 98
Race B: 68, 78, 88

The latter is probably not as accurate numerically, but I think it does a better job of showing what really happened in a race.

[MPRanger]

Equibase uses the formula - ( 80/number of furlongs) / 5

[cj]

Equibase uses the formula - ( 80/number of furlongs) / 5

For what, final time?

[MPRanger]

For what, final time?

That's a good question. They don't give you a grid like Beyer does where you know the value for every number at a specific time. I sent Trackmaster an email and asked them what the values were for a 90 at each distance. The response was, "We don't have any customer facing products for that."

But David Siegle has told us the values for 4f, 6f, and 1mile in his youtube video.

A SF of 90 at 1 mile = 1:37.00
A SF of 90 at 6fur = 1:10.00
A SF of 90 at 4fur = :45.00

They don't say what a 90 is at 1/4 mile but I deduced it as :22

So, if you know what a 90 is at those times you can then use the formula to deduce the raw figure for different times at those distances if you need to.

None of that matters to me. I use their first call pace figure to deduce their variant and apply it to the second call and just work with the time vs a pace figure. So I have variant adjusted time which I use with the SF in my version of Handicapping Magic to create a pace balanced speed figure and PPF. I use a finger chart and Excel so it goes pretty quick. Especially since I only do a workup on my contenders.

[cj]

That's a good question. They don't give you a grid like Beyer does where you know the value for every number at a specific time. I sent Trackmaster an email and asked them what the values were for a 90 at each distance. The response was, "We don't have any customer facing products for that."

But David Siegle has told us the values for 4f, 6f, and 1mile in his youtube video.

A SF of 90 at 1 mile = 1:37.00
A SF of 90 at 6fur = 1:10.00
A SF of 90 at 4fur = :45.00

They don't say what a 90 is at 1/4 mile but I deduced it as :22

So, if you know what a 90 is at those times you can then use the formula to deduce the raw figure for different times at those distances if you need to.

None of that matters to me. I use their first call pace figure to deduce their variant and apply it to the second call and just work with the time vs a pace figure. So I have variant adjusted time which I use with the SF in my version of Handicapping Magic to create a pace balanced speed figure and PPF. I use a finger chart and Excel so it goes pretty quick. Especially since I only do a workup on my contenders.

Thanks.

Personally I think Equibase numbers are horrible. Just one example, they have American Pharoah as the seventh fastest three year old this year, tied with Donworth. Elnaawi is the second fastest horse in the country behind Ashleyluvssugar. I could never trust the variants that made those numbers.

[MPRanger]

Thanks.

Personally I think Equibase numbers are horrible. Just one example, they have American Pharoah as the seventh fastest three year old this year, tied with Donworth. Elnaawi is the second fastest horse in the country behind Ashleyluvssugar. I could never trust the variants that made those numbers.

Interesting. I only play sprints so maybe those route numbers get skewed. It works great for me.

[cj]

Interesting. I only play sprints so maybe those route numbers get skewed. It works great for me.

That is all that matters!

[classhandicapper]

Also, they adjust for pace now a lot. I hate that. Let the number be based off of final times/adjusted for variants. I used to love seeing a great horse get a low number because they closed into a super slow pace with a fast last quarter. Now they adjust the figure up to reflect the slow pace.



That typically happens on turf where the interrelated complexities of extremely slow paces, changing rail settings, changing run ups, and a limited sample of turf races on the day make calculating the true track speed more of a idealistic fantasy than a real possibility. So he sometimes gives the race a figure that best reflects the ability of the horses in isolation. You can still determine who ran best within the race given the race flow.

[Tom]

If you use Beyer's, I recommend that you sign up for Simulcast Daily and check the Winner's Book's. You can quickly see what races he "adjusts" and which distances he might not be using the charts fro the book for. I think you will find Aqu 6.5 furlongs is not from the book charts - there are others.

[Capper Al]

I was never able to make much sense out of it years ago. The ones in Steve Davidowitz's book made even less sense to me. I'll have to go back and re-read those sections again someday.

I think honestly, if I were starting over, I'd use the same value for time I use for beaten lengths at the finish for all the pace calls. In other words, if a .10 seconds is worth a point for a 6f race, .10 would be worth a point at 2f and 4f as well.

I had the same experience with their pace.

[EMD4ME]

That typically happens on turf where the interrelated complexities of extremely slow paces, changing rail settings, changing run ups, and a limited sample of turf races on the day make calculating the true track speed more of a idealistic fantasy than a real possibility. So he sometimes gives the race a figure that best reflects the ability of the horses in isolation. You can still determine who ran best within the race given the race flow.

I agree and understand but I have also seen the upwards/downwards adjustment for dirt races, sprints and routes due to slow paces. The turf, I can totally understand. The dirt, I'd rather they leave alone. Whatever, just like anything else, we adjust and work accordingly.

Yes, through replays and race flow you can still decipher who ran best in those cases.