PDA

View Full Version : 2015 Saratoga Daily Variants


highnote
07-25-2015, 12:52 AM
I finished up my dirt par charts today and thought I'd give them a whirl by trying to make daily variants for opening day at Saratoga. If anyone else has done the daily variants today I'd be interested in seeing what you came up with!

There were 5 dirt sprints and 1 dirt route.


SPRINT VARIANT:
2F 4F FT
0.49s 1.04s 2.18s

ROUTE VARIANT:
4F 6F FT
1.28s 0.93s 1.07s


I'm kind of guessing at the route variant. Hard to say with only 1 sample.

Overall, the races generally ran slower than par at every fraction with the exception of two sprint races that were less than a 1/5 of a second fast at the first fraction.

Thoughts?

CosmicWon
07-25-2015, 01:38 AM
I usually leave the figure making to the experts but just wanted to make sure you knew that the track was sealed after the Schuylerville because of the tornado warning but apparently it never actually rained.

Not sure if that affected the track much at all but might be something to consider just in case :)

Godspeed with your variants! There's a bad pun in there somewhere lol

pandy
07-25-2015, 06:33 AM
I finished up my dirt par charts today and thought I'd give them a whirl by trying to make daily variants for opening day at Saratoga. If anyone else has done the daily variants today I'd be interested in seeing what you came up with!

There were 5 dirt sprints and 1 dirt route.


SPRINT VARIANT:
2F 4F FT
0.49s 1.04s 2.18s

ROUTE VARIANT:
4F 6F FT
1.28s 0.93s 1.07s


I'm kind of guessing at the route variant. Hard to say with only 1 sample.

Overall, the races generally ran slower than par at every fraction with the exception of two sprint races that were less than a 1/5 of a second fast at the first fraction.

Thoughts?


what does this mean? For sprints, are you saying the track was 2 fifths of a second slow, or 2 seconds slow?

The track was definitely slow, it appeared like the rail was dull, and the track definitely favored stalkers, closers, and outside paths, a tiring track. Last year's meet I had it favoring closers 7 days, favoring speed one day and the rest no bias.

Some_One
07-25-2015, 09:28 AM
I finished up my dirt par charts today and thought I'd give them a whirl by trying to make daily variants for opening day at Saratoga. If anyone else has done the daily variants today I'd be interested in seeing what you came up with!

There were 5 dirt sprints and 1 dirt route.


SPRINT VARIANT:
2F 4F FT
0.49s 1.04s 2.18s

ROUTE VARIANT:
4F 6F FT
1.28s 0.93s 1.07s


I'm kind of guessing at the route variant. Hard to say with only 1 sample.

Overall, the races generally ran slower than par at every fraction with the exception of two sprint races that were less than a 1/5 of a second fast at the first fraction.

Thoughts?

For the Sprint variant, the 4f to finish is the final ~2f correct? and the route variant 6f to finish is the final 3f? If so how can the route variant for almost the same piece of track be one second faster the sprint?

cj
07-25-2015, 11:14 AM
I personally had the variant split after (I think) the first two dirt races.

highnote
07-25-2015, 11:44 AM
My variants are in seconds. So I have the sprint daily variants 1/2 second slow at the first call, 1 second slow at the second call and 2 and 1/5 second slow at the finish, (i.e., 0.49s, 1.04s, 2.18s).

Here are the variants for each dirt race:


2F 4F 6F Final
2. 0.32s 1.07s 0.83s MSW Fillies 2yo state bred 5.5 furlongs
4. 0.01f 0.21s 1.59s STK Fillies 2yo 6 furlongs
6. 0.98s 1.72s 2.00s MSW Fillies 2yo 5.5 furlongs
8. 1.37s 2.06s 2.32s AOC NW3 $62.5k 6 furlongs
10. 0.19f 0.13s 4.16s Claiming $16k NW2 7 furlongs
----------------------------------------
0.49s 1.04s 2.18s SPRINT DAILY VARIANT
========================================

2F 4F 6F Final
1. 1.28s 0.93s 1.07s AOC NW3 $40k 9 furlongs
----------------------------------------
???? ???? ???? ROUTE DAILY VARIANT?
========================================


It looks like the sprint final time variants got slower with each race.

It isn't clear to me what the route DV should be, but it looks like it should be slower than par.

classhandicapper
07-25-2015, 11:52 AM
Based on what I am reading here and what Beyer did I'm going to guess is was a tough day. ;)

highnote
07-25-2015, 12:52 PM
For the Sprint variant, the 4f to finish is the final ~2f correct? and the route variant 6f to finish is the final 3f? If so how can the route variant for almost the same piece of track be one second faster the sprint?


The route race was the first race of the day.

Someone said the track was sealed after the 2nd race.

CJ said he did a split variant.

If the track got slower as the day went on (and that seems to be the case judging from my sprint variants for races 2-10) then it makes sense that the route daily variant could be 1.00 and the sprint daily variant 2.00.

But if the track was getting slower as the day went on then it is possible that there should be separate daily variants made for each race.

cj
07-25-2015, 03:35 PM
There were my figures for the winners on the day pasted from Twitter. Keep in mind pace can be and often is a factor in the figures and the winner isn't necessarily the highest:

Winning Saratoga @TimeformUS Speed Figures from Fri:

Sea Raven 105
Lemon Liqueur 63
Khaleesi Kat 77
Off the Tracks 95
Kelly's Prize 84
Just Wicked 86
Bishop's Pond 110
Pulling G's 104
Mrs McDougal 116
Big Blue Talent 98

highnote
07-27-2015, 02:14 AM
My worksheet for Daily Variants at Saratoga for Sunday, July 26 appears below.

The track appeared to be running faster today. The pace times were a touch slower than par, but 4 out of 5 the final times were faster than par.

The 6.5f MSW race for state bred fillies was the quickest. I'd watch for the winner of the race next time out to see if she can step up in company.

7/26/2015 ACTUAL PARS DIFFERENCES
race# 2f 4f 6f ff | 2f 4f 6f ff | 2f 4f 6f ff class
1 23.26 46.11 83.27 | 22.72 45.61 83.89 | -0.54 -0.5 0.62 7f MSW F 3+
3 22.65 46.23 76.83 | 22.32 45.55 76.43 | -0.33 -0.68 -0.4 6.5f c25 3yo
4 21.86 45.13 75.9 | 22.54 45.97 77.9 | 0.68 0.84 2 6.5f MSW F 3+ SB
8 48.94 73.32 109.79 | 48.21 72.93 111.14 | -0.73 -0.39 1.35 9f ALW 3+ NW1
10 46.99 71.27 109.74 | 48.16 72.63 110.34 | 1.17 1.36 0.6 9f STK F 3yo

AVG DV -0.06 0.02 0.49 0.83
MED DV -0.33 -0.50 0.49 0.62

highnote
07-27-2015, 10:16 PM
The July 27 variants are in the table below.

Today was a little tricky because there was only one dirt route race and it had the biggest final time variant of all the dirt races -- 2.54 seconds slow. However, the route 4 furlong variant was right in line with the sprint 4 furlong variants.

This was a card of slower horses, though... maiden claimers, fillies and state breds. So it is hard to say if the track was slow or if the horses were slow.

7/27/2015 ACTUAL PARS DIFFERENCES
race# 2f 4f 6f ff | 2f 4f 6f ff | 2f 4f 6f ff class
1 23.49 47.38 71.18 | 22.66 46.13 70.4 | -0.83 -1.25 -0.78 6f MCL 30 F 3+
2 22.42 45.83 70.9 | 22.45 45.77 69.48 | 0.03 -0.06 -1.42 6f c25 3+ SB
3 21.74 44.71 76.16 | 22.19 45.27 75.88 | 0.45 0.56 -0.28 6.5f AOC F 3+ c62.5
5 47.09 71.84 110.89 | 47.45 71.85 108.35 | 0.36 0.01 -2.54 9f AOC F 3+ c62.5
6 22.2 45.68 77.26 | 22.62 46.1 76.75 | 0.42 -0.51 6.5f MCL 40 3+ SB
| |
AVG DV -0.12 0.01 0.01 -1.11
MED DV 0.03 0.36 0.01 -0.78

highnote
07-30-2015, 01:20 PM
The Wednesday dirt races look pretty formful. My Daily Variant worksheet appears below.

The most interesting race to me was #6 -- 6.5f claiming race. The winner ran almost to par at 2f. Exceeded par by three fourths of a second at 4f and then finished the race about a fifth of a second faster than par. The winner could be worth a closer look next time out depending on where he's placed.

Race #4 was a GII handicap. The winner ran about a fifth second faster than par after running about par at the second call. That would seem to indicate a good closing ability. She ran a little slower than par at the first fraction which might explain why her final time was better than par. Still, she may be worth a look next time out.

7/29/2015 ACTUAL PARS DIFFERENCES
race# 2f 4f 6f ff | 2f 4f 6f ff | 2f 4f 6f ff class
2 22.61 45.5 69.71 | 22.34 45.5 69.14 | -0.27 0 -0.57 6f c25 3+
4 22.47 45.03 69.14 | 21.93 44.94 69.31 | -0.54 -0.09 0.17 6f STK F 3+
6 22.44 44.7 76.2 | 22.32 45.55 76.43 | -0.12 0.85 0.23 6.5f c26NW2L 3+
10 23.58 46.63 83.77 | 23.05 46.24 82.8 | -0.53 -0.39 -0.97 7f c12.5 3+

AVG DV -0.40 0.09 -0.28
MED DV -0.40 -0.05 -0.20

cj
07-30-2015, 01:31 PM
I thought yesterday's dirt races were all a good fit with a single variant as well.

Highly unlikely we'll be so lucky today!

highnote
07-31-2015, 12:41 AM
Overall, I'm pretty happy with the Daily Variants I made at Saratoga today -- July 30.

For the Off The Turf claiming races I used par times that were two classes lower because I've always felt the claiming prices on the turf are more generous than on the dirt. It didn't make much of a difference. The running times were still much slower than par.

It looks like the first two furlongs were run about a fifth of a second faster than par, but as the horses ran farther into the race the muddy track produced an exaggerated telescoping effect. They ran about 2/5 slow at the half and by 3/4 they were running about a second behind par in routes and a 2 seconds behind in sprints. At the finish of the routes they were running a full 3 seconds slow.

The most impressive race of the day was the 3rd race -- The MSW race for 2 year old state breds. The winner almost ran to par over a presumably slow track. Definitely worth watching next time out. The second place finisher lost by a few lengths, but beat the third place finisher by 10. So the second place horse could be a good bet in the MSW next time.

7/30/2015 ACTUAL PARS DIFFERENCES
race# 2f 4f 6f ff | 2f 4f 6f ff | 2f 4f 6f ff class
2 49.35 75.05 115.7 | 48.21 73.37 111.39 | -1.14 -1.68 -4.31 9f c16 NW2L F 3+ sloppy sealed
3 22.67 46.06 65.14 | 22.44 46 65 | -0.23 -0.06 -0.14 5.5f MSW 2yo SB
4 22.03 45.2 77.69 | 22.19 45.27 75.9 | 0.16 0.07 -1.79 6.5f STK 3+SB
5 22.04 45.85 65.87 | 22.13 45.55 64.63 | 0.09 -0.3 -1.24 5.5 MSW F 2yo off turf
6 49.15 74 113.91 | 48.06 73.07 110.78 | -1.09 -0.93 -3.13 9f c50 F 3yo off turf use c30
7 21.87 46.03 65.82 | 22.54 45.9 63.54 | 0.67 -0.13 -2.28 5.5f c40 NW2L F 3+ off turf use c25
8 47.97 73.08 112.91 | 48.31 72.93 111.14 | 0.34 -0.15 -1.77 9f MSW F 3+
9 49.11 73.97 112.85 | 47.76 72.47 109.59 | -1.35 -1.5 -3.26 9f AOC62.5 F 3+ off turf use c40 muddy sealed
10 22.29 45.72 71.64 | 22.55 45.97 69.88 | 0.26 0.25 -1.76 6f MCL25 F 3+ SB
RTE DV -0.81 -1.07 -3.12
AVG DV 0.19 -0.38 -1.07 -2.19
MED DV 0.16 -0.13 -1.22 -1.79

Lemon Drop Husker
07-31-2015, 12:45 AM
[QUOTE=highnote]Overall, I'm pretty happy with the Daily Variants I made at Saratoga today -- July 30.

Hate to ask, but did the daily variant win you anything today?

highnote
07-31-2015, 12:51 AM
[QUOTE=highnote]Overall, I'm pretty happy with the Daily Variants I made at Saratoga today -- July 30.

Hate to ask, but did the daily variant win you anything today?

No is the short answer because I just finished making them for today's races.

The long answer is that my plan is to finish writing my speed figure software and then make variants for the past Belmont meet and hopefully by the mid August or September I will have enough daily variants to make good bets. But the daily variants that produce speed figures are only one small piece of a big handicapping pie. So ask me again in two or three weeks and you might get a different answer. Or you might get the same. I don't know.

steveb
07-31-2015, 06:03 AM
may i ask why the variant(s) appear to be so important to you?
i have never placed any importance on them, other than to tell me how fast a track is racing compared to the neutral.
am i missing something?

DeltaLover
07-31-2015, 08:04 AM
may i ask why the variant(s) appear to be so important to you?
i have never placed any importance on them, other than to tell me how fast a track is racing compared to the neutral.
am i missing something?

Speed handicapping is still (and will forever remain), the most predictive approach.

Since speed figures depend on track variants, it is easy to realize why they are so important.

cj
07-31-2015, 08:51 AM
[QUOTE=highnote]Overall, I'm pretty happy with the Daily Variants I made at Saratoga today -- July 30.

Hate to ask, but did the daily variant win you anything today?

Do you even know what the purpose of a track variant is? How can something calculated after the day is over win you anything that day? What it can help do is win money in the future.

highnote
07-31-2015, 09:15 AM
may i ask why the variant(s) appear to be so important to you?
i have never placed any importance on them, other than to tell me how fast a track is racing compared to the neutral.
am i missing something?

DeltaLover answered your question more eloquently in #18 than I could have.

Let me ask you a question...

Do you use speed figures?

peteman
07-31-2015, 11:43 AM
We can also see it as moving away from pars,speed and pace figures,and
the variants that everyone else has,and using your own information.

A big tell here is that,based on your free variant server idea.and lack of
response,that there are only a few that have the time or inclination to
use such a service,regardless of your honorable intentions.

pandy
07-31-2015, 12:03 PM
Speed handicapping is still (and will forever remain), the most predictive approach.

Since speed figures depend on track variants, it is easy to realize why they are so important.


I agree with you but there are several on this board who believe that there are other more predictive win factors. If we're talking about win percentage, I've never seen proof that anything can beat speed figures for win predictability.

classhandicapper
07-31-2015, 12:36 PM
I agree with you but there are several on this board who believe that there are other more predictive win factors. If we're talking about win percentage, I've never seen proof that anything can beat speed figures for win predictability.

I'm one of those guys. My data suggests a high quality class figure will outperform speed figures in both win% and ROI.

There are several major problems though.

1. There are many hundreds of different class conditions, seasonal issues, age development issues, statebred issues etc... you have to cope with. So assigning high quality values to every class at every track would be a monumental task and the file would constantly be growing.

2. The short cut class abbreviations in the PPs often don't provide an adequate description of the class. You have to manually read the details and then assign the value.

3. With so many horses shipping, if even you lock down a single circuit (which is not that tough), there are numerous races from out of town that you have to evaluate.

4. Class is dynamic. Changes in purse structure and other factors could change the quality of a circuit. So you have to stay on top of that.

Those are the problems with just creating the structure.

Then you still have to develop a methodology that can differentiate between strong and weak fields at a specific level.

Then you have to have a structure that allows you to compare horses that ran at different class levels. For example, horse A won by 4 lengths at level X and horse B ran 3rd by 3 lengths at level Y. How do they compare?

No one person can do all that nationally and it doesn't lend itself to easy automation. It's way easier for most people to just look at the speed figure. But since everyone is looking at the speed figure, it's going to be hard to find value.

steveb
07-31-2015, 08:30 PM
Speed handicapping is still (and will forever remain), the most predictive approach.

Since speed figures depend on track variants, it is easy to realize why they are so important.

yes but your variant is a product of your race speeds, it's not a separate process.
one defines the other.
doesn't it?

Lemon Drop Husker
07-31-2015, 08:45 PM
[QUOTE=Lemon Drop Husker]

Do you even know what the purpose of a track variant is? How can something calculated after the day is over win you anything that day? What it can help do is win money in the future.

I was more alluding to the track variant from the day before helping someone on the next day of racing on the same track.

I certainly appreciate the information, but from a handicapping perspective, track variants are included in any speed number or handicapping exercise anybody of any credibility puts forth.

Outside of real times, any handicapping number and venture is a logical guess. And as for real times, we all know those are many times useless due to track variants, track conditions, and the incredible differences from track to track across the country.

I poorly worded a short question. I was more asking as to how one uses previous day variants to help in handicapping future days at the same track.

DeltaLover
07-31-2015, 10:29 PM
yes but your variant is a product of your race speeds, it's not a separate process.
one defines the other.
doesn't it?

I am not sure I understand what you are saying here..

The TV is a derivative of the final time (or race speeds as you put it) but is also affected by other factors, mainly those who try to cluster and classify each race in a meaningful way to produce an optimized curve of variants, that when applied to the whole universe of past performances are satisfying a specific set of restrictions (for example to minimize the discrepancies for certain subsets of performers)...

The whole process of track variant estimation, is part science and part experience. It is impossible to tackle all the related challenges using a purely algorithmic approach... Besides this, it is also correct that all track variants are not created equally and definitely there exist good and bad methodologies and developing a superior one, is one of the main challenges of a handicapper.

highnote
08-01-2015, 02:02 AM
The worksheet to make Saratoga variants for July 31 races appears below.

The simple average of all the variants seems to fit the best and this makes the daily track variant about 1/5 of a second slow at each call.

The hottest race of the day was the 5th -- a 9 furlong MSW race for 3 and up. I had it running almost a second and a half faster than par. There are always good maidens at Saratoga. So it's not surprising to see some fast ones. The winner of this race ran faster than the 3 year olds in the 4th race -- a stakes race.

The 7th race was a MSW and was about a half second faster than par. So not too bad, either.

The 80k allowance optional claimers in the 9th ran the fastest of the three 9 furlong races on the card, but still much slower than par. So my pars could be wrong for this class. I haven't figured out yet which class those Allowance Optional Claimers belong in. My studies showed that the average times for those races are about the same as the claiming class listed in the conditions. But the quality of the horses in the race can vary since not all of them are entered with claiming tags.

7/31/2015 ACTUAL PARS DIFFERENCES
race# 2f 4f 6f ff | 2f 4f 6f ff | 2f 4f 6f ff class
3 22.57 46.01 77.19 | 22.55 45.97 76.48 | -0.02 -0.04 -0.71 6.5f c25 vF 3+ SB
4 47.7 72.14 110.4 | 48.01 72.33 109.74 | 0.31 0.19 -0.66 9f STK 3yo
5 48.79 73.12 109.71 | 48.31 72.93 111.14 | -0.48 -0.19 1.43 9f MSW 3+
7 23.05 45.75 83.02 | 22.62 45.41 83.49 | -0.43 -0.34 0.47 7f MSW 3+
9 47.67 71.35 109.36 | 47.35 71.66 107.97 | -0.32 0.31 -1.39 9f AOC 80 3+
RTE DV -0.17 0.10 -0.17
AVG DV -0.23 -0.17 -0.17
MED DV -0.23 -0.32 -0.66

cj
08-01-2015, 09:59 AM
[QUOTE=cj]
I poorly worded a short question. I was more asking as to how one uses previous day variants to help in handicapping future days at the same track.

That makes sense, thanks for explaining. I don't use a variant that way other than noting when there is an extreme change from day to day that isn't weather related, i.e. mud, slop, extreme wind, etc. I don't think there is much to gain other than in helping make pace and speed figures.

steveb
08-01-2015, 08:03 PM
DeltaLover answered your question more eloquently in #18 than I could have.

Let me ask you a question...

Do you use speed figures?

you already know that i do.
i also know that i do it better than most.
so let me repeat the question, why are variants so important to you?

ronsmac
08-01-2015, 08:42 PM
I am not sure I understand what you are saying here..

The TV is a derivative of the final time (or race speeds as you put it) but is also affected by other factors, mainly those who try to cluster and classify each race in a meaningful way to produce an optimized curve of variants, that when applied to the whole universe of past performances are satisfying a specific set of restrictions (for example to minimize the discrepancies for certain subsets of performers)...

The whole process of track variant estimation, is part science and part experience. It is impossible to tackle all the related challenges using a purely algorithmic approach... Besides this, it is also correct that all track variants are not created equally and definitely there exist good and bad methodologies and developing a superior one, is one of the main challenges of a handicapper.I agree with you, I think the experience is actually more important than the science. Plain logic says a track at 5pm can't be the same speed at 1pm over a 1 mile distance, even if the difference is minimal.

highnote
08-01-2015, 09:20 PM
you already know that i do.
i also know that i do it better than most.
so let me repeat the question, why are variants so important to you?


Not to sound like a smart ass, but here is what DeltaLover wrote and it explains exactly why they are important to me:

Speed handicapping is still (and will forever remain), the most predictive approach.

Since speed figures depend on track variants, it is easy to realize why they are so important.

ronsmac wrote the following and I agree with him:

Plain logic says a track at 5pm can't be the same speed at 1pm over a 1 mile distance, even if the difference is minimal.

So if you can measure the amount of time that a track speeds up or slows down the final time of the race then you should have a better idea of how fast a horse can run on a "normal" track.

Now, there's no guarantee that you can measure with a great degree of preciseness how fast or slow a track is on a given day, but unless the "projection" method is used, I do not know of a better way to measure track speed than by looking at the deviation from par for a particular class of horses for a particular group of races on a given day.

To reiterate, other than the final odds of the betting public, I have not found any single factor that predicts the outcome of a wide range of races at varying classes and distances than speed figures. And since track variants are an important component of speed figures track variants are important to me.

Does that answer your question? :)

ronsmac
08-01-2015, 09:29 PM
Not to sound like a smart ass, but here is what DeltaLover wrote and it explains exactly why they are important to me:



ronsmac wrote the following and I agree with him:



So if you can measure the amount of time that a track speeds up or slows down the final time of the race then you should have a better idea of how fast a horse can run on a "normal" track.

Now, there's no guarantee that you can measure with a great degree of preciseness how fast or slow a track is on a given day, but unless the "projection" method is used, I do not know of a better way to measure track speed than by looking at the deviation from par for a particular class of horses for a particular group of races on a given day.

To reiterate, other than the final odds of the betting public, I have not found any single factor that predicts the outcome of a wide range of races at varying classes and distances than speed figures. And since track variants are an important component of speed figures track variants are important to me.

Does that answer your question? :)There's so many factors when computing the speed of a track that have been covered a million times here, The best any of us can do is make an educated guess. When I was at my speed figure peak, I doubt I ever had 1 variant for a particular day. One of the beauties of racing, everyone has their own reasoning and methodology when computing the speed of the track.

steveb
08-01-2015, 09:45 PM
Not to sound like a smart ass, but here is what DeltaLover wrote and it explains exactly why they are important to me:



ronsmac wrote the following and I agree with him:



So if you can measure the amount of time that a track speeds up or slows down the final time of the race then you should have a better idea of how fast a horse can run on a "normal" track.

Now, there's no guarantee that you can measure with a great degree of preciseness how fast or slow a track is on a given day, but unless the "projection" method is used, I do not know of a better way to measure track speed than by looking at the deviation from par for a particular class of horses for a particular group of races on a given day.

To reiterate, other than the final odds of the betting public, I have not found any single factor that predicts the outcome of a wide range of races at varying classes and distances than speed figures. And since track variants are an important component of speed figures track variants are important to me.

Does that answer your question? :)

you can be a smart arse if you want, it matters not.
one never gets anywhere if they don't ask questions.

the variants are basically the remainder after one has figured the speeds is all they are to ME; i was just wondering what they are to others.

that the track can change during the course of a meeting is neither here nor there as far as my question goes.
you can solve for that using iteration and/or regression and various other means to get the most accurate racespeeds(or whatever one likes to call them).

so your answer is simply that you try to measure the difference between what actually happened and what you think should have happened.
that makes sense to me, although i don't believe 'projection' is the best means of doing so.
it still does not mean that the variant in and of itself is a means of making money, which is what i was wondering.

highnote
08-01-2015, 10:22 PM
it still does not mean that the variant in and of itself is a means of making money, which is what i was wondering.

I do not think the variant in isolation is a means of making money. It's one step in the process, but not a step that can that can be used by itself.

As far as I am aware, the track variant by itself is probably meaningless for predicting the winner of a future race. It's only value is that it is used to adjust the final time of races that were run on a given day to what the final times would have been if those races would have been run on a "normal" track.

But even when it is applied to the final times to produce speed figures those speed figures probably won't be profitable, on the whole, because so many other people use speed figures made by various methods.

Speed figures will tell you which horses are fastest and can separate the contenders from the pretenders, but the odds on the fastest speed figure horses will probably be too low to make a long term profit by betting them. But if you don't use them you will surely miss a lot of winners.

So my strategy is to use them, but also use additional handicapping angles and factors. My use of angles would be reserved for special situations whereas factors are to be used for nearly every race.

For example, "Pop and Stop" is an angle I use in Maiden races. Let's say a horse has only one prior race. In that race, it jumped out to an early lead in the first 2 or 4 furlongs, but then it stopped badly or just ran out of gas because it wasn't fit. Hence, it popped and stopped. Well, next time the horse races it should be fitter and hopefully it will pop out to an early lead and won't stop. Many people see that running line in the past performances and discount the horse. I see that running line and upgrade the horse.

To my way of thinking, a speed figure is a factor because it can be generated for nearly every horse that has at least one prior race.

That's more info than you asked for, but once I get started, it's hard to shut me up. :D

highnote
08-02-2015, 12:14 PM
My daily variant worksheet for Aug 1, 2015 at Saratoga appears below.

The sprint and route variants for the sectional times are almost zero and the final time variant is under 1/5 of a second. It looks like the track was running pretty normal, timewise.

The most fun I've had so far making variants at Saratoga was on this set of races -- good stakes races and all the races had almost as little variance from par as they do in California. So it was interesting to make the variants and also fairly easy.

It was interesting because the 3rd race which was run over 6 furlongs by NW2 lifetime older state breds had a faster 2 furlong time than the 8th race which was a Grade One 6 furlong stakes race for older horses. In fact, the state breds ran the first 2 furlongs nearly a second faster than the Grade 1 horses.

This raises a couple of questions about timings of the races and gate placement. Len Ragonzin wrote in his book that Saratoga is a casual meeting and sometimes the gate placement differs from race to race for the same distance race. This would cause the runups to change which affects the timing of the race. I'm not saying this is the case, but it's an interesting point.


20150801 ACTUAL PARS DIFFERENCES
race# 2f 4f 6f ff | 2f 4f 6f ff | 2f 4f 6f ff class
2 22.23 45.66 71.39 | 22.31 45.61 70.68 | 0.08 -0.05 -0.71 6f MSW 2yo
3 21.68 44.54 69.98 | 22.34 45.57 70.1 | 0.66 1.03 0.12 6f ALW NW2L 3+ SB
5 22.17 44.83 75.38 | 21.88 44.84 75.91 | -0.29 0.01 0.53 6.5f STK 3yo
8 22.74 45.31 68.75 | 21.83 44.74 68.91 | -0.91 -0.57 0.16 6f STK 3+
10 48.15 71.92 108.77 | 47.96 72.23 109.54 | -0.19 0.31 0.77 9f STK 3yo
RTE DV 0.05 0.31 0.17
AVG DV -0.12 0.05 0.17
MED DV -0.11 -0.05 0.16

cj
08-02-2015, 12:24 PM
My daily variant worksheet for Aug 1, 2015 at Saratoga appears below.

The sprint and route variants for the sectional times are almost zero and the final time variant is under 1/5 of a second. It looks like the track was running pretty normal, timewise.

The most fun I've had so far making variants at Saratoga was on this set of races -- good stakes races and all the races had almost as little variance from par as they do in California. So it was interesting to make the variants and also fairly easy.

It was interesting because the 3rd race which was run over 6 furlongs by NW2 lifetime older state breds had a faster 2 furlong time than the 8th race which was a Grade One 6 furlong stakes race for older horses. In fact, the state breds ran the first 2 furlongs nearly a second faster than the Grade 1 horses.

This raises a couple of questions about timings of the races and gate placement. Len Ragonzin wrote in his book that Saratoga is a casual meeting and sometimes the gate placement differs from race to race for the same distance race. This would cause the runups to change which affects the timing of the race. I'm not saying this is the case, but it's an interesting point.


20150801 ACTUAL PARS DIFFERENCES
race# 2f 4f 6f ff | 2f 4f 6f ff | 2f 4f 6f ff class
2 22.23 45.66 71.39 | 22.31 45.61 70.68 | 0.08 -0.05 -0.71 6f MSW 2yo
3 21.68 44.54 69.98 | 22.34 45.57 70.1 | 0.66 1.03 0.12 6f ALW NW2L 3+ SB
5 22.17 44.83 75.38 | 21.88 44.84 75.91 | -0.29 0.01 0.53 6.5f STK 3yo
8 22.74 45.31 68.75 | 21.83 44.74 68.91 | -0.91 -0.57 0.16 6f STK 3+
10 48.15 71.92 108.77 | 47.96 72.23 109.54 | -0.19 0.31 0.77 9f STK 3yo
RTE DV 0.05 0.31 0.17
AVG DV -0.12 0.05 0.17
MED DV -0.11 -0.05 0.16


The run ups are listed in the charts these days, at least the ones on the Equibase site, and are available in all the chart data files available for download. The run up was consistent between races 2 and 8, I always confirm races that look suspect via video. It was just a matter of jockeys taking a hard hold in R8, far from unusual these days.

highnote
08-02-2015, 12:45 PM
The run ups are listed in the charts these days, at least the ones on the Equibase site, and are available in all the chart data files available for download. The run up was consistent between races 2 and 8, I always confirm races that look suspect via video. It was just a matter of jockeys taking a hard hold in R8, far from unusual these days.

Apparently, I don't use equibase charts. ;)

I wonder if they measure the run ups before every race or just assume they are always the same?

Those younger horses probably have a tendency to break out and run as fast as they can for as long as they can, which would explain why they run faster pace times. But the G1 horses crush them on the final time.

Reminds me of my maiden attempt at the 440 yard dash in high school. I thought I was going to die at the end of the race. Pace was not in my wheelhouse.

cj
08-02-2015, 01:11 PM
Apparently, I don't use equibase charts. ;)

I wonder if they measure the run ups before every race or just assume they are always the same?

Those younger horses probably have a tendency to break out and run as fast as they can for as long as they can, which would explain why they run faster pace times. But the G1 horses crush them on the final time.

Reminds me of my maiden attempt at the 440 yard dash in high school. I thought I was going to die at the end of the race. Pace was not in my wheelhouse.

The run ups are reported by the track to Equibase and reported as such. They are not always accurate, and I've had some talks with Equibase about ensuring they don't echo the track report without checking they are accurate. It takes what, 2 seconds a race to verify? They are working on improving in this area. There were two bad ones recently on Belmont Stakes day and Belmont Derby day, which doesn't lend a lot of faith to trusting run up reported from a small track on a Tuesday.

But like I said, they are working to better the process. I'm hopeful.

highnote
08-02-2015, 02:13 PM
The run ups are reported by the track to Equibase and reported as such. They are not always accurate, and I've had some talks with Equibase about ensuring they don't echo the track report without checking they are accurate. It takes what, 2 seconds a race to verify? They are working on improving in this area. There were two bad ones recently on Belmont Stakes day and Belmont Derby day, which doesn't lend a lot of faith to trusting run up reported from a small track on a Tuesday.

But like I said, they are working to better the process. I'm hopeful.


Well at least someone is on top of this -- even if it isn't the people generating and SELLING the initial data. :D

highnote
08-03-2015, 12:32 PM
It looks like the track was playing fairly normal yesterday, Sunday, Aug 2 at Saratoga and was one of the fastest tracks I've seen at the meet. Of course, this could be because better horses ran closer to par more often this weekend than on other days.

The 5th race was a 7 furlong sprint and was the one anomaly of the day. It had a very slow pace and final time which dragged the average sprint variant down about a 1/5 of a second compared to the median sprint variant. So I'll probably use the median as the variant in sprints.

The worksheet that I use to make pars appears below.

20150802 ACTUAL PARS DIFFERENCES
race# 2f 4f 6f ff | 2f 4f 6f ff | 2f 4f 6f ff class
2 22.45 45.75 70.39 | 22.41 45.85 71.08 | -0.04 0.1 0.69 6f MSW F 2yo
5 24.04 47.23 83.2 | 22.85 45.85 82 | -1.19 -1.38 -1.2 7f c50 N2L 3+
8 21.93 45.27 76.84 | 22.39 45.67 77.1 | 0.46 0.4 0.26 6.5f ALW NW1 3+ SB
9 21.66 44.45 75.7 | 22.18 45.25 75.85 | 0.52 0.8 0.15 6.5f c32 3+
10 47.79 71.83 109.37 | 47.96 72.23 109.54 | 0.17 0.4 0.17 9f STK F 3+
RTE DV 0.02 0.40 0.01
AVG DV -0.06 0.02 0.01
MED DV 0.21 0.17 0.17

highnote
08-04-2015, 02:38 AM
For posterity's sake and in case anyone actually reads this, I'll attempt to make variants for Monday, Aug 3 variants at Saratoga.

There were two races that ran just about to par and two that were way below par.

The 2nd race of the day was run slow, but I think my par is correct.

The 8th race is the one I'm not sure about because it is an Allowance Optional Claiming for $80k claiming with a purse of $93,000. I have a hard time knowing what the actual class is, but given that the purse is nearly 100 grand, maybe these really are $80k horses? If they are, they're slow $80k horses. If they're worth less than $80k then I need to use a class with slower pars to evaluate this group of horses which would then move the final time towards par, but given that these horses ran a second and a half slower than $80k horses I would need to drop them to the $10,000 claiming level to get them to par!

So I'll just use the average variant for the day and then wait to see how they run next time out.

20150803 ACTUAL PARS DIFFERENCES
race# 2f 4f 6f ff | 2f 4f 6f ff | 2f 4f 6f ff class
2 22.5 45.85 72.23 | 22.64 46.09 70.34 | 0.14 0.24 -1.89 6f c16 F 3+ NW2L
4 22.56 46.2 70.75 | 22.44 45.77 70.9 | -0.12 -0.43 0.15 6f MSW 3+ SB
5 22.05 44.89 70.64 | 22.71 46.28 70.51 | 0.66 1.39 -0.13 6f c12.5 F 3+
8 22.49 45.14 83.02 | 22.72 45.57 81.45 | 0.23 0.43 -1.57 7f AOC 3+ c80

AVG DV 0.23 0.41 -0.86
MED DV 0.19 0.34 -0.85

DeltaLover
08-04-2015, 08:49 AM
My variants so far:

SAR Dirt

20150803 => -1
20150802 => -4
20150801 => 0
20150731 => -2
20150730 => 2
20150729 => -5
20150727 => -4
20150726 => -1
20150725 => -3
20150724 => 1

highnote
08-04-2015, 10:50 AM
My variants so far:

SAR Dirt

20150803 => -1
20150802 => -4
20150801 => 0
20150731 => -2
20150730 => 2
20150729 => -5
20150727 => -4
20150726 => -1
20150725 => -3
20150724 => 1

Do you make pace variants, or only final time variants?

I'll post mine this evening.

reckless
08-04-2015, 10:52 AM
The run ups are reported by the track to Equibase and reported as such. They are not always accurate, and I've had some talks with Equibase about ensuring they don't echo the track report without checking they are accurate. It takes what, 2 seconds a race to verify? They are working on improving in this area. There were two bad ones recently on Belmont Stakes day and Belmont Derby day, which doesn't lend a lot of faith to trusting run up reported from a small track on a Tuesday.

But like I said, they are working to better the process. I'm hopeful.

CJ, at what point --if any-- is the run up distance the most important?

The first fraction? Final time?

When you make your figs is the additional run up distance factored in every incremental fraction or just once?

Thanks.

cj
08-04-2015, 10:58 AM
CJ, at what point --if any-- is the run up distance the most important?

The first fraction? Final time?

When you make your figs is the additional run up distance factored in every incremental fraction or just once?

Thanks.

First fraction without a doubt. Outside of the effect of the first fraction, there is also an effect on final time as run up can shorten or lengthen the race. It depends on the pace scenario. But most of the difference is going to be seen in the first fraction.

That said, it is something I've only been able to use with general rules by dividing run up into classes---

1. 5 feet or less
2. 6 to 30
3. 31-80
4. >80

Even then, you can have a 280 foot run up but if the jockeys all hold out of the gate, it won't matter.

highnote
08-04-2015, 11:09 AM
My variants so far:

SAR Dirt

20150803 => -1
20150802 => -4
20150801 => 0
20150731 => -2
20150730 => 2
20150729 => -5
20150727 => -4
20150726 => -1
20150725 => -3
20150724 => 1

...and also... what is the value of your variants -- fifths, tenths, other?

reckless
08-04-2015, 12:11 PM
First fraction without a doubt. Outside of the effect of the first fraction, there is also an effect on final time as run up can shorten or lengthen the race. It depends on the pace scenario. But most of the difference is going to be seen in the first fraction.

That said, it is something I've only been able to use with general rules by dividing run up into classes---

1. 5 feet or less
2. 6 to 30
3. 31-80
4. >80

Even then, you can have a 280 foot run up but if the jockeys all hold out of the gate, it won't matter.

When I was doing my pace-speed figs whenever I was able to ascertain the run-up distance, I included it in my pace figure, and made a minor final time adjustment.

I don't ever recall the run-up distance being as short as 5 feet or less at the tracks I monitored, except at Pimlico, where a distance of 5 feet plus meant the horses actually started on the sidewalk outside the racetrack near that nice homeowner's living room. :)

Thanks again, CJ.

cj
08-04-2015, 12:28 PM
When I was doing my pace-speed figs whenever I was able to ascertain the run-up distance, I included it in my pace figure, and made a minor final time adjustment.

I don't ever recall the run-up distance being as short as 5 feet or less at the tracks I monitored, except at Pimlico, where a distance of 5 feet plus meant the horses actually started on the sidewalk outside the racetrack near that nice homeowner's living room. :)

Thanks again, CJ.

Gulfstream mile races have virtually no run up now either. There are a few other places with similar situations, mostly smaller tracks.

DeltaLover
08-04-2015, 12:32 PM
Do you make pace variants, or only final time variants?

I'll post mine this evening.

Only final times. No pace, trip or anything else is reflected on it

DeltaLover
08-04-2015, 12:38 PM
...and also... what is the value of your variants -- fifths, tenths, other?

These variants do not express time or lengths (although they can be adjusted for)..

Instead they reflect the variant in actual speed figure deltas.

Every winner's time regardless of the distance, is expressed as a raw speed figure and the TV is used to correct it.

The average figure is 100. G1 horses can run a 115+ while very weak MCL can go as low as 85- ..

There is some weakness in my algorithm that has to do with days when the traditional racing form speed figure is very high (especially on the turf)..

These figs are not cross track...

highnote
08-04-2015, 12:43 PM
These variants do not express time or lengths (although they can be adjusted for)..

Instead they reflect the variant in actual speed figure deltas.

Every winner's time regardless of the distance, is expressed as a raw speed figure and the TV is used to correct it.

The average figure is 100. G1 horses can run a 115+ while very weak MCL can go as low as 85- ..

There is some weakness in my algorithm that has to do with days when the traditional racing form speed figure is very high (especially on the turf)..

These figs are not cross track...

Have you tried combining a few days of turf racing to form a larger sample? That might help if the speed figure is high over the course of a few days.

cj
08-04-2015, 01:19 PM
Have you tried combining a few days of turf racing to form a larger sample? That might help if the speed figure is high over the course of a few days.

This is possible if you keep note of the various rail settings and gate placements.

DeltaLover
08-04-2015, 01:22 PM
Have you tried combining a few days of turf racing to form a larger sample? That might help if the speed figure is high over the course of a few days.

Not for this specific track variant I am posting here

DeltaLover
08-04-2015, 01:23 PM
This is possible if you keep note of the various rail settings and gate placements.

I think the impact of the rail setting and the gate placement, is already embedded in the final times, although I cannot make a string case here, since I have not done enough related research

cj
08-04-2015, 01:27 PM
I think the impact of the rail setting and the gate placement, is already embedded in the final times, although I cannot make a string case here, since I have not done enough related research

Well yes, it is, just saying you have to factor that in if lumping days together because they can and do change often. At some tracks, they change in the same day.

Tom
08-04-2015, 01:42 PM
I treat the various combinations as separate tracks.
I have two different speed charts for Rail = 0 and Rail =24 feet, for example.

cj
08-04-2015, 01:43 PM
I treat the various combinations as separate tracks.
I have two different speed charts for Rail = 0 and Rail =24 feet, for example.

I do as well, and of course *about* distances get the same treatment.

DeltaLover
08-04-2015, 01:47 PM
I treat the various combinations as separate tracks.
I have two different speed charts for Rail = 0 and Rail =24 feet, for example.

In theory this is the best approach... The problem I can see with it, lies in the fact that for several distances (especially on the turf) we do not have enough races to break them down in such a way...

Tom
08-04-2015, 02:37 PM
When that happens, I combine, say 24 and 27 feet together, until I get enough. Better than nothing.

highnote
08-04-2015, 03:42 PM
If you're making variants, wouldn't the variants have the rail movement factored in?

If the race is longer on Day 2 than on Day 1 because of the rail placement then wouldn't the variant on Day 2 would be larger than on Day 1 -- all else being equal?

A longer race has the same effect on final time as a slower surface, right?

For example, a one mile race with rail at zero feet might run, on average, in 1:38.00.

And a one mile race with the rail at 24 feet might run, on average, in 1:39:00.

Therefore, if you don't factor in the rail then on days when the rail is at 24 feet the variant will average about 1 second slow.

Or am I overlooking something?

cj
08-04-2015, 03:47 PM
If you're making variants, wouldn't the variants have the rail movement factored in?

If the race is longer on Day 2 than on Day 1 because of the rail placement then wouldn't the variant on Day 2 would be larger than on Day 1 -- all else being equal?

A longer race has the same effect on final time as a slower surface, right?

For example, a one mile race with rail at zero feet might run, on average, in 1:38.00.

And a one mile race with the rail at 24 feet might run, on average, in 1:39:00.

Therefore, if you don't factor in the rail then on days when the rail is at 24 feet the variant will average about 1 second slow.

Or am I overlooking something?

It doesn't work like that. Whether the rail is at 0 feet or 50 feet or 100 feet, if the distance is reported as one mile then it is supposed to be a mile race. The rail is factored in. (I'm not saying this is always accurate, but usually it is!) If it isn't exactly a mile, it is supposed to be reported as About One Mile.

But these moving rails are what cause timing to be so iffy for turf races. The points where fractional timings are taken shouldn't be stationary. I have my doubts about that though.

highnote
08-04-2015, 04:01 PM
It doesn't work like that. Whether the rail is at 0 feet or 50 feet or 100 feet, if the distance is reported as one mile then it is supposed to be a mile race. The rail is factored in. (I'm not saying this is always accurate, but usually it is!) If it isn't exactly a mile, it is supposed to be reported as About One Mile.

But these moving rails are what cause timing to be so iffy for turf races. The points where fractional timings are taken shouldn't be stationary. I have my doubts about that though.

I'll have to think about it some more, but, intuitively, it seems like the variant would capture the rail movement.

For example, I remember nights at Mountaineer where the rail on the dirt track was extremely dead because the mud was thick and sticky. I would hear the jockeys talking downstairs by the claim box how they were avoiding the rail. So this had the same effect as a temporary rail, but the variant would have captured that effect. At least it seems like it should have.

The race distance didn't change, except that there were no horses on the rail getting the shortest trip. So the horses were running in the 3 through N paths. Maybe that made the race a little longer?

I'm trying to think why this wouldn't work. I'm not saying it will work, but as long as the race is the correct distance the variants should capture the differences from normal.

cj
08-04-2015, 04:14 PM
I'll have to think about it some more, but, intuitively, it seems like the variant would capture the rail movement.

For example, I remember nights at Mountaineer where the rail on the dirt track was extremely dead because the mud was thick and sticky. I would hear the jockeys talking downstairs by the claim box how they were avoiding the rail. So this had the same effect as a temporary rail, but the variant would have captured that effect. At least it seems like it should have.

The race distance didn't change, except that there were no horses on the rail getting the shortest trip. So the horses were running in the 3 through N paths. Maybe that made the race a little longer?

I'm trying to think why this wouldn't work. I'm not saying it will work, but as long as the race is the correct distance the variants should capture the differences from normal.

It would, but I was talking about when grouping races together, that was all. You will have trouble if you group races from a Wednesday with the rail at 30 feet and a Thursday with no temp rail. I personally don't trust the timings when the rail moves around, but they are consistent in each setting. But even forgetting that, moving the rail changes the ease of running around the turns which changes the speed of the course IMO.

cj
08-04-2015, 04:14 PM
Here is an example of a mile race at Gulfstream with no temp rail:

F5B52pWiemc

Here is an example of a mile race at same track with temp rail at 120 feet

9fGUuXy7gdc

Notice the difference in how long it takes to get to the red mark on the temp rail in both cases.

The run up is listed at 189 feet in the 1st example and 32 in the second. Trakus does pretty good with the run ups and it is pretty clear the difference to reach that red mark is a lot more than 157 feet to my eye. That makes it obvious the timing isn't being started in the same spot.

But of course Gulfstream uses Trakus. Other tracks do not and you have to trust that they know how to time when the rail is moved. I know better.

highnote
08-04-2015, 04:37 PM
It would, but I was talking about when grouping races together, that was all. You will have trouble if you group races from a Wednesday with the rail at 30 feet and a Thursday with no temp rail. I personally don't trust the timings when the rail moves around, but they are consistent in each setting. But even forgetting that, moving the rail changes the ease of running around the turns which changes the speed of the course IMO.

that makes sense that you can't group them.

highnote
08-04-2015, 06:41 PM
Here is an example of a mile race at Gulfstream with no temp rail:

F5B52pWiemc

Here is an example of a mile race at same track with temp rail at 120 feet

9fGUuXy7gdc

Notice the difference in how long it takes to get to the red mark on the temp rail in both cases.

The run up is listed at 189 feet in the 1st example and 32 in the second. Trakus does pretty good with the run ups and it is pretty clear the difference to reach that red mark is a lot more than 157 feet to my eye. That makes it obvious the timing isn't being started in the same spot.

But of course Gulfstream uses Trakus. Other tracks do not and you have to trust that they know how to time when the rail is moved. I know better.


Great examples! Thanks!

If knowing how to time the race correctly is a problem because of rail movement then what about measuring distance correctly?

cj
08-04-2015, 10:02 PM
Great examples! Thanks!

If knowing how to time the race correctly is a problem because of rail movement then what about measuring distance correctly?

No guarantees on that either, but the raw times tell you they are in the ballpark! :)

sammy the sage
08-04-2015, 10:26 PM
No guarantees on that either, but the raw times tell you they are in the ballpark! :)

yeah...that would be the left field bleachers on SOME days ;)

highnote
08-05-2015, 12:36 AM
Here are my variants in comma-delimited for if anyone wants them.

All times are in seconds. A negative time means that the track was slow. So you would add the variant to get an adjusted time that is lower (faster) than the published time.

SAR,*,*,*,*,*
date,race,2f,4f,6f,ff
20150724,1,*,-1.06,-1.23,-2.25
20150724,2,-0.37,-1.06,*,-2.25
20150724,4,-0.37,-1.06,*,-2.25
20150724,6,-0.37,-1.06,*,-2.25
20150724,8,-0.37,-1.06,*,-2.25
20150724,10,-0.37,-1.06,*,-2.25
20150725,1,-0.4,-0.46,*,-0.51
20150725,4,-0.4,-0.46,*,-0.51
20150725,5,-0.4,-0.46,*,-0.51
20150725,7,-0.4,-0.46,*,-0.51
20150725,9,-0.4,-0.46,*,-0.51
20150725,11,-0.4,-0.46,*,-0.51
20150726,1,-0.06,0.02,*,0.83
20150726,3,-0.06,0.02,*,0.83
20150726,4,-0.06,0.02,*,0.83
20150726,8,*,0.02,0.49,0.83
20150726,10,*,0.02,0.49,0.83
20150727,1,-0.12,0.01,*,-1.11
20150727,2,-0.12,0.01,*,-1.11
20150727,3,-0.12,0.01,*,-1.11
20150727,5,*,0.01,0.01,-1.11
20150727,6,-0.12,0.01,*,-1.11
20150730,2,*,-0.81,-1.07,-3.12
20150730,3,0.16,-0.13,*,-2.19
20150730,4,0.16,-0.13,*,-2.19
20150730,5,0.16,-0.13,*,-2.19
20150730,6,*,-0.81,-1.07,-3.12
20150730,7,0.16,-0.13,*,-2.19
20150730,8,*,-0.81,-1.07,-3.12
20150730,9,*,-0.81,-1.07,-3.12
20150730,10,0.16,-0.13,*,-2.19
20150731,3,-0.23,-0.17,*,-0.17
20150731,4,*,-0.17,0.1,-0.17
20150731,5,*,-0.17,0.1,-0.17
20150731,7,-0.23,-0.17,*,-0.17
20150731,9,*,-0.17,0.1,-0.17
20150801,2,-0.12,0.05,*,0.16
20150801,3,-0.12,0.05,*,0.16
20150801,5,-0.12,0.05,*,0.16
20150801,8,-0.12,0.05,*,0.16
20150801,10,*,-0.05,0.31,0.17
20150802,2,0.21,0.17,*,0.17
20150802,5,0.21,0.17,*,0.17
20150802,8,0.21,0.17,*,0.17
20150802,9,0.21,0.17,*,0.17
20150802,10,*,0.17,0.3,0.17
20150803,2,0.23,0.41,*,-0.86
20150803,4,0.23,0.41,*,-0.86
20150803,5,0.23,0.41,*,-0.86
20150803,8,0.23,0.41,*,-0.86

no breathalyzer
08-05-2015, 12:59 AM
sorry i don't know how to start a new thread yet... or if its even thread worthy. but can some one tell me why the official charts and trackus times are not in line with each other, particularly on Aug. 1st ?

my only guess is because of the run-up.. can some one verify this please.

no breathalyzer
08-05-2015, 01:19 AM
sorry i don't know how to start a new thread yet... or if its even thread worthy. but can some one tell me why the official charts and trackus times are not in line with each other, particularly on Aug. 1st ?

my only guess is because of the run-up.. can some one verify this please.

i thought about deleting this since i actually caught up with this thread . but i'm still confused wouldn't the chart times be faster since i thought the timer don't start till after the run -up?

I guess what i really want to know is trakus more accurate? now i'm seeing times swinging back in forth.. some faster some slower.

cj
08-05-2015, 08:52 AM
i thought about deleting this since i actually caught up with this thread . but i'm still confused wouldn't the chart times be faster since i thought the timer don't start till after the run -up?

I guess what i really want to know is trakus more accurate? now i'm seeing times swinging back in forth.. some faster some slower.

Trakus only reports the times from the official starting point as well. They have the information, the time and distance of the run up portion, but it is deducted from what you see on the site.

highnote
08-05-2015, 11:03 AM
Trakus only reports the times from the official starting point as well. They have the information, the time and distance of the run up portion, but it is deducted from what you see on the site.

Any idea why they don't report all the info they're collecting?

Is someone using it to bet with?

cj
08-05-2015, 11:20 AM
Any idea why they don't report all the info they're collecting?

Is someone using it to bet with?

Last I checked, the tracks asked them not to publish it publicly. I feel the same way, somebody could certainly be using it for an edge.

Tom
08-05-2015, 11:49 AM
Yeah, sell it to some high rollers - why would the tracks give a crap about their customers - they never have.

highnote
08-05-2015, 01:10 PM
Last I checked, the tracks asked them not to publish it publicly. I feel the same way, somebody could certainly be using it for an edge.

What reason could the tracks possibly have for not wanting the information published?

cj
08-05-2015, 01:45 PM
What reason could the tracks possibly have for not wanting the information published?

I can't answer that. Probably worth asking one of the tracks that employ Trakus. It seems odd to me that all the tracks using Trakus's services don't want this, but that is what I was told. It has been a while though, maybe it has changed.

highnote
08-05-2015, 01:54 PM
I can't answer that. Probably worth asking one of the tracks that employ Trakus. It seems odd to me that all the tracks using Trakus's services don't want this, but that is what I was told. It has been a while though, maybe it has changed.


Sounds like something HANA might be interested in looking into.

cj
08-05-2015, 04:38 PM
You guys should check out the variants at Parx on turf this week. Especially yesterday, fastest variants I've ever seen.

highnote
08-05-2015, 07:08 PM
Do you think it is a timer issue -- or just a fast surface?

cj
08-06-2015, 01:59 AM
The course.

highnote
08-06-2015, 02:42 AM
Race #2 at Saratoga on Wednesday, Aug 5 is interesting because the winner exceeded the pace pars by the largest amount on the day. The winner's final time was about in line with the rest of the dirt races, but also was the farthest below par. So I'm thinking the horses in the 2nd race simply ran too fast early and then ran out of gas. Which means I won't award the 9 furlong races a daily variant of 1 second fast for the 6 furlongs pace. Instead, I'll probably lower it to 0.60 (or maybe lower). This would bring the route pace pars closer to the sprint pace pars. The big pace variant in the second race skews the averages quite a bit. However, I don't want to simply ignore the race. So I'll use the 2nd race variant, but only a little.

Using -0.60 for the final time daily variant looks about right because all the winners ran slower than par.



20150804 ACTUAL PARS DIFFERENCES
race# 2f 4f 6f ff | 2f 4f 6f ff | 2f 4f 6f ff class
2 46.89 71.52 111.34 | 47.99 72.92 110.49 | 1.1 1.4 -0.85 9f 3+ c12.5
3 48.7 72.63 111.9 | 48.46 73.23 111.74 | -0.24 0.6 -0.16 9f msw 3+ SB
7 21.95 45.28 71.34 | 22.18 45.44 70.91 | 0.23 0.16 -0.43 6f msw F 3+
9 22.31 45.44 82.9 | 22.47 45.11 82.29 | 0.16 -0.33 -0.61 7f stk F 3+
10 22.6 45.78 70.89 | 22.54 45.89 69.94 | -0.06 0.11 -0.95 6f c16 3+ NW2L
RTE DV 0.16 1.00 -0.60
AVG DV 0.11 0.16 -0.60
MED DV 0.16 0.11 -0.61

DeltaLover
08-06-2015, 08:09 AM
You guys should check out the variants at Parx on turf this week. Especially yesterday, fastest variants I've ever seen.


Cannot say the same.. Here is a graph with all the variants for PARX for the year so far (blue line is daily variant, red is moving average for last 6):

cj
08-06-2015, 09:41 AM
That is for turf? Even just looking at the raw times and the quality of horses tells you the track is super fast.

DeltaLover
08-06-2015, 10:41 AM
That is for turf? Even just looking at the raw times and the quality of horses tells you the track is super fast.

Sorry, I've missed that you were referring to TURF.. These are dirt vars

cj
08-06-2015, 11:04 AM
Sorry, I've missed that you were referring to TURF.. These are dirt vars

No worries...I deleted the image since it was so big. Sorry to highnote for getting off topic. Back to the Spa.

DeltaLover
08-06-2015, 11:10 AM
No worries...I deleted the image since it was so big. Sorry to highnote for getting off topic. Back to the Spa.

OK, I will still review my turf figs for philly later tonight to see how they look..

highnote
08-06-2015, 01:46 PM
I only maker Dirt variants. Hope to get Turf variants later this year, but it's a lot of work as most of what I do is done manually by reading result charts.

I'd love to automate the making of Dirt variants, but I don't think I can program a computer to make creative decisions.

A computer can do more repetitive tasks more quickly that I can, but it is hard to program it to make subtle judgments. Although, it might be the case that my subtle judgments are wrong and that a computer algorithm might be better because in the long run it is consistent.

To be decided...

cj
08-06-2015, 01:54 PM
I only maker Dirt variants. Hope to get Turf variants later this year, but it's a lot of work as most of what I do is done manually by reading result charts.

I'd love to automate the making of Dirt variants, but I don't think I can program a computer to make creative decisions.

A computer can do more repetitive tasks more quickly that I can, but it is hard to program it to make subtle judgments. Although, it might be the case that my subtle judgments are wrong and that a computer algorithm might be better because in the long run it is consistent.

To be decided...

Later today I'll post an example of one of my worksheets. I still have to make some judgments, but I have the computer do it first and go from there. I would NEVER trust the computer to do it alone. Too much happens in this game.

cj
08-06-2015, 07:29 PM
I think the 10th today is a must "break out" race from the others. Curious to see how you see it highnote.

Not sure, do you split variants at times?

highnote
08-07-2015, 01:57 AM
I think the 10th today is a must "break out" race from the others. Curious to see how you see it highnote.

Not sure, do you split variants at times?

I will split daily variants on occasion. I frequently split up the sprints and routes. Occasionally, I will separate the early sprints from the late ones and the same with routes.

Race #10 was the fastest on the day based on the differences from my pars. It was slower than my par by half a second.

The interesting thing about today is that last Thursday was also a slow day and then Saturday the track ran fast. So maybe the track wasn't slow today -- maybe the horses are slow on Thursdays?

If you look at my variants for each race of the day from the 2 furlong variant all the way to the final time variant you see the classic telescoping effect -- that is, the differences from par get wider as the distances get longer.

2furlongs..4furlongs..6furlongs..final time
0.45fast, -0.76slow, -0.83slow, -2.01slow

But would the track have been tiring to any class of horses, or does it just appear to be a slow, tiring track because today's runners are just naturally slow.

Projecting the variants from a good set of past figures would probably be the best way to answer the question.

20150806 ACTUAL PARS DIFFERENCES
race# 2f 4f 6f ff | 2f 4f 6f ff | 2f 4f 6f ff class
2 49.97 74.52 113.35 | 47.92 47.92 72.79 110.22 | -2.05 -1.73 -3.13 9f c25 F 3+
3 22 45.46 78.84 | 22.7 46.27 77.08 | 0.7 0.81 -1.76 6.5f mcl25 3+ SB
5 49.97 73.89 113.47 | 48.06 73.09 110.79 | -1.91 -0.8 -2.68 9f c16NW2L
8 22.21 45.67 71.42 | 22.41 45.63 69.41 | 0.2 -0.04 -2.01 6f c40 F 3yo
10 48.87 73.43 112.16 | 48.26 73.47 111.59 | -0.61 0.04 -0.57 9f mcl30 3+
RTE AVG DV -0.76 -0.83 -2.03
RTE MED DV -0.61 -0.80 -2.01
SPR AVG DV 0.45 -0.76 -2.03
SPR MED DV 0.45 -0.61 -2.01

steveb
08-07-2015, 06:25 AM
I will split daily variants on occasion. I frequently split up the sprints and routes. Occasionally, I will separate the early sprints from the late ones and the same with routes.

Race #10 was the fastest on the day based on the differences from my pars. It was slower than my par by half a second.

The interesting thing about today is that last Thursday was also a slow day and then Saturday the track ran fast. So maybe the track wasn't slow today -- maybe the horses are slow on Thursdays?

If you look at my variants for each race of the day from the 2 furlong variant all the way to the final time variant you see the classic telescoping effect -- that is, the differences from par get wider as the distances get longer.

2furlongs..4furlongs..6furlongs..final time
0.45fast, -0.76slow, -0.83slow, -2.01slow

But would the track have been tiring to any class of horses, or does it just appear to be a slow, tiring track because today's runners are just naturally slow.

Projecting the variants from a good set of past figures would probably be the best way to answer the question.

20150806 ACTUAL PARS DIFFERENCES
race# 2f 4f 6f ff | 2f 4f 6f ff | 2f 4f 6f ff class
2 49.97 74.52 113.35 | 47.92 47.92 72.79 110.22 | -2.05 -1.73 -3.13 9f c25 F 3+
3 22 45.46 78.84 | 22.7 46.27 77.08 | 0.7 0.81 -1.76 6.5f mcl25 3+ SB
5 49.97 73.89 113.47 | 48.06 73.09 110.79 | -1.91 -0.8 -2.68 9f c16NW2L
8 22.21 45.67 71.42 | 22.41 45.63 69.41 | 0.2 -0.04 -2.01 6f c40 F 3yo
10 48.87 73.43 112.16 | 48.26 73.47 111.59 | -0.61 0.04 -0.57 9f mcl30 3+
RTE AVG DV -0.76 -0.83 -2.03
RTE MED DV -0.61 -0.80 -2.01
SPR AVG DV 0.45 -0.76 -2.03
SPR MED DV 0.45 -0.61 -2.01


it's very easy to see what you are doing wrong, and WHY you are getting your 'telescoping effect'.
so long as i understand what you are doing, which i may have wrong.
if you are simply subtracting a class par from the race time(or vice versa) then of course as the distances increase they are going to look slow normally.
1 second over 6 furlongs has a much different meaning that 1 second over 9 furlongs.

for instance in your 1st race you have -3.13 seconds and only -1.76 seconds for the second one which is a difference of 1.37 which is humungous(excuse my english), but my way there would be less than 2 lengths difference between them(about 5.5 metres), or less than 6 points(and that's assuming that your pars are right).

obviously too, you think my one time for each track and distance is not the right way to go! :)

and the main problem in dealing with time in my humble opinion, is using time to make your calculations.

i would also guess that a meeting with mainly sprints would give you faster variants than one with a mix or more routes.

highnote
08-07-2015, 09:27 AM
The 10th race route was run closer to par than either of the sprint races, so it is not always the case that sprints are run closer to par than routes, but I see your point.

I just use the method Quirin talks about in his book. It probably makes more sense to make separate sprint and route variants to avoid having too big of a sprint variant or two small of a route variant.

Yesterday it just looked like you got the telescoping effect when sprint and route variants were combined. But when you made separate sprint and route variants the effect disappeared.

But I did these at 1 in the morning so maybe I didn't analyze them as closely as I should have!

highnote
08-08-2015, 01:49 AM
it's very easy to see what you are doing wrong, and WHY you are getting your 'telescoping effect'.
so long as i understand what you are doing, which i may have wrong.
if you are simply subtracting a class par from the race time(or vice versa) then of course as the distances increase they are going to look slow normally.

The first time I read your reply I tended to agree, but now that I've had a day to think about it, I'm not so sure. I'll think about it some more and reply because if I can't explain it that means I don't understand it. And right now, I don't understand it well enough.

1 second over 6 furlongs has a much different meaning that 1 second over 9 furlongs.

I agree with this.

for instance in your 1st race you have -3.13 seconds and only -1.76 seconds for the second one which is a difference of 1.37 which is humungous(excuse my english), but my way there would be less than 2 lengths difference between them(about 5.5 metres), or less than 6 points(and that's assuming that your pars are right).

I need to check my variants again because those do seem like big differences.


obviously too, you think my one time for each track and distance is not the right way to go! :)

Not really. Nick Mordin wrote in his book, "Mordin On Time" about how he uses a "standard time" to make his figures. His standard time was the time a theoretical Grade or Group 1 horse could run. Then he compared the actual times to the standard times to arrive at a daily variant. Dr. Charles Carroll in his book, "Handicapping Speed" also used "one time" to make his track "constants".


i would also guess that a meeting with mainly sprints would give you faster variants than one with a mix or more routes.

This is probably true when using "one time", but not sure about class pars times. We know that sprinters tend to run closer to track records than horses running 12 furlongs. If the track record for 6 furlongs is 68 seconds and the track record for 12 furlongs is 144 seconds a horse that runs 69 seconds is only one second off the track record, but is still average. A horse that runs 12 furlongs in 145 seconds is world class.

Now when using class pars there is probably more variance around the sprint class par times than the route class par times, but I would think that the variances in times around class pars are less than the variances around "one times". And, of course, as the distances get longer the variances around one times will get larger, too.

I wrote software for Nick to make speed figures using his method and he picked a lot of good horses with his figures, so I know "one times" can work well.

I use Quirin's class par method with some of my own twists because I figure the way to add value is to do something different.

Nick showed me a good method for making "handicap" figures where the method does not rely on time at all, but instead is a "who beat who" method. I think this would work very well on American Turf races and is something I may add to my "bag of tricks".

highnote
08-08-2015, 01:37 PM
Friday, Aug 7 was another tricky day at the Spa, variant-making-wise.

The first race had a race shape of Average/Average/Slow.

The third race is the one that made things tricky because it was extremely slow run at the pace calls, but the final time was slightly faster than par. It was the only dirt route on the day -- a 9 furlong race for 14k claimers. The race shape was SLOW/SLOW/AVERAGE. The winner had been dropped from 40k turf to 14k dirt. A Bill Mott class dropper was the favorite and finished third. There were two claims from this race, but the winner did NOT get claimed. The winner was the interesting horse just because of how far it dropped and is the horse I would have claimed (red board :D ).

But what makes this day tricky is that the first and third races both were run slower than par to pace calls. The first race was run just a tick below par at the pace, but was a second and half slower than par at the finish. While the third race was run a second and a half slow to the pace par, but finished a tick faster than par. The third race had a lot of class droppers, but the eventual winner ran near the lead to a slow pace and then was able to pull away. I would have thought these big class droppers could have run faster than the 14k pace pars. I think what happened is that they just slowed down the pace.

Bottom line... I won't use the pace variants from this race to calculate the daily variants.

The fourth and sixth race both had race shapes of AVERAGE/AVERAGE/FAST and they both were run about a tick faster than par at the pace and 2 or 3 ticks faster than par at the finish.

In the end, it looks like the track was running about normal and I'll probably make the daily variant 0-0-0 for sprints and routes.

20150807 ACTUAL PARS DIFFERENCES
race# 2f 4f 6f ff | 2f 4f 6f ff | 2f 4f 6f ff class
1 22.75 46.18 71.29 | 22.49 45.85 69.86 | -0.26 -0.33 -1.43 6f c14 NW3L 3+
4 21.96 45.1 69.94 | 22.08 45.24 70.51 | 0.12 0.14 0.57 msw 3+
6 22.28 45.87 64.63 | 22.54 46.2 65.4 | 0.26 0.33 0.77 5.5f msw F 2yo SB
3 49.52 74.47 111.47 | 47.95 72.85 111.75 | -1.57 -1.62 0.28 9f c14 NW3L 3+
RTE AVG DV 0 0 0.28
RTE MED DV 0.14 0.57
AVG SPR DV 0.04 0.05 -0.03
MED SPR DV 0.12 0.14 0.57

highnote
08-09-2015, 06:02 PM
Saturday, Aug 8 -- It looks like the sprint and route variants need to be split. The routes were run fast early and average late, while the sprints were slow early and slow late.

20150808 ACTUAL PARS DIFFERENCES
race# 2f 4f 6f ff | 2f 4f 6f ff | 2f 4f 6f ff class
3 22.5 45.79 71.01 | 22.31 45.61 70.68 | -0.19 -0.18 -0.33 6f msw 2yo
6 23.4 46.34 82.75 | 22.76 45.66 81.63 | -0.64 -0.68 -1.12 7f AOC 62.5 3+
8 23.21 46.21 82.81 | 22.57 45.31 82.69 | -0.64 -0.9 -0.12 7f stk F 3yo
1 46.53 71.09 110.1 | 47.69 72.32 109.3 | 1.16 1.23 -0.8 9f c32 3+
9 46 69.72 107.82 | 47.81 71.93 108.94 | 1.81 2.21 1.12 9f stk 3+
RTE AVG DV 1.49 1.72 0.16
RTE MED DV 1.49 1.72 0.16
AVG SPR DV -0.49 -0.59 -0.52
MED SPR DV -0.64 -0.14 -0.68

steveb
08-09-2015, 06:48 PM
Saturday, Aug 8 -- It looks like the sprint and route variants need to be split. The routes were run fast early and average late, while the sprints were slow early and slow late.

20150808 ACTUAL PARS DIFFERENCES
race# 2f 4f 6f ff | 2f 4f 6f ff | 2f 4f 6f ff class
3 22.5 45.79 71.01 | 22.31 45.61 70.68 | -0.19 -0.18 -0.33 6f msw 2yo
6 23.4 46.34 82.75 | 22.76 45.66 81.63 | -0.64 -0.68 -1.12 7f AOC 62.5 3+
8 23.21 46.21 82.81 | 22.57 45.31 82.69 | -0.64 -0.9 -0.12 7f stk F 3yo
1 46.53 71.09 110.1 | 47.69 72.32 109.3 | 1.16 1.23 -0.8 9f c32 3+
9 46 69.72 107.82 | 47.81 71.93 108.94 | 1.81 2.21 1.12 9f stk 3+
RTE AVG DV 1.49 1.72 0.16
RTE MED DV 1.49 1.72 0.16
AVG SPR DV -0.49 -0.59 -0.52
MED SPR DV -0.64 -0.14 -0.68


aside from the fact you are doing it wrong(quirin or not), you also need to think about races when they are restricted by age.

especially 2yo's which are going to vary hugely during the course of a season.

3yo's will be the same, although higher class 3yo races will scale higher(get faster quicker during course of year) than weak ones.

before i would split any variant i would iterate through the races, splitting them into 2 groups

1,2 / 3,4,5,6,7,8,9,10
1,2,3 / 4,5,6,7,8,9,10
1,2,3,4/5,6,7,8,9,10
etc etc

at each iteration making 2 deviations(expected less actual) and then seeing if the sum of the 2 is less than the deviation from actual to expected for all races.
if it is then you can be fairly sure that's when the change happened, and when splitting is better then not splitting.

there are other ways too, but that's just one.

highnote
08-09-2015, 07:15 PM
aside from the fact you are doing it wrong(quirin or not), you also need to think about races when they are restricted by age.

How do you know if it's wrong?

especially 2yo's which are going to vary hugely during the course of a season.

3yo's will be the same, although higher class 3yo races will scale higher(get faster quicker during course of year) than weak ones.

I make separate par tables for 2 year olds and the pars for 3 year olds are seasonally adjusted. Altogether I use 12 different sets of pars.

before i would split any variant i would iterate through the races, splitting them into 2 groups

1,2 / 3,4,5,6,7,8,9,10
1,2,3 / 4,5,6,7,8,9,10
1,2,3,4/5,6,7,8,9,10
etc etc

at each iteration making 2 deviations(expected less actual) and then seeing if the sum of the 2 is less than the deviation from actual to expected for all races.
if it is then you can be fairly sure that's when the change happened, and when splitting is better then not splitting.

there are other ways too, but that's just one.

That's a good idea. I will look into it.

For Saratoga there are only 3 to 6 dirt races per day, so it's not too hard to grasp the totality of all the variants.

SteveR who also posts here gave a good method using linear regression in the American Pharoah Triple Crown thread.

steveb
08-09-2015, 07:29 PM
How do you know if it's wrong?




i know it's wrong because of the way you simply subract one from the other.
i don't care if quirin says that's the go, it's not right.

there is no way that a second over 5f can be treated the same as a second over any other distance, and the bigger the difference then the more wrong on average it will be.

i have not read any quirin books, but if that is how he reckons, then he is wrong.
i know it's wrong and i would think that lots of people on this forum know it's wrong too.
the difference is that i don't mind telling you.

the same as beyer was wrong with his beaten margin charts in the book i read of his in the long ago.
his mistake was treating the also rans the same way as the winners of the races.

highnote
08-10-2015, 12:40 AM
i know it's wrong because of the way you simply subract one from the other.
i don't care if quirin says that's the go, it's not right.

there is no way that a second over 5f can be treated the same as a second over any other distance, and the bigger the difference then the more wrong on average it will be.

i have not read any quirin books, but if that is how he reckons, then he is wrong.
i know it's wrong and i would think that lots of people on this forum know it's wrong too.
the difference is that i don't mind telling you.

the same as beyer was wrong with his beaten margin charts in the book i read of his in the long ago.
his mistake was treating the also rans the same way as the winners of the races.

A few points:

It sounds like we are talking about two different things. I'm talking about a process to measure the speed of a DIRT racing surface, not a grass surface. Grass racing in the U.S. is very different than dirt racing. The process I'm doing will yield individual race variants which can then be used to gauge the speed of the track on a given day, hence, daily variants.

The daily variants are then used to create pace and speed figures by adding or subtracting the daily variants to the actual times of races and then adjusting each horse's times by their beaten lengths at the various sectional times. This is the stage where a "proportional" beaten length chart is used to make the speed figure. I alluded to this a few posts back and this was the key insight Beyer got from his classmate at Harvard.

At shorter distances the value of a length is different than at longer distances because horses run faster at the end of 5 furlongs on DIRT than they do at 10 furlongs on DIRT. The same might not be true, or truer to a lesser extant, on U.S. grass races because horses in U.S. grass races tend to run slower and tactfully in the early part of the race and then sprint to the wire. Whereas, on U.S. dirt racing horses tend to run faster early and then decelerate nearing the wire.

So when making track variants I think Quirin, Beyer, Quinn, Brohammer, Carroll and others who wrote about using this method are correct. It makes sense to compare the actual time of a race to the table of class pars and determine the difference from par that a race was run. Then, regardless of distance (with some exceptions), the deviations from par for a group of races on the same surface can be used to gauge the speed of a track.

I'm not saying this is the only way to do it for Dirt racing, but of all the methods I've read, this one makes the most sense to me.

Quirin admits in his book that this method is not really applicable to grass racing. Beyer has been adjusting his grass method for years and still has not achieved satisfactory results. He wrote in his last book about using his method in Australia. It's very interesting. What he found was that the Australian handicappers were very sophisticated and bet huge sums of money. They have to bet huge in order to make decent money because their edge is so small because the competition is so good.

Quinn also wrote an entire book about making speed figures with a few chapters dedicated to making grass figures. I have never had good results with his grass method, but I may revisit it once I've finished the software for dirt racing.

steveb
08-10-2015, 02:10 AM
A few points:

It sounds like we are talking about two different things. I'm talking about a process to measure the speed of a DIRT racing surface, not a grass surface. Grass racing in the U.S. is very different than dirt racing. The process I'm doing will yield individual race variants which can then be used to gauge the speed of the track on a given day, hence, daily variants.

The daily variants are then used to create pace and speed figures by adding or subtracting the daily variants to the actual times of races and then adjusting each horse's times by their beaten lengths at the various sectional times. This is the stage where a "proportional" beaten length chart is used to make the speed figure. I alluded to this a few posts back and this was the key insight Beyer got from his classmate at Harvard.

At shorter distances the value of a length is different than at longer distances because horses run faster at the end of 5 furlongs on DIRT than they do at 10 furlongs on DIRT. The same might not be true, or truer to a lesser extant, on U.S. grass races because horses in U.S. grass races tend to run slower and tactfully in the early part of the race and then sprint to the wire. Whereas, on U.S. dirt racing horses tend to run faster early and then decelerate nearing the wire.

So when making track variants I think Quirin, Beyer, Quinn, Brohammer, Carroll and others who wrote about using this method are correct. It makes sense to compare the actual time of a race to the table of class pars and determine the difference from par that a race was run. Then, regardless of distance (with some exceptions), the deviations from par for a group of races on the same surface can be used to gauge the speed of a track.

I'm not saying this is the only way to do it for Dirt racing, but of all the methods I've read, this one makes the most sense to me.

Quirin admits in his book that this method is not really applicable to grass racing. Beyer has been adjusting his grass method for years and still has not achieved satisfactory results. He wrote in his last book about using his method in Australia. It's very interesting. What he found was that the Australian handicappers were very sophisticated and bet huge sums of money. They have to bet huge in order to make decent money because their edge is so small because the competition is so good.

Quinn also wrote an entire book about making speed figures with a few chapters dedicated to making grass figures. I have never had good results with his grass method, but I may revisit it once I've finished the software for dirt racing.

it is irrelevant if it is turf or dirt insofar as what i am saying is concerned.
japan has dirt races
hong kong has dirt races
korea has dirt races.
and it is true of those and it will be true of your fine country too.

all i am doing is trying to help you, but i know some people don't want help, and that is fine
maybe others are taking something from what i say, who knows?

there is no way beyer is doing what you say he is, so that may mean that the others are not doing it that way ether, and you are interpreting what they say wrongly.

your software will only be as good as the logic behind it, and i know your logic is wrong, so.......

and i will never write a book, but that does not mean i don't think i am better at it than they are. :)

highnote
08-10-2015, 07:16 AM
it is irrelevant if it is turf or dirt insofar as what i am saying is concerned.
japan has dirt races
hong kong has dirt races
korea has dirt races.
and it is true of those and it will be true of your fine country too.

all i am doing is trying to help you, but i know some people don't want help, and that is fine
maybe others are taking something from what i say, who knows?

there is no way beyer is doing what you say he is, so that may mean that the others are not doing it that way ether, and you are interpreting what they say wrongly.

your software will only be as good as the logic behind it, and i know your logic is wrong, so.......

and i will never write a book, but that does not mean i don't think i am better at it than they are. :)

I appreciate your help. I think the way you are doing it makes sense, but it's just not the way I chose to do it and if I find out that my figure do not work then I will change what I'm doing. But until I know if they work or not then I have no incentive to change.

As far as Beyer, I'm doing it exactly the way he does it -- using class pars. The biggest difference is that he does not make pace figures -- only final time figures.

I don't see the faulty logic in Quirin's method. Given that he's a college professor and has a doctorate in mathematics, there is a good chance his logic is better than mine. But it does not mean his method is better than yours -- only different.

Nick Mordin and Charles Carroll wrote about a method similar to yours. So I know people make figures the way you do and use them successfully, but your way and their way is not the only way.

Charles Carroll wrote that one of the things he learned from Beyer is to do things his own way. Cary Fotias made his Equiform figures in a way that no one else was doing. My method of producing figures adds a couple new twists, but you have never seen my figures -- only my daily variant worksheets.

The main value of making figures is making them unique enough that they have added value. If I make mine exactly like Beyer then they will probably not add much value.

The one thing I do like about your method is that it is much easier. Maybe I will make a second set of figures using your method. Maybe by combining them I can come up with a "super" figure? :ThmbUp:

steveb
08-10-2015, 08:02 AM
my last bit on the subject(unless i change my mind!) as i probably talk too much.
but the attached may help explain why i know you are doing it wrong.
i should say that i had already done this, as it sometimes helps me to clarify things in my own mind.
i don't mind admitting that i am probably past it, and although i like studying racing, i would rather be watching birds(feathered endangered plovers) on the beach.
pity the weather is so bad lately that i can't, and thus i start raving on fora instead.

i am yet to read your last reply.

steveb
08-10-2015, 08:52 AM
ok, one more!
just because i have used beyer to demonstrate, don't think it is copying him, it's not.

and i have now read.
the better your figures are the more chance you have in the long run.
my figures were probably learned from wanting to understand what the beyer points actually were.
so i can credit him with my interest in times, when once upon a long time ago, i thought they(times) were meaningless.
but there has been a lot of water under the bridge since then.

and i don't see how any of that would stop me knowing the importance of section times.
i simply did not mention them at all, but that does not mean i don't use them.
it's another topic i think.

highnote
08-10-2015, 09:33 AM
my last bit on the subject(unless i change my mind!) as i probably talk too much.
but the attached may help explain why i know you are doing it wrong.
i should say that i had already done this, as it sometimes helps me to clarify things in my own mind.
i don't mind admitting that i am probably past it, and although i like studying racing, i would rather be watching birds(feathered endangered plovers) on the beach.
pity the weather is so bad lately that i can't, and thus i start raving on fora instead.

i am yet to read your last reply.


Good stuff.

In your example, you are making the 6 furlong race 1/2 second faster and the 8 furlong race a 1/2 second slower. That is a difference of 1 second -- which is a lot.

Surely, Beyer understands this. To me, it is not so much a flaw in his method as it is a flaw in the application of his method.

Perhaps he would have split the variant and made a separate one for the 6f race and another for the 8f race? Or perhaps he would have projected the variant? Or perhaps after the horses ran their next race he would have gone back to this race and changed the variant?

Your method looks easier to implement and that in and of itself might make it superior.

Ed Thorp's original blackjack system was fairly difficult to implement, but when used properly was a great system. I found other blackjack systems easier to use and therefore I made fewer mistakes. So for me, the easier blackjack systems were superior, but only because they were easier, not because they were better when used properly.

steveb
08-10-2015, 09:40 PM
Good stuff.

In your example, you are making the 6 furlong race 1/2 second faster and the 8 furlong race a 1/2 second slower. That is a difference of 1 second -- which is a lot.

Surely, Beyer understands this. To me, it is not so much a flaw in his method as it is a flaw in the application of his method.

Perhaps he would have split the variant and made a separate one for the 6f race and another for the 8f race? Or perhaps he would have projected the variant? Or perhaps after the horses ran their next race he would have gone back to this race and changed the variant?

Your method looks easier to implement and that in and of itself might make it superior.

Ed Thorp's original blackjack system was fairly difficult to implement, but when used properly was a great system. I found other blackjack systems easier to use and therefore I made fewer mistakes. So for me, the easier blackjack systems were superior, but only because they were easier, not because they were better when used properly.


no i am not, i am understanding the fact that 61 seconds with a par of 60 over 5 furlongs , is not as good as a time of 101 seconds with a par of 100 seconds.
you obviously can't see(or understand) that distinction, which has me surprised.
beyer has faults for sure, but this case is not one of them.
anyway that's it for me, i have decided it's an exercise in futility.

highnote
08-11-2015, 01:02 AM
no i am not, i am understanding the fact that 61 seconds with a par of 60 over 5 furlongs , is not as good as a time of 101 seconds with a par of 100 seconds.
you obviously can't see(or understand) that distinction, which has me surprised.
beyer has faults for sure, but this case is not one of them.
anyway that's it for me, i have decided it's an exercise in futility.


If you think I don't understand this then you must not have read what I wrote in post #94:

We know that sprinters tend to run closer to track records than horses running 12 furlongs. If the track record for 6 furlongs is 68 seconds and the track record for 12 furlongs is 144 seconds a horse that runs [6 furlongs in] 69 seconds is only one second off the track record, but is still average. A horse that runs 12 furlongs in 145 seconds is world class.


I understand the concept. You gave an example of how averaging the variants from a 6 furlong race and an 8 furlong race produce dubious speed figures. I agreed.

So I'm not sure where your frustration with me is coming from.

I do appreciate your time and you've given up more of your time than is necessary. I'll go back and re-read your posts because I must be missing something. I was in bed all weekend with the flu. So maybe that is part of the problem.

cj
08-11-2015, 01:06 AM
If you think I don't understand this then you must not have read what I wrote in post #94:



I understand the concept. You gave an example of how averaging the variants from a 6 furlong race and an 8 furlong race produce dubious speed figures. I agreed.

So I'm not sure where your frustration with me is coming from.

I do appreciate your time and you've given up more of your time than is necessary. I'll go back and re-read your posts because I must be missing something. I was in bed all weekend with the flu. So maybe that is part of the problem.

I use points based variants instead of time, and of course they treat time differently based on distance. That said, the route races have been playing faster at Saratoga recently by my point system.

highnote
08-11-2015, 01:41 AM
I use points based variants instead of time, and of course they treat time differently based on distance. That said, the route races have been playing faster at Saratoga recently by my point system.

The past few days the route variants look faster. Why do you think that might be? Maybe because the clubhouse turn is faster than the far turn? Or maybe the routers are just better than the sprinters at Saratoga this week?

highnote
08-11-2015, 02:16 AM
It looks like using the median pace variants for sprints is the way to go for the Sunday, Aug 9 daily variants because the average is skewed by the 4th race in which the pace variants for the 4th were much slower than the rest of the races.

The route variant looks like it needs to be split from the sprints.


20150809 ACTUAL PARS DIFFERENCES
race# 2f 4f 6f ff | 2f 4f 6f ff | 2f 4f 6f ff class
1 22.72 45.96 70.84 | 22.44 45.69 69.54 | -0.28 -0.27 -1.3 6f c16 3+
4 23.91 47.24 82.89 | 22.57 45.31 83.09 | -1.34 -1.93 0.2 7f alw nw1 3+
6 22.22 45.88 64.65 | 22.13 45.55 64.63 | -0.09 -0.33 -0.02 5.5f msw F 2yo
7 22.7 46.05 77.04 | 22.49 45.87 77.5 | -0.21 -0.18 0.46 6.5f alwnw1 F 3+ SB
9 46.84 70.94 108.92 | 47.81 71.93 108.94 | 0.97 0.99 0.02 9f stk 3+
RTE AVG DV 0.97 0.99 0.02
RTE MED DV 0.97 0.99 0.02
AVG SPR DV -0.48 -0.68 -0.17
MED SPR DV -0.24 -0.30 0.09

highnote
08-11-2015, 03:07 AM
The variants for Monday, Aug 10 are fairly difficult to discern -- for me -- probably because 3 of the 4 dirt races were for state breds.

I'll probably split the sprints and routes and use the simple average.

20150510 ACTUAL PARS DIFFERENCES
race# 2f 4f 6f ff | 2f 4f 6f ff | 2f 4f 6f ff class
1 22.61 45.71 57.8 | 22.15 45.87 58.59 | -0.46 0.16 0.79 5f msw 2yo sb
7 22.02 45.27 70.3 | 22.37 45.6 69.15 | 0.35 0.33 -1.15 6f aoc c40 3+ sb
2 47.94 72.18 109.39 | 47.54 72.01 108.68 | -0.4 -0.17 -0.71 9f c50 3+
4 49.37 73.97 110.84 | 47.69 72.31 109.28 | -1.68 -1.66 -1.56 9f alw nw1 sb
RTE AVG DV -1.04 -0.91 -1.14
RTE MED DV -1.04 -0.91 -1.14
AVG SPR DV -0.05 0.24 -0.18
MED SPR DV -0.05 0.24 -0.18

cj
08-11-2015, 09:46 AM
The past few days the route variants look faster. Why do you think that might be? Maybe because the clubhouse turn is faster than the far turn? Or maybe the routers are just better than the sprinters at Saratoga this week?

I don't think it is that the routers are just better because I use projections, not pars. I really haven't tried to figure out they why, but it has been consistent for a few days now.

highnote
08-11-2015, 02:27 PM
no i am not, i am understanding the fact that 61 seconds with a par of 60 over 5 furlongs , is not as good as a time of 101 seconds with a par of 100 seconds.
you obviously can't see(or understand) that distinction, which has me surprised.
beyer has faults for sure, but this case is not one of them.
anyway that's it for me, i have decided it's an exercise in futility.


Now that I am starting to recover from the flu I understand better what you are saying. You are correct that averaging the sprint and route differences from those two races and then calling that a daily variant would lead to a less accurate speed figure than your method.

I agree with what you wrote.

Thanks!

highnote
08-13-2015, 12:18 AM
Saratoga daily dirt variants for Wednesday, Aug 12:

The sprint variants make sense today. Nothing out of the ordinary. There were about par for the pace calls and about 0.8 seconds slow at the finish.

The one route race on the card -- a 14 furlong stakes race -- is not a typical distance run on the dirt at the Spa. The splits were about 26 seconds -- which seem slow to me.

I have a par time for 14 furlongs of 171.94 seconds. The race was run in 179.15 seconds -- over 7 seconds slower than par.

The big question is whether or not my par time is correct for 14 furlongs on the dirt at Saratoga. I'll guess at the par and then may update it later after these horses run again.


20150812 ACTUAL PARS DIFFERENCES
race# 2f 4f 6f ff | 2f 4f 6f ff | 2f 4f 6f ff class
2 22.5 45.48 76.59 | 22.24 45.38 76.1 | -0.26 -0.1 -0.49 6.5f c40F 3+
3 22.55 45.99 71.41 | 22.77 46.4 70.75 | 0.22 0.41 -0.66 6f MCL40 F 3+ SB
4 21.96 45.03 70.81 | 22.31 45.43 69.01 | 0.35 0.4 -1.8 6f c40 3yo
5 22.18 46.14 64.89 | 22.13 45.55 64.63 | -0.05 -0.59 -0.26 5.5f msw F 2yo
8 49.92 76.16 179.15 | 47.81 71.93 171.94 | -2.11 -4.23 -7.21 14f STK 3+
RTE AVG DV -2.11 -4.23 -7.21
RTE MED DV -2.11 -4.23 -7.21
AVG SPR DV 0.06 0.03 -0.80
MED SPR DV 0.08 0.15 -0.58

steveb
08-15-2015, 01:53 AM
I don't think it is that the routers are just better because I use projections, not pars. I really haven't tried to figure out they why, but it has been consistent for a few days now.

what does 'use projections' mean, if you don't mind me asking?
if you think 'so and so' race should have run an 80, how do you figure if it has or not, unless time is involved somehow?

highnote
08-15-2015, 02:05 AM
Hey, steveb -- glad you're back.

I have taken your advice and am in the process of redoing my daily variants using your method.

I'll let you know how it goes.

The scale that I use is different from the example you showed, so I won't post my daily variants anymore because they only have meaning to someone using my scale.

Dark Target
08-15-2015, 02:35 AM
Presumably he means that how fast / slow the track is gets worked out by individual horse projections, not how fast or slow each race was compared to the average for that "class" of race (which is still a varient of sorts anyway).

steveb
08-15-2015, 02:41 AM
Presumably he means that how fast / slow the track is gets worked out by individual horse projections, not how fast or slow each race was compared to the average for that "class" of race (which is still a varient of sorts anyway).

strewth,talk of the devil i just sent you an email two ticks ago!

steveb
08-15-2015, 03:05 AM
Hey, steveb -- glad you're back.

I have taken your advice and am in the process of redoing my daily variants using your method.

I'll let you know how it goes.

The scale that I use is different from the example you showed, so I won't post my daily variants anymore because they only have meaning to someone using my scale.

i never went anywhere highnote i just thought it was pointless continuing in the same vein.

highnote
08-15-2015, 04:39 AM
i never went anywhere highnote i just thought it was pointless continuing in the same vein.

I can understand why you thought it was pointless!

What you wrote makes perfect sense.

I was sick with the flu and wasn't able to focus very well.

I re-read Beyer's book which helped me to understand what you were illustrating.

My figures look much better now. THANKS!

EMD4ME
08-17-2015, 07:01 PM
Question for all here:

Aug 15th, 2015. Beyer had 2 vastly different variants. Had races 9 & 11 as run on deeper/slower tracks, hence in adjustment in numbers for those races as opposed to races 1,3,4 and 6.

I see race 9 had an abnormally slow Q(Quarter)1 & Q2 but had an excellent 3rd quarter and a good last 1/8. My opinion is that the track did not change after race 6 and the pace of 9 caused the slower overall time. (partially caused by a shorter run up)

I bring this is up for the following. Race 11 was also adjusted (obviously by the slower variant) and I don't see Aleander running an 81 Beyer Fig (even though he raced in the better part of the track, had a nice trip etc.). I feel that the Beyer team over credited the 11th race (due to their decision that race 9 and 11 were effected by a change in variant) and there shouldn't be a seperate variant as the pace of race 9 caused the abnormal time.

I hand timed race 9 and it seemed right. Plus that race did have horses rating. Run up was kind of short at 34 feet vs. 60 feet for the 6 1/2 races that day.

CJ, what was your assessment? Am I crazy to feel this way?

Everyone else, thoughts?

cj
08-17-2015, 08:55 PM
Question for all here:

Aug 15th, 2015. Beyer had 2 vastly different variants. Had races 9 & 11 as run on deeper/slower tracks, hence in adjustment in numbers for those races as opposed to races 1,3,4 and 6.

I see race 9 had an abnormally slow Q(Quarter)1 & Q2 but had an excellent 3rd quarter and a good last 1/8. My opinion is that the track did not change after race 6 and the pace of 9 caused the slower overall time. (partially caused by a shorter run up)

I bring this is up for the following. Race 11 was also adjusted (obviously by the slower variant) and I don't see Aleander running an 81 Beyer Fig (even though he raced in the better part of the track, had a nice trip etc.). I feel that the Beyer team over credited the 11th race (due to their decision that race 9 and 11 were effected by a change in variant) and there shouldn't be a seperate variant as the pace of race 9 caused the abnormal time.

I hand timed race 9 and it seemed right. Plus that race did have horses rating. Run up was kind of short at 34 feet vs. 60 feet for the 6 1/2 races that day.

CJ, what was your assessment? Am I crazy to feel this way?

Everyone else, thoughts?

I can give more detail tomorrow, but these are the raw final time figures and variants I used for that day on dirt:

1 113, v = 29
3 101, v = 29
4 120, v = 29
6 95, v = 29
9 106, v = 9
11 105, v = 9

To me it was very clear the track changed between races 6 and 9.

As for R11, Aleander, he got a 96 which the best I can figure these days equates to about a 76 Beyer...until Beyer figs shrink some more next month.

EMD4ME
08-17-2015, 09:09 PM
I can give more detail tomorrow, but these are the raw final time figures and variants I used for that day on dirt:

1 113, v = 29
3 101, v = 29
4 120, v = 29
6 95, v = 29
9 106, v = 9
11 105, v = 9

To me it was very clear the track changed between races 6 and 9.

As for R11, Aleander, he got a 96 which the best I can figure these days equates to about a 76 Beyer...until Beyer figs shrink some more next month.


I appreciate the time CJ. I neglected to mention. Race 11 had a really strong 1st quarter (fastest of the day), which if the track had really slowed, I don't see them running that hot 1st quarter. Race 1 had the second fastest and the leader was running on a negative rail for the second time in a row, I think he's an improving maiden. I can't see the broken down fav in race 11 running a supersonic first quarter (if the track was indeed slower).

The 4th race was a good group of 2YO F in a G2. Granted 2 year olds and fillies run slower but I see the time differential as normal (6/5th's of a second difference). To say the track was slower later, makes Aleander and the improving (but this much?) Indian Stones look like solid older claimers (way above their 16,000 N2L level).

Aleander looked to hang as usual at the 1/16 pole, yet another reason why I question it.

Again, I really think the super slow pace in race 9 caused any thoughts of a variant change.

I could be 100000% wrong. :lol: Not debating you. Just trying to hear other's opinions.

Hence, why I want other's opinions to set me straight if my tired eyes are imagining things.

Thanks in advance in CJ.

cj
08-17-2015, 09:16 PM
I do account for pace when doing the variants (both fast and slow) so that isn't much of an issue for me. It is all right there in black and white. I'm attaching my worksheet for the dirt races. I'll explain more tomorrow.

cj
08-17-2015, 09:17 PM
1C 2C F PF Proj Var (Last 4 Races ) RVar
1 FT- 5.5-D MSW 121 117 112
2 52 7 114 116 112 113 ( - - - -)
6 120 117 110 113 ( - - - -)
3 121 117 109 112 86 27 ( 86 80 83 -)
4 53 84 109 107 ( - - - -) 27

3 FT- 6.0-D MCL 129 105 101
2 44 3 101 99 101 101 ( 66)( 35) ( 66 - - -)
5 114 102 99 100 ( 89)( 12) ( 89 80 78 -)
9 113 97 91 93 51 41 ( 51 - - -)
6 123 102 90 95 74 21 ( 74 - - -) 33

4 FT- 6.5-D STK 130 119 121
2 F 60 7 112 117 121 120 89 32 ( 89 87 - -)
5 130 119 118 119 86 33 ( 86 - - -)
1 105 106 104 104 84 20 ( 84 - - -)
4 100 111 103 104 83 21 ( 83 91 97 -) 28

6 FT- 6.0-D MSW 117 108 95
2 F 44 8 81 92 95 95 ( - - - -)
9 108 105 95 99 ( - - - -)
1 114 108 94 100 ( - - - -)
10 117 108 92 100 62 39 ( 62 - - -) 39

9 FT- 7.0-D ALW 100 97 106
3^F 34 9 84 94 106 103 (100)( 6) ( 100 106 101 109)
10 77 89 103 100 94 9 ( 94 92 86 94)
3 60 77 101 99 97 4 ( 97 80 111 112)
11 94 86 101 95 94 7 ( 94 93 87 86) 7

11 FT- 6.5-D CLM 131 111 105
3^ 60 6 112 107 105 105 92 14 ( 89 92 69 74)
4 125 110 105 107 ( 93)( 14) ( 93 76 76 -)
3 116 106 100 102 96 7 ( 86 96 87 95)
10 22 80 99 97 77 22 ( 77 81 77 82) 13



Raw Figures

SAR 08/15/2015 1 121 117 112 x
SAR 08/15/2015 3 129 105 101 x
SAR 08/15/2015 4 130 119 121 x
SAR 08/15/2015 6 117 108 95 x
SAR 08/15/2015 9 100 97 106 x
SAR 08/15/2015 11 131 111 105 x

cj
08-17-2015, 09:18 PM
That is what the attachment looks like...looking again I may re-visit the last two, I may need to deduct more than 9 from races 9 and 11.

EMD4ME
08-17-2015, 09:19 PM
Thanks CJ. I appreciate it. I'll take a look at it and look forward to your feedback tomorrow.

Like I said, I could be looking at this with tired eyes :blush: . So, before I book the notes, I wanted to see what you thought.

highnote
08-18-2015, 01:33 PM
Here is part of my worksheet for Aug 15:

DIFFERENCES
2f 4f 6f ff class
-0.21 -0.23 -0.08 5.5fmsw 2yo
0.24 -0.18 -0.75 6f mcl65 2yo
-0.09 -0.16 -0.03 6.5f stk f 2yo
-0.19 -0.09 -0.32 6f msw f 2yo SB
-1.11 -1.8 -1.08 7f alw NW1 F 3+
0.24 -0.12 -1 6.5f c16 NW2L 3+


This sheet reflects the differences from par in seconds.

The final time for Race 3 was slow compared to par and it wasn't that much different from Races 9 and 11. So I am not too certain the track changed from Race 6 to 9.

This next worksheet shows the speed figure differences from par:


1 5.5fmsw 2yo
10 6f mcl65 2yo
0 6.5f stk f 2yo
5 6f msw f 2yo SB
13 7f alw NW1 F 3+
13 6.5f c16 NW2L 3+
7 average figure diff


On my scale, Race 3 was 10 points slower than par and Races 9 and 11 were 13 points slower. So, again, I am not certain the track changed after Race 6.

highnote
08-18-2015, 02:26 PM
Also... those fillies ran such a slow pace in the 9th race that they probably did not have enough time to run a fast final time.

Here is part of my worksheet for Aug 15:

DIFFERENCES
2f 4f 6f ff class
-0.21 -0.23 -0.08 5.5fmsw 2yo
0.24 -0.18 -0.75 6f mcl65 2yo
-0.09 -0.16 -0.03 6.5f stk f 2yo
-0.19 -0.09 -0.32 6f msw f 2yo SB
-1.11 -1.8 -1.08 7f alw NW1 F 3+
0.24 -0.12 -1 6.5f c16 NW2L 3+


This sheet reflects the differences from par in seconds.

The final time for Race 3 was slow compared to par and it wasn't that much different from Races 9 and 11. So I am not too certain the track changed from Race 6 to 9.

This next worksheet shows the speed figure differences from par:


1 5.5fmsw 2yo
10 6f mcl65 2yo
0 6.5f stk f 2yo
5 6f msw f 2yo SB
13 7f alw NW1 F 3+
13 6.5f c16 NW2L 3+
7 average figure diff


On my scale, Race 3 was 10 points slower than par and Races 9 and 11 were 13 points slower. So, again, I am not certain the track changed after Race 6.

cj
08-18-2015, 05:42 PM
Using projections (I don't even have pars) that was about as clear a variant change as I see. The one race that would give me pause was the 6th as there was very little to go on and it borders "the split." Given that it was NY Bred 2yo MSW fillies and that I did have one horse that had run in the top 4, I'm pretty confident that using the faster track variant was the way to go for that one.

EMD4ME
08-18-2015, 06:38 PM
Thanks for your time/feedback CJ. I'll PM you if you don't mind, if I have a question or two.

highnote
08-18-2015, 08:11 PM
Using projections (I don't even have pars) that was about as clear a variant change as I see. The one race that would give me pause was the 6th as there was very little to go on and it borders "the split." Given that it was NY Bred 2yo MSW fillies and that I did have one horse that had run in the top 4, I'm pretty confident that using the faster track variant was the way to go for that one.

I'm glad your variants differ from mine. That is where pari-mutuel value comes from. If I'm wrong then you'll get more value and vice-versa if you're wrong.

The 2 furlong and 4 furlong splits were all about +/- a fifth of a second different from par for all the races except the 9th. Even the 11th race, which followed the slow run 9th, reverted back to the +/- fifth of a second range.

So that makes me think that those fillies in the 9th just ran too slow early to make an impact late.

The 11th race final time was 1 second slower than par, but the final time of the 3rd race was also slow -- about 3/4s of a second.

I like your projection method -- it looks similar to the way Nick Mordin did his projections. So there is a good chance you are more accurate. However, in the event that your variants are not as accurate as mine, I plan on using mine at the betting window.

Cratos
08-18-2015, 09:18 PM
Using projections (I don't even have pars) that was about as clear a variant change as I see. The one race that would give me pause was the 6th as there was very little to go on and it borders "the split." Given that it was NY Bred 2yo MSW fillies and that I did have one horse that had run in the top 4, I'm pretty confident that using the faster track variant was the way to go for that one.
Will you explain in general your use of projections?

I am not asking for any confidential info, but I am curious about your use of projections. Also are these multivariate projections?

Again, this is just a general inquiry.

cj
08-18-2015, 09:34 PM
Will you explain in general your use of projections?

I am not asking for any confidential info, but I am curious about your use of projections. Also are these multivariate projections?

Again, this is just a general inquiry.

Based on the top four finishers' last four races, I project a number they might have run today based on those and assign it a confidence rating. The confidence rating looks at surface, distance, recency, finish position and a few other factors. If it is too low, I toss it from consideration. I build in things like projected improvement based on age, distance, and finish position. For example, I wouldn't trust a figure for a horse the finished 8th last time and won today as being representative. I look at all that I think are relevant and base my variant on those.

Cratos
08-18-2015, 10:18 PM
Based on the top four finishers' last four races, I project a number they might have run today based on those and assign it a confidence rating. The confidence rating looks at surface, distance, recency, finish position and a few other factors. If it is too low, I toss it from consideration. I build in things like projected improvement based on age, distance, and finish position. For example, I wouldn't trust a figure for a horse the finished 8th last time and won today as being representative. I look at all that I think are relevant and base my variant on those.
Thanks; from your response I see your projections as probability multivariate projections.

cj
08-18-2015, 11:24 PM
Thanks; from your response I see your projections as probability multivariate projections.

I would agree with that.

highnote
08-23-2015, 03:45 PM
steveb --- if you're still listening ---

Inyour experience is it better to find the average beyer deviation for all the races on a card whether they are sprints or routes, or is it better to mix them all together in order to get a bigger sample size?

For example, there are 3 six furlong races and 3 ten furlong races on the day. would you find the combined average deviation for all of them or would you find the average of the six furlong races and then find the average of the ten furlong races?

steveb
08-24-2015, 04:41 AM
steveb --- if you're still listening ---

Inyour experience is it better to find the average beyer deviation for all the races on a card whether they are sprints or routes, or is it better to mix them all together in order to get a bigger sample size?

For example, there are 3 six furlong races and 3 ten furlong races on the day. would you find the combined average deviation for all of them or would you find the average of the six furlong races and then find the average of the ten furlong races?

i don't do beyers highnote, mine are strictly my own, even if they were the inspiration.

i don't have any personal experience with your racing, so am not qualified to comment, except that i always uses all distances for myself wherever i have done it.

but i would first compensate for the way the race was paced before i would worry about the variant.
(and before dark target says anything, this is AFTER all that DT, when we already know the times and classes and whatnot)

or i might weight the races as to how much i think they contribute to the variant.
a race paced much slower than i would expect would get less weighting.
the faster the pace the higher the weighting.

so i would do those sorts of things before i decided if i needed a split variant or not.

but if you have the right method i can't personally see any reason why you need to split it between sprints and routes.

and of course my way will not be like yours or others ways so it may not be the right answer for you.

highnote
08-24-2015, 12:29 PM
i don't have any personal experience with your racing, so am not qualified to comment, except that i always uses all distances for myself wherever i have done it.

I have seen people normalize all the times and distances to one mile and then make the variants. I'm not sure it makes sense to normalize, though, because it might add more noise?

but i would first compensate for the way the race was paced before i would worry about the variant.

or i might weight the races as to how much i think they contribute to the variant.
a race paced much slower than i would expect would get less weighting.
the faster the pace the higher the weighting.

I have noticed, especially on grass route races, that the pace can be extremely slow in one race on the card and then much faster in the next race. You will see the same thing on dirt, but to a lesser extent. I don't know how you could make an accurate assessment of the speed of the surface under those conditions, unless you project the variant. So one concern I have about projecting the variant is that you move away from time based speed ratings into a method that resembles handicap ratings. However, if the handicap ratings produce more accurate predictions then the concern is unwarranted.

and of course my way will not be like yours or others ways so it may not be the right answer for you.

Maybe not, but I like your idea of analyzing the pace.

James Quinn writes about an interesting pace method in his book "Figure Handicapping" where he uses utilizes a "pace range" to rate the time of the final fraction in grass races. If my memory is correct, he says that if the pace is slow run, you would expect the final fraction to be quick, but the final time will suffer because there is not enough distance remaining in the race after the pace call for the horses to run a fast final time.

On the other hand, if the pace is too fast then the final fraction will tend to be slower which will cause the final time to be slower.

So if the pace time is run in the "sweet spot" range then the final fraction and final time will be fast.

Now, if the pace is slow and the final time is fast or if the pace is fast and the final time is fast then you might have an extraordinary horse.

Quinn creates a set of "pace range" pars.

In Beyer's most recent book, he discusses Quinn's method and points out some of the flaws as well as the good points.

Also, Cary Fotias wrote in his book "Blinkers Off" about a figure handicapping pattern called "Turf Decline Line" (C) where over the past three races the pace figure declines, but the final figure is about the same. This shows that the horse's closing speed is increasing because the horse is getting fitter.

I noticed yesterday at Saratoga the winners of the turf races all had "Turf Decline Line" over their past two or three races even though their final speed figures were not as high as others. And a given horse that had the highest pace and speed figure did not win. Whereas, on dirt, the horses with highest pace and final figures ran well.

Tom
08-24-2015, 12:34 PM
When I see pace variants for, say, sprints, of

F5
F9
F7
S6
F8
F5

I just ignore the S6 completely.
Obviously, the track did not cause the slow pace, so why bother figuring it in?
I make the track F7 for the day, and the S6 race is now S13.

highnote
08-24-2015, 01:20 PM
When I see pace variants for, say, sprints, of

F5
F9
F7
S6
F8
F5

I just ignore the S6 completely.
Obviously, the track did not cause the slow pace, so why bother figuring it in?
I make the track F7 for the day, and the S6 race is now S13.


That makes sense.

highnote
04-06-2018, 07:22 PM
Steveb -- if you're listening... are you familiar with Steven Brecher's book "Beating the Races with a Computer"?

He has a method of making speed ratings that looks interesting.

steveb
04-06-2018, 08:15 PM
Steveb -- if you're listening... are you familiar with Steven Brecher's book "Beating the Races with a Computer"?

He has a method of making speed ratings that looks interesting.

you rang!

never heard of him highnote.
why don't you critique it and i might have a read depending on if you reckon it's worthwhile or not.

i just searched for that book and it's from 1980!!
might be a bit dated?

highnote
04-07-2018, 11:24 PM
you rang!

never heard of him highnote.
why don't you critique it and i might have a read depending on if you reckon it's worthwhile or not.

i just searched for that book and it's from 1980!!
might be a bit dated?

Woods would have been familiar with the book and it's speed ratings as it is referenced in his former partner's paper.

I've read the details he spells out for making figures, but his process has not yet become clear to me.

You're right. The book is from the 1980s. There are probably better methods available that are easier to follow.

I'll try to write up a summary about the process for you.

steveb
04-08-2018, 02:08 AM
Woods would have been familiar with the book and it's speed ratings as it is referenced in his former partner's paper.

I've read the details he spells out for making figures, but his process has not yet become clear to me.

You're right. The book is from the 1980s. There are probably better methods available that are easier to follow.

I'll try to write up a summary about the process for you.


thank you.
so you are saying benter references it?

the only copy i could find was on abe books.
where i guess if i was still caring about racing, i would willingly buy it, if i thought it would help me, but i am too old now.
my time is done.
the jigsaw puzzle that is racing interests me still, but the racing itself bores me.

and, if i was honest, the only books that i thought worthwile, were the early beyer books, which got me on my way a zillion years ago.
have never seen a good one since those days.
not that i have read too many on times.
don scott and rem plante being best of the others, but they reckoned (wrongly) that time was not much good.
but their other stuff was good for somebody learning their way, as i was in the long ago.

a quick google say some nice things about him(brecher), although apparently wong is better according to some(on PA too)
i have the wong book and his speed stuff is colour by numbers.
some of wong's other stuff is a good read, but the times stuff is nonsense.

give me some clues highnote as regards brecher, you obviously have the book and think it ok.
you could always scan the relevant pages and email them to me, and i could then offer you some opinion/advice.

highnote
04-08-2018, 08:14 AM
Brecher's approach to making speed ratings is interesting, but I think there are better ways.

First, he finds the final time of the race.

Second, he uses a multiple regression equation to predict the time of the race.

Third, he finds the difference between the actual time and his predicted time. He calls this difference the track variant. He sums the variant for all the races and then divides by the number of races to find the daily variant. The daily variant will be added to all the races on the day to create a final time that takes into account the speed of the track.

Fourth, he makes a Base Time for the race by setting all the factors in his multiple regression equation to a "detent" level. For example, if purse value is one of the multiple regression factors and it varies from $10,000 to $100,000 depending on the quality of the race, he will use a detent purse level of 10,000.

Fifth, he creates a standardized time by dividing the final time of the race plus the daily variant and dividing it by the base time. Then subtracts 1. This standardized time can be positive or negative where negative times are faster.

It's an interesting and unique approach, but I'm not convinced it is any better than any of the typical approaches to making speed figures.

steveb
04-08-2018, 05:57 PM
Brecher's approach to making speed ratings is interesting, but I think there are better ways.

First, he finds the final time of the race.

Second, he uses a multiple regression equation to predict the time of the race.

Third, he finds the difference between the actual time and his predicted time. He calls this difference the track variant. He sums the variant for all the races and then divides by the number of races to find the daily variant. The daily variant will be added to all the races on the day to create a final time that takes into account the speed of the track.

Fourth, he makes a Base Time for the race by setting all the factors in his multiple regression equation to a "detent" level. For example, if purse value is one of the multiple regression factors and it varies from $10,000 to $100,000 depending on the quality of the race, he will use a detent purse level of 10,000.

Fifth, he creates a standardized time by dividing the final time of the race plus the daily variant and dividing it by the base time. Then subtracts 1. This standardized time can be positive or negative where negative times are faster.

It's an interesting and unique approach, but I'm not convinced it is any better than any of the typical approaches to making speed figures.

bugger you highnote!!!
it is interesting when you say it like that.
i might have to buy a copy.

nothing unique about using regression to figure the time stuff, as i do it all the time.

your third paragraph is the correct way to do it imo, just that it needs more steps, and checks and balances which he probably had anyway.

can you enlarge on the bold bit??

and finally i had to get my dictionery out
'detent' , never heard the word before!

highnote
04-08-2018, 06:56 PM
can you enlarge on the bold bit??

and finally i had to get my dictionery out
'detent' , never heard the word before!

He calls his speed ratings "Standardized Normalized Times" or "SNT".

The bold bit describes in words a simple formula he shows:


(Final Time + Variant)
SNT = ---------------------- - 1
(Base Time)


He wrote:

"The first term in the formula may be thought of as a multiple of the time a mature male Base Purse class of horse would take to run the race. SNTs can be positive or negative. Negative is faster and approximates the logarithm of the ratio of the horse's time to the base time."

I think his approach is novel, but so is James Quinn's approach outlined in his book "Figure Handicapping". Quinn describes the best system I've seen for making speed ratings for grass races where horses are jockeyed for position early and then sprint to the wire -- as opposed to dirt races where the horses run as fast as they can for as far as they can and are decelerating as they approach the wire. As a general rule, on dirt, the horse that expends its energy optimally over the course of the race is the most likely winner.

Brecher's approach could be improved upon. He created his system sometime around 1980. It was designed to be an automated system where the figures were created by an automated computerized process. Nowadays, there are better ways to do this. Jerry Brown of Thorograph, and CJ of TimeformUS use a method where many horses from the same racecard are used to make projections rather than just using the winners' times.

Brecher uses just the winner's times and averages the individual race variants from the par time. His way is too general and misses the subtleties of a race. Plus, he does not take into account the pace of the race, whereas, James Quinn's method on grass racing does account for pace.

Have a look at the terrific race from 1993 called the Grade 1 Turf Classic Invitational Stakes run at Belmont Park on Long Island, not far from New York City at the link below. (scroll ahead to 8:50 for the start of the race) The winning horse was trained by Andre Fabre. Its name was Apple Tree. You might be familiar with the trainer and the horse.

The pace was dawdling for a 12 furlong grass race and the final time was 2:28. (For comparison, Secretariat ran 12 furlongs on the dirt in 2:24. Secretariat also ran 12 furlongs on turf in 2:24 and 4/5.)

With such a slow pace and final time how do you make an accurate speed rating for a race like the Turf Invitational? The horse on the lead in this race should have been able to win after setting such a slow pace. (For example, Buck's Boy set a slow pace when he won at 12 furlongs on grass in the 1998 Breeders Cup Classic.)

https://www.youtube.com/watch?v=pJUs7eDgH-0&t=531s

lansdale
04-08-2018, 07:50 PM
Thought it might be worth re-posting this venerable Benter piece, since Steve B is in this thread, but also to emphasize first, that the models created by himself and Woods were much more complex than this discussion implies, that they never relied on single figures, and that final time was only one among a huge array of handicapping factors.

Although it's clear he took some ideas for factor selection and defnition from Brecher, there's no suggestion here or anywhere in his writings that he relied on him for any overall analysis, as he did Bolton and Chapman. In any comparison between the ideas/techniques of Benter/Woods and Brecher, whose work was entirely theoretical, I would tend to rely on the former pair.

Cheers,

lansdale



https://www.scribd.com/doc/166556276/Benter

highnote
04-08-2018, 09:26 PM
Thought it might be worth re-posting this venerable Benter piece, since Steve B is in this thread, but also to emphasize first, that the models created by himself and Woods were much more complex than this discussion implies, that they never relied on single figures, and that final time was only one among a huge array of handicapping factors.

Although it's clear he took some ideas for factor selection and defnition from Brecher, there's no suggestion here or anywhere in his writings that he relied on him for any overall analysis, as he did Bolton and Chapman. In any comparison between the ideas/techniques of Benter/Woods and Brecher, whose work was entirely theoretical, I would tend to rely on the former pair.

Cheers,

lansdale



https://www.scribd.com/doc/166556276/Benter

Agreed.

Speed figures help to predict race outcomes, but there are so many inaccuracies about them that you still need a comprehensive approach to handicapping in order to win.

However, there may be some people who can win using speed figures only. My guess is that in order to win using speed figures only they would have to be very selective of the types of races they bet. And even then they would probably have to handicap a little. I doubt that there are any speed figures that could be used blindly that could win consistently and produce a profit.

steveb
04-09-2018, 05:34 AM
thank you highnote

just come home so i will have study of you post and get back to you tomorrow maybe.

steveb
04-09-2018, 10:59 PM
Have a look at the terrific race from 1993 called the Grade 1 Turf Classic Invitational Stakes run at Belmont Park on Long Island, not far from New York City at the link below. (scroll ahead to 8:50 for the start of the race) The winning horse was trained by Andre Fabre. Its name was Apple Tree. You might be familiar with the trainer and the horse.

The pace was dawdling for a 12 furlong grass race and the final time was 2:28. (For comparison, Secretariat ran 12 furlongs on the dirt in 2:24. Secretariat also ran 12 furlongs on turf in 2:24 and 4/5.)

With such a slow pace and final time how do you make an accurate speed rating for a race like the Turf Invitational? The horse on the lead in this race should have been able to win after setting such a slow pace. (For example, Buck's Boy set a slow pace when he won at 12 furlongs on grass in the 1998 Breeders Cup Classic.)

https://www.youtube.com/watch?v=pJUs7eDgH-0&t=531s

let me work backwards!

my opinion is that you can't, so why bother?

you can though hypothesize on how fast they COULD have run the race.
what about if you run a regression of all the expecteds, less the actuals as one variable, and the other the pace percent of the races, where 100% is best way to pace?
either side of that is either too fast or too slow depending on the %.

i already know that the correlation will be very very high.
it still does not translate to nags that finished close in the slow pace being able to do the same in a fast or evenly paced race.
but it could be one factor in a model.

but i would never dream of projecting myself, so the speed would stay what it was, regardless of how fast or slow it was paced.
i would just be aware of it.
if my pace was maybe .97 for a particular point of the race, and the speeds were slow, then i know why, and that's enough for me.
i don't believe in changing stuff to suit expectation.

that some do, is neither here nor there, but the fact is they are no longer speeds, they are something else.
they may work fine for those that do, but they are still not speed figures as such.

one number to sum up a race may well work for some, but not for me.
times are just one part of the jigsaw puzzle that is racing. so that there are many other things that would need to be accounted for.

highnote
04-10-2018, 12:20 AM
let me work backwards!

my opinion is that you can't, so why bother?

you can though hypothesize on how fast they COULD have run the race.
what about if you run a regression of all the expecteds, less the actuals as one variable, and the other the pace percent of the races, where 100% is best way to pace?
either side of that is either too fast or too slow depending on the %.

i already know that the correlation will be very very high.
it still does not translate to nags that finished close in the slow pace being able to do the same in a fast or evenly paced race.
but it could be one factor in a model.

but i would never dream of projecting myself, so the speed would stay what it was, regardless of how fast or slow it was paced.
i would just be aware of it.
if my pace was maybe .97 for a particular point of the race, and the speeds were slow, then i know why, and that's enough for me.
i don't believe in changing stuff to suit expectation.

that some do, is neither here nor there, but the fact is they are no longer speeds, they are something else.
they may work fine for those that do, but they are still not speed figures as such.

one number to sum up a race may well work for some, but not for me.
times are just one part of the jigsaw puzzle that is racing. so that there are many other things that would need to be accounted for.

Agreed

steveb
04-10-2018, 01:12 AM
Agreed

i have more to come highnote, but have other stuff to do thate takes precedence.
that formula of brecher's you posted has me scratching my chin!
i need to figure(for my own satisfaction) what it is actually doing before i can comment or offer opinion.

highnote
04-10-2018, 01:29 AM
i have more to come highnote, but have other stuff to do thate takes precedence.
that formula of brecher's you posted has me scratching my chin!
i need to figure(for my own satisfaction) what it is actually doing before i can comment or offer opinion.

I can't figure out what it does. I may try using his method just so that I can understand it better. It's hard for me to think it through. I need to actually see the data and then see the results.

highnote
02-28-2019, 06:09 AM
i have more to come highnote, but have other stuff to do thate takes precedence.
that formula of brecher's you posted has me scratching my chin!
i need to figure(for my own satisfaction) what it is actually doing before i can comment or offer opinion.

I'm still waiting... ;)