Speed Figures Revisited

[classhandicapper]

My typical speech for 1000th time.

The numerous complexities and subjective element of making figures impact their accuracy. On top of that, pace, race development, and rider aggressiveness depending on how the track is playing all impact time.

In the end pace and speed figures as ballpark guesses. That's why if you line up 3-4 sets of figures you will constantly see very significant differences from race to race and horse to horse.

All that said, it's often hard to tell the difference between the quality of one field and another or to understand a result without knowing how fast the horses ran, especially with very lightly raced horses in maiden/limited winner allowance races and sometimes shippers.

After 45 years of struggling with this (an ongoing process lol); I think the best approach is to understand the class pecking order at your track and just rate races as very fast, fast, average, slow, and very slow for that class without getting too exact. After awhile you'll get a feel for when a field is better or worse than the "number" based on the quality of the horses actually in the race and you'll be able to tell by how fast the race was whether it was a strong or weak race for the class when it would have been hard to tell otherwise.

I think you just want to avoid starting with the exact number and assuming it's accurate when it's close between horses. That will tend to drag you towards favorites or false overlays that may not even be faster or better.

[cj]

Originally Posted by classhandicapper

My typical speech for 1000th time.

If you know it, why do it?

[classhandicapper]

Originally Posted by cj

If you know it, why do it?

In case someone missed it the first 999 times.

[MJC922]

Originally Posted by classhandicapper

My typical speech for 1000th time.

The numerous complexities and subjective element of making figures impact their accuracy. On top of that, pace, race development, and rider aggressiveness depending on how the track is playing all impact time.

In the end pace and speed figures as ballpark guesses. That's why if you line up 3-4 sets of figures you will constantly see very significant differences from race to race and horse to horse.

All that said, it's often hard to tell the difference between the quality of one field and another or to understand a result without knowing how fast the horses ran, especially with very lightly raced horses in maiden/limited winner allowance races and sometimes shippers.

After 45 years of struggling with this (an ongoing process lol); I think the best approach is to understand the class pecking order at your track and just rate races as very fast, fast, average, slow, and very slow for that class without getting too exact. After awhile you'll get a feel for when a field is better or worse than the "number" based on the quality of the horses actually in the race and you'll be able to tell by how fast the race was whether it was a strong or weak race for the class when it would have been hard to tell otherwise.

I think you just want to avoid starting with the exact number and assuming it's accurate when it's close between horses. That will tend to drag you towards favorites or false overlays that may not even be faster or better.

It was a conclusion I reached long ago, wind and all kinds of environmental variables regularly have a significant impact on time to the point where I'd split variants 40% of all racing days and then to top that off if you're using splits you have massive numbers of inaccurate times to deal with annually, those will be timer malfunctions, typos, run up variability, you name it. Anyone can feel free to use any speed figures with a traditional variant, take a 1500 race sample and compare the results to my numbers which no longer use variants and they will see there's nothing to be had for increased accuracy (win rate) with a traditional speed figure. The issue you mentioned with the lightly raced and what amounts to regional pockets of inaccurate numbers, again largely specific to very lightly raced isolated populations is 100% legitimate and I could fix that issue with some code if I ever had the time to write another line of code someday or remember how. In short it's as simple as starting firsters off with a REAL par instead of no par, and if you have accurate pars (which I'm confident I do) then this is very likely a fixable issue. Being a perfectionist with the numbers I guess it's a bucket list item for me at this point, like maybe before I kick the bucket.

[classhandicapper]

Originally Posted by MJC922

It was a conclusion I reached long ago, wind and all kinds of environmental variables regularly have a significant impact on time to the point where I'd split variants 40% of all racing days and then to top that off if you're using splits you have massive numbers of inaccurate times to deal with annually, those will be timer malfunctions, typos, run up variability, you name it. Anyone can feel free to use any speed figures with a traditional variant, take a 1500 race sample and compare the results to my numbers which no longer use variants and they will see there's nothing to be had for increased accuracy (win rate) with a traditional speed figure. The issue you mentioned with the lightly raced and what amounts to regional pockets of inaccurate numbers, again largely specific to very lightly raced isolated populations is 100% legitimate and I could fix that issue with some code if I ever had the time to write another line of code someday or remember how. In short it's as simple as starting firsters off with a REAL par instead of no par, and if you have accurate pars (which I'm confident I do) then this is very likely a fixable issue. Being a perfectionist with the numbers I guess it's a bucket list item for me at this point, like maybe before I kick the bucket.

I'm glad someone whose opinion I've long respected agrees with me on this. I'm kind of in the minority on this issue at this point. Most people just trust the figures they like best even if inside they know there are occasional issues.

I played a couple of races this week where the figure I had suggested a horse fit perfectly or should be favored, but I was familiar with the horses in the race and was convinced the horse did not stack up. They all finished up the track.

One the flip side, there was another race where a relatively lightly raced improving ALW winner that was fast but didn't "class up" beat me and ran a big new top.

It's still hard for me to separate the wheat from the chaff because so many things impact the times (and ultimately the figures) it's hard to know what you are dealing with unless you have other clues.

[GMB@BP]

I use to rely on speed figures much more for my selections but over the years I have found that there is less and less value with them.

I trust the Timeform figures but for the most part I use them to figure out who are the contenders that can win the race, even then I have to put other figures into context when its lightly raced horses or new few races on the surface.

Most the time my biggest context is race shape and other factors that pull people away from betting a horse who is a strong figure contender.

There is also little value with the superior figure horses, it can happen, but its rare to pick a card and find a race where the figure horse is value. I suppose if you were patient and were able to look at all races across all cards you would find some but thats not my cup of tea.

[46zilzal]

Originally Posted by Jeff P

HDW

the same

[classhandicapper]

I'm surprised we got so many "homemade" responses. I didn't think that many people had the time and energy to make their own figures.

[mikesal57]

Originally Posted by classhandicapper

I'm surprised we got so many "homemade" responses. I didn't think that many people had the time and energy to make their own figures.

Can be done if your concentrating on a track or two....

but what to do with shippers???

[Tom]

Originally Posted by mikesal57

Can be done if your concentrating on a track or two....

but what to do with shippers???

Combine Winners Books with Dave Schwartz par times and you can
make a pace figure for tracks you do not do full figures for in a couple of minutes. Beyer describes a method in his last book, but it neglects the pace variant, so it is not that accurate. You need to factor in the pace variant or you might as well forget the whole idea. But you can use the Beyer variant for final times.

[mikesal57]

Originally Posted by Tom

Combine Winners Books with Dave Schwartz par times and you can
make a pace figure for tracks you do not do full figures for in a couple of minutes. Beyer describes a method in his last book, but it neglects the pace variant, so it is not that accurate. You need to factor in the pace variant or you might as well forget the whole idea. But you can use the Beyer variant for final times.

Who is Dave Schwartz?

[Tom]

Recently,.. we had a discussion about Beyers for a horse beaten by a good margin. Unfortunately, I can't find it now, but did send the question we had to Andy Beyer and his reply is in the DRF this week:


https://www.drf.com/news/ask-beyer-a...l-handicapping

[ReplayRandall]

Originally Posted by Tom

Recently,.. we had a discussion about Beyers for a horse beaten by a good margin. Unfortunately, I can't find it now, but did send the question we had to Andy Beyer and his reply is in the DRF this week:


https://www.drf.com/news/ask-beyer-a...l-handicapping

Sign-in is required, Tom....For those who don't have an account.

[classhandicapper]

Originally Posted by Tom

Recently,.. we had a discussion about Beyers for a horse beaten by a good margin. Unfortunately, I can't find it now, but did send the question we had to Andy Beyer and his reply is in the DRF this week:


https://www.drf.com/news/ask-beyer-a...l-handicapping

I think he gave too generic an answer, most likely because he didn't want to write another book. I assume he assumed all else being equal.

IMO, if the well beaten horse actively engaged in the flow of the race before dropping back he will typically be better because he is clearly coming out of the much faster race against better horses. He should run a little faster in a softer spot, not only because he may not have been all out, but because the weaker horses will set a softer pace and the overall trip will tend to be easier.

IMO, if he just sat in the back and did no real running on his way to that 85, he may actually face a tougher trip against weaker horses when asked to compete up front and actually engage with similar horses.

[highnote]

Originally Posted by classhandicapper

I'm surprised we got so many "homemade" responses. I didn't think that many people had the time and energy to make their own figures.

I wrote computer software to make automate the process.