Velocity and Energy

[km]

T'Brown asked the questions below in the thread about the Derby files. I think that discussion has run its course and his question will be better served in a new thread here.

Tom, you'll be happy to know i'm bringing back the PRAT (whole number velocity figs) for the seminar update this July. Hope to see you and Bill there again.


Which leads to another question - when choosing manual pacelines, which screen is the bet to use? I would say using Ap on fps, based on this thread, would be betther than Fig2???

Digging deeper into velocity:

Two horses, A and B , both Early horses.

A has AP of 53.45, F3 of 51.22 and a 52.89% EE
B has AP of 53.34, F3 of 52.24 and a 51.03% EE

Can B adjust his EP enough to run in h is comfort one and still outfinish A?

B will have to increase his EE% to keep his EP in line with A, but will it kill him or does he have enough left?

If stuff like that could be somewhat mechanized, WOW! Kind of a "what if" template?
PL-0
The screen choice is a personal thing, but the focus with PL-0 should be on finding the line that best pre-positions the horse against today's field. For me, that usually means searching for a good effort that mimics today's dist/surf/class situation, so the figures don't matter as much as the running line. I try never to use < 6 line if today is 6.0f or further, and to avoid wire to wire romps in maiden claimers. Turf sprints should always be avoided for dirt races and turf routes.

Example race "A" vs. "B"
If this is a dirt race for older claimers, "A" has all the advantages. "B" is forced to run harder early to keep up and that will sacrifice his final fraction. Or so the theory of 'unrecoverable energy loss" goes. But if this is a grass, 2yr or 3yr, stakes race, etc, the analysis would be different.

[tbrown]

PRAT! Whoopie!
Guess I'd better start working on a new tag line!

You know what, Ken, I STILL have an old version of HTR and use the X Screen - the one with the toggle for pacelines and the whole number Crammer pace figs. I love using that one to seperate early runners.

[km]

In another thread Donnie says:

I just use the figures available with no consideration of variants being applied. If there is a variant applied to Ken's number, then I assume that like many others who construct numbers, he probably uses half the total variant at the PAC call. Ken may add more on this since it is his number.

HTR has always used a home grown variant algorithm that can be applied with our without the availability of a 3rd party speed figure/variant. Cramer numbers strengthen that formula, but it is not dependent on them.

With the velocity numbers, variant application is based on the segment of the four primary fps ratings: Fr1-Fr2-Fr3 and E/P according to the amount of ground covered on its own as well as a percentage of the whole race.

One of the great aspects to velocity numbers over speed figures is their resilience to variants. Usually the rank of the horses in the various fps ratings remain the same no matter how much we tinker with the adjustments.

This is why I was not that concerned about the changes to the Cramer numbers in the Derby file update.

The PAC and PER ratings are derived from the same methodolgy as the fps. But the ratings have to be manipulated to conform to each other and the EPS/SOR indexes (Quirin style). So they are more sensitive to the variant because the numerical range is tighter (typically 080 - 115).

But the PAC/PER are not nearly as sensitive as final time speed figures which are directly adjusted from the variant and can be impacted dramatically if there is a major change as we saw with that file difference. Final time speed figures have always been a secondary aspect to HTR. The velocity numbers are the backbone.

[hracingplyr]

I for one really like it, would hate to see you get rid of it

[km]

FIG2 is untouchable.

I might replace FIG1 or PP1

[Donnie]

Glad to hear FIGs2 is untouchable.
Before I throw out a suggestion....are there many subscribers who use FIGs1? Better hear from them first....
Now a suggestion Ken--
Would it be possible to fit the DFIGs chart next to the FIGs2 chart if you removed everything to the right of the chart in FIGs2? This could possibly replace FIGs1...unless there is a chorus of "Boo!"'s. Not looking to stir up controversy.
Many times the patterns are the same, just off aways from the par line a tad further. But what I am seeing is there is a major change when the horse has run a higher pace number than maybe what the horse ran in the past. Some of these horses can be immediately discounted, probably because the inflated number was caused by them "drafting" with the leaders. But, like the example I posted yesterday to compare the two FIGs with one another in another thread, even with the inflation by the pace numbers, the #3 at AP was much the best in DFIGs and ran away from the field in the stretch...paid $8.00. Dunno if anyone else would be interested, or even if it can be done (space considerations and such), but I would pay for you to custom code such an animal. I would liken it to people who like to compare PL choices back and forth. But I think there may be room on one page for both. Just a thought if you have dead area and want to experiment. No problem sharing DFIGs with others if there is an interest......

[tbrown]

Donnie, I love the DFIGS.
What does the "D" stand for? :rolleyes:

[Donnie]

if you want a copy, let me know...I'll send the latest version along....to the far right I had Ken drop the weight from each race and put in the Purse from each race!

[MVM]

Des Moines;)

[km]

I can replace FIG1 with the DFIGs easily enough. But i haven't tested it yet to see if they can out perform the PAC by itself. I'll report back on the DFIGS in a side by side test with Random#, PAC, PER, Cramer Speed Fig (sheet and regular).

Heck, some speed fig methods people have sent me don't even out perform random numbers any more. To test them, i'll use all 5 PL modes and look at the Win% and ROI.

I realize that the key to the display is an understanding of the form-cycle and pattern. But if a rating (and its top ranked horses) can't even beat random numbers in a mechanical test, i won't add it to HTR.

========== Read below for interesting story about Speed Figures ==============

Jim Bayle "Sports Stat" was running a side-by-side test of all the popular speed figures around 5 years ago. Beyer, Rag, T'Graph, BRIS, TrackMaster, and Cramer, among others. Each participant was to submit their last race fig for each horse in all the races at designated tracks. The side-by-side test was to find out "who really has the most best speed figure" and would be conducted for a set period of time and include at least 1000 races.

They all promised to abide by the unbiased results which would be published in a book and obviously create a lot of buzz on the bbs,etc. It would have been a big deal because everyone brags that their figs are the best. While no one expected last race figure to show a profit, it would have been a reasonable comparison as to the accuracy of the ratings at face value.

Suddenly, near the end of the test period. Bayle - bailed out! Cramer was sure he was winning the challenge because he was tracking his figures vs. a couple of the others. He called Bayle and received evasive responses "things didn't work out" - "had to stop for personal reasons". The rumors started and Bayle dissappeared. He has not been heard from in public since.

The most obvious theory is that one of the participants was told that his ratings would finish poorly in the test and that he paid off Bayle to stop the presses. Who had the most to lose? Would it have hurt the reputation of one of these vendors so bad that they were willing to pay silence money?

I know Donnie would never do that - but on the other hand, a couple of 20 oz grilled premium steaks with all the fixins might just do the trick for me and keep the DFIGS alive!

[Donnie]

Well Buddy...you just jinxed me!
Heck, some speed fig methods people have sent me don't even out perform random numbers any more. To test them, i'll use all 5 PL modes and look at the Win% and ROI.
Hopefully everyone understands this is just another use of the numbers....I don't know if these are meant to outperform chance either....I just wanted a number that is impacted by the Pace figure. When the PAC number is high, the number goes ballistic....but like the example yesterday, that #3 in the 3rd at AP finally found a softer pace to run against....many would have thrown him out on form. He actually probably got a breather at AP. I personally look for horses just like this....everyone is well below par with the DFIG, one horse although doesn't look the prettiest on form, actually is now running with a much cheaper field.
Please put up on the Header....Caution!! Use at your own risk!!
They will always be alive in my own custom program....whether they pass muster for the public.....dunno? I may be joining the MIKE fig here huh?? :cool:

[Lee Carson]

Since I have only been around since 04' can you old timers define what those term mean?

On another note - I have been noticing on the VEL screen, when you run your eye down the FR3 column until you hit the best FR3 to that point, and go across to the FR1 column and that fig isn't too far from the top, you have probably found the #1 FX horse. I haven't tested the FX as I write this, but simply locating the highest FR3 and corresponding FR1 as described above seem to win often. (Sometimes. coming to an improved FR3 and looking perhaps down one or two more lines where a Large fr3 appears - that is the one to use if the FR1 is not too far out. Too far out would seem to mean where thare are many 59's and 58's and the corresponding FR1 is in the low 58's, as an example)
Two items to consider:

1 How would one go about testing for the above decribed Fr3 - Fr1 combo? and,

2. How about Ken doing a study on the FX?
Lee

[BillW]

Since I have only been around since 04' can you old timers define what those term mean?


Ditto here, I've been around since Tuesday :D

Bill

[km]

Lee - good idea, i'll run some F/X stats in the June newsletter. F/X = (Fr1 + Fr3) / 2

PRAT = an acronym for whole number velocity figures
example: Fr1 - Fr2 - Fr3
fps = 58.00 - 56.00 - 52.00
PRAT = 95 - 75 - 35
(that is not an accurate corresponding PRAT figure, just an example)

Some people prefer PRAT over fps because it is easier to eyeball and determine de-acceleration. The 1-2-3 ranks remain the same for Fr1, etc. with either approach.

DFIGS = Donnie's custom indexed number based on PAC and adjusted w/ Cramer sheet fig. Placed into a FIG type graph format. This number is not found in HTR software ---- yet!

MIKE = a figure that MikeDee thought up combining K and Pscan. Turns out the MIKE could not out-perform the (K). Shown in the TLC screen, but 'Dee would prefer all references to him be dropped from that rating!

Donnie, I'm not knocking anyone's methodology. Won't argue with success.

As an example of eating humble pie and preventing me from being a "know it all". Yesterday in the 9th at BEL there was a horse that was 0 for 35 in Mcl and the trainer was 0 for the year with the minimum 050 rating. A user emailed me and said, "this horse gets the highest PAC, PER and CLA ratings, shouldn't he be used in the p4?" I said "no way", a 5yr maiden has lost 35 in a row with a zero trainer - toss him. The horse won easily and paid $13.

[tbrown]

He'd have paid $4.20 at Finger Lakes! :eek:

[Donnie]

I always try to remember how stupid some of these animals really can be. They don't know that they are running for a bottom level trainer.... we were on the opposite end of that scenario 2 years ago when we took 7th at the GC Summer Classic.....I played a horse who was going for a trainer of 0-24. Paid $24 or something.

....not to mention..."every dog has his day!"

Wasn't taking your comments as a knock at all! I will probably end up making the same request as Mike...."What? Who? Me? I have no idea what that DFigs number is or where it came from......" :o

The people who use it will probably have a good basis in reading the sheets, or are already using the FIGs2 successfully. Hopefully people will share their successes here on the board if you do include them. Success begets success.

[Donnie]

ohhhh...in case you were wondering....no they haven't paid my mortgage yet....so don't bet the farm on them! :D

[njcurveball]

Ken,

Interesting story on Sport Stats. I still remember their "super" play which was basically use the 1 or 2 post in route races with a date requirement and something else. Many "studies" shown here on the BBS outperforms them on their best day.

My 2 cents is that ratings have to be complimentary. Perhaps it is that Brohamer modeling thing, but I really do not want to track 6 or 7 ratings that measure the same thing to see which is the best if they are all close.

The best model comes from different factors that perform well at different times. F1, F2, F3 can make the perfect model at some tracks. Is F1 better or F2 or F3? That is why we keep the model.

At the abbreviated Atlantic City meet, first fraction was deadly in their turf routes. many of these winners had poor final time in their last race. It is great having the best number, but if the horse will still be rated 5th or 6th there won't be much benefit.

I think HTR has a great mix of traditional factors, as well as, bells and whistles to add even more profit. I am sure whatever you add will be tested thoroughly and make for a better product in the end.

Thanks for your professionalism!

Jim

[hracingplyr]

Donnie for that idea on using the pac and figs 2 idea, nice 30.00 horse yesterday at belmont

bob

[Donnie]

Bob,
I would bet you will find a few more, especially if Ken puts it into the FIGs2 format for everyone....congrats! Nice score!
Went out to check out the race....he was definitely playable!! In the FIGs2 you would assume he had paired up his last 2....in DFIGs he paired up the 2nd and 3rd back and had broken thru that pair up. A definite play at 15-1!! I attached the DFIG for that race.

[Donnie]

....again...this was a case of improving pace numbers affecting the "sheets" figs.

3rd race back.....ran a 93 pac but finished well (21)
2nd race back....ran only 90 pac, but finished strong again (18).
last race....ran a 100 pac, although his running line shows a HUGE tiring effort, his final time equaled the race before (18)! Maybe the public thought there was something wrong since they put him on the turf (tender feet??) and then he doesn't run.

[jbshoulda]

Donnie,
Don't you get the same results just by adding the pac & per number and then stick it in the graph format? I've had this in my program but never had Ken put it in the pp's for me. If you just put the "total" (pac+per) after these numbers in the ppq screen you would get the same results I think. Or just use the fig screen and then you get the graph visual.

[Donnie]

John-
you may get the same type of results, I have never added pace and per (??). But the patterns from the normal Figs screen can be quite a bit different from the DFigs screen. Many times this is the function of a high pace number making the difference. I like the visual of the graph layout in the Figs screen. Many horses can be thrown with a glance.

I just went to look at the numbers in the PPQ screen....

AP R1 Sunday....here is the first horse's (Resteam) last 3 numbers.....

PPQ (PAC + PER) (John...please advise if I am looking at the wrong numbers here!!)
Last Race...180
2nd back....190
3rd back.....197

DFIGS (PAC - CRMR Speed, the higher the number the better)
64
74
81

FIGS (Straight up from HTR2, lower the number the better)
27
20
20

It appears you do get the same patterns from PAC + PER, but the pattern is different from the Figs screen. Those two 20's aren't showing the 7 point spread from either the DFigs or your method of PAC + PER. Being a visual person, it is a HUGE benefit to see these in the graph view. Like minds....I guess!! :D

[jbshoulda]

You got it. Basically just use the fig2 sreen as the pac+per #'s are there also. If you substitute the dfig # with the pac+per # in the graph and move the cramer # you have the same spread as your numbers but on the graph. I agree it would be a better visual that way.

[jbshoulda]

Also like the purse value you have. It paints a good picture of the placement by the trainer.

[Donnie]

...you mean you don't watch weight changes??:eek:

[jbshoulda]

Just the 15 #'s up and down on me

[km]

I'll put the DFIG ratings into FIG1 slot in HTR2 and send to Rick asap so everyone can have a look and beta-test it. Includes the replacement of the weight column with the purse value as mentioned above. If everyone likes it, we'll leave it in there forever.

Also, will put the PRAT ratings into the slot for PP1 into this beta test version.

File name = HTR2BETA.exe

The Robot and Export will NOT function in the beta version, so continue to use HTR2 for all other screens and for the regular FIG1 and PP1.

Will run a test for the next newsletter that attempts to compare the accuracy of the figures --- but not the ability to assess form-cycle with them. Will test last race figure side-by-side to see which one, if any, has an advantage.

DFIG
PAC
PER
PAC + PER
Cramer Speed Fig
IMP (Total)
PRAT (1+2+3 rating)

The test will be non-maiden dirt races only, 6.0f - 9.0f. This avoids unratable horses. Which figure do you think will show the best Win% / ROI ?

[km]

ps: i'll set up the beta sometime next week, not available right now, will post announcment here.

[Donnie]

Ken--- :cool:
John-- the reason I wanted a 2 digit number was because when Ken originally coded this for me he was showing me the leading zeros (065) and when the numbers were close in proximity to one another, it was harder to read. The 2 digits are easier on my eyes. Another price we pay as we age. <sigh>

[tbrown]

Holy Cow! Christmas comes early this year.
Awsome, Ken - thanks!

[Huguenot]

Donnie et al,

the number on the far right -- where the weight used to be -- what does that do? Is that the "Mike" number?

[jbshoulda]

Purse value