Beyers make no sense

[Amazin]

looking at the form for BMF 8/16,race #2 you have a horse named Barefoot Caity who on 8/2 at S.R. 9th race finished 3/4 length behind the winner's time of 1:40 and 2/5 for 1 mile for a beyer of 44.
Next entry is Foolish Times who also ran at S.R. on 8/2 but in the 13th race finishing 3-1/2 lenghts behind the winner's time of 1:40 and 4/5 for a beyer of 46.
So Barefoot Caity ran approximately a full second faster than Foolish times at the same track on the same day at the same distance and class but in a different race yet is rated 2 points lower via Beyer's.I dont think this is a misprint as I have seen this before.
Does anyone have an explanation as to this gross misrepresentation or is this just another proof of the unreliability of Beyer's.

[BillW]

Amazin,

The most ingenious use of Beyers I've seen yet is people looking for inequities and playing those (I believe a couple of people on this board do this).

Bill

[freeneasy]

its probably due to a varient that puts the faster race behind the slower race timewise. it can get screwey I know. I still use em cause their about the closest thing to the figures I used to make

[Amazin]

Free

DRF gives both races a 22 variant.These 2 races were run approxametly 1-1/2 hours apart on a clear day.Doubt there would be a change in surface in that time to a Beyer making person that would cause him to rate a horse who is slower by five lengths as faster than the the other unless a tidal wave hit that I don't know about.

[Tom]

Two possible reasons com eto mind

1. Split variant
2. Error
3. Tinkering to make figs fit.
Hmmm, that's three things.

What track is SR? I'll check this out. I am one that looks for Beyer errors and finds them more than you would think. CJ does this too.

[andicap]

Maybe our friend Mark ("we kid because we love") at the DRF can help us out here.

[Amazin]

Tom

S.R. is Santa Rosa,part of the summer fair circuit that they run here in Northern Cal.It is only a two week meet.

[freeneasy]

SR is Santa Rosa in Cal.
could be that ol Andy had a few to many rips off the Jack Daniels when he was making the vars. maybe he was having a bachelor party and wasnt even gettin married, strippers n all, btw is that guy married, and how much does the drf pay him for those numbers of his. personally I dont agree with the way he makes his variants and speed # as I feel a few more ingredients could put out a tighter fitting number. But hey if I knew anything about how to write a computer program Id be using my own numbers by far. but since I dont, ol Andy is all I got. I tried it once a long time ago, making up varients by pencil and paper, for all the tracks in the nation, but after three weeks of going crosseyed I turned into a wide eyed zombie, then If I remember correctly I woke up in a sylum for the insaneacappers, O but once there I learned how to knit, make ice cream stick houses and tie my shoelaces all over again. I escaped tho and Lordybe here I am. hehaw

[Derek2U]

Just 1 thought on the Beyers: I think the Beyers are too Fitted
to Who the Horse is. Does the horse have Back Class?
What race class is the horse at? Now of course, Beyers manage'
to get away with that because almost always 20K winners will
run faster than $10K winners, but I would like to see ONLY
speed figs and not so much mingling w/ class structures.

[Jaguar]

Amazin,

the title of your post says it all; the Beyers are off.

The best work I have seen on correcting the Beyers was done a few years ago, by the developer of the "Stretch Run" handicapping program. He is a mechanical engineer by training and a sharp handicapper. He has moved from Kansas City to Florida and I do not have his new address.

All the best,

Jaguar

[andicap]

The guy I know who developed "Stretch Run" lives in Washington State. Are there two "Stretch Run" programs? The program I'm familiar with is rather recent.

[cj]

As Tom stated, I look for Beyer errors and try to exploit them. For future reference, the NoCal fair circuit (SR, Sol, Pln, etc.) consistently has the worse made Beyers of all tracks covered by Simulcast Weekly. I don't know why, I just don't take ANY figures earned at those tracks very seriously.

As for the races in question, the Beyer variant for the first three routes that day, race 9, 10 and 12, was around 23, while the 13th was given a 12 for no apparent reason. The speed figures should have been around 42 for Barefoot City and 35 for Foolish Times. That said, speed figures don't mean too much at this lowly level.

It also shows you the Beyers are changed after the fact, as the winner of the Barefoot City race, Whirlwyn, was originally given a 43 in Simulcasat Weekly, and now would appear to have been changed to 45 based on the fact Barefoot was given a 44, higher than the winner originally received.

CJ

[Lefty]

The "Stretch Run" I remember was authored by Dan Pope of Royal Turf. Don't know where he is now.

[Jaguar]

Dan Pope is the guy. He developed the original Stretch Run method, selling it pre-programmed on a Sharp handheld.

Good program, but you had to pick the right paceline. Also, his speed figures were excellent- so good that I always thought Andy would catch on and improve his figs-.

Curiously, Andy never bothered to wake up and smell the coffee.
I have always said that if Andy had asked Davidowitz to re-do the numbers, the resulting speed ratings would be dead on.

If you want to see superb speed and pace ratings, check out Cramer's numbers in Fastcapper, -uncanny.

Since Dan Pope left the market, there is another guy out there, calling himself "Stretch Run". You gotta have considerable intestinal fortitude to use another program's name. Amazing.

But, then again, maybe Dan never registered the name.

All the best,

Jaguar

[MarylandPaul@HSH]

Amazin, that race sounds like a good illustration of the danger of relying on "projected" times when making figures. As far as I know, Beyer believes that making varients based on projected figures is a more accurate approach than the traditional method of comparing times to pars. To echo CJ, I've seen Beyer varients that change several times throughout a card, with no changes in weather. Seems like each race is projected independently on some days.

It sounds good, but it has a way of compounding errors. You make one bad projection, then when horses run back, you make projected figures based on the bad number, and so on. Also, if your projections tend to be too forgiving, or not forgiving enough, you'll find your numbers "creeping". Pretty soon, you have $5k claimers running the $8500 par.

When I made my own numbers, I'd always start with my par charts, then use projections to help clarify confusing days. You'd better be sure when you're giving a Maiden Claimer a number that would win a NW1. Interestingly enough, one of the great advantages of making your own numbers is that you can put a big red "?" next to a day, reminding yourself to proceed with caution. Can't do that when you're selling them :)

I don't know how Andy does it now, but years ago he had 7 people making figures for every track in North America. Doesn't leave much time for careful consideration.

[Derek2U]

For simplicity, let's assume ALL tracks are the typical 1-mile layout.
(Forget about all 1-1/8 ovals & the Bullrings). OKay? Now,
measure ALL speed, at ALL distances, using the same Chart.
That's as close to measuring SPEED as I can think of. Now,
someone develop a measure of CLASS, 98% independent of Speed. The trouble I see with Variants (as I know of them) is
that they co-mingle Speed with Class in weird ways. Beyer's
Projection method included. Then, we could have Very Decent measures of: Running Style x Speed x Class as Free of each other as possible. To me, that's the key to getting the field down to 2 horses. Now anyone with a large database could easily do
this ..hehe ... The 1 Beyer idea I think is 100% essential for
Speed Figs is this: The Value of a length varies with Distance ...
but I do think his values OverValue a 6f race.

[JimG]

All this talk about the Beyer speed ratings got me wondering. Anyone on the board subscribe to Henry Kuck, Thorograph, or Ragozin figures with regularity? If so, why?


Jim

[Jaguar]

JimG,

Used the Sheets for a while, Thorograph and Ragozin, went to 2 Ragozin seminars in Manhattan.

Eagerly searched the day's entries for horse's with those magic 10 and 11 speed ratings- which happened to appear 7 running lines ago. Horse's most recent race is a problematic 14, run 14 days ago.

What's that horse going to do today? Who knows?-- And the sheets aren't going to grab the form cycle patterns or the trainer patterns and tell you what the horse is likely to do today.

The Ragozin algorithm is excellent, but the Sheets aren't truly a predictive tool, unless you're the Amazing Kreskin and can predict the future.

Now let's see, where did I put Kreskin's phone number?

All the best,

Jaguar

[Derek2U]

I have looked at both Ragozin's sheets & Jerry Browns sheets,
but I have NOT used them. At meadowlands it's hard not to
sit next to guys using them & so my limited knowledge of
them is this: they stink. They both have a static view of racing;
the PAST = NOW. Also, they both Over-Adjust their SpeedFigs
for every conceivable situation. And both of them, and their
disciples, rationalize race outcomes NOT predicted by their
Figs. The word Prediction contains the Prefix PRE --- a fact that
both them players seem to forget. And I'm not sure if any program should have a algorithm for "going wide" or "bumped" ...
that's seems like more of an artful, brainy thing to do after more
analysis. I just flag such trouble races after I'm sure a Horse
deserves that break. With Figs, it's just another example of
tinkering to me.

[Lefty]

Is someone still putting out Kuck's figs? Pretty sure I read that he's dead.

[andicap]

www.woodsideassociates.com

[Lefty]

andicap, thanks, most interesting.

[Amazin]

One thing I got to say to Andy Beyer is that his complaint about not getting overlays anymore due to his Beyer numbers being out in the open to the public is unreal .On the contrary,they seem inaccurate enough to regurally attract money to a false hi Beyer horse and create an overlay on an underated Beyer horse. Can't bame Beyer himself because I remember reading an article a few years ago about the process Beyer makers use.Not quite the original process he put forth in his book. Maybe that's how he makes his money now at the track. Betting against Beyers he knows are inaccurate. And why should he criticize how they're made?He's getting paid for them.

[Barremill]

Derek2U,
i think it is the SPEED of the race rather than the distance. at Delmar
where the running time for the race might be 1:09 and at Suffering Downs
in the winter running the same distance at 1:17 (no exaggeration!)
will give greatly disparate readings in computing the 'length' of the horse. in each of these races.

barremill

[superfecta]

Originally posted by Amazin
One thing I got to say to Andy Beyer is that his complaint about not getting overlays anymore due to his Beyer numbers being out in the open to the public is unreal .On the contrary,they seem inaccurate enough to regurally attract money to a false hi Beyer horse and create an overlay on an underated Beyer horse. Can't bame Beyer himself because I remember reading an article a few years ago about the process Beyer makers use.Not quite the original process he put forth in his book. Maybe that's how he makes his money now at the track. Betting against Beyers he knows are inaccurate. And why should he criticize how they're made?He's getting paid for them.
Your right about Beyer betting against figs he thinks are wrong,he mentioned this factor in his book Beyer on Speed.Many times the Beyer fig doesn't work,all of the time it doesn't tell you how it was earned.Pace is a integral factor in determining how fast a horse will go or can go.He tries to incorporate pace into his fig scale but when it doesn't work,he blames the theory of pace,not on his scale.But if your scale isn't working, everything you place into the equation isn't going to work.
I saw this litle gem recently-same track same day same distance.
6 furlongs final time-1:12.92
Beyer- 77

1 1/2 hour later:

6 furlongs final time-1:12.89
Beyer- 97

no noticable change in track or conditions,only change I can see is the class of the horses involved.First race is for 20k claimers fillies and mares.
Second race is for 75k optional claimers.
Now if your speed fig factors a class element,its not a speed fig,its a handicapping program.I know the varient is subjective to who is making the fig,but you can't tell me these horses are 20 points apart.Then your telling me every point then is worth .00015 of a second?
If these two horses met ,the classier horse would most likely win.But I don't want my speed fig to tell me that,I want to know how fast the horse ran compared to the track surface.

[aaron]

I have been using the xtra's from the book "Blinkers Off"with some success for about 2 months.The numbers are Beyer like and incorporate patterns using pace numbers.They seem to be accurate and if you wait for overlays you should do okay with them.I'm still learning and trying to be more patient,but I really believe these numbers give you an edge.-

[andicap]

When you say "OK," is it because the cycles aren't working? I have read the book and he's big on betting on horses with "Pace tops."
Do these only do OK? Are you not waiting for good value? What's the problem?

[aaron]

I'm not having any problem winning with pace tops,but if you read the book you should realize that a horse that has a new pace top still has to be competitive on the final numbers.There are races where there are more than one pace top and there are also other patterns that must be considered.
The xtra's are not a magic potion,but I consider them the best new product put on the market in years.
This is one product you can win with and it is fun to use.