PDA

View Full Version : Bris Prime Power number


Milleruszk
12-22-2003, 11:20 AM
Do any of the punters on this board pay attention to the Bris Prime Power rating? The info given out by Bris is that a horse that has the top Prime Power rating of 3 points higher than the any other horse will win 39% of the time; 6 points higher 46% of the time; 10 pts 55%! The Prime Power rating takes Bris Speed, Class, Pace, form, weight, distance are all combined by a computer algorithm into an rating.
I don't imagine you will find many 2-1 horses that 10+ points higher most of those 10+ horses will probably be 1-5. However, the Prime Power rating should occasionally point out horses that the public overlooks? Anybody track these? Thanks.

Tom

JustMissed
12-22-2003, 11:58 AM
Miller, I posted this at the software area 12/10/03:

" Prime Power Strikes Again!!!
My last post got me curious so I am showing last nights results from the Mountain:

R1 #9 1st Prime Power paid 4.20

R2 #4 2nd Prime Power paid 3.60

R3 #2 1st Prime Power paid 4.40

R4 #7 1st Prime Power paid 7.00

R5 #3 1st Prime Power paid 6.20

R6 #4 6th Prime Power paid 25.20(ouch)

R7 #8 3rd Prime Power paid 26.40(way to go prime power)

R8 #8 4th Prime Power paid 36.40(way to go prime power)

R9 #4 4th Prime Power paid 11.00

R10#4 1st Prime Power paid 6.60

Of course this is a small sample but I have been using TSN Ultimate PP's w/Quick Play Comments for over a year and the prime power figures work pretty good. Tuesday night was foggy and muddy as it gets but the prime power top four horses won 9 out of 10 races. Not too shabby huh.

I won't comment any further except to say that if your current software isn't getting you the top 4 contenders that win 70-80% of the time then you probably got a filter or switch or something set wrong and probably need a checkup.

JustMissed"

Miller, You can see from the payoffs that Prime Power can put you on some good payouts.

I used to keep up with Prime Power success on a regular basis but after I was convinced of its power I stopped keeping track.

You should print off Bris/TSN pp's for a week or two of your favorite track and check it out. I bet you will find that the winner comes out of the top 4 Prime Power horses 70% or more.

If you are a throw out capper, start with one half the field plus one and throw out the lower Prime Power horses. That should give you about six to look at. Just throw out two more and you are probably looking at your win & place horses.

Let us know what you think.

JustMissed
:)

bettheoverlay
12-22-2003, 08:56 PM
I've used the Prime Powers as a guide for about 2 years. Its tough to make money on the top choice but if it is 4/1 or up I will give it a look. I usually concentrate on the 2nd and 3rd choice at 6/1 or better and occasionally look at the next 3 at 8/1 and up. I ignore the numerical gap between the horses although I have noticed a top horse with an advantage of 5 or more is dangerous to bet against. They usually pay $3.40 or so. Also the gaps are much larger in turf, maiden and stakes races.

I look for horses that have good recent form or good early pace #s mostly. I played Hollywood this weekend. The 2nd horse won the 10th on Saturday and payed $17.20. In Sundays 5th race I got real lucky with the 3rd horse who was in good form but never on the turf before. Payed $34.60 although 6/1 in morning line. In the 7th I hit the 3rd horse Unfurl the Flag, in good form, paid $23.60. I passed up the 3rd choice in the G1, a maiden off a layoff who payed $14.60. I also had a bunch of losers but it was a good weekend. They're not all like that unfortunately.

I do much better at the major tracks probably because I emphasize good recent form. The #s have simplified the handicapping process for me and simplicity (and finding overlays) is my goal.

Derek2U
12-22-2003, 09:28 PM
hello Bet ... I think you're right in ignoring the numerical diffs
between the horses; the rankings are really what matters, even
though I'm NOT acquainted with your particular approach. The
reason 4 this is purely a math-measurement problem. Also,
because RANKINGS & NOT SCORES matter, going to the 2nd/3rd
choices for value makes great sense. In reality, the diffs between Top Rank & Others doesnt matter anyways. Hence,
is the Eventual Winner ranked among the top or not thats what
matters. In my approach, which finds plays in 6/10 races, the
Winner is right there among the top 3 ~~92%.

bettheoverlay
12-22-2003, 09:46 PM
I neglected to mention in my original post that a few months ago I started downloading the DRF files into the PP Custom Generator that allow you to print the PPs in Prime Power ranking order. Has helped my focus for sure and hopefully caught a few more winners I might have passed up.

nomadpat
12-22-2003, 10:35 PM
Just for kicks, I queried my CRC database to see the results of this question about the effectiveness of BRIS Prime Power. Here are the results:

All Dry Dirt Sprints: 868 races.

Horse ranked 1st win %: 31%.
Horse ranked 1st or 2nd win %: 52%.
Horse ranked 1st, 2nd or 3rd win %: 62%.

Let's add some profitability numbers to this:
Betting the top Prime Power resulted in a return of $1.60 for every $2.00 bet.
Second Prime Power choice returned $1.55.
Third Prime Power choice returned $1.40.

Non maiden dirt sprints: 451 races.
Horses with an advantage in Prime Power, listed by amount of edge, win %, ROI and ITM %. Number of horses listed in ()
1 - 33% winners, -20% ROI, 70% ITM (383)
2 - 33%, -21%, 71% (299)
3 - 35%, -19%, 75% (210)
4 - 38%, -19%, 74% (138)
5 - 42%, -15%, 79% (106)
6 - 49%, -6%, 81% (63)
7 - 49%, -12%, 81% (47)
8 - 41%, -29%, 76% (29)
9 - 39%, -33%, 74% (23)
10 - 50%, -16%, 79% (14)

The way to read the above is that in 451 races, there were 210 horses that had a 3 point edge in Prime Power. They won 35% of the time, for a -19% ROI and ran in the money 75% of the time.

Enjoy
:cool:

VetScratch
12-23-2003, 12:42 AM
Nomadpat,Horse ranked 1st win %: 31%.A rating like top Prime Power should be like best Beyer and other obvious positive factors... characterized by a very high Impact Value but unprofitable when played blindly in every race.

At 8.5 starters per race (about right for CRC), a win percentage of 31% yields an Impact Value of 2.64. This is a respectably high I.V. for any ranked data factor that is present in each race. For your 868 race sample (all dry dirt sprints), some winners must have been 1st-time-starters (without a rating), so the I.V. would need tweaking to measure precise impact among horses with ratings.

When I last looked at top Prime Power in about 4,500 races, it was about as valueless as the betting favorite... both returned a $2Net "around" $1.65, which is less than a randomly selected wager.

However, one virtue of Prime Power ratings is that they are available far in advance of the final odds which determine favoritism. This suggests that rather than focus on ordinal rankings for Prime Power, a computer study of Prime Power frequency distributions in the manner final odds have been studied postmortem may produce some financially useful findings... with the goal of identifying races with the greatest probability of offering attractive win payoffs and/or exotic payoffs.

dav4463
12-23-2003, 02:36 AM
How does using the Prime Power number compare with using the Daily Racing Form consensus or any other public handicapper consensus? Has anyone ever done a study of the DRF consensus win percentage, profitable odds, etc. I know Mark Cramer did a study of the second or third consensus choice vs. the top choice as long as at least one handicapper picked the second or third consensus choice on top. I believe the second and third choices showed a profit at 4-1 and higher.

JustMissed
12-23-2003, 10:17 AM
Nomad, You said:"The way to read the above is that in 451 races, there were 210 horses that had a 3 point edge in Prime Power. They won 35% of the time, for a -19% ROI and ran in the money 75% of the time."

You guys with data bases are dangerous.

All the other post above were about how players use Prime Power to select contenders and/or pass or play races and then you come along with your study doesn't mean squat.

Your study makes the FALSE assumption that a player would play every race in your sample. This just does not happen in the real world and you know it.

I'm sure your little test and post was for the purpose of discrediting Prime Power and its users and I'm sure you are like some other here who get a real kick out of typing "-19%".

JustMissed

:(

lsbets
12-23-2003, 10:27 AM
Nomadpat,

What I would be curious about is if you broke those horses down into odds ranges. I would imagine that a large number of them went off at less than even money, but what would the results be for horses over 1-1, 3-1, 4-1 (how many would even be 4-1 or higher). I would be curious as to the ROI for the horses with a 3 point edge as their odds get higher. That could be very useful information.

BillW
12-23-2003, 10:30 AM
Originally posted by JustMissed
Nomad, You said:"The way to read the above is that in 451 races, there were 210 horses that had a 3 point edge in Prime Power. They won 35% of the time, for a -19% ROI and ran in the money 75% of the time."

You guys with data bases are dangerous.

All the other post above were about how players use Prime Power to select contenders and/or pass or play races and then you come along with your study doesn't mean squat.

Your study makes the FALSE assumption that a player would play every race in your sample. This just does not happen in the real world and you know it.

I'm sure your little test and post was for the purpose of discrediting Prime Power and its users and I'm sure you are like some other here who get a real kick out of typing "-19%".

JustMissed

:(

JustMissed,

Are you dataphobic? nomadpat only posted data. Data by itself is harmless (i.e. it doesn't bite) if you came up with a conclusion you consider dangerous based on that data, don't shoot the messanger, examine your conclusion.

Bill

JustMissed
12-23-2003, 10:44 AM
Bill, Nomadpat started his post with this"Just for kicks, I queried my CRC database to see the results of this question about the effectiveness of BRIS Prime Power. Here are the results:"

When he said "to see the results of this question about the effectivenss of BRIS Prime Power" I would expect to see his results show the 'effectiveness' of Prime Power.

The use of Prime Power does not result in a -19% ROI. I know it, you know and Nomadpat know it.

If you want to see the proper use of databases and the proper way to state results you should take a look at Ken Masas work over at the HRT sight.

Most of the so called DB studies and results I see posted here are most almost a negative to someone's attempt to be a winner player. In other words, they are as useless as tits on a boar.

JustMissed

alysheba88
12-23-2003, 10:58 AM
Prime Power is the greatest. Never need to look at anything else. God's gift to handicapping. The Rosetta Stone. Will let you retire a millionaire.

That better?:)

BillW
12-23-2003, 12:02 PM
Originally posted by JustMissed
Bill, Nomadpat started his post with this"Just for kicks, I queried my CRC database to see the results of this question about the effectiveness of BRIS Prime Power. Here are the results:"

When he said "to see the results of this question about the effectivenss of BRIS Prime Power" I would expect to see his results show the 'effectiveness' of Prime Power.

The use of Prime Power does not result in a -19% ROI. I know it, you know and Nomadpat know it.

If you want to see the proper use of databases and the proper way to state results you should take a look at Ken Masas work over at the HRT sight.

Most of the so called DB studies and results I see posted here are most almost a negative to someone's attempt to be a winner player. In other words, they are as useless as tits on a boar.

JustMissed

My point was that you were criticizing your own conclusion and jumping on nomad for it. Seemed weird to me.
While averages are difficult to derive any conclusion from other than results of blindly betting a single parameter, if by saying "The use of Prime Power does not result in a -19% ROI. I know it, you know and Nomadpat know it. " you are implying that betting the top prime power number on dry dirt at CRC across the board would produce a different results, you would be mistaken.

Bill

VetScratch
12-23-2003, 01:28 PM
Alysheba88,Prime Power is the greatest. Never need to look at anything else. God's gift to handicapping. The Rosetta Stone. Will let you retire a millionaire.How true!

However, some handicappers do not know how Prime Power is actually derived or understand how to use it. This unfortunate circumstance dates back to the PBS radio broadcast of a mock debate over the meaning of entropy that pitted Von Neumann and Shannon against Epstein and Sengbush. Scholarly thespians were recruited to portray the principals through the course of pre-scripted debate authored by graduate students at MIT and presented before a live audience at the 1992 Conclave of Handicapping Beta Testers.

The pre-scripted debate concluded with Sengbush's formal explanation of a value-added derivative that he called Prime Power and with Epstein's demonstration of using Prime Power entropy calculations to produce expectation density profiles that often reveal positive expectation plays in races that correspond to specific entropy values computed from Prime Power ratings.

Unfortunately these astonishing discoveries were never presented or broadcast because the debate ended prematurely when it erupted into a wild melee resulting in several injuries and arrests. From the beginning, the thespian who portrayed Shannon had courted audience attention by balancing on a unicycle while juggling billiard balls. These theatrics infuriated the actor playing Sengbush, who finally jammed one of Sengbush's seismic instruments into the spokes of Shannon's unicycle, catapulting Shannon's character into the orchestra pit. A rampaging brawl ensued, and the handicapping world was denied exposure to the magic of Prime Power ratings.

Legend has it that the original debate script circulated in a secret MIT student society known as The Whales until 2001 when a copy inadvertently found its way onto the BetaTesters message board moderated by Jaguar. The rest of the story is well known... and some astute handicappers are living large as they accumulate their fortunes in Swiss bank accounts.

Trijack
12-23-2003, 01:36 PM
As a Always user for years and making my own profiles I always (pun intended) compare my profile numbers with Prime Power. As I mostly play tri's and like to put 3 horses on top and usually not the favorite has led me to some good tri's but not many big ones. Finished in the black in the recent Hollywood meet. I have the winner on top in more than 80% of my tri's but the 2nd and 3rd horses are another matter.

While at Toga for the week with the PA gang Bris Power did great but I still only managed to pick 5 tris's out of about 23 races.. Since I had not played Toga before I did not have a large data base to figure from. Will do better next year..

I'm sure when people post something they use, it works for them but not for everyone. I would never try to change the way someone picks their horses. I think it is best through trial & error to come up with your own way but keep an open mind to some elses suggestions.

One final thought. What works at one track for you does not mean it will work at all tracks.

Jack

Milleruszk
12-23-2003, 01:51 PM
Bettheoverlay seems to have the right approach in using the Prime Power ratings. Use them as a way to find value. After all that is what this game is all about......finding hidden value that is not apparant to the general public. Don't use these ratings as the holy grail but they can be useful in pointing out horses that you may overlook. Seems like a rational approach to me!

Tom

nomadpat
12-23-2003, 02:21 PM
Hi Just Missed,

The point of my post wasn't to knock them at all. I merely provided a look at how they have performed at CRC in my database. As Trijack noted, they can be a great tool for predicting the ITM horses and for completing trifectas.
There is no question in real life that no one would blindly bet the top one or two or whatever. I don't see how that information was totally useless. When someone is considering just "one" factor like that, wouldn't you want to know some kind of measure of how it performs? Then leave the rest to your judgment as the "rational approach" as the last poster suggests?

VetScratch, I got a great laugh out of your response! Were you secretly conducting laboratory reserach for the Whales group? :p

JimL
12-23-2003, 04:48 PM
Vet, Say What?????!!!!!

JustMissed
12-23-2003, 05:47 PM
Hey Bill W,

You said "My point was that you were criticizing your own conclusion and jumping on nomad for it. Seemed weird to me."

This was addressed to me, JustMissed.

Are you one of those guys who, in grammar school, got an A+ in Reading and a D- in Reading Comprehension?

Would you care to point out exactly where in this thread I criticized "my own conclusion". Could be you're having trouble following the conversation old buddy.

JustMissed
:mad:

BillW
12-23-2003, 06:41 PM
Originally posted by JustMissed
Hey Bill W,

You said "My point was that you were criticizing your own conclusion and jumping on nomad for it. Seemed weird to me."

This was addressed to me, JustMissed.

Are you one of those guys who, in grammar school, got an A+ in Reading and a D- in Reading Comprehension?

Would you care to point out exactly where in this thread I criticized "my own conclusion". Could be you're having trouble following the conversation old buddy.

JustMissed
:mad:


Let's start from scratch

-> You stated that the Bris Prime power numbers were good for identifying contenders

-> nomadpat posted data supporting your contention (top 3 Prime power numbers win 62% of the time) making no attempt to draw any conclusion from it.

-> You then replied "Your study makes the FALSE assumption that a player would play every race in your sample. This just does not happen in the real world and you know it. " As this interpretation was not offered before in the thread I assumed it was yours.

What am I not comprehending?

Bill

JustMissed
12-23-2003, 07:02 PM
Hey Bill,

You know what happens when you make assumptions.

Merry Christmas to you.

JustMissed
:)

Tom
12-23-2003, 07:27 PM
Do they have Christmas on your planet?:D

If, have a merry one, If not, hang in there until NewYears.;)

BillW
12-23-2003, 07:50 PM
Originally posted by JustMissed
Hey Bill,

You know what happens when you make assumptions.

Merry Christmas to you.

JustMissed
:)

I understand :D

Bill

VetScratch
12-23-2003, 08:02 PM
Originally posted by Tom
Do they have Christmas on your planet?:D

If, have a merry one, If not, hang in there until NewYears.;) Merry Christmas, Tom.

You should know better than most of us that peanuts come in shells! :)

VetScratch
12-24-2003, 04:22 PM
Here is boring rendition of how to look at Prime Power or any similar hybrid derivative that is designed to gauge Pace, Speed, Class, Form, Surface, and Distance qualifications relative to today's race.

Imagine that you had final odds hours prior to post. Numerous database studies of final odds have been conducted, and authors like Epstein (pages 294-295) have proposed sophisticated means of finding value pockets in final odds lines. Several of the noted value handicapping authors have also provided guidelines or computations to help players spot odds distributions that foreshadow large payouts (on the basis of statistical studies).

The rub is that a lot goes on between two-minutes to post-time and one-minute after post-time when odds are final. This is crunch time for most players... and hardly a good time to wrestle with "new math" in your head. However, if you had hours to anticipate prime opportunity races, you would have already spent more time handicapping, considered wagering options, and have an edge on other players.

Some players use Morning Lines to anticipate prime opportunity races. I have found hybrid derivatives like Prime Power to be more consistently predictive of prime opportunity races, and you have access to them many hours before post time.

When you think about it, hybrid derivatives like Prime Power reflect consideration of the same factors that primarily influence public wagering. While such derivatives do not reflect privileged (unpublished) information, they are good enough to quickly screen many races without missing too many prime opportunities.

If you have a database that includes Prime Power (and you do your own programming/analysis), pull out your literature about final odds analysis and noodle your way to using Prime Power ratings to predict prime opportunity races and find value pockets (i.e., positive expectation plays).

andicap
12-24-2003, 11:40 PM
Is that your high-falutin' way of saying Prime Power does a better job than the morning line of predicting which horses are going to be bet, so you can scope out the best overlays way ahead of time rather than worry about late odds changes?

If not, I have no idea what you meant.

VetScratch
12-25-2003, 12:32 AM
Andicap,

Sort of... Morning Lines are not standardized and they vary from track to track... but your computer can rapidly scan ten complete racecards for races with specific Prime Power distribution characteristics that statistically foreshadow value opportunities for straight and/or exotic payoffs. If you find ten such races, you concentrate on them... and simply ignore the other 80+ races. The subtle part is that what you look for may not be intuitively obvious, especially when you play exotics and want to optimize wagering costs.

Tom
12-25-2003, 12:58 PM
...will you be posting real-time examples of how this is done and what the 10 or so focus races are ahead of time?

andicap
12-25-2003, 01:57 PM
VS,
What do you mean "sort of."?
It appears that is EXACTLY what you meant. (or do you exalt in your obliqueness?). You are using the Prime Power as a substitute for the ML -- because the ML varies too much f rom track to track. You raise a valid point. The Prime Power is more consistent than the ML and could be a better indicator of where the public is going to go, assuming the Prime Power generally points to the horses the public is headed toward.

Right???

If so, that's a fine idea if it works (if the PP really does point to horses the public does).

VetScratch
12-26-2003, 10:26 AM
Originally posted by andicap
VS,
What do you mean "sort of."?
It appears that is EXACTLY what you meant. (or do you exalt in your obliqueness?). You are using the Prime Power as a substitute for the ML -- because the ML varies too much f rom track to track. You raise a valid point. The Prime Power is more consistent than the ML and could be a better indicator of where the public is going to go, assuming the Prime Power generally points to the horses the public is headed toward.
Right???
If so, that's a fine idea if it works (if the PP really does point to horses the public does). Sorry, but I meant "sort of." Right now, Prime Power ratings are not offered in the cheapest products where they could be used to save money on the more expensive products... and nothing about the way I think Prime Powers are valuable has anything to do with traditional handicapping (i.e., picking horses). The idea is to pick positive expectation races first, then pick positive expectation wagers for these races.

I have done enough preliminary research with Prime Power to see how it can be valuable to me... but I am not going to incorporate it into my race selection system until I can be sure my implementation effort won't be wasted.

My personal (privateer) handicapping system got launched about six years ago when my dad challenged me (and a working friend) to surpass what he and his cronies had started on mainframes in the 70's and eventually ported onto personal computers. Our work focuses on the creation of accurate value odds-lines and then the drudgery of monitoring toteboards from many races for value plays... or (more recently) mining many odds-lines (in advance) against criteria developed from database studies to pluck out races that have the highest probability of large straight and or exotic payouts. None of what we are trying to do lends itself to discrete examples like you find in handicapping books... which impart very valuable information but confine their procedural examples mostly to stuff that we can follow concurrently and replicate with pen and paper, like Brohamer's pace calculations.

When BRIS introduced both Prime Power and Custom Cards, it occurred to us that instead of downloading and handicapping tons of racecards to find a few uniquely promising races, it might be possible to use Prime Power ratings instead of handicapped odds to do the same job. After all the hoopla over Prime Power, we expected BRIS to start offering it across the board in their product line. Our plan was to buy Prime Power ratings cheap, use them to pre-select races, then use the Custom Card feature to grab these individual races rather than download lots of expensive racecards... in other words, download maybe ten Custom Card races per day instead of many complete racecards.

We did enough preliminary analysis to see great promise in this approach (and we expect to get better at it as our database of racecards with Prime Powers expands). Unfortunately, BRIS has so far restricted Prime Powers (billed as the most exciting handicapping factor ever devised) to their premium product lines, so we will have to wait for competitive pressures to induce BRIS to offer Prime Power ratings in products priced at a $1.00 or less (e.g., $.50 at TSN).

BTW, advanced All-Ways users already follow a similar approach to economizing on download costs by using the Custom Card feature. Some on this board may want to explain how they use the All-Ways analysis features to determine exactly what races they choose to download (but I rather doubt anyone will volunteer because an attempt at explanation will soon become an exhausting excercise and end up being meaningful to a limited audience of other advanced All-Ways users).

There are many software developers looking for answers to the same problem statement: how to find the shortest and cheapest route to the promised land. No one wants to manually monitor toteboards for 50,000+ races to find the key value opportunities that will make or break a year... and there are additional considerations that make this approach questionable even if it is automated.

The fully-automated brute force approach to monitoring toteboards (and triggering wagers) has been widely talked about (and implemented to varying degrees). However, just like the current fuss over the disappearance of free toteboards and video feeds, the online providers will have to respond if numerous users implement software that swamps them with activity. The principle of "you get what you pay for" eventually rears its ugly head.

Thus, "sort of" meant that I am confident that standardized value-added derivatives like Prime Power will become valuable race selection variables when they become available at commodity prices (for individuals willing to tackle the up-front research and development efforts). I would gladly pay $.25-$.50 per card to get nothing but date, track, race numbers, race conditions/classifications, and horses with their Prime Power ratings. I already know that good handicapped odds lines are valuable precursors of race-level value opportunities (especially for exotics) but getting there was hard and time consuming at considerable cost since no data-cost-savings are realized. The next step is to attack cost.

Beyond race selectivity, there is another dimension of statistical research that seeks to predict which odds values within specific races are "blackbox" positive expectation plays. And beyond that will be other challenges to find ways to leave the herd. The most enjoyable part of individual effort is the work itself... the universe expands fastest late at night!

I don't mean to be "high-faluting," but describing exactly how any complex problem solution has been programmed would not even excite other developers since they will be inclined to tackle the same problem, follow a different path to resolution, get better results, and dance a late-night jig when they are finally done.

wolsons
12-26-2003, 10:43 AM
In case you weren't aware, TSN's Procaps files have the Prime Power ratings in them, and you can get unlimited racefiles from them for $59.95 per month - all tracks, every day. It's one hell of a bargain, and includes unlimited Exotic results files and instant race charts as well. Might be worth your checking them out!
:)

VetScratch
12-26-2003, 10:56 AM
Yes, and I see they are now available in the BRIS DRF PP downloads, but we are currently are set up to handicap BRIS premium racecards... but, thank you, $59 per month to get all Prime Powers from TSN is tempting.

Can anyone explain the occasional big price differentials between BRIS and TSN since the product lines are almost identical?

VetScratch
12-26-2003, 12:05 PM
Wow! I just went through the actual table of contents specs for the ProCaps files on TSN... $59/month is really an outstanding bargain for unlimited downloads that come very close to the value-added content in BRIS premium downloads that cost $5.60 each after all Gold Club and BrisBet discounts are applied.

I think BillW posted awhile back that he converted from BRIS to TSN... sure looks like he got his money's worth for going through the details of a conversion effort.

Right now, it appears that TSN is the best value/cost alternative among all the providers for anyone who intends to do their own programming.

JustMissed
12-26-2003, 02:14 PM
Originally posted by VetScratch
Wow! I just went through the actual table of contents specs for the ProCaps files on TSN... $59/month is really an outstanding bargain for unlimited downloads that come very close to the value-added content in BRIS premium downloads that cost $5.60 each after all Gold Club and BrisBet discounts are applied.

I think BillW posted awhile back that he converted from BRIS to TSN... sure looks like he got his money's worth for going through the details of a conversion effort.

Right now, it appears that TSN is the best value/cost alternative among all the providers for anyone who intends to do their own programming.

Vet, I have touted the use of TSN Prime Power ratings on this board for over a year. There has been a fair amount of postive support but by and large there has been a lot negative comments that I suspect were proferred by the software/data sellers.

Anyway, my son and I play Mountaineer 4 nignts a week and usually Tampa every day of live racing and other tracks occasionally. We use Prime Power for our basic contender selection with final picks made by the use of basic handicapping principles: speed, pace, trainer intent, etc.

As you know, the Prime Power is an exceptional handicapper factor to put you on the finishing horses.

My son, for several months, has been after me to come up with a way to query all the tracks available to us via simulcast in order to come up with additional plays. We believe we need to push over $1 million through the window each year to make what we need to support our life style, and as I am sure you know, you need a lot of playable races to bet that kind of money.

Anyway, after reading your post about using Prime Power to identify value plays, I got to thinking that Prime Power may be the answer to our query problem.

I thank you for turiing on that "light bulb".

Sincerely yours,

JustMissed
;)

takeout
12-26-2003, 02:22 PM
Originally posted by VetScratch
Can anyone explain the occasional big price differentials between BRIS and TSN since the product lines are almost identical?
My guess has always been the DRF "fly in the ointment," meaning BRIS's contract with DRF.

As to why BRIS is the ONLY reseller that can't seem to live without going through the DRF middleman, the only thing that I can think of, is that being originally a breeding stats outfit, BRIS has to maintain a tie to DRF's database which goes way back in time. That's just a guess.

All I know for SURE is that I used to pay twice as much for the BRIS DRF file just to end up with a bunch of wrong trainer names. I'm sure it's much better now but I just don't trust DRF or BRIS anymore. A lot of that stuff stayed wrong for well over a decade and may still be.

VetScratch
12-26-2003, 06:22 PM
Originally posted by takeout
My guess has always been the DRF "fly in the ointment," meaning BRIS's contract with DRF.

As to why BRIS is the ONLY reseller that can't seem to live without going through the DRF middleman, the only thing that I can think of, is that being originally a breeding stats outfit, BRIS has to maintain a tie to DRF's database which goes way back in time. That's just a guess.

All I know for SURE is that I used to pay twice as much for the BRIS DRF file just to end up with a bunch of wrong trainer names. I'm sure it's much better now but I just don't trust DRF or BRIS anymore. A lot of that stuff stayed wrong for well over a decade and may still be. Folks who "should" know have said BRIS stayed with DRF because of economic advantages grandfathered to them in their original deal with DRF... going back before the Internet when they and DRF were the only games in town. Makes me wonder though... if they have a cheap grandfathered deal, why aren't they the one with the cheapest prices?

I don't find many trainer problems anymore... used to find a lot of synonyms. When Equibase took over reporting results from the tracks, I think BRIS had a lot of difficulties pairing the Equibase names with the DRF names in order to compile all of their various categories of trainer statistics. Why EquiBase and DRF names seem to have been compiled from different data sources in the first place is a bit puzzling... the differences seem to extend beyond one spelling out first/middle names and the other using initials. The use of pet names instead of proper names (e.g., Dick for Richard) can also be overcome. I actually went out and bought a book called The New American Dictionary of Baby Names in order to build a cross-reference between pet names and proper names. This little library utility comes in real handy in a program that helps compose Internet searches as well as for matching names and avoiding synonyms in my own horseracing database. When Anastasia Reynolds gets listed as Stacey Reynolds, it won't give me any transparent problems. :)

takeout
12-27-2003, 01:36 AM
Originally posted by VetScratch
Makes me wonder though... if they have a cheap grandfathered deal, why aren't they the one with the cheapest prices?
That would've been my first question also.
IMO nothing about that BRIS & DRF thing has ever passed the smell test.

VetScratch
12-27-2003, 11:50 AM
JustMissed,

It all boils down to analyzing the distribution curves for Prime Powers within races. Note that most folks only go as far as simple ordinal rankings for Prime Powers (1st, 2nd, 2rd, etc.).

Especially for exotics, the distribution curve can be meaningful since players turn exotics into an underlay trap in many races.

jk3521
12-28-2003, 10:25 PM
Let me throw this one out there, does anyone have any data suggesting whether the BRIS Prime Power ratings work better in sprint or route races ?

Rick
12-29-2003, 03:38 PM
VS,

So, are you looking for positive expectation situations using Prime Power numbers to calculate the entropy of a race in a similar manner to the method Epstein proposed (using actual odds) in his book "Theory of Gambling and Statistical Logic"?

VetScratch
12-30-2003, 01:49 AM
Rick,

Epstein popped up as a support character in a whimsical earlier post, but we do use the entropy calculation as one of several distribution descriptors when we analyze our database.

We use the first part of what Epstein describes to compute one of several handicapped odds-line descriptors that get stored with other handicapping results in our database:
(1) All original and validated race/horse data.
(2) All synthetic race/horse derivatives (what gets handicapped).
(3) Handicapped results (odds-lines and lots of stuff for later analysis).
(4) Race results (payoffs, final odds-line, etc).

With BRIS Prime Powers we looked at computing descriptors for the rating-lines just as we do for our handicapped odds-lines. The idea was to cull out races that have the highest statistical probability of large exacta/trifecta/superfecta payoffs. If this can be done, you would only want to download and handicap maybe 12 races per day (as a Custom Card) instead of up to 200+ races per day (to use the handicapped odds-lines as exotic value precursors).

However, only TSN currently offers a package ($59/month) that is attractive for this purpose (i.e., get all the Prime Powers with some other stuff, then go back for specific races to download and handicap, probably from the pseudo-All-Ways files for Millennium). After I get re-united with my working partner, we plan to look at converting from BRIS to TSN (and even evaluate using the ProCaps files as our primary handicapping files).

If you ask why we screen for exotics in view of takeout... the public often does a much worse job of making overlay/underlay wagering decisions than with straight action.

So what we did to test Prime Powers is approximate what we do with handicapped odds. Of course the Prime Powers do not lend themselves to the Epstein entropy computation or other descriptive derivatives like odds do, so we fiddled with ways to approximate the same results from Prime Powers... even if less than the full distribution range was as meaningful as our handicapped odds (of course, in all references to odds, I mean underlying probabilties).

In short form, for just the Epstein derivative, we:
(1) Convert Prime Powers to a relative index, .01 to 1.0.
(2) Raise all values lower than the harmonic mean to that value.
(3) Score the index with an exponential co-efficient-decimal (.xx) that varies by number of horses (i.e., score = index^(1+co-efficient-decimal) (e.g., index^1.xx).
(4) Convert the scores to their pseudo-probability values(y).
(5) Compute the (-)Sum of all(y * Ln(y)).

When you analyze this, we have focused on the distribution characteristics for the Prime Powers above the harmonic mean (used as the most representative indicator of central tendency for small samples like Prime Powers, and always less than arithmetic mean).

We know this is different than using odds, but it yielded a derivative (along with other statistical descriptors) that allowed us to find a promising degree of overlap with the population of races culled via our existing method of using handicapped odds.

What we are looking for is a quick and cheap method to handicap and monitor toteboards for a small but promising subset of all exotic opportunities. With 50,000+ races a year, you can go nuts watching toteboards.

Finally, many of the qualifiers mentioned by others apply. While, Prime Powers may actually be better indicators of who will win races for young (lightly raced) horses and odd-ball fields, for our purposes we screened for common race conditions for older horses where all horses are likely to have ten or more performances (or we could really screen by switching to ProCaps files).

For simply picking winners, Prime Powers seem like other high I.V. derivatives... too obvious to be valuable.

Right now, we still handicap a lot of races and cull with our handicapped odds.

Rick
12-30-2003, 01:59 AM
If anyone's interested, I did some limited testing many years ago of the "entropy" concept and got some positive results over a few hundred races. But, the win % was only a little over 12% for the best category so I never pursued it any further. The most profitable races were those with a relatively high entropy value (as Epstein suggested) and at odds of 8-1. In fact, when I went back over the calculations I found that just using the simple rule of playing only races with 9 or more entries where the favorite is at least 3-2 odds selected all of the high entropy races without having to go through the calculations. I retested using the simple version a couple of times and came up with positive results both times (on different samples several years apart).

That would lead me to guess that the median ranked Prime Power horse in large fields might be a good bet, especially if the top ranked horse doesn't have a big edge over the second rated horse.

That's my two cents worth on the subject.

Rick
12-30-2003, 02:04 AM
VS

VetScratch
12-30-2003, 02:12 AM
Rick,

What you have done is similar to my dad's approach (which started our endeavors as a family challenge that has become an obsession). His odds sniff out overlays best in the 6/1 to 12/1 range of track odds. We can't believe the nights we have spent exploring different angles. Every point above breakeven becomes more difficult when you are looking to fully automate handicapping. I question our sanity about twice a year.

Rick
12-30-2003, 02:21 AM
VS,

It makes sense that exotics would offer more possibilities, especially in large fields. Even something as simple as wheeling short priced favorites on the bottom of exactas in very large fields is surprisingly effective. The more possible combinations the better I guess. Those 20 entry Kentucky Derby fields should offer some huge overlays somewhere.

Also, it always seemed to me that anyone doing a study of various odds ranges should relate it to the number of horses in the race. My guess is that in fields of n horses, those in the odds group that wins with a probability of about 1/n should lose less than the other groups, at least in large fields anyway.