bris prime power into odds line

[acorn54]

for you mathematically inclined members, is there a mathematical function that can be used to convert the bris prime power into an odds line.
i asked bris staff about this and they were very hush hush about revealing any info that might have helped me in this regards.
it seems the prime power number has good predictive power and i think that is a good start towards converting the prime power number of a horse into probability for the horse.
acorn (guy)

[HTRFGuy]

Originally posted by acorn54
for you mathematically inclined members, is there a mathematical function that can be used to convert the bris prime power into an odds line.
i asked bris staff about this and they were very hush hush about revealing any info that might have helped me in this regards.
it seems the prime power number has good predictive power and i think that is a good start towards converting the prime power number of a horse into probability for the horse.
acorn (guy)

I make an odds line and after a lot of testing of factors I settled on a line based upon 47 factors before I was satisfied with the results! BRIS prime power is not one of the 47 but the BRIS speed rating (or equivalent) is along with the 'BRIS formula' from Nunamakers work on 'Modern Impact Values.' An odds line based upon only one factor like prime power would not be viable it seems to me.

[Overlay]

I agree that no odds line should ever be based on just one overriding factor. (Even if the probabilities associated with the factor remain valid, it will eventually lose its pari-mutuel value when enough people start keying on it and driving the odds on every horse with the factor down below fair-odds level.) I would imagine that the more factors you include, the more accurate your line would be (as long as your math is sound). And computers certainly take the drudgery out of the number-crunching, once you've arrived at your basic algorithm. However, in my own experience, I've found that it's possible to arrive at an acceptable combination of precision and complexity with a much smaller number of factors (a maximum of seven), even allowing calculations to be done without the need for a computer.

[DJofSD]

Instead of trying to convert the BRIS Prime Power numbers into odds, try comparing the rank order of the odds to the rank order of the BRIS PP ratings. Then look to see if there are large discreptancies.

DJofSD

[ceejay]

Originally posted by acorn54
for you mathematically inclined members, is there a mathematical function that can be used to convert the bris prime power into an odds line.
i asked bris staff about this and they were very hush hush about revealing any info that might have helped me in this regards.
it seems the prime power number has good predictive power and i think that is a good start towards converting the prime power number of a horse into probability for the horse.
acorn (guy)

Guy,

I think thats what the Brisbet "profit line" (under tools) basically is. You can also buy them seperate I think. http://www.brisnet.com/cgi-bin/trk_report.cgi?plr

[headhawg]

Originally posted by ceejay
Guy,

I think thats what the Brisbet "profit line" (under tools) basically is. You can also buy them seperate I think. http://www.brisnet.com/cgi-bin/trk_report.cgi?plr

I have a Brisbet account and Profit Line is free. And it's also pretty pathetic.

Make your own line.

[betovernetcapper]

Sort the numbers from low to high, and subtract each rating from the top rating.
Add 1 to each number.
Devide 1 by each number and then total that number.
example (where there are 5 horses with a one point diference between them)

0=1=1
1=2=.5
2=3=.33
3=4=.25
4=5=.2

add subtotals (1=.5+.33+.25+.2)=2.28

then devide each number by the total

1/2.28=.43
2/2.28=.22
3/2.28=.14
4/2.28=.11
5/2.28=.09

then devide 1 by each % and subtract 1 to get an odds line.

1/.42-1=$1.32 to 1
1/.22-1=$3.55 to 1
1/.14-1=$6.14 to 1
1/.11-1=$$8.09 to 1
1/.09-1=$10.11 to 1

I don't think this will produce a particularly good odds line but it should produce an odds line that will refllect the ralationship of the ratings.

[formula_2002]

Originally posted by acorn54
for you mathematically inclined members, is there a mathematical function that can be used to convert the bris prime power into an odds line.
i asked bris staff about this and they were very hush hush about revealing any info that might have helped me in this regards.
it seems the prime power number has good predictive power and i think that is a good start towards converting the prime power number of a horse into probability for the horse.
acorn (guy)

To begin with, go to my home page for a very good summary of the Bris Power rankings

The Bris power ranking is very well correlated to the odds.
I have not been able to show a profit when using the BP.
Under cetain conditions I can knock the effect of the take out down to about 10%. (thats an ROI of .90

Know you know why I switched to craps;)
roi is 98.6

[traveler]

I think Betover's approach will find you the odds of each horse relative to their Prime numbers. What I have found is that in "playable" races(based on the types of races I play) that the top 3 Prime horse win about 75% of the time. So I assign 75% to them and allocate 25% to the other horses - this isn't set in stone as a 6 horse field is different that a twelve horse field. You then addjust the odds for each horse within those paramaters.
I have no idea how to set an accurate odds line by a computer formula - I am sure some can but not me, I have enough trouble with above procedure

[Shacopate]

Acorn,

The answer to your question is yes.

Try putting the prime power numbers into CJ's linemaker.

I downloaded the Bris Free PP's for MNR tonight (sunday 12th, R6) and did just that. Top Prime Power horse was 3-1 ML. CJ's linemaker made him 4-5. Won by 2 and paid just about that. Pretty accurate! Don't forget to include the feild size, very important.

[betovernetcapper]

Cooll tool-CJ's link is
http://www.pacefigures.com/

[bettheoverlay]

I apply angles to the top3 PP #s. I have best success with wos/recency. 2 wos in 10 days, bullet wo in 10days, 4f wo in last 4 days, at cheapie tracks, race in 14 days. I used to require 6-1 but have dropped to 4-1 lately. The top PP is rarely at those odds. If the 4th PP is within 1 pt of the 3rd I'll look at him too.

[acorn54]

Originally posted by Shacopate
Acorn,

The answer to your question is yes.

Try putting the prime power numbers into CJ's linemaker.

I downloaded the Bris Free PP's for MNR tonight (sunday 12th, R6) and did just that. Top Prime Power horse was 3-1 ML. CJ's linemaker made him 4-5. Won by 2 and paid just about that. Pretty accurate! Don't forget to include the feild size, very important.

thanks shacopate
i'll play around with cj's odds line tool and see what happens
acorn (guy)

[Exactaman]

once you do that try making your own line acorn. if you have a good feel for the accuracy of those figures it won't be hard. you may find the thinking process you go through to do it helps your capping. i have an excel sheet to help you do it if you want.

[acorn54]

Originally posted by Exactaman
once you do that try making your own line acorn. if you have a good feel for the accuracy of those figures it won't be hard. you may find the thinking process you go through to do it helps your capping. i have an excel sheet to help you do it if you want.

hi exactaman
i don't have excel in my computer. i have staroffice.will your excel sheet work in that?
acorn

[JustMissed]

Hey Acorn,

Once you come up with your line, what are you going to do with it?

JM

[Exactaman]

i don't know Acorn, does it generally accept excel files?

there used to be a freeware on the delmar site that did the same thing, but it doesn't seem to be there anymore. maybe there are others out there.

[Shacopate]

JM,

He's probably gonna do what everyone else does with an oddsline. Try to find value.

Do you have an idea or suggestion?

[acorn54]

Originally posted by JustMissed
Hey Acorn,

Once you come up with your line, what are you going to do with it?

JM

exactaman
i don't know i don't think i ever imported excel programs before

just missed
if i can get an accurate odds line it would be sort of a black box wouldn't it. just bet the overlays?
acorn

[Equineer]

Even though BRIS Prime Power is compiled from a "secret" blend of several factors, don't expect that you will systematically overcome the pitfall inherent to odds-lines that depend upon either a single factor or only a handful of factors.

What I mean is that the nature of racing data is such that limiting the scope of your odds formulation invites a lot of false underlay/overlay signals unless you diligently evaluate everything you excluded to estimate the likelihood that public favor or disdain has reasonable merit.

But if you proceed with BRIS Prime Power (BPP), you could do much worse, since BPPs do produce impact values that plot somewhat "closer" to a linear function than most other discrete or hybrid factors.

Excluding fields that include one or more absent/missing BPP figures, here are some BPP impact value examples for a broad and sizeable sample of tracks and sprints for 3&UP Non-Maidens using all such sprint race types not contested on "off" tracks.Races _1st _2nd _3rd _4th _5th _6th _7th _8th _9th 10th 11th 12th
918 2.19 1.11 0.71 0.63 0.37
2756 1.98 1.42 1.01 0.74 0.53 0.31
3488 2.07 1.55 1.23 0.79 0.62 0.47 0.26
3049 2.47 1.54 1.16 0.86 0.78 0.52 0.39 0.27
2139 2.35 1.73 1.42 1.07 0.80 0.62 0.36 0.38 0.27
1950 2.55 1.84 1.42 0.97 0.87 0.69 0.57 0.45 0.38 0.27
650 3.03 1.78 1.43 1.01 0.80 0.97 0.66 0.43 0.51 0.23 0.14
512 3.15 1.62 1.69 1.50 1.24 0.74 0.56 0.57 0.30 0.24 0.23 0.16Bear in mind that samples for larger fields were more difficult to collect since many races were apt be excluded (i.e., because of missing BPPs), and all impact values are relative to the random chance of winning (i.e., 1/5, 1/6,... 1/11, 1/12, which decreases as field size increases).

Also, serious handicappers might prefer to compile a large enough sample database to facilitate a more granular analysis (i.e., by individual track, and even various sprint distances and race classes). This of course, always entails more data, time, and investment.

To convert any impact value to a "probability representation" divide it by field size, such as 3.15/12=.2625, but this again will invite distortion since impact values are based on linear BPP rankings, first through last, while actual BPP values will be non-linear (i.e., unlike the ordinal rankings).

[cj]

You can apply thousands of factors to generate an odds line, and it may not be any better than one generated from just a few factors. The key is to know which factor(s) are important in selecting the winner of TODAY'S race. That is the key to the mint.

[Exactaman]

well if you think you can use it let me know acorn. it does make use of some special excel functions though.

Originally posted by Equineer
Even though BRIS Prime Power is compiled from a "secret" blend of several factors, don't expect that you will systematically overcome the pitfall inherent to odds-lines that depend upon either a single factor or only a handful of factors....

Cmon you're going to scare the guy before he even starts :) i think this kind of argument really misses the mark. certainly there are many things a line can be. one thing however is simply a tool for translating handicapping opinion into a method for pursuing value. i contend that Acorn who attacks the races with a sheet of brisbet ratings--if that is his chosen method, whatever his method is is really irrelevant to this argument--is not going to do as well as Acorn who has taken the step of translating his ideas into a concrete opinion about value. at the very least he is going to be attempting to bet on horses that represent value and gaining insights into that process. can that be wrong?

[Equineer]

Exactaman,

It's not really that confusing. Given that I only meant to illustrate an appraoch by way of example, it is easy to convert probabilities represented by impact values into track odds (using for instance 17% as average takeout).

The previous impact values converted to 17% takeout track odds:_1st _2nd _3rd _4th _5th _6th _7th _8th _9th 10th 11th 12th
0.89 2.74 4.85 5.59 10.22
1.52 2.51 3.93 5.73 8.40 15.06
1.81 2.75 3.72 6.35 8.37 11.36 21.35
1.69 3.31 4.72 6.72 7.51 11.77 16.03 23.59
2.18 3.32 4.26 5.98 8.34 11.05 19.75 18.66 26.67
2.25 3.51 4.85 7.56 8.54 11.03 13.56 17.44 20.84 29.74
2.01 4.13 5.38 8.04 10.41 8.41 12.83 20.23 16.90 38.70 64.21
2.16 5.15 4.89 5.64 7.03 12.46 16.79 16.47 32.20 40.50 42.30 61.25So now you have a lookup table for an estimated odds-line that corresponds to BPP rankings for the various field sizes, bearing in mind the caveats posted previously for using ordinal BPP rankings, ranked from first (highest) to last (lowest).

Notice the anomalies for several rankings in 11-horse fields.

I will go back to my data sources and verify that BPPs for this field size were computed/reported/copied correctly (although the probabilities did add up to 100%).

[JustMissed]

Originally posted by acorn54
just missed
if i can get an accurate odds line it would be sort of a black box wouldn't it. just bet the overlays?
acorn

1.Acorn, first let me say I use Prime Power numbers every day and they are a great handicapping tool. Many different ways to use them and if you have not done so, go to the Bris library and read and study about them.

2. I would not want to use just one factor as a betting line under any circumstances.

3. I'm not really sure there is such a thing as an "overlay".

Remember that your line is just a mathematical expression of your opinion. Same as the tote line is to the public opinion or the morning line is to the track handicapper's opinion. All are just expressions of the same future event.

In fact if you bet your line based on a comparison with the tote, your opinion immediately becomes part of the public opinion and you in fact are contributing to reducing the alleged "overlay".

For example, if you make a horse 5-2 and the tote shows 3-1, as soon as you purchase your ticket you are helping to send the tote down closer to 5-2, where is the value in that?

4.Let me suggest another way rather than expending the time and effort required to make your line.

a. Based on your handicapping method, eliminate all horses that you think cannot finish 1st thru 4th.

Visualize how you think the race will finish and rank your horses as to their predicted order of finish. Label a,b,c, etc or 1,2,3, etc.,
what ever you like.

If you have a very strong opinion on the A horse, label him ++ or whatever to indicate a gap between the A and B. If you think A, B and or C all have an equal chance to run 1st, draw a circle around all three or a bracket or whatever. Do this for all your contenders.

You should already have a written betting plan so you will know how to struture you bet(s).

For example, if you think A,B,C will finish in the exacta but your not sure of the order and you want to play a box-just look at the probable exacta payoffs shown on the TV screen to see if they meet your betting minimums. Some guys use $30 for an exacta box. Use whatever your records indicate will put you in the black.

For the trifecta, since they don't show probable payoffs, you could use the rule of 70 or some such thing to decide if you want that bet. If you study the charts for the track you play they will give you a reasonable indication of payoffs at various odds levels. If you do this long enough pretty soon you can guess the payoff as soon as they cross the wire within 20 bucks or so.

For win bets, if you have a very strong opinion and your records show you have a high hit rate at those odds-just bet it and bet it heavy.

Just my take on making lines-if if works for you, great.

I have not yet seen anyone cash an "overlay" at the track. They only cash winning tickets.

Good luck to you.


JM

[Exactaman]

Originally posted by JustMissed

I have not yet seen anyone cash an "overlay" at the track. They only cash winning tickets.



Whether one is cashing an overlay, an underlay or a winning ticket, the ultimate test of course is long term profits. Good luck to you there too Acorn!

[acorn54]

equineer
thank you for your study on bris impact values relative to the prime number. i might be able to use it.
and thanks to rest of you for some intelligent input into the conversation on odds lines.
its conversations like these that makes this website the best bar none on handicapping the horses
until later acorn