DRF Speed Rating [Archive] - Horse Racing Forum - PaceAdvantage.Com

KyRacer

12-16-2002, 03:28 PM

The stats Barry gave us was for the DRF speed rating without variant. The page before gives the DRF Speed Rating with Track Variant and shows a profit at 4-1 and up. Sample sizes were around 800 for last race and best of last three and just under 400 for 4-1 and up. The Bris rating were similar to the Beyer's with a slight edge to the Beyer's and the best of last three almost identical.

Daily Racintg Form Speed Ratings plus Track Variant

Best Last Race Speed Rating:
DRF 21.97% for -0.17, Beyers 25.2% for -0.17

Best Speed Rating Last Three:
DRF 21.31% for -0.04, Beyers 24.4% for -0.17

Best Speed Rating Last Three 4-1 & Up only:
DRF 13.72% for 0.13, Beyers 11.4% for -0.10

A larger sample size might bring these closer together. It would be interesting to know how much an accurate variant added to the DRF speed rating might improve them.

Looking at all the different rating as a whole, Best of Last Three gets a slight edge over Best Last Race figure in win % and a few more points in roi. Taking the Best of Last Three at 4-1, cuts the win % in half but improves roi quite a bit and showed a profit for several or the ratings, and appears the way to go, again with a small sample. The Speed Ratings study appears on EBay from time to time.

KyRacer

andicap

12-16-2002, 04:08 PM

Here's why the DRF variant is so poor. I looked at three days at Aqueduct on Oct 27, 30 and 31st.
The DRF variant for routes

Oct. 27 12
Oct. 30 28
Oct. 31 17

Now check out the average winning Beyer figure for routes

Oct. 27 84
Oct. 30 69
Oct. 31 85

Using the Beyer/route adjust of about 1.8-1.9 per length, the races on Oct. 31 were a good 9 lengths better than on the 30th.
(or something like that). That's why the DRF variant on the 31st was so much lower than on the 30th. The competition was much better!!

Oct. 30th was obviously slower than Oct. 27 -- you can tell by eyeballing the raw times. But was it 11 pts slower than Oct. 31? No freakin way.

Oct. 30 1 mile NW1 F 136.97
Oct. 31 1 mile NW2 F 136.69

Both were for 3 yrs old and up. The classier race ran a tick or so faster. Like it should be. That doesn't mean the track was exactly the same as the day before, but it sure as hell indicates it was pretty close, certainly there wasn't a difference of 11 points!!

My suggestion: if you must use the DRF variants. Average the winning Beyers from that day and adjust accordingly.
You won't be as accurate as if you compute your own, but you'll be a lot closer with a lot less work.

So if you adjust the Oct 30 variant by 9 lengths DRF style (1 pt = 1 length), you'll get a 26, close to a 28.

If you subscribe to Tom Brohamer's theory of using 2 DRF variant points to adjust 1 length, you'll have a difference of only 1 length.

Simulcast Weekly prints the winning Beyers for almost every track each week. You can average the Beyers for sprints and routes in about 5 minutes and adjust the printed version accordingly.

Suff

12-16-2002, 07:06 PM

nice thread... good read.

Dick Schmidt

12-16-2002, 07:42 PM

Andy et al,

Though I no longer use any part of the Racing Form (haven't bought one for years), this discussion does take me back. When Tom Hambelton and I brought out Total Pace Ratings, we were roundly criticized for using the daily variant from the Form as the basis of our adjustments. I had people call to tell me how much better our numbers would be if only we used better adjustments before they finished reading the book! Jim Quinn told me that while he was flipping through his pre-publication review copy.

Well, neither Tom or I are dummies, and Tom at least isn't afraid of hard work, so before we released the book, we tried "good" speed ratings, Beyer numbers, and, in Southern California, the best speed ratings available anywhere; those hand crafted by Tom Brohamer. The results? Well, the critics were right, we won more races using "good" numbers, but in every case we won more money using the much maligned Racing Form variants.

Mostly I suspect because they are so ignored by "serious" handicappers. Once a number gets popular, it stops working (remember, Andy once made a living with Beyer numbers). The Form variants are over the "good enough" line when you value them as we did and apportion them correctly to the race segments. Handicappers routinely make two mistakes about numbers: simple can't be good, and the more exact a number appears to be, the better it will work. Yet above we see a 13% profit laying on the table while the best the vaunted Beyer numbers can do is lose 10%.

Time and time again, handicappers get bogged down in the process and ignore the outcome. If carefully hand crafted speed numbers are more complex, more accurate looking, are computed to more decimal places and are harder to do, they must be better. Ask yourself this, after spending hours "making numbers," are your figs showing a flat bet 13% profit with no other handicapping? Are you finding the winner in your top 5 at least 85% of the time? Maybe simple can be best.

Dick

Fastracehorse

12-16-2002, 08:28 PM

I believe speed figures are a superior tool to class handicapping.

Fascinatingly, the Beyer has a pace fig already incorporated in itself.. Why?? Well look at sprints. The quarters get progressively slower. What does this mean?? It means that the better horses quit slower, hence, better horses attain a higher fig.. And that is what a Beyer measures: A pace/time/variant factor.

Further, if a horse comes out of a better group, and races
against inferior stock, he will probably beat them. But with class handicapping U often only know there is a drop. U don't know the horse's true ability [ true Beyer ] because horses rarely run a good fig if they are compromised by factors that lead to the drop [ poor trip, off track, poor surface, inappropriate distance, bad break, etc, etc ].

To summate,

Class is kind of a very generalized handicapping tool that isn't very precise at determining a horse's true ability. Beyers that are adjusted, are very accurate. So with a dropper, U would know how the horse fit by adjusting it's previous speed fig.. And pace figs, aren't t really that important, because they are already innate to the Beyer.

Now, I want to repeat this: Adjusted Beyers lead to many overlays. Fixed Beyers in DRF do not.

fffastt

Doug

12-16-2002, 08:41 PM

Fastracehorse,

When you talk about adjusting the Beyer for class are you just adjusting the Beyer by the amount of class drop or are you say taking the lengths behind at a different call (assuming a better class race) and adjusting the Beyer that way?

Doug

Fastracehorse

12-16-2002, 08:59 PM

Doug,

I'm sure U would agree with me when I say: A 98 Beyer run in a higher class will more often be better than the 98 Beyer run in a lower class. We assume the pace is tougher in the higher class.

To answer U're question, my derived figures are not specifically based on the class drop or rise. However, I did mention that indirectly pace is part of the Beyer. So, I can assume that class has some affect on the figure.

That is why reference races are so popular [ key races ]. The theory behind these is that there are superior races to others. And if the 1st, 2nd, and, 3rd placed horses all come back to win, maybe the 4th place horse has a shot just based on the key race merit. That is, maybe the key race has a time/pace element [ Beyer ] that is admirable to handicppers.

My figure is based on how a horse fairs in a key race-type event. I am so thankful that I don't have to do chart work any longer [ extrememly tedious ].

I became addcited to key races when I saw a horse named Prosperous Bid win a key race. 4 horses came out of that race to win at odds of 20-1 or more. I have never seen a key race like that again but U get my point.

My point is, there is something inherent in the pace/time element of the race that makes horse's performances look much better than they appear by un-adjusted speed figures. At least their performances will imporve dramatically if they don't meet up with Tigers like Prosperous Bid.

fffastt

BillW

12-16-2002, 09:11 PM

Here is some data from my records. This is for all classes and types. Only restriction is all data is for dry dirt.

TSN Speed last race: (9789 races total)

25.03% win 2.11 IV 0.85 ROI 1.04 A/E

TSN Ave. last 45 days

23.69% win 2.00 IV 0.83 ROI 1.01 A/E

TSN Ave. last 90 days

24.21% win 2.04 IV 0.86 ROI 1.05 A/E

BRIS Speed last race: (14436 races total)

26.72% win 2.26 IV 0.875 ROI 1.07 A/E

BRIS Ave. last 45 days

25.61% win 2.17 IV 0.86 ROI 1.05 A/E

BRIS Ave. last 90 days

25.62% win 2.17 IV 0.865 ROI 1.05 A/E

The above datasets do not contain the same mix of tracks (TSN probably averages higher quality)
Approximately 300 races are common to both datasets, otherwise all races are different..

A/E is the Actual/Expected mutuel ratio (Greater than one indicates an overlay on average)

ROI: $1.00 breakeven

The 45 and 90 days averages are the closest thing I have to "last 3" without some work.

As usual, remember CAVEAT EMPTOR!. These are averages and have little significance while standing alone (relative to the quality of one fig vs. another).

Bill

GR1@HTR

12-16-2002, 10:34 PM

Ranchwest (or any other Trackmaster subscriber), do you have stats on Trackmaster last fig? I am curous how those are.

MV McKee

12-16-2002, 10:37 PM

I really don't use the Beyer speed figures to handicap the horses thenmselves, but I do think they are very usefull for handicapping the way the "crowd" is likely to bet. So, in that sense, I do take them into account. A 20-1 ML horse with a high last Beyer(s) is obviously going to be a much lower price, and of course, the inverse is true.

I think Beyer figures are relatively reliable in comparing 2 horses at the same track at the same distance. But, from the research I have done, they are not very reliable in comparing figures across distances, and in some cases, from track to track. The track to track issue is a tough nut to crack, and has nothing to do with why I am posting this. The distance issue is another question.

About an eon ago I read a Beyer book, I think it was called "Picking Winners". The only reason I think it was an eon ago is because I was playing Yakima Meadows at the time, and it hasn't been open for @ an eon. Anyway, I remember reading about how he and a partner (I think his name was Sheldon) arrived developed something akin to a parallel time chart which made the figures from various distances (somewhat) interchangable. The chart obviously makes it necessary that the value of 1/5th of a second or a beaten length differ, depending on the distance run. The chart also appeared to be mathmatically sound, in that the value of a length (or 1/5th second) was a linear function. For example, if Horse A ran 3/5ths faster than Horse B at 6 furlongs (3/4 mile) then he would be expected to run 4/5ths faster at One Mile (one third further). Of course, everyone knows that some horses don't "stretch out" as well as others, and in essence, each individual horse has what could be considered it's own "parallel time chart" as it relates to performance figures at various distances. But, as a whole, this linear relationship seemed sound enough to provide an accurate enough figure for comparison of the average across distances. I used this assumption in the making of my figures for 15 years, although my figures were actually what are reffered to as Quirin style, a different numeric scale, but still makes the relationship between distances linear.
Long story short, over time I saw a lot of things that made me question this assumption (that the distance relationships were linear). I did some initial research using 3 years of data from Hollywood Park and came to one important conclusion, that I will attempt to explain, though I likely will not do it very well.

The basic premise is that the relationship(s) between time, distance, (class) and figures is not at all linear. If the relationship were linear (as it would be with cars for instance) one would expect that in 6 furlong races, on the average, 1/3rd of the time difference between two races would be made up in the 1st quarter, 1/3rd in the 2nd quarter, and 1/3rd in the final 1/4.
As I am certain most of you know, this is not the case. In looking at 3 years of data each, from 16 different tracks, I have found the following to be true in every case:
On the average, in 2 6 furlong races whose final times are .20 seconds apart, the faster of the 2 races will be:
.03 - .04 faster for the first 1/4
.05 - .06 faster for the second 1/4
.10 - .12 faster for the final 1/4
Now obviously, a horse's final time in a race is a function of both his speed and his stamina/ability to ration his speed. The Beyer's (through the speed rating/parallel time charts) and Quirin figures are an expression of the speed element solely. This is why you see insanely high figures in routes at high levels (Swept Overboard), and ridiculously low figs in some of the cheaper routes. The distance translation is accurate within only a very narrow scope of ratings/classes (10k Claimers for example).

As a handicapping tool, my figures have become much more usefull to me since I have incorporated this logic into them. But lest anyone think that I am saying they are more usefull to me because they are more accurate than the Beyer's or Quirin style, I am not. They are usefull because they are different from them, not necessarily better.

Lindsay

12-16-2002, 11:45 PM

MV McKee.

Thank you. This is one of the most interesting posts I've seen on this board. A couple of things:

"Anyway, I remember reading about how he and a partner (I think his name was Sheldon)"

Yes. Sheldon Kovitz. There's a Sheldon Kovitz who teaches math at UMass, but I don't know if it's the same man.

"If the relationship were linear (as it would be with cars for instance) one would expect that in 6 furlong races, on the average, 1/3rd of the time difference between two races would be made up in the 1st quarter, 1/3rd in the 2nd quarter, and 1/3rd in the final 1/4."

The disparity is not so severe if you look at it in terms of percentage of time rather than distance. In other words, since horses are usually running much slower in the third quarter than in the first, it would make sense for variances to be greater in the third quarter, and indeed they are. There is more room for variance in a slow quarter than in a fast quarter.

".03 - .04 faster for the first 1/4
.05 - .06 faster for the second 1/4
.10 - .12 faster for the final 1/4"

My results were almost identical to these. I had the second quarter .01 faster than you did.

"The Beyer's (through the speed rating/parallel time charts) and Quirin figures are an expression of the speed element solely. This is why you see insanely high figures in routes at high levels (Swept Overboard), and ridiculously low figs in some of the cheaper routes. The distance translation is accurate within only a very narrow scope of ratings/classes (10k Claimers for example)."

Now we get to the heart of the matter. This problem, if it exists, is a sitting duck to recognize and correct. It is not necessary to make such charts using solely cheap horses. You can go all the way up to Grade One stakes races. It is hard to believe Beyer and Quirin missed this all these years.

Fastracehorse

12-17-2002, 01:31 AM

I'll start with MV,

While I'm not arguing about the non-linearity of parallel time charts I will suggest there is alot of universiality in the game.

By universiality, I mean that thorughbreds react to similar situations in surprisingly universal manners.

This universiality has helped me adjust numerically, and with uncanny accuracy, the projected Beyer a horse will run when stretching out.

Being that it is true that not every thoroughbred will route, or take to the lawn, the universiality amongst trainers, will see to it that most that try these distances and surface switches will be on the whole, successful.

So yes, I have an accurate projection of a horses's ability when going from a dirt sprint to a route lawn.

Now MV, I will also suggest that the Beyer is not just a pure speed figure and will in fact, incorporate elements of a horse's ability to stay.

Lindsay, I have not noticed the ridiculous separation in Beyer figures like U two have. I have seen truths of the game in adjusted Beyers.

Remember, my derivation is based on sound logic using the most popular tool in thoroughbred horse racing. I rarely use the final fig in the form which lends me to capitolize on many overlays.

I innovated the adjusted Beyer- I stole the bullwork from Andy.

fffastt

Lindsay

12-17-2002, 03:09 AM

Fastracehorse wrote:

"Lindsay, I have not noticed the ridiculous separation in Beyer figures like U two have. I have seen truths of the game in adjusted Beyers."

Fastracehorse,

I'm sorry I wasn't clear. I disagree with MV McKee on this, but I do admire his independent streak. Say what you want about him, and there is much to say, Len Ragozin has been studying this type of question for 40 years, and his fanaticism is legendary. Here's a quote from his book:

"But our study showed a linear fall-off, with this fatigue factor very minor until the races get longer than 1 1/2 miles."

Fastracehorse wrote: "Remember, my derivation is based on sound logic"

On a different string, your comments about measuring the effects of drugs strike me as being fanciful. What you're doing with your Beyer adjustments might be based on sound logic, but I'm going to withhold judgment until you tell me exactly what the hell you're doing. An email would be fine.

Tee

12-17-2002, 05:38 AM

Originally posted by Lindsay

What you're doing with your Beyer adjustments might be based on sound logic, but I'm going to withhold judgment until you tell me exactly what the hell you're doing. An email would be fine. [/B]

Good luck getting exactly(key word being exactly) what the hell he's doing. Fast is very secretive with his adjusted fig lol.

andicap

12-17-2002, 11:08 AM

Fast,
I certainly don't mind you not giving away the store on your adjusted Beyer, but why mention you have had success with an adjusted Beyer in THREE messages without giving anyone a clue as to how you do this. What purpose does it serve?

Fastracehorse

12-17-2002, 02:31 PM

Do U think I am being unreasonable with this claim on drugs??

Other topic, I am not argiung the non-linearity of parallel time charts.

Excellent post by U and MV. It leads to contention fo an intellectual and healthy kind.

fffastt

Fastracehorse

12-17-2002, 02:32 PM

Just curious,

fffastt

Lindsay

12-17-2002, 02:53 PM

Fastracehorse wrote:

"Other topic, I am not argiung the non-linearity of parallel time charts."

Right. You and I have no disagreement here.

"Excellent post by U and MV. It leads to contention fo an intellectual and healthy kind."

Thank you.

"Do U think I am being unreasonable with this claim on drugs??"

I'll stick with fanciful. Please tell me this: Forget the average and the range. What is the median improvement? How do you isolate drugs from the other factors that affect horses? How do you know who is getting what drug--Lasix and Bute aside?

andicap

12-17-2002, 03:04 PM

Originally posted by Fastracehorse
Just curious,

fffastt

I'm just wondering if you could give us a bit more of a clue as to how you adjust the Beyer figs in the DRF.

TomC of SnapCapper sells something called "True Tenders" which adjusts the Beyers based on stretch gain and early speed (using running lines). I tested it (someone sent it to me to test) and found it miserable at picking contenders.

I don't adjust the Beyers but use them in a different way that helps me identify potential longshots as contenders. Works much better than TrueTenders. My final selection is never based on the Beyers but used correctly they can point to long odds contenders.

Big Bill

12-17-2002, 03:42 PM

Dick Schmidt,

The following was copied from your post on this thread:

"The Form variants are over the "good enough" line when you value them as we did and apportion them correctly to the race segments."

I gave my copy of Pace Makes the Race to a friend several years ago and he has lost it. I can't remember whether you and Tom included how you valued and apportioned the DRF variant to the race segments in the book or not. If you didn't, is the method of apportionment considered proprietary?

Big Bill (bwb900@cox.net)

Doug

12-17-2002, 05:12 PM

Reading the above comment about True Tenders adjustment to the Beyer brings up a thought that maybe I could get some input on.

Using Beyer, DRF speed rating, or basically any other final time fig, what if you used a different call for a horse that maybe had trouble, drops in class, etc and assigned a rating to that call? For example:

Horse A has a runing line that looks like this

2nd 2 2nd 2 6th 6 8th 9m final rating 80

Now in the trip notes it says pressed pace early, steadied hard 1/4.

So if you took the lengths behind at the 2nd call and subtracted from 89 (80 + 9 lengths behind at finish_ the horse would have an 87 instead of an 80.

Horse B has the same running line has no trouble, but is dropping from a Grade 2 race to Alw nw2.

Don't use anything like this, never have. Just curious if anyone here had any thougts one this type of time adjustment?

Thanks

Doug

superfecta

12-17-2002, 10:26 PM

As stated before,I have a problem with the way Beyers are figured,not with the math theory behind it.When I look at a Beyer fig,i do not know how the fig has or has not been made(with projected times or literal times).I need a fig that has a standard method of computation,not one that may be trimmed or doctored because the race was too fast or the race too slow.Quick example :
One mile

1st race time 137.76 beyer=69

2nd race time 137.65 beyer= 80

Races were one day apart,same track,same track condition.How do you compare these horses if they ran against each other?The DRF track varient was the same,but we do not know how or how much of the varient was assigned to each horse.Too many questions and leaps of faith to get me excited about the "Figs".

Fastracehorse

12-17-2002, 10:57 PM

Lindsay,

To enjoy or not to enjoy.

http://forums.prospero.com/dr-general/messages?msg=6918.16

fffastt

Fastracehorse

12-17-2002, 11:06 PM

Interesting that U use Beyers in U're way.

Maybe our insights share a common ground-Who knows??

I don't use stretch gain but like many handicappers I love horses with early speed.

I essentially adjust for a host of factors including: Offtrack, surface switches, distance switches, 2nd-time out, out off of first lay-off, trouble lines, 1st-time legal drugs, 1st-time illegal drugs, and indirectly, pace pressure.

fffastt

Dick Schmidt

12-17-2002, 11:29 PM

Big Bill,

"I gave my copy of Pace Makes the Race to a friend several years ago and he has lost it. I can't remember whether you and Tom included how you valued and apportioned the DRF variant to the race segments in the book or not. If you didn't, is the method of apportionment considered proprietary?"

You GAVE AWAY the best racing book ever written?!?!?!?!?! Or at least the best racing book I ever wrote. Lucky for you it is still in print and available to PA posters for half price. Of course we spent a lot of space on how to use the Racing Form variant. It is central to the method. On top of that, we spent a couple of years perfecting how to use and distribute it. We finally got a real rocket scientist (from JPL) to help us with the work and it's held up ever since.

Good read, that book.

Dick

Rick

12-18-2002, 12:28 AM

The treatment of the variant in "Pace Makes the Race" is a good example of how such things should be applied. That is, use whatever works best rather than how you think it should be in theory. If the same thing was done with beaten lengths and distance and surface comparisons in speed ratings, they would work much better.

Lindsay

12-18-2002, 02:06 AM

Fastracehorse wrote: "Lindsay,

To enjoy or not to enjoy.

http://forums.prospero.com/dr-gener...ges?msg=6918.16"

Fastracehorse,

You seem like a decent guy. Please don't take this the wrong way. What you wrote is mindless, libelous crap. What you need to do is lose the paranoia and start fresh.

andicap

12-18-2002, 05:41 AM

If you can be specific about the dates -- maybe something recently, I can reply after checking it out. Give me dates, man, give me dates and tracks.

(BTW, I'm not saying the Beyer figs are the be-all. Sometimes the BRIS figs are better. And I'm sure Colt's Neck and some Sheets are better. I'm just saying the Beyers are good enough for me to pick contenders.)

Originally posted by superfecta
As stated before,I have a problem with the way Beyers are figured,not with the math theory behind it.When I look at a Beyer fig,i do not know how the fig has or has not been made(with projected times or literal times).I need a fig that has a standard method of computation,not one that may be trimmed or doctored because the race was too fast or the race too slow.Quick example :
One mile

1st race time 137.76 beyer=69

2nd race time 137.65 beyer= 80

Races were one day apart,same track,same track condition.How do you compare these horses if they ran against each other?The DRF track varient was the same,but we do not know how or how much of the varient was assigned to each horse.Too many questions and leaps of faith to get me excited about the "Figs".

delayjf

12-18-2002, 11:37 AM

MV MCKEE
Given your post on the incremental differences between the 6 furlong pace segments. Can I assume you apply your varients to the pace segments in the same percentages as follows:

17.5% to the first 1/4

27.5% to the second 1/4

55% to the final 1/4

I''ve heard other speed / pace handicappers on this board state that most of the varient should be applied to the final quarter, but I've never heard in what percentages. the fact that Lindsey came up with the same info adds validity to your research.

From a mathimatical point of view, I don't know how Raggozin could conclude that the decelaration in a horse race is linear. I know Doc Sartin had an program that looked at races from and exceleration / decelaration stand point. and I personally know two former PIRCO guys who are developing a pace program using non-linear math. It would seem to me that they could more accurately predict how a horse may or may not stretchout. Well see.

I think Distance equalization maybe the toughest adjustment to figures of them all. Sartins programs did a poor job of it, made every horse rated off a sprint that was now stretching out look like Secretariat and vise versa.

ANDICAP
I believe you posted that you look for form patterns using pace and speed numbers. Are your pace numbers adjusted for distance or normalized? I've heard the argument that distance equalization / normalization would affect the true relationship between pace and speed. I know equiform author Cary Fotias doesn't adjust fractions to distance simply lets the velocity (after track varient) stand on its own, which makes sence to me. If you do adjust your figure by distance, do you still see the same relationship described in "Blinkers Off"?

If I've confused anybody, what I mean by distance equilization is the adjustment to a pace / speed figure to todays distance. Normalization is the attempt to equate or assign the same number to related pace and speed figures as with par times. Quirin speed / pace figures are an exsample of this.

Rick

12-18-2002, 11:52 AM

I've always thought that maybe one could use a distance and track adjusted 4 furlong time to compare horses. It's available in nearly every race, isn't affected nearly as much as final time by off tracks, and probably is a fairly good measure of the class of the race.

TenZin

12-18-2002, 12:14 PM

What's this talk about making horses normal? I don't think I
would like meeting a person who was really a horse normalized.

andicap

12-18-2002, 12:27 PM

Originally posted by delayjf

ANDICAP
I believe you posted that you look for form patterns using pace and speed numbers. Are your pace numbers adjusted for distance or normalized? I've heard the argument that distance equalization / normalization would affect the true relationship between pace and speed. I know equiform author Cary Fotias doesn't adjust fractions to distance simply lets the velocity (after track varient) stand on its own, which makes sence to me. If you do adjust your figure by distance, do you still see the same relationship described in "Blinkers Off"?

If I've confused anybody, what I mean by distance equilization is the adjustment to a pace / speed figure to todays distance. Normalization is the attempt to equate or assign the same number to related pace and speed figures as with par times. Quirin speed / pace figures are an exsample of this.

Good question. I use the BRIS figs and find them useless in comparing sprints to routes and vice versa. They are useless with sprint to route comparisons.
Even one turn to two turn routes is suspect. But my figs work OK with what I do with them in the same distance structures. It's a problem.

I don't do pace form cycles (tho I read "Blinkers Off") just speed.

Lindsay

12-18-2002, 03:13 PM

Delayjf wrote:

"From a mathimatical point of view, I don't know how Raggozin could conclude that the decelaration in a horse race is linear."

Delayjf,

Ragozin is talking about relationships between final times. He has studied hundreds of thousands of races. He is not talking about interior fractions.

Rick

12-18-2002, 04:37 PM

Lindsay & Delayjf,

There's a good analysis of that in "Handicapping Speed" by Charles Carroll. The relationship is nearly linear overall, but there are some interesting glitches in it, especially at about 7 furlongs.

superfecta

12-18-2002, 07:25 PM

Originally posted by andicap
If you can be specific about the dates -- maybe something recently, I can reply after checking it out. Give me dates, man, give me dates and tracks.

(BTW, I'm not saying the Beyer figs are the be-all. Sometimes the BRIS figs are better. And I'm sure Colt's Neck and some Sheets are better. I'm just saying the Beyers are good enough for me to pick contenders.) The example I gave was for April 17 and 18th of this year at Aqueduct.First race was a MSptW(2nd race on the card) and the second was for a 14k claiming race(4th race on the card).I just had the simulcast weekly for that race at hand to get an example from.Thats one of the problems ,I can find similar or worse examples too many times from PP.Why is there 11 points diff in rating and only .09 diff in time?You can't tell me Beyer is trying to be that accurate with his points system.There seems to be a Class or pace or Varient adjustment to his figs,and since I don't know why or which it is,that casts doubt on his methodology.
I don't begrudge anyone from using the figs to isolate contenders,sometimes its enough,just not for me.

Fastracehorse

12-18-2002, 07:30 PM

Of course U have.

I like to think I do what works.

Further, others have different opinions than U:

http://forums.prospero.com/dr-general/messages?msg=6918.20

http://forums.prospero.com/dr-general/messages?msg=6918.38

http://forums.prospero.com/dr-general/messages?msg=6918.44

http://forums.prospero.com/dr-general/messages?msg=6918.24

Lindsay,

Not to enjoy I'm sure.

fffastt

BTW,

Re: Mindless- this is another quote from a post to U by me:

I think getting too precise is dangerous as it takes alot of creativity out of the game [ a lack of imagination limits one's scope ].

delayjf

12-18-2002, 09:51 PM

Rick thanks for the comment with regards to Carrols Handicapping speed. Almost bought the book once but after a quick review I waved off. Now I will have to take a look at it again. Without having done that I wonder If he was refering to a horse races being linear but not the individual horses themselves. I know Mahl graphed several horse's velocity both front runners and closers and they were not linear. I'll check it out, thanks again.

The problem with distance equalization is that from a energy perspective it blurs what happened in the adjusted race. If thats not your bag, it might not be very detrimental at all.

Not sure what the point is other than convenrance to normalizing speed and pace figures. The reality is that a horse's velocity is greater at the pace call than at the finish. Why not let the figures reflect that reality? adjustments to pace / final times are difficult at best , why complicate the matter more.

Lindsey
Thanks for clearing that up.

Andicap
I've not had the best luck using the bris numbers. But, it's been a while since I last used them perhaps they've improved. Good to hear you've broken the code. I thought I'd read a post by you on the HTR board about form cycles, my bust.

MIKE D

12-18-2002, 10:02 PM

Originally posted by hurrikane
The best use of speed I've seen is from Massa. Jim Cramers best number in the last 90 days.

rC90 rank 1 27% win .88 ROI IV 1.94

this is pretty good for a speed fig.
I use to live and die by speed figs..now I use them just for appraising form.

As for the DRF numbers...if you find anything of value in them..you're way ahead of me. They don't move from track to track, and that's just the nature of the beast. So unless you are in a circuit like fla without many shippers they seem useless. Hi I read this board often ,never posted before,However let me say that I have never seen a use for speed figs.Maybe winners and horses finishing in close contention,but a horse not in contention will not be pushed to run all out so what use could be made of a speed fig?

Fastracehorse

12-18-2002, 10:59 PM

So what use could there be for a speed fig??

I like to think of the adjusted Beyer as a numerical value indicating a horse's ability.

Now, knowing a horse's ability has to be considered as powerful knowledge. Doesn't it??

Can U think of a more important factor than: Which animal is projected to be the best?? Even if U aren't sure if a projection will be realized, U would certainly want to know who has the highest potential.

Still, the best horse doesn't always win the race. We might all know this. And, just because a horse's past figs are the best does not mean he'll run back to those lofty numbers.

So, while one powerful handicapping tool in someone's arsenal is preferable, it is somewhat meaningless in the absence of a deeper knowledge of horse racing.

To re-capitulate,

Speed figs, at least the one I am familiar with [ the adjusted Beyer ] is a lethal tool. But, without the skill of handicapping, no singular tool is enough.

fffastt

MIKE D

12-18-2002, 11:07 PM

Hi I have been reading this board for quite some time now and would like to give my thoughts on speed figs.I think that they are pretty much useless due to the the nature of the sport.Trainers would not want a horse pushed into contention if for some reason,Bad Start,Bumped hard,trapped on rail.Save the horse for another day.eronious speed fig here.Best fig last time.will he repeat?Best speed last 3 races.My data base shows 24% win on this fig with a 0.90 roi.not very usable.Maybe the way to use the speed figs if we feel they mean anything,would be to average the last 10 o15 races.then look at the last race.was he above or below avg?If below maybe he is due to run back to his avg this time out.and how does that compare with the rest of the field,rated in the same manner.Its all pretty hypothetical.I think the only accurate speed fig is the winners time.I would love some input on this.Mike D

andicap

12-18-2002, 11:19 PM

Originally posted by delayjf

Andicap
I've not had the best luck using the bris numbers. But, it's been a while since I last used them perhaps they've improved. Good to hear you've broken the code. I thought I'd read a post by you on the HTR board about form cycles, my bust.

Well, I wouldn't go that far. It's just gotten me to a point where I have reliable contenders. I can't say yet I have broken the code. But I feel I'm on the right track.

Fastracehorse

12-18-2002, 11:36 PM

Find a serious overlay running against that horse with the best average of his last 3 speed figs.??

U are talking about the fixed fig in the form.

I am talking about adjusting figs.

The adjustment creates the overlay, not the fixed fig itself.

Notice this overlay beating a high fig horse.

http://forums.prospero.com/dr-general/messages?msg=4959.551

fffastt

smf

12-19-2002, 12:49 AM

Originally posted by MIKE D
Hi I have been reading this board for quite some time now and would like to give my thoughts on speed figs.I think that they are pretty much useless due to the the nature of the sport.

Mike D,

Welcome to the board. With comments like the one above, I think you'll get some vigorous opposition at times from some posters, but I hope you stick around. I think your view on figs has a lot to offer.

Keep the thick skin, and keep posting.

MV McKee

12-19-2002, 03:52 AM

Much of what is being discussed here relates to "figures", etc. that are being widely distributed. I think much of the logic that goes into products if this type is done as something of a matter of convenience, in order to make the product (that is what we are dealing with) palatable to a broad audience. An obvious example of this is the equating of 1 length to 1/5 of a second. It is not reality, but it is relatively close, and it lends itself to a lot easier scaling and neatness when using pen and paper to calculate figures (as was the case when the Beyer's were conceived). I feel that much of the basic logic behind Beyer and Quirin numbers, most pace figures and like products was arrived at through thorough research, but I also feel that the formulas and scales used to produce these numbers are vastly oversimplified and only produce accurate comparisons within a very narrow range.

It would be a very long, dry mathematical explanation to try and explain why I think these types of products are "homogonized" versions of what they could (should?) be. So it would be much easier to explain the problem I have with this type of thing by using a "plain english" type of product, the BVA's (Betting Value Averages) now being distributed by DRF.
There is a lot of sound logic in this product. They are gleaned from a substantial database with a sufficient number of events (races) to produce the numbers. The various categories are valid, time tested handicapping factors that handicappers may want to be aware of. The numbers are based on betting value and not WPS% (after all, most of us are in this for monetary gain, right?). And the numbers are weighted for recency (what have you done for me lately?). But therein lies the rub. The overall numbers (3 years) and the recent numbers, whether they are composed of the last X number of starts of the last X number of days, are two distinct data abstractions and should be viewed accordingly (as 2 seperate numbers). But in order to mass market this product (successfully), you just about have to roll it up into one neatly packaged and market number. In order to do this, a rolling average was incorporated. But what weighting do you give the most recent starts/time frame when calculating the BVA? In sound statistical analysis, this would be wholly dependant on the siutation. Racing just shifted to Track A to Track B and Trainer Bob, who is stabled at Track B suddenly wins 5 races in the past week. I think I'll weight his last 5 days heavily. Podunc Downs is closing as racing shifts to BigPurse Meadows, and trainer Joe is prepping all his good horses for the big meet and letting his tired, overraced cheap claimers fail on the drop so he can get rid of them and has gone 0-20 the last month...how do I weight that? There are a myriad of other regularly occuring circumstances that certainly drive home the point that there can be no reasonable way to accurately assess how a sample should be weighted. I am certain that the professional statisticians involved in producing the logic behind this number were aware of this, but they had to make the product a)a reasonably static calculation (there had to be some constants) and b)make a singular number that would be marketable to a wide audience.
Once again, that was long-winded, but I am trying to point out that much of what is marketed in the way of racing information is, by necessity, a simplified version of what it is actually meant to represent.
This entire thread has been very interesting to me, and some great discussion has taken place.
I put a great deal of research into the methods I use to produce my speed figures, and the actual calculations behind them are far too complex for me to do by hand in an efficient manner, but the PC makes it very quick and easy. But my figures are only a measuring stick for me, so I can compare apples to apples. In an earlier post someone mentioned that their figures were a measurement of a horse's ability. I take that down a notch and simply view my figures as a measurement of what a horse's ability WAS at that point in time under those particular circumstances.
I had a couple of scenarios/questions I wanted to pose, both to provide fodder for further conversation and to hear others views.

1)Scenario:
2 horses...both won last race at 6 furlongs with similar trips.
Fractions and Final Times of the horses.
HorseA :22 :23 :26 (1:11)
HorseB :23 :23 :25 (1:11)
The average winner of a 6 furlong race in 1:11 runs :22 :23 :26
Which horse do you think ran a better race?

2)Question
In races at the same class level, on the same surface, why is the average "ahead lengths" (as seen in a standard chart, not PP) between horses 1&2, 2&3, 3&4 and 4&5 the same regardless of whether the race is a route or a sprint?
And why is the same average in a turf race at the same class distance less than 1/2 that of a dirt race?

3)Question
You are making pace figures and are devising your scale for the figures. If you (for instance) assign a rating of 100 to a :44 half mile and a rating of 99 to a :44.10 half mile (one point for every .10 seconds), would you also assign an 80 to a :46 half mile and a 79 to a :46.10?

Just a couple questions for those of you still awake. I have my own opinions, but wanted to hear others.

Thanks

Michael

Rick

12-19-2002, 11:13 AM

delayjf,

I think Carroll was referring to world records at different distances rather than individual horses. It's certainly true that individual horses could have very different distance characteristics. But those kind of adjustments might be OK in the absence of any other evidence.

But I still say that whatever works best is what's right to do regardless of any theories. I corresponded with a guy once who had developed a brilliant mathematical theory of how horses should slow down as the distance of the race increases. I was really impressed until I asked him how it had done in practice using simple rules for race selection. His angry response that I was being simplistic to even consider testing it with no additional handicapping involved assured me that it didn't really work. Testing a small sample myself confirmed that what seemed like an excellent theory just didn't hold up in the real world. And I was sure in advance that it would have some value.

andicap

12-19-2002, 11:29 AM

It's silly to say final figures don't work because they don't hold up in a static statisitical sample that says very little. Are you saying horses that drop in class is a useless measure because these horses also show an ROI loss? How about a horse being claimed by a top trainer like Jerry Hollendorfer? Another flat ROI loss. I guess that means it's not really meaningful.

All these Sports Stat type of studies are silly and meaningless because no one handicaps that way. Or they shouldn't. I will pick the horse here with the best Beyer in his last three races. or in his last race. Of course you'll lose that way! Just like you'd like taking any single factor in isolation.

That doesn't mean they are unimportant.

What it does mean is the Beyers/BRIS figs, etc. are an important part of a larger mosaic. I incorporate them into pace figures and use a bit of common sense handicapping besides.

Example.
horse in the 9th race at Aqu on Sunday had a 92 Beyer in his last race -- top Beyer. Was 15-1. Horse had an outside post position in a two turn race that started close to the turn with a couple of speed horses inside of him (he was an E/P type who wants to be close to the lead).
The horse finished a very game 4th less than a length back in a blanket finish.
The Beyer is worthless! The horse lost. This horse with the 4th post position would more than likely have won (albeit not at 15-1).

delayjf

12-19-2002, 11:55 AM

Keep in mind that the BEST LAST RACE BEYER angle picked a 40.00 winner in the Kentucky Derby.

Rick

12-19-2002, 12:05 PM

andicap,

You're right, but studies can be very useful in identifying factors that are promising to be used in combination with others. If the very best speed rating loses 12%, that would be a positive thing to me since it probably could be combined with several other factors and get close to breaking even on all races. If one can get to that level he should then be able to find overlays in somewhere between 1/3 and 1/2 of the races.

I once ran a test of DRF handicapper's selections where I'd play any handicapper's top selection if the morning line odds were 4-1 or higher. There were only a few each day (I think about 3), but they actually showed a little profit, albeit over a pretty small sample (about 100 races I think). They don't show their ROI any more, but I think for the best handicappers it was about 0.9 in the past, or 10% loss playing every race. So, there may be value even using something that just loses less than the average as long as you pick your spots carefully.

andicap

12-19-2002, 12:13 PM

agreed.

Lindsay

12-19-2002, 02:07 PM

MV McKee wrote:

"I think much of the logic that goes into products if this type is done as something of a matter of convenience, in order to make the product (that is what we are dealing with) palatable to a broad audience."

Excellent point. The man who invented the Nautilus exercise machines, Arthur Jones, used to say that he could make the machines one percent more efficient but chose not to. Why did he choose not to? Because the additional efficency would come at great cost: People would no longer be able to get into the machines.

"1)Scenario:
2 horses...both won last race at 6 furlongs with similar trips.
Fractions and Final Times of the horses.
HorseA :22 :23 :26 (1:11)
HorseB :23 :23 :25 (1:11)
The average winner of a 6 furlong race in 1:11 runs :22 :23 :26
Which horse do you think ran a better race?"

I think I know where you're going. We have something of a paradox here. Horse A's faster fractions were both evidence that a stronger performance would be coming that day (in terms of final time), and an inhibitor to that performance (in terms of final time). I would disregard the question of who ran a "better race." I believe it's irrelevant in the sense you mean it here. Instead, I would ask how today's fractions are most likely to differ from yesterday's and give the edge to the horse that is more likely to run more-even fractions today, keeping in mind the paradox. The ideal situation is having a horse who is in condition to produce the extra early speed that is evidence of good condition, but, due to the circumstances of today's race, is not obliged to produce it.

More later on all your questions. They are excellent. Got to go.

Fastracehorse

12-19-2002, 02:23 PM

I read how BVA's are computated and U are right: There is a blending of the long-term and short-term values, which to me is useless.

Completely useless.

What works for me is knowing two things: 1) the individual trainer for his most consisitent moves and; 2) how the barn is doing currently.

For me, much info in DRF is not needed. But by proving U're own hypotheticals wrong a successful player will learn what really works.

I also concede to U that there is convenience in computing figs. I am glad this game is not so precise. A range is all one needs sometimes to identify overlays and find prime bets.

Also, what good does exactness do a handicapper anyways??Especially after considering all the probabilites against the animal running the perfect race.

Paradoxically, my cohesiveness to the useage of figs has actually taught me that many races will beat me if I rely on precision. Hence, I have a newer angle, called magic horses.

These magic horses show one or more factors that indicate trainer intent. However, this horse does not have my top figs necessarily. Maybe they do deserve it but I can't find the reason.

Usually magic horses pay well anyways because most people are fooled by horses chosen only by merits of trainer intent.

Answer to your questions:

1) I would adjust their Beyers with confidence.

2) No idea. Good question.

3) I don't devise pace ratings that way. I believe the Beyer is a pace/variant/time numerian.

fffastt

kitts

12-19-2002, 03:49 PM

"Betting Value Averages"
Sorry to be out of touch here, but what are they and where can I read up on them(search drf.com did not help)?

Believe it or not it is hard to get a DRF in Las Vegas other than at a (shudder) racebook.

Lindsay

12-19-2002, 05:11 PM

Kitts writes: ""Betting Value Averages"
Sorry to be out of touch here, but what are they and where can I read up on them(search drf.com did not help)?"

turfday.com

kitts

12-19-2002, 05:28 PM

Lindsay-
Thanks for the help. Keep the faith!

Lindsay

12-19-2002, 05:39 PM

MV McKee writes:

"3)Question
You are making pace figures and are devising your scale for the figures. If you (for instance) assign a rating of 100 to a :44 half mile and a rating of 99 to a :44.10 half mile (one point for every .10 seconds), would you also assign an 80 to a :46 half mile and a 79 to a :46.10?"

No. The tenth of a second from 44.00 to 44.10 is much more powerful than the tenth of a second from 46.00 to 46.10. Something similar can be seen in Beyer final figures. Maybe one horse is capable of running a 130, five horses can run a 125, 50 horses can run a 120, etc., just to make up a few numbers.

This is not a problem if one is aware of it, in which case the point-per-tenth thing can be seen as simply a convenient way of throwing clockings into a figure scale.

delayjf

12-19-2002, 09:10 PM

Lindsey,
Not sure I follow you as to the value or of .10 seconds with regards to 4 furlong clocking of 44.10 and 46.10. I can understand why the .10 would have more or less weight at diff distances, but diff velocities at the same distance??? Interesting, Could you elaborate,

Mckee
Love to hear your anaylsis.

Fastracehorse
I tend to mirror your connections approach, but I also ask the question, has this horse shown the ability to compete at this level in the past. Thats were speed / pace handicapping comes in for me. I think you'll find that in most longshots PPs you will find a race that is competitive (speed figure wise) with todays field. Lightly raced horse require a different approach.

Lindsay

12-19-2002, 09:31 PM

Delayjf writes:

"Not sure I follow you as to the value or of .10 seconds with regards to 4 furlong clocking of 44.10 and 46.10. I can understand why the .10 would have more or less weight at diff distances, but diff velocities at the same distance??? Interesting, Could you elaborate,"

Sure. There is a point of diminishing returns. It's easier to improve from a half in 49 to a half in 48 than it is to improve from a half in 44 to a half in 43. Think of a weightlifter. When last I checked, there was one man in the world who could bench press 700 pounds, but thousands who could bench press 300 pounds. An extra pound is an extra pound, yes, but what's easier, to go from 299 to 300, or to go from 699 to 700?

There is also the minor question of what a fast horse will do in the .10 versus what a slower horse will do in the .10.

Tom

12-19-2002, 10:27 PM

1)Scenario:
2 horses...both won last race at 6 furlongs with similar trips.
Fractions and Final Times of the horses.
HorseA :22 :23 :26 (1:11)
HorseB :23 :23 :25 (1:11)
The average winner of a 6 furlong race in 1:11 runs :22 :23 :26
Which horse do you think ran a better race?"

You have to know what the running styles of each horse is to answer this. If both are early horses, B is dead-he will be out of touch and A will proabably beat him by daylight. If B is a presser, and A is early, he has a better chance. If Both are pressers, A may not take the lead and run a better final fraction. The fact that both ran 1.11 last time is not what you need to know today. Many races are over well before the finish line. The faster final fraction of B is worthless if he is already out of the match up by the time they run it.

Tom

12-19-2002, 10:40 PM

I think you have to look at the horse's whole speed fig history, not just one fig. Better horses will run lesser figs at times. You have to figure out which figure they will deliver today. The 92 top fig on the outside horse at two turns is good example. I don't know what number he ran finishing fourth, but say it was an 84.
Next week, if he is in the same class and draws the 3 hole, which number is better to use? The 92 or the 84? Was 92 his top? Is he capable of running a 98? Is he likely to improve or decline off the last race? You have to handicap the horse, not just a number. Speed figs are just milepoles on his career-the are a baseline to judge performances by. They help put performances into some perspective. If you have pace figs, they will help explain some variations in final figs. But no number is the final answer. And you can't query a db to get the real impact of a number. Say the top beyer horse wins 25% of the time. So what? What is the percantage of horses with the top Beyer, third off a layoff, with a career top 10 points higher than the last race high number, and dropping in class today after an even effort last out 7 days ago? As opposed to a last out top beyer who won a maiden claimer wire-to-wire in the slop and today steps into a restricted stakes?
Both meet the definition of the query-last out best beyer.

(andicap>long post - check out the spelling! <G>)

MarylandPaul@HSH

12-19-2002, 10:55 PM

superfecta

12-20-2002, 12:33 AM

Originally posted by delayjf
Keep in mind that the BEST LAST RACE BEYER angle picked a 40.00 winner in the Kentucky Derby. Also keep in mind the man who(allegedly) came up with that fig did not bet that horse.Also keep in mind that the DRF speed figure was also the highest as well.

superfecta

12-20-2002, 12:52 AM

Originally posted by MV McKee

1)Scenario:
2 horses...both won last race at 6 furlongs with similar trips.
Fractions and Final Times of the horses.
HorseA :22 :23 :26 (1:11)
HorseB :23 :23 :25 (1:11)
The average winner of a 6 furlong race in 1:11 runs :22 :23 :26
Which horse do you think ran a better race?

Tom wrote the best answer on this one.

2)Question
In races at the same class level, on the same surface, why is the average "ahead lengths" (as seen in a standard chart, not PP) between horses 1&2, 2&3, 3&4 and 4&5 the same regardless of whether the race is a route or a sprint?
And why is the same average in a turf race at the same class distance less than 1/2 that of a dirt race?

I don't see that in my charts,I see varying differences between the horses,but I am looking at the winners position at those points of call, not the horses setting the pace.

3)Question
You are making pace figures and are devising your scale for the figures. If you (for instance) assign a rating of 100 to a :44 half mile and a rating of 99 to a :44.10 half mile (one point for every .10 seconds), would you also assign an 80 to a :46 half mile and a 79 to a :46.10?

Yes,simply because it gives equal weight to each horse.Taking into account the accuracy of these calls(which is a whole other subject)We must have some universal scale of comparing horse to horse.Once we have that measure then we have to factor in the horses running style,appropriate distance and current form(much of which some call a trainer angle).And some of that is purely subjective,no way to measure accurately.Those who say they can are not being honest with themselves.

keilan

12-20-2002, 01:46 AM

The questions were a bit obscure. It appears MV McKee wanted to stimulate discussion rather than first make some point. Hopefully MV will get involved in the discussion or his scenario's will soon come to a close.

Scenario #1 - if both horse's used the same amount of "try". Horse A ran the better race. If these two horse's ran only against each other and the races in scenario #1 are indicative of both horse's then again Horse A wins. However if these horse's ran in a field of 8 horse's no one would be surprised if Horse B finished ahead of Horse A if the fractions were quick.

Scenario #2 - First a 80k clmg race on the dirt and a 80k clmg race on the turf are not considered to be the same class of animal. If I understand the question you poise though, I would explain that relationship "ahead lengths" to be a by product of how dirt and turf races are run. For dirt races the emphasis is on early speed and the ability to carry that speed the distance, turf racing requires a delayed or more gradual disbursement of energy. Of course these are generalities and exceptions can be found.

Scenario #3 - looks like I'm missing something here, Sorry I don't buy the argument of diminishing returns nor the weight lifter theory. A bottom level claimer has as much difficultly shaving .10 seconds off his pace number and carrying his speed the distance as does the better class of animal. Pace Figures of 100 and 80 are miles apart and so are the class levels.

Lindsay

12-20-2002, 02:08 AM

Keilan wrote:

"A bottom level claimer has as much difficultly shaving .10 seconds off his pace number and carrying his speed the distance as does the better class of animal."

We can play a game, Keilan. You name horses who improved from 43.00 to 42.90, and I'll name horses who improved from 49 to 48.90. We'll see who comes up with a longer list, and it will be interesting to see if you come up with any list.

keilan

12-20-2002, 02:25 AM

Lindsay - love to play your little game, you list three first then it's my turn. Oh lets just stick to fast tracks!!

Lindsay

12-20-2002, 02:36 AM

Shamrock Blues, Miss Ivy Hilton, Frisky Kitty. (I did some rounding, but I got all three from back-to-back pages in yesterday's DRF. Try that.) We also need a denominator, Keilan. How many horses tried, how many succeeded?

keilan

12-20-2002, 09:45 AM

Brilliant Lindsay -- Now grow up!!

12-20-2002, 09:53 AM

Originally posted by Lindsay
Keilan wrote:

We can play a game, Keilan. You name horses who improved from 43.00 to 42.90, and I'll name horses who improved from 49 to 48.90. We'll see who comes up with a longer list, and it will be interesting to see if you come up with any list.

Try this Lindsay, find 3 horses who even ran a 43.00 half in that form...your statement makes no sense at all.

CJ

12-20-2002, 09:59 AM

Originally posted by MV McKee

1)Scenario:
2 horses...both won last race at 6 furlongs with similar trips.
Fractions and Final Times of the horses.
HorseA :22 :23 :26 (1:11)
HorseB :23 :23 :25 (1:11)
The average winner of a 6 furlong race in 1:11 runs :22 :23 :26
Which horse do you think ran a better race?

2)Question
In races at the same class level, on the same surface, why is the average "ahead lengths" (as seen in a standard chart, not PP) between horses 1&2, 2&3, 3&4 and 4&5 the same regardless of whether the race is a route or a sprint?
And why is the same average in a turf race at the same class distance less than 1/2 that of a dirt race?

3)Question
You are making pace figures and are devising your scale for the figures. If you (for instance) assign a rating of 100 to a :44 half mile and a rating of 99 to a :44.10 half mile (one point for every .10 seconds), would you also assign an 80 to a :46 half mile and a 79 to a :46.10?

Just a couple questions for those of you still awake. I have my own opinions, but wanted to hear others.

Thanks

Michael

I'll give my thoughts:

#1 I'll take the first horse every time...speed rules American dirt racing

#2 Part 1-The first part of a race is about establishing position regardless of the distance. The second part is that you can't compare dirt to turf, they are totally different

#3 If making pace ratings, .10 seconds should have the same value regardless of the actual time.

CJ

delayjf

12-20-2002, 11:54 AM

MV MCKEE - LINDSAY

Curious to know if what you did the same study of route races as you did with sprints. If so, what relationships did you discover.

Lindsay
Gave your response to my questions some thought last night. I agree that the ability to run 43.00 flat and faster 1/2 miles is usually limited to the best horses. But it seems to me that if a horse was only capable of running a 48 half (using your example) than the jump to 47.9 is just as tough for HIM as the jump to 42.9 is to the speed demon. Using your weight lifting example, if your genetic / physical limit is 300 pounds, then your the jump to 310 or 320 will be just as tough as the Olympic lifters jump from his genetic / physical limit of 700 to 710 or 720.

Not many horses run 43 flat, that alone would make the attempt at 42.9 rare. But almost all horses can run 48 flat even the slowest on the grounds. Meaning you'd see a lot of horses (except those who have reached their pace limits) make big jumps in pace numbers to their individual physical limits.

That's my take, am I missing something?

Lindsay

12-20-2002, 04:01 PM

CJ wrote: "Try this Lindsay, find 3 horses who even ran a 43.00 half in that form..."

That is the point, CJ. And that was precisely the point of my "denominator" line. I wish you had followed my comments through the entire string. As I wrote earlier, if all you want to do is convert fractions into convenient figures, then make one point equal .10 at all rates of speed. But if you want to give extra credit for a horse doing something that it is exceedingly unlikely that another horse can do, then adopt my suggestion. My analogy to Beyer final times was a good one, I think. Almost every horse in training who runs a 45 will eventually run a 55. How many horses who run 125s will eventually run 135s?

Keilan wrote: "Brilliant Lindsay -- Now grow up!!"

I will read your old posts and work toward your level of maturity, Keilan. But I want you to promise to be less hackneyed next time.

Delayjf wrote: "Curious to know if what you did the same study of route races as you did with sprints. If so, what relationships did you discover."

Yes, but riders do more rating in routes, which increases the disparity in early fractions and reduces my confidence in the results. A difference in .20 in final time translates to a difference of about .10 at the 6F call, but, again, this is based on a sample with wide variances.

"Gave your response to my questions some thought last night. I agree that the ability to run 43.00 flat and faster 1/2 miles is usually limited to the best horses. But it seems to me that if a horse was only capable of running a 48 half (using your example) than the jump to 47.9 is just as tough for HIM as the jump to 42.9 is to the speed demon. Using your weight lifting example, if your genetic / physical limit is 300 pounds, then your the jump to 310 or 320 will be just as tough as the Olympic lifters jump from his genetic / physical limit of 700 to 710 or 720."

We don't know what a horse's potential is. Potential is what he will become capable of doing. That is why we generalize based on statistics. Sure, there might be horses running 43s who can go on to run 42.90s, but that is not the way to bet. On the other hand, it is a rare horse in training who can't go from a 49 to a 48.90. If he didn't figure to have that kind of potential, he wouldn't be in training. The same thing applies to weight lifting, but I'd like to add one thing. Let's take two weight lifters who have reached their potential. One can lift 300 pounds, the other 700 pounds. The man who lifts 700 pounds is training MUCH more intensely--regardless of potential. Therefore, he will require more recovery time and sustain more injuries. This intensity--potential aside--taxes his reserves more. I believe this is why a horse is more likely to bounce off a 130 than off a 95--regardless of potential.

delayjf

12-20-2002, 09:25 PM

Lindsay,

I Agree with you that a horse running 43 flat is doing more absolute work as is a weight lifter who is benching 700 lbs compare to another only benching 300. Interesting theory that a horse who competes in the higher (faster) classes of racing require more time to recover due to his higher energy output. I think I recently read this particular theory as it applies to weightlifting by Peter Sisco. Does his theory on weight lifting have anything to do with yours on horses?

Consider this, The 700lb weight lifter is doing more work in an absolute sence (ie, he's lifting more weight) but in a relative sence (what is he capable of lifting) the weaker lifter is straining just as hard (if he's working at his physical limits). But again, aren't their levels of exertion already reflected in the higher pace figure or weight total.

I can understand how a super effort like a earning a 135 beyer would lead to a bounce, but how many horses run those figures, off the top of my head I can think of one, other than Secretariat.
I think it was General Assembly that ran an unbelievable beyer when he won the Travers way back when.

I'm not sure Pace figures (in this example) are a measure of potiential rather than a measure of race exertion at a specific point in the race. 49 to 48.90 is not the best analogy as your correct, almost all race horses can go that fast. My feeling is that regardless of time, once a horse reaches its physical limit, be that 49 flat, 45 flat or 43 flat, to exceed that level is just as taxing to that horse as it is it would be to a higher quality horse that has exceeded its physical abilities.

I figured as much concerning the route pace segments, thanks for the reply, interesting topic.

Lindsay

12-20-2002, 10:29 PM

Delayjf wrote: "I think I recently read this particular theory as it applies to weightlifting by Peter Sisco. Does his theory on weight lifting have anything to do with yours on horses?"

I'm sure they're similar, but my theory came from Mike Mentzer, a bodybuilder who won the Mr. Universe contest in 1979 and retired in 1980 after losing, in a pathetic decision, to Arnold Schwarzenegger in the 1980 Mr. Olympia contest. Mike and his brother, Ray (Mr. America 1980) did extensive research on training methods, intensity, overtraining, etc. I studied their work quite a bit several years ago when I was involved in weight lifting, and it has stood up quite well. Mike died recently, at a very young age, and Ray, who was even younger, died two days later. This was quite a jolt. I had talked to Ray a few times and was quite fond of him. Anyway, I am an inveterate maker of analogies between horse racing and weight lifting.

"The 700lb weight lifter is doing more work in an absolute sence (ie, he's lifting more weight) but in a relative sence (what is he capable of lifting) the weaker lifter is straining just as hard (if he's working at his physical limits)."

He is straining just as hard, yes, but he is not training with the same intensity; hence he will recover faster. Muscles recover quickly, but the system as a whole does not, and the drain on the system is greater in the case of the man who lifts 700 pounds. That's why Mike Mentzer, as he progressed, went from training six days a week to training perhaps one day in seven.

"But again, aren't their levels of exertion already reflected in the higher pace figure or weight total."

If by "exertion" you mean "force brought to bear," yes. What we don't know is how much this will take out of them or how close to their potential this will take them. But we can make estimates based on horses and weight lifters who came before them.

"I can understand how a super effort like a earning a 135 beyer would lead to a bounce, but how many horses run those figures, off the top of my head I can think of one, other than Secretariat.
I think it was General Assembly that ran an unbelievable beyer when he won the Travers way back when."

Yes. Secretariat ran one in the 140s (his Belmont) and General Assembly ran one in the 130s in the Travers, where he had an easy lead on a wet track. Feel free to knock my knockout numbers down into a more plausible range, say 115 or 120. The principle doesn't change.

"My feeling is that regardless of time, once a horse reaches its physical limit, be that 49 flat, 45 flat or 43 flat, to exceed that level is just as taxing to that horse as it is it would be to a higher quality horse that has exceeded its physical abilities."

Two things: 1: We don't know the outer limits of the horses' abilities. We do know, through statistics, that it is more likely a horse can go from a Beyer of 95 to 100 than from 100 to 105 than from 105 to 110, etc. etc., getting exponentially harder the faster the Beyer gets. And I believe the same thing applies to pace. 2: The effect on the horse's system is greater when numbers get faster--regardless of potential (this is the weight lifting analogy again).

"I figured as much concerning the route pace segments, thanks for the reply, interesting topic."

Thank you. I've enjoyed it.

Tom

12-21-2002, 10:59 AM

Too bad there isn't a way we could all look at some races and find a way to pick our horse that examplify our various opinions and then somehow run an experiment to see whose ideas do the best. And if there were only a way to keep score.
Oh, wait....we could bet races! And count our winings.
Never mind.

delayjf

12-21-2002, 07:48 PM

Lindsay,

Peter Sisco often quotes Mentzer in is weightlifting books. Didn't His brother commit suicide over Mike's death? I know Mentzer was an intensity freak. I've been looking for other strength experts to chime in on Sisco's theories as well. I believe in his latest book, Sisco claims that the human body can take up to 2 weeks to fully recover from a single intense workout.

Can't say that I disagree with your theory that greater exertion requires more recovery time, just not sure if I understand with how you apply it as to adjustments.

If your theory is correct, then stakes horses should bounce more often after a "top" effort than a 10,000 claimer. Is this what your research has found?

Fastracehorse

12-21-2002, 08:41 PM

I lift wieights via a program called Periodization.

The founder of the program is named Tudor Bompa.

It is extremely scientific ( Bompa has his Doctoral and is famous ).

Anyways, the program is very intense and does promote a period of downtime. This downtime has a regneratory effect on the body and mind.

Horses need this too.

Side-note: The milk-shake for horses contains mostly sodium bi-carbonate ( baking soda ). It absorbs lactic acid production during the race to delay the onset of fatigue. Horses can improve their performance by 5 lengths just with the tubing of this delicious cocktail.

Optimum recovery time is obviously a goal of triners. There is a super-chargeing effect too with optimum times in betwen work-outs and races.

fffastt

Lindsay

12-21-2002, 10:01 PM

Delayjf wrote:

"Peter Sisco often quotes Mentzer in is weightlifting books. Didn't His brother commit suicide over Mike's death?"

It was reported as a suicide, but someone close to Mike and Ray insisted that Ray died of "a broken heart." It's easy to see where the latter could lead to the former. Ray had been having severe health problems before Mike's unexpected death.

"I believe in his latest book, Sisco claims that the human body can take up to 2 weeks to fully recover from a single intense workout."

Yes. The thinking on recovery has changed a lot over the last 30 years. Mike Mentzer and Arthur Jones (Jones invented the Nautilus machines) deserve much credit for this, and it sounds as if Sisco deserves some too.

"Can't say that I disagree with your theory that greater exertion requires more recovery time, just not sure if I understand with how you apply it as to adjustments."

I mentioned this earlier in the string, but it got buried under the hail of replies. It is not necessary to adjust the scale as long as one understands what is happening. As fractions or final times get really fast, they should be considered rarefied, obtainable by only a select few. Thus, a horse who has already run a freaky number, say a 125 Beyer, is more likely to run another 125 than is a horse whose previous best is a 117. This is the case even if the faster horse ran the 125 several months ago and seems to be going off form, and the slower horse seems set to run the race of his life. This is because the horse with the old 125 has already proven that he's one of the minuscule percentage of horses who can run a 125, while the slower horse (even if he's only a little slower) has not. And a layoff for the horse with the old 125 would make it even more likely that he'll get back to the 125. Weight lifters are fond of saying that muscles have memories, and they do. Someone who has bench pressed 500 pounds may quit training for a year to allow his injuries to heal. The day he launches his comeback, he can probably lift only 250 pounds. But he can get back to 500 in just a matter of months--even if it took him five years to get from 250 to 500 the first time.

"If your theory is correct, then stakes horses should bounce more often after a "top" effort than a 10,000 claimer. Is this what your research has found?"

I can't call this my research. I started using sheets (usually Thoro-Graph) about three years ago, and as I was studying form cycles, naturally I began to think of many analogies between horse racing and weight lifting. The Beyer example I gave above is nothing more than Jerry Brown's (the Thoro-Graph founder) research converted to Beyers and put into my words. To answer your question, yes and no. Stakes horses get much more rest and much better care than 10,000 claimers, which generally results in them bouncing less. But once their numbers start getting very fast, they become great candidates to bounce despite the rest and care, and even to bounce right into career-ending injuries. One thing that really impressed me when I started using sheets is how sheet handicappers could occasionally look at a freaky fast number and predict that a horse would be hurt by it and possibly never run again. Point Given is one example, Left Bank another. There are many more.

delayjf

12-22-2002, 09:09 AM

Lindsay
thanks for the reply, the Scott Lake horse (Thunderello) that recently finshed 2nd in the breeders cup sprint and then broke down in his next start three weeks later would be another good example.

Didn't know you used the sheets, have to ask because its come up in other treads if you have used the Xtras, which are sheets style numbers with a pace number as well. The Developer was a long time sheets user. Basically he incorporates pace "tops" to predict improvement in the final number. Some have tried to incorporate his theories with various pace / speed figures, not to much out there as to how they are doing. They are fairly new.

Lindsay

12-22-2002, 07:30 PM

Delayjf wrote:

"thanks for the reply, the Scott Lake horse (Thunderello) that recently finshed 2nd in the breeders cup sprint and then broke down in his next start three weeks later would be another good example."

You're quite welcome. Thanks for the Thunderello example. Of all the horses who ran on the rail on BC day, he ran the biggest race, a huge race.

"have to ask because its come up in other treads if you have used the Xtras, which are sheets style numbers with a pace number as well."

I haven't used them, but I read about them when they came out. In addition, I had long discussions about them with a friend of mine (also a Thoro-Graph user) when Blinkers Off came out. I can't stand the way Fotias makes his numbers, pace or final time, and I don't believe his explanation for why he doesn't incorporate ground loss. It's similar to the explanation Trackmaster gives on its site, and I don't believe that explanation either. It costs hundreds of thousands of dollars a year to gather all the information (ground, wind, run-up distances, etc.) necessary to make good figures; so when Fotias and Trackmaster say they CHOOSE not to do something, not because it's expensive, but because they think it's unnecessary, it's all I can do to keep from laughing. Also, I think velocity numbers (Fotias and Sartin) are pure nonsense dressed up to look like impure nonsense. Having said that, I do think Fotias's ideas on combining pace with final time are outstanding. I've been doing more or less the same thing for many years. I just wish Fotias made his numbers differently, and so does my friend. He stopped using them many months ago.

delayjf

12-22-2002, 08:20 PM

Lindsay,

As someone with experience with the sheets, how good are they, perhaps you've seen posts were some handicappers quote win % for best last race beyer or best beyer in last 90 days etc. Pretty much all speed figures have about the same win % and ROI within a few percentage points. Any idea what the sheets win % is?

Concerning why Fotias doesn't account for ground loss. I think he has a valid point with regard to the physics of running wide, horses or human cannot run as fast going around a turn as they can on a straight. In that respect horses running wide have a easier time running than those on the rail. BUT, DOES THAT MAKE UP FOR THE GROUND LOSS??? I'll leave that to the rocket scientist, (any rocket scientist out there???) as I have no clue. Other than the conveniance of looking at whole numbers verses integers, I have no problem with velocity. there's a lot to be learned from guys like Ken Massa at HTR.

The pace performance patterns to me are the heart and soul or his approach. I think he can explain some performances better than the sheets by incorporating pace. Time will tell, I've heard from some who love his numbers.

I've always felt the big edge in the sheets was that they hand-timed the races and used that time vice the Official time. For the life of me I can't underdstand why tracks don't time races out of the gate. Been toying with the idea of hand timing races off a TV replay, wonder if it would be as accurate.

Lindsay

12-23-2002, 04:16 AM

Delayjf wrote:

"As someone with experience with the sheets, how good are they"

I think both Thoro-Graph and The Sheets are excellent. The moment when I switched to sheets and started studying form cycles was a momentous one for me. In fact, I would put it up there with my first glimpse of Halle Berry.

"Any idea what the sheets win % is?"

Most sheet players, me included, will reject this sort of question. We are inclined toward pattern reading, not best last-race figure. However, Sport-Stat did a study that showed that betting the horse with the best Ragozin figure of the last three races produced a flat-bet profit if the horse went off at 4-1 or higher. I don't think this says much, if anything, but there it is.

"Concerning why Fotias doesn't account for ground loss. I think he has a valid point with regard to the physics of running wide, horses or human cannot run as fast going around a turn as they can on a straight. In that respect horses running wide have a easier time running than those on the rail. BUT, DOES THAT MAKE UP FOR THE GROUND LOSS???"

I'm certain that it doesn't. There is a flow to sheet numbers that I've never seen in any other numbers. Horses work up to tops, bounce, recover gradually, get back to their tops, bounce again, etc. I believe this "dutiful march" would be impossible if the ground-loss adjustment were off by much. And if you could get Fotias on a polygraph machine, I bet he would say he omits ground loss because getting it costs too much money.

"Other than the conveniance of looking at whole numbers verses integers, I have no problem with velocity."

I certainly respect your opinion, but I have to disagree. Say a horse goes the half in 45.00. This is 58.66 feet per second. These are simply two ways of saying exactly the same thing. So far, so good. But if someone goes on to average velocity numbers, as the Sartin people do, he will end up with an artificial number. In real life, a horse who goes 21.80 44.80 109.80 will lose to a horse who goes 23.00 46.00 109.70, and none of Sartin's contrivances can change that. And if they have a rematch and everything--including the fractions--remains the same, that horse will lose again.

"The pace performance patterns to me are the heart and soul or his approach. I think he can explain some performances better than the sheets by incorporating pace."

Agreed. I think the best approach is to combine Thoro-Graph or the Sheets with good pace numbers.

"For the life of me I can't underdstand why tracks don't time races out of the gate."

It might once have had something to do with an ego thing between thoroughbred owners and quarter horse owners. With a running start, thoroughbreds can often record a first quarter that is faster than quarter horses can record, since quarter horse races are timed from the gate. And the best quarter horses often race at a quarter mile. This is just a guess.

"Been toying with the idea of hand timing races off a TV replay, wonder if it would be as accurate."

Interesting idea. You might want to practice on quarter horse races, since they're timed from the gate. Comparing your times to the official times will tell you how good you are.

There is one thing you can do even if you don't want to handtime the races: Watch the starts carefully. Occasionally, a horse will get a flyer, breaking a length or two faster than the field and thus activating the timer sooner. This affects final figures and absolutely destroys pace figures. Knowing about such races can save you from being misled by the occasional bad number.

karlskorner

12-23-2002, 09:19 AM

Your thoughts on "hand timing" from TV replay. both TG and the Sheets hand time from the gate, we kicked this around several months ago on this board. Your problem will be as to how far the gate is set back from the "light". it's not the same everytime. What I do is assume the "down poles" from the rail are approximately 10" apart and adjust from there. Some times the gate is set back 30' and I have seen it as far back as 100", makes a big difference from the "Offical time" Get a "good" stop watch so that you can do the 1/4's and 1/2's at the same time. different story from what you read in the PP's.

Karl

delayjf

12-23-2002, 11:11 AM

Lindsay,

Given the scenario you decribed, I'd agree with you both horses run that same race, the samething would happen. the fastest horse would win. But if the first horse now gets away with a slower 1/4 or 1/2, pace handicappers would argue that he now has the advantage and would prevail. Same scenario, but not horse b gets sucked into a faster 1/4 or 1/2, he will not have the same closing kick.

Why is computing ground loss so expensive, compared to the other factors the Sheets use?

Good point about the Flyer.

Karskorner
What I had in mind was to time from the gate. Problem with timing some fractions is that from your vantage point you might not be able to see exactly when the horses cross the beam as you would at the finish. Are you saying that some tracks are not consistant with where it lines up the gate for a specific distance?

andicap

12-23-2002, 11:25 AM

Lindsey wrote:
[B]
I certainly respect your opinion, but I have to disagree. Say a horse goes the half in 45.00. This is 58.66 feet per second. These are simply two ways of saying exactly the same thing. So far, so good. But if someone goes on to average velocity numbers, as the Sartin people do, he will end up with an artificial number. In real life, a horse who goes 21.80 44.80 109.80 will lose to a horse who goes 23.00 46.00 109.70, and none of Sartin's contrivances can change that. And if they have a rematch and everything--including the fractions--remains the same, that horse will lose again. [B/]

Are you talking about what a horse is showing in the PPs? Or how they actually run the race. If the latter, yes, you are right. But if the former, I disagree.
Why would the horse that went in 109.70 with the slower fractions win every time? The horse that went in 44.80 could just slow down to 45.40 and finish in a faster time. Of course that's assuming the horse isn't an E horse that is being pushed by two other E horses to go that fast. It depends on the overall pace picture. But assuming a moderate/honest pace, the 44.80 horse will win if he is allowed to slow it down.

Show Me the Wire

12-23-2002, 12:21 PM

State of confusion:

Lindsay wrote in response to Delayjf:

“I certainly respect your opinion, but I have to disagree. Say a horse goes the half in 45.00. This is 58.66 feet per second. These are simply two ways of saying exactly the same thing. So far, so good. But if someone goes on to average velocity numbers, as the Sartin people do, he will end up with an artificial number. In real life, a horse who goes 21.80 44.80 109.80 will lose to a horse who goes 23.00 46.00 109.70, and none of Sartin's contrivances can change that. And if they have a rematch and everything--including the fractions--remains the same, that horse will lose again.”

Lindsay I have been following the thread and the thought processes behind the logic. Now, I am confused as to the logic applied, because of the above quote. If I have quoted of context this may explain my lack of understanding the thought process.

A: Earlier a position was stated that a faster performance is stronger, basically few horses have the potential to run a 45 half.

“The tenth of a second from 44.00 to 44.10 is much more powerful than the tenth of a second from 46.00 to 46.10. Something similar can be seen in Beyer final figures. Maybe one horse is capable of running a 130, five horses can run a 125, 50 horses can run a 120, etc., just to make up a few numbers.”

I understand the logic behind this statement. I understood it to mean the runner is exhibiting more natural ability and is pushing the upper limits of its potential. Therefore, I understood any runner capable of this improvement is a superior runner as compared to the majority. Based on this progression the next logical step would be the next statement:

B: “The ideal situation is having a horse who is in condition to produce the extra early speed that is evidence of good condition, but, due to the circumstances of today's race, is not obliged to produce it.”

I understood this statement to imply a superior runner as described above could defeat average runners any time the superior runner could conserve its initial energy i.e. relax.

Reading the response to Delayjf, I understand the response to mean the average runner, the one that never extends itself would always win. This is the basis of my confusion regarding the thought process quoted earlier.

If I am not understanding correctly, could you clarify my misunderstanding?

Regards,
Show Me the Wire

Too much color blinds us.

Show Me the Wire

12-23-2002, 01:18 PM

[QUOTE]Originally posted by Show Me the Wire
[B]State of confusion:

A: Earlier a position was stated that a faster performance is stronger, basically few horses have the potential to run a 45 half.

“The tenth of a second from 44.00 to 44.10 is much more powerful than the tenth of a second from 46.00 to 46.10. Something similar can be seen in Beyer final figures. Maybe one horse is capable of running a 130, five horses can run a 125, 50 horses can run a 120, etc., just to make up a few numbers.”

So as not to add to the mix I meant to type 44 and not 45 half.

Regards,
Show Me the Wire

Too much color blinds us.

Lindsay

12-23-2002, 07:50 PM

Karlskorner wrote:

"Your thoughts on "hand timing" from TV replay. both TG and the Sheets hand time from the gate, we kicked this around several months ago on this board. Your problem will be as to how far the gate is set back from the "light". it's not the same everytime."

I agree. Knowing the "real" time is of little use if we don't know the real distance. Delayjf will need to check to see if a particular track plays fast and loose with the run-up distances.

Delayjf wrote:

"But if the first horse now gets away with a slower 1/4 or 1/2, pace handicappers would argue that he now has the advantage and would prevail. Same scenario, but not horse b gets sucked into a faster 1/4 or 1/2, he will not have the same closing kick."

Absolutely. That is the heart of pace handicapping. What I'm opposed to is a contrived rating that gives an edge to the quicker horse regardless of today's pace situation. Another problem I have with velocity is that it can be a lazy way of making comparisons across distances. It's too easy and far too clumsy, I think. At least it was in Brohamer's book. It can be used in a more sophisticated manner, but I still see no advantage for it other than in something like Sartin's "percent early," in which case it is quite handy.

"Why is computing ground loss so expensive, compared to the other factors the Sheets use?"

Thoro-Graph and the Sheets pay trackmen to be at every major track every day. These men get the ground loss (often using the head-on shot, I think), handtime the races, watch for horses in trouble, estimate the wind, and check the run-up distances. I don't think Fotias has trackmen at every track. I don't know that he has them at any track. He could be getting his wind estimates from nearby airports. Getting ground loss for every horse is a ton of work. I did it myself for a while and hated it beyond belief.

Andicap wrote: "Are you talking about what a horse is showing in the PPs? Or how they actually run the race. If the latter, yes, you are right. But if the former, I disagree.
Why would the horse that went in 109.70 with the slower fractions win every time?"

I was talking about how they actually run the race. It was a tautology. If nothing changes, everything remains the same. I think we agree that today, the frontrunner will probably face fractions that are either slower or faster than he did in his last race, and the closer will largely be at the mercy of these fractions.

Show Me the Wire wrote:

"Reading the response to Delayjf, I understand the response to mean the average runner, the one that never extends itself would always win. This is the basis of my confusion regarding the thought process quoted earlier."

I'm sorry I wasn't clear. All I was saying was that no matter how talented he is, a frontrunner who got fried on the lead last time is not a good bet if he figures to get fried on the lead again this time. For his talent (a half in 44.00, say) to get a payoff in final time, he will need a pace that is kinder, and judging whether today's pace will be kinder is a central question in pace handicapping.

"I understood this statement to imply a superior runner as described above could defeat average runners any time the superior runner could conserve its initial energy i.e. relax."

That is how I meant it, though I would say "usually" instead of "any time." Thanks for following this string so closely, Show Me the Wire." Due to its length, it has gotten kind of crazy.

delayjf

12-23-2002, 09:19 PM

LINDSAY,

Thanks for the insight to the sheets. Went back and re-read "Blinker Off" as to how he constructs his figures. Fotias, claims that the sheets don't factor in track bias but his numbers do. He makes a logical case where that would create inflated numbers if that were not considered into the ground loss equation.

Do you attempt to construct you pace numbers simular to the sheet numbers.

You'd better sit down for this one, but it just dawned on me that using the scenario and the pace fractions you listed in your earlier thread, that you and old Doc might actually agree as to who is the better horse. I don't have a FPS chart here with me, but it would be interesting to see what the Avg pace velocity would be for those two horses in your example. Showmethewire, can you give us a hand with that one??

Perhaps this tread has gotten long, but I've enjoyed it. Learned something. Which is why I'm here.

Tom

12-23-2002, 09:35 PM

Velocity of the race in question:

A
21.8 60.55 F-1
23.0 57.39 F-2
25.0 52.80 F-3
----------------------------------------------------------
69.8 170.74 Total Energy
69.08% Percent Median Energy

B
23.00 57.39
23.00 57.39
23.70 55.70
----------------------
69.70 170.48
67.33%

Horse A has a huge advantage in the first fraction. B runs the same second fraction as A, therefore, B DOES NOT GAIN ON A until the last fraction. By that time, the race is probably over. A will run many more times than not against B. B cannot win this match up-A can lose it, by dueling with other horses, but B cannot win it. B needs something else to help him win.
A horse behind at the half needs to make up twice the deficinecy
inthe third fraction to win (HUEY factor). In this case. B needs to make up at least 3.1 units to win-he only makes up 2.9. Close, but this is why so many fast closers run second and third-just a bit to much to do before the wire gets there.
This is the edge pace has over final times.

delayjf

12-23-2002, 10:07 PM

Tom,

Thanks for going to the trouble, but I got to say I don't get it.
and I can just hear Lindsay:

"If these two horses are racing and one finishes in 109.8 and the other in 109.7, how did the horse with the faster time lose this race?"

Could you elaborate on the Huey factor?

keilan

12-24-2002, 01:35 AM

Jeff - - Just catching up with this thread. Tom has given you the FPS calculation for the 1st, 2nd and 3rd fractions. The Average pace numbers are 56.91 fps (%E 52.74 )and 56.83 fps ( %E 50.75 ) respectively. My calculation for energy dispersion differs from Tom's as we are probably measuring at different points in the race.

Based on the fractions Lindsay's provided ,Tom's race analysis is absolutely correct. Simply put horse B starts his run too late.

Lindsay -- I apologize for my comment to you the other day. I would like to wish you a " Merry Christmas ".

Dick Schmidt

12-24-2002, 02:32 AM

Tom,

Good post. You got it right in one.

Delay:

You asked:

"If these two horses are racing and one finishes in 109.8 and the other in 109.7, how did the horse with the faster time lose this race?"

You're looking at this wrong. They ran those times LAST TIME OUT, not today. Obviously the horse with the fastest final time wins every race, but we are trying to use history to predict the future. If best final time won every race, it wouldn't be much of a sport, would it. Also the times given are very unrealistic. Very few horses would ever run the times of "B". If they were accurate, and they're not totally impossible, then I would be willing to bet "A" every time and expect to cash 90% of my tickets.

This is what Pace Handicapping IS! The idea that a slower horse with better internal fractions can beat a horse with a faster final time is offensive to many people, but any quick look at a Racing Form will provide multiple examples. I had this same argument with Andy Beyer once, and he finally snorted and walked away, muttering that fast horses always won. Two years later he admitted in print that pace was real, and a valid aspect of handicapping. You get beat often enough, it finally gets through. You keep betting on the "B" horse and watching him almost get up time after time and it may sink in.

Dick

Lindsay

12-24-2002, 03:12 AM

Dick Schmidt (one of my favorite people), you have completely misunderstood what Delayjf was saying, probably because you didn't read what came before.

Lindsay

12-24-2002, 03:54 AM

Keilan wrote:

"Lindsay -- I apologize for my comment to you the other day. I would like to wish you a " Merry Christmas "."

Keilan: Thank you, but no need to apologize. Merry Christmas to you, too.

Delayjf wrote:

"Fotias, claims that the sheets don't factor in track bias but his numbers do. He makes a logical case where that would create inflated numbers if that were not considered into the ground loss equation."

The Ragozin Sheets ignore biases. Ragozin's right-hand man, Len Friedman, believes that while biases pop up from time to time, they don't occur with the frequency that many people think. He believes that many people attribute to biases what they ought to attribute to randomness. Thoro-Graph, on the other hand, does look for, and flag, bad-rail biases. This was done to great effect in last year's BC, where, according to Thoro-Graph, the rail was death and the two-path was near death. An amazingly high percentage of horses who ran near the rail that day improved in their next start.

"Do you attempt to construct you pace numbers simular to the sheet numbers."

No. But I do pay attention to weight and ground loss, as the sheets do.

"Thanks for going to the trouble, but I got to say I don't get it.
and I can just hear Lindsay:

"If these two horses are racing and one finishes in 109.8 and the other in 109.7, how did the horse with the faster time lose this race?"

Well said, Delayjf. Assuming the same fractions next time, as my example did, why will the result change? Undergirding pace theory is the belief that fast fractions are destructive. This is why, as Dick Schmidt wrote below, slower horses with faster internal fractions can often be expected to beat faster horses. But key to this is that the destructive fractions not be repeated.

Now I'd like to do an experiment with velocity. Warning, I am exaggerating for effect. I realize horses don't run like horse B.

Horse A: 22.00 45.00 109.00. Horse B: 24.00 47.00 109.00.

Who will get a higher rating on avg. pace, the frontrunner or this deepest of all closers?

Richard

12-24-2002, 08:32 AM

I recall a two-part article (ATM, Jan-Feb 2000)by Howard Sartin where he recommended using the best of last three races adjusting as necessary the speed number depending on how high or low the variant was(17,18,19 =0.Up to,say,10 points over,add to speed rating.Down to 10 points below subtract points from speed rating).I have used this this with some success.I still feel however,that some subjective judgement is involved.I would still would want to use the same class of track or better than the track upon which the race is to be run upon.The article in question can be found at the ATM website if anyone is interested.

Richard

12-24-2002, 08:38 AM

My apologies.The Sartin article is located at:
www.americanturf.com/atm/atmindex.htm

karlskorner

12-24-2002, 09:30 AM

An old race track saying "Time is only important when you are in jail".

You can beat the questions mentioned above to death and still not come up with an answer. Why ? Becasue as I have stated many times before, the race you are looking at today was NEVER run before. You can compare the times of 2 horses if they came out of the same race, or a different race and still not come up with an answer. Why ? Because there are 100's of different scenarios in TODAY'S race. Even if the 2 horses you are looking at run again today, they will never run the same as their previous race. Horses are not "mechanical wind ups". Compare any horses last three 1/4 or 1/2 times, for the same distance, are they the same ? The answer, Form.

Big Bill

12-24-2002, 09:52 AM

Richard,

I've read Sartin's two-part article in ATM and his method of using the DRV Variant for making adjustments. If you remember, in Brohamer's MPH he describes another approach of Sartin's for the use of DRF's variants with his fps (velocity) methodology. He (Sartin) discoved that .08 fps was equivalent to approximately one point of DRF Variant. Thus, a horse running on a day with a variant three points lower than another contender, receives a negative adjustment of -.24 (3*.08) to each of its fractional components.

William Scott, in his Total Victory as the Track, explained a method similar to what Sartin described in the ATM article of using the DRF Variant to make adjustments.

For those who don't have par times and/or the time to make their own variants and track-to-track adjustments, the above methods used to make daily track variant adjustments work better than making no adjustments at all. Also, ATM's Track Equalization Chart can be used to make track-to-track adjustments that would result in better figures than not making any at all.

Big Bill

keilan

12-24-2002, 10:50 AM

Lindsay wrote:

Now I'd like to do an experiment with velocity. Warning, I am exaggerating for effect. I realize horses don't run like horse B. I realize horses don't run like horse B.

Horse A: 22.00 45.00 109.00. Horse B: 24.00 47.00 109.00.

Who will get a higher rating on avg. pace, the frontrunner or this deepest of all closers

And the answer is -- Both horse's receive an AP figure of 57.46 FPS. Fractions were 22.0 23.0 24.0 for Horse A and then applied in reverse order for Horse B.

GR1@HTR

12-24-2002, 01:56 PM

Originally posted by karlskorner
You can beat the questions mentioned above to death and still not come up with an answer. Why ? Becasue as I have stated many times before, the race you are looking at today was NEVER run before. You can compare the times of 2 horses if they came out of the same race, or a different race and still not come up with an answer. Why ? Because there are 100's of different scenarios in TODAY'S race. Even if the 2 horses you are looking at run again today, they will never run the same as their previous race.

You mean like, luck? or...
chance, accident, break, destiny, fate, fifty-fifty, fortuity, fortune, hap, hazard, occasion, occurrence, toss-up, unforeseen event, affliction, bad break, bad news, bummer, calamity, catastrophe, clutch, cold, contretemps, crunch, difficulty, disaster, distress, downer, drag, evil eye, grabber, hard knocks, hard time, hard times, hardship, holy mess, hurting, ill fortune, jam, jinx, misery, misfortune, mishap, pickle, poison, reverse, rotten break, rotten luck, scrape, sorrow, suffering, the worst, tough break, tough luck, trial, trouble, unholy mess

Just kidding...Happy holidays to all...

Lindsay

12-24-2002, 03:37 PM

Keilan wrote:

"And the answer is -- Both horse's receive an AP figure of 57.46 FPS. Fractions were 22.0 23.0 24.0 for Horse A and then applied in reverse order for Horse B."

Thanks, Keilan. Why does this method stop giving an edge to frontrunners over closers when closers close really fast? In other words, doesn't averaging velocity numbers reward horses who run uneven fractions, regardless of whether they're frontrunners or closers? If so, how does this blend with all the talk about the advantage of frontrunners?

karlskorner

12-24-2002, 04:29 PM

The word Variant will just not die, we kicked it around in the past. My question is: You have a variant for 12/07/02 for your track which may be minus 3, plus 3 or #17 (what ever system you use, split for sprint/routes, etc.)

Now you open your PP's for your favorite track for today and there are 2 horses who ran on 12/07/02 and the varaint was one of the above. What is the variant for today ? YOU DON'T KNOW You will know when the races are over and done, than you will do your math and give a variant to today's races, but you do know that on 12/07/02 the track was either slow or fast by your variant.
What are you using for a bench mark for today's races. Is today's track fast/slow ? How can you use a 17 day old variant on today's track ?

Words such as Variant, Pars and Bias, all invented by book writers to add mystery to racing and they sure are keeping a lot of people busy

smf

12-24-2002, 05:38 PM

Originally posted by GR1
You mean like, luck? or...
chance, accident, break, destiny, fate, fifty-fifty, fortuity, fortune, hap, hazard, occasion, occurrence, toss-up, unforeseen event, affliction, bad break, bad news, bummer, calamity, catastrophe, clutch, cold, contretemps, crunch, difficulty, disaster, distress, downer, drag, evil eye, grabber, hard knocks, hard time, hard times, hardship, holy mess, hurting, ill fortune, jam, jinx, misery, misfortune, mishap, pickle, poison, reverse, rotten break, rotten luck, scrape, sorrow, suffering, the worst, tough break, tough luck, trial, trouble, unholy mess

Just kidding...Happy holidays to all...

Not to mention 'Baby Arm' or 'Lobster Claw'...

Just my take on this stuff....Reading this thread makes me think that most cappers look at races (sprint for example) being run on a monopoly game pad where there are 3 panels to place down a "piece" . As if nothing happens between the starting gate and the first qtr, where a piece is laid out on panel #1. Then the piece is lifted to panel #2 as if nothing happens from the first call to the second. Only time is taken into consideration.

Personally, I think watching races is more important. Seeing how the runner "goes" is more important than what the pp' info shows. Things like a jock steering a horse away from flying dirt or holding him back to secure position isn't taken into consideration and it should......Merry Xmas to all, btw.

Lindsay

12-24-2002, 06:24 PM

Dick Schmidt wrote:

"I had this same argument with Andy Beyer once, and he finally snorted and walked away, muttering that fast horses always won."

Dick, I'm confused by this story. It makes Beyer look like an idiot, but I doubt that that's your intention. According to Brohamer, you had no handicapping experience before 1985. Beyer did his mea culpa on pace in the Washington Post in 1981, and he discussed the importance of internal fractions at great length in his book that came out in 1983. In addition, he made it clear in 1975 (Picking Winners) that he was not dumb enough to think the horse with the best last-out figure always wins today. What he was doing with you, I think, was giving the truism that the best time TODAY will win TODAY'S race. From this it follows that previous final times have at least some importance in predicting today's winners. This, after all, is why Brohamer, Cramer, etc. put so much work into making final-time figures. It is also why there are many people who succeed using sheets and ignoring pace altogether. And it might have something to do with why the Sport-Stat study I mentioned earlier showed pace figures taking a comprehensive drubbing when compared to final-time figures, though, again, I am not a big believer in these studies.

Fastracehorse

12-24-2002, 08:25 PM

The words of Karls Korner are:

What are you using for a bench mark for today's races. Is today's track fast/slow ? How can you use a 17 day old variant on today's track ?

Todays variant is not important for today. Why U ask?? Because I say: We are evaluting horses based on past performnce, ala the mighty PP's.

So, what U would need to look at for todays races are any apparent biases. Biases are real. They are powerful. I admit that they are often over-emphasized like class is but they affect a horse's ability to win.

fffastt

BTW,

Merrry X-Mas.

Fastracehorse

12-24-2002, 08:51 PM

Final times ala the teletimer?? I hope that's not what U mean.

Internal fractions ala the teletimer, again they are not good indicators of pace.

Pace is important but it is highly difficult to measure without incorporating a variant: And how or why would U do that??

That is why the Beyer is such an important fig-it does include pace and a variant. And if every horse could run in a perfect world, then, the final Beyer given in DRF would be very reliable.

Final times are nowhere near as powerful as good speed figs at predicitng the outcome of races ( U might have suggested this but either U omitted this point or don't believe it ).

People who use good speed figs aren't ignoring pace. Pace is part of a good speed fig. It is just that some people think it is important to open speed figs up and put them on the operating table. I WAS one of those people but now I have innovated an adjusted fig, thank the horse gods.

Cramer's # is apparently very complex. I would put mine vs. his any day-I have that much confidence.

BTW,

I was not interested when this thread took an ugly turn by analyzing internal fractions because I understand the frivolty of this obsession.

Now don't go and say that Fast doesn't think pace is important, he does, he just has an easier and more relaible way of measuring it. Allbeit, my concepts were derived via hard work.

Oh yah, Merry Christmas.

fffastt

karlskorner

12-24-2002, 11:52 PM

It's 11:30PM 12/24/02 and we are anxiously awaiting the arrival of Santa, so I have a few minutes to continue with you.

You did'nt answer my question on variants, but you stated that "Biases are real. They are powerful"

At what point in the day do you suddenly notice that there is a "bias" on the track you are playing. Is it before the races start, after the 1st race has been run, or maybe after the 3rd race or later on in the day, maybe after the 7th or 8th race you suddenly realize that the so called "bias" exists, at that time do you adjust for this "bias", or are you carrying over a "bias" you thought you found yesterday and therefore if a "bias" existed yesterday it certainly should be there today ?

I repeat my above statement, the words , variant, pars and bias was invented by book writers to add to the mystery of horse racing.

I often fall back on my favorite slogan

Never take investment advice from someone who is working.

Wishing you and yours a Merry Christmas

Fastracehorse

12-25-2002, 01:02 AM

I'm not trying to sell U anything.

Don't fall too hard ( I often fall back on my favorite slogan- U said this ).

I didn't say it was always easy but yes, sometimes yesterday's bias carries over.

It obviously is preferential to determine if the bias is real immediately, ie, after the 1st race, or even before.

It can be done very easily with some practice: Identifying early leaders or pace pressers. If they duel and hold- U have U're answer.

Merriest of Christmas' to U again.

Don't fall la la la la, la la too hard.

fffastt

delayjf

12-27-2002, 01:40 AM

Lindsay,

Not sure I understand your point concerning Avg pace. In your example, they came out the same (energy wise), one ran fast early one late, but the same none the less.

On another note, one reason some pace handicapper might prefer the early "slower" early pace horse over the faster Closer is the belief held by some pace handicappers that energy lost early does not result in a linear loss of energy late. Ken Massa at HTR believes that "in general" its a two to one relationship, one part to fast early results in two parts too slow late. Dick Schmidt, perhaps you can offer some insight here??

Dick Schmidt

12-27-2002, 03:25 AM

Delay,

You asked:

"On another note, one reason some pace handicapper might prefer the early "slower" early pace horse over the faster Closer is the belief held by some pace handicappers that energy lost early does not result in a linear loss of energy late. Ken Massa at HTR believes that "in general" its a two to one relationship, one part to fast early results in two parts too slow late. Dick Schmidt, perhaps you can offer some insight here??"

Ken has it about right. In general, closers don't win races unless the frontrunners lose them. Of course, there are other factors. What track are we talking about? Some have such a distinct bias that certain running styles have almost no chance. I've seen Santa Anita go a month with not ONE horse coming from more than 2 lengths back at the second call in a sprint. Not one, and there were some high class closers in some Grade I stakes that month.

The big mistake most people who don't use Pace Handicapping make is believing that there are universal "rules" for pace handicapping. I used to tell people "Don't trust what you read, don't trust what I say, trust your model." It is the keeping of a Brohamer Model that elevates pace to winning heights. Otherwise, it's just another speed number.

A lot of "database guys" will tell you that pace doesn't work. No it doesn't if you stick every horse with the last line and do no modeling. I know many people are searching for a "hard and fast" method that isn't as slippery as Pace, but believe me when used by someone who does the work, it flat outperforms everything else.

Dick

Lindsay

12-27-2002, 03:26 AM

Delayjf wrote:

"Not sure I understand your point concerning Avg pace. In your example, they came out the same (energy wise), one ran fast early one late, but the same none the less."

That was my point, Delayjf. This method that purports to punish closers doubles back on itself when it comes to extreme closers. It rewards them just as it rewards frontrunners. In truth, what it does is reward horses who run uneven fractions--whether they're frontrunners or closers.

12-27-2002, 09:00 AM

Lindsay,

The difference is, no horse runs like your sample closer. So in theory it would reward the horses the same, but in reality, it doesn't happen, so it doesn't matter.

CJ

Tom

12-27-2002, 09:07 AM

I don"t see many horses win races unless they run uneven fractions. By saying that two horses are equal becasue they ran the same average pace is to say two horses are equal becasue they ran the same speed figure. This the whole secret of pace handicapping-they aren't equal, adn depending on the race match up or the track bias, one will outperform the other becasue it is better suited to today's condition.

As for the 2-1 one thing, that is the HUEY I refered to - =horse's ultimate energy yield.
Take this race - 6 furlongs, 46 112
Horse A is back 5 at the pace call and wins the race.

Pace of Race Velocity EP = 57.39 F-3 - 50.77
Pace of Horse EP = 56.30 F-3 - 52.69
(assuming 10 feet per length for simplicity)
The horse is defiecient 1.09 at the second call.
He runs 1.92 faster that the POR in the last fraction to win.
1.92/1.09 = 1.76, or about 2 times he deficiency.

In TPR numbers, Horse A is an early horse = 90/85 =175
Horse B is a closer = 80/95 = 175. Which is the more likely winner at Turf Paradice? At Fairgrounds? At Belmont? At Calder?
On the turf at Churchill? In the mud at Saratoga?
In a 5 furlong dash at Hollywood?

I think final time is acknowledged to the degree that to be a contender, a horse should be within 5 points of the top TPR or ranked 1 early or late (although these guys usually are within the 5 points anyway). It is just that final time alone is not always the answer.

More Beer!

delayjf

12-27-2002, 09:52 AM

For anybody who wants to read more about late pace effectiveness in handicapping, I'd refer you to
www.homebased2.com
click into HTR and check one of the recent newsletters that deals with late pace.

Ken Massa also addresses other pace ratings (including avg pace) as to its effectiveness. For those who don't use or believe in its value, might surprize you.

Lindsay

12-27-2002, 06:56 PM

CJ wrote:

"The difference is, no horse runs like your sample closer. So in theory it would reward the horses the same, but in reality, it doesn't happen, so it doesn't matter."

I stated that I was exaggerating for effect. I wanted to show the artificiality of averaging velocity numbers. Beyer gives a more plausible example on page 139 of Beyer on Speed. Here's another example: LA to San Diego is 100 miles. If I drive the first 50 miles at 60MPH and the second 50 miles at 70MPH, what was my average MPH for the trip? What would "average pace" say it was? Why are these two numbers different?

Tom wrote:

"I don"t see many horses win races unless they run uneven fractions."

If running uneven fractions is the best way to win races, why give extra credit (through an artificial velocity average) to a horse who ran uneven fractions? The advantage of running the uneven fractions will be reflected in his final time. To give extra credit is to double count. The answer, of course, is that running uneven fractions is a disadvantage, which is why the key question is whether a horse will run more-even fractions today than he did yesterday. This is the central question. Everything else is artificial and meaningless.

"In TPR numbers, Horse A is an early horse = 90/85 =175
Horse B is a closer = 80/95 = 175. Which is the more likely winner at Turf Paradice? At Fairgrounds? At Belmont? At Calder?
On the turf at Churchill? In the mud at Saratoga?
In a 5 furlong dash at Hollywood?"

Here I agree with you, Tom.