Form from Bris

[sidvicious]

does anyone use the bris pace figures to determine form? i'm refering particularly to the E1, E2, & Late figures?

i'd appreciate some tips here if you're willing to share them.

sv-

[mhrussell]

I have been using something I call 'integral total energy', which is a function of E2,LP and the distance of the race. I calculate this as follows:

IntEnergy = 2*(dis2* E2 + disLP * LP)/(dist)

where:

E2: BRIS 2nd call pace figure
LP: BRIS late pace figure
dis2 = distance from the start to the 2nd call
disLP = distance from the 2nd call to the finish
dist = distance of the race

So, this is a distance weighted sum of E2 and LP which I treat as total energy expended during a race. By looking at how this varies, race to race, it helps me see which horses are improving, declining or staying about the same. I have noticed in doing this approach for about 9 months now, that good improvement or regression often correlates to about +/-11 points in integral energy over a horses 'rated race'; where the rated race is either the best race in it's form cycle window (see handicapping magic related threads) or something between that figure and the figure corresponding to the horses last race. There are no fixed rules for this, it depends. I also will take a poorer run race in the FCW if the horses figures to 'bounce' (for any of 5-6 bounce 'reasons' I consider) or otherwise indicates regression due to a poor trainer claim, poor (or no) recent works..layoff pattern.. whatever. I also tend to be more 'forgiving' if the horse figures to be a price..I may stick with a better FCW figure in that case.

I am still experimenting with this idea, so I also would be interested in what others are doing in regards to 'form' analysis using the BRIS pace/speed figures.

best,

[garyoz]

does anyone use the bris pace figures to determine form? i'm refering particularly to the E1, E2, & Late figures?
sv-


You might want to check out the literature on pace figures and figure handicapping. For example, Brohamer's Modern Pace Handicapping, more advanced but probably the best, Fotias Blinkers Off. There are plenty of other books. Ken Massa also makes his newsletters available at the HTR web site and does great research, often focusing on pace variables.

You can also do a search in the Handicapping Library Forum on this site for some other ideas.

This is not to suggest that the Bris numbers are very good. That is a whole other discussion.

[NateBracey]

Personally, I do not like the accuracy of the Bris pace figs at all, but don't know how that would releate to using them to determine form. I make my own
figs and put a lot of weight on last race pace for my selections, but the
Bris figs do nothing for me.

[Big Bill]

Matt,

In a post to another thread awhile back you used the same formula for calculating a "pace figure", except it did not multiply the result by two as you have indicated you do to calculate your "integral total energy". Is there a reason other than obtaining a larger number?

Big Bill

[mhrussell]

Big Bill-

The factor of 2 is needed to get a distance weighted average total energy (E2+LP). The other post referred to a total pace number, to taken/compared as a "speed" type figure.

[Zaf]

I have been using something I call 'integral total energy', which is a function of E2,LP and the distance of the race. I calculate this as follows:

IntEnergy = 2*(dis2* E2 + disLP * LP)/(dist)

where:

E2: BRIS 2nd call pace figure
LP: BRIS late pace figure
dis2 = distance from the start to the 2nd call
disLP = distance from the 2nd call to the finish
dist = distance of the race



Very Interesting Matt,

I programed your formula into a template in the formula view in Equisim. I used the horses last 4 past performances and then compared the change from race to race. There are some pretty huge swings from negative to positive & vice versa. Could you give a bit more detail on how to interpret the swings ?

ZAFONIC

[mhrussell]

Zafonic-

What I've noticed in looking at a bunch of pp's, is that you will see the older horses tend to 'cycle' or stay within +/- 8 Integral Energy points; the 3-4 year olds are often still 'improving' and may easily run a much better IntE. But it is uncanny how this 'improvement' / regress cycle runs about 11 points. So if I figure a horse to improve, I usually project about 11 points to the last total E figure. Then, by looking at the pace scenario (and the fulcrum horse) I 'divy' up the energy points such that the resulting E2 figure places the horse positionally (and beaten lengths wise) 'correctly' according to the horses running style and the remaining points are then assigned as the LP rating (LP = IntE - E2).

For horses on the decline, if you expect regression, I will assign an Integral E from one of the poorer performances; and not the 'best' rating within its form cycle window. Then I will 'project' based on that integral energy.

Like all the paceline selection/modification methods; you get better with lots of practice. I have found that this does improve/builds upon the classic HM form cycle window approach, and if restricted to races you have all ready deemed 'worthwhile to handicap'; the extra time needed to do this paceline energy projection is well worth it!

With EquiSim, you might be able to run a bunch of simulations and see how this energy parameter of mine correlates, not so much with only winning, but improvements/declines... and get some data on that 'magic' +11 number too!

I will be interested if you or some of your other EquiSim cohorts could look into this.
Might be enlightening for us all.

best,

[Zaf]

Matt,

Thanks for your response. I have added it on to an Equisim template that I use regularly.

ZAFONIC

[grahors]

Gentlemen,
A quick question...I to use bris numbers..am I right in asuming the idea of 1=2 with regards to EP vs LP. An incr in EP of 1 pt would require a decrease of 2 pts in LP? I thought I read this somewhere in another forum but don't recall if it applied to Bris numbers.
Grahors.
I like what I'm reading here and will incorporate in my hcping.
Matt...thanks for that speed like equation in my other bris thread.

[michiken]

I also have been using bris pace and speed figs for about 5 years. I find that they have there idiosynracies so I also calculate turn time and final fractions from them i.e.

2f = 95 4f = 98 6f = 72

Turn Time = (98-95) +3 Final Fraction = (98-72) = -26

To hunt down better prices, I look at the final fraction of the horses that posted the higher speed figs. If their Final Fraction is -15 or lower, the horse tends not to run back to his fig. If the horse tends to have a negative turn time as well as final fraction, I either don't use the paceline or throw the bum out of my contenders list.

I often see cheap claimer races where all of the horses have nasty final fractions and rank accordingly.

[bobhilo]

michiken......

i use bris ultimate pps(free computer ratings no program to buy) lots of data in the race summary...lots of ratings...keeping records of what wins at what track and distance

i bet MNR Pen SA(so calif hol dmr) FG and a few others

turn time calculation i get

last fraction minus second call i get(yours comes out usually minus)

what, if i may ask do you do with the figures?
does it work best in sprints? routes?

a quick race I did using your figuring

SA 4/3/05
race 1 6furlongs

I used last pace line most recent

last fraction ratings

#1 -40
2 -22 wins pays 9.80 was 3rd bet horse behind #7(favorite) #5(DNFinish)
3 -22
4 -36
5 -14
6 -43
7 -8 (won last race,,morn line 5/2 today)
8 +7 last line (134 day layoff)routed 8.5f E1 1/2 mile call 72 6f call 79

result 287 $9.80 win exacta $144 tri $402

what do you do with ties in numbers? # 2 and #3 both -22 last fraction

throw out # 1 #4 #6(bums?) leaves 5 horses

what are you looking for? win bets or exotics?

maybee it is a poor example to use....i don't know....

it did get the big exacta $144 in best 5 and $402 tri

Michiken....please clarify a bit or use more examples

I too am a new board member....used sartin's Synthesis until I see similar ratings for free from Bris....

I really do not want to use some program (equism hsh htr synergy6 as discussed on this board.)..and lose money despite having a program....

any further input would be greatly appreciated

[exactajack]

Didn't know if you guys that are calculating POR from bris numbers are aware but the E2 number for races <6 furlongs is calculated from the 3/8 pole.
So if you use the 2cBL*2+E2 for any dist <6 you'll get different POR's for horses that have come out of the same race.

[kgonzales]

I also have been using bris pace and speed figs for about 5 years. I find that they have there idiosynracies so I also calculate turn time and final fractions from them i.e.

2f = 95 4f = 98 6f = 72

Turn Time = (98-95) +3 Final Fraction = (98-72) = -26

… I often see cheap claimer races where all of the horses have nasty final fractions and rank accordingly.

michiken,
I'm not sure if by the 6f=72 you are referring to LP. There isn't a 6f pace rating in sprints in the BRIS files. In routes <10f the 6f pace rating is the E2 rating. If you do mean "LP", technically the LP number already is measuring the last fraction (2nd call to finish) only. The calculation you're doing is finding the dif between two distinct sections of the race and comparing them, I guess. I'm not familiar with the more advanced pace programs, but maybe this is more akin to a deceleration number(???). But hey if it's working call it whatever you like :)
I would think the (improving) turn time number would work better in cheap claimers. Early speed seems to be more important at cheaper levels, but maybe I'm being too conventional in my thinking.

k
g

[Light]

This E=MC2 is above my head,but I can relate to the original question. My Excel spreadsheet program has about Eleven seperate categories(columns) extracted from the Bris form. Among them I use a category which gives a total pace rating using the Bris E2+LP for the paceline I picked. At first this was an experimental category. I was getting most of my winners out of my "speed" and "overall totals" category. My speed category is also a total pace rating but,Sartin/Brohamer based.

What I've found is that the "Sartin/Brohamer" based speed rating and the Bris "E2+LP" rating give me about the same number of winners on their own. What I find advantageous about this is that sometimes one will rank a horse poorly and the other highly. It's sort of a checks and balances,because one process is catching something the other isn't. That signals me to look at the horses form and decide which category is telling the truth. The strong part is when both categories rank a favorite low,I can throw him out with confidence as this has a high sucess rate. And when you can throw out a favorite succesfully,you have the edge over the house.

[michiken]

My method of using the pace figs is not well defined and requires out of the box thinking on some parts.

In the program that I wrote, I calculate the following rankings:

An average pace rating (2nd call + Speed Fig)/2
An ability time rating (2nd call + turn time)
My 'Final Fraction' or Deceleration Number.

In my program, I have found it best for myself to search only the last (2) pacelines. Depending on how the track is running, I will choose one of the following reports:

Sorted by Pace Call - Front Runner or Early Horse Day

Sorted by Ability Time - Front Running/Close to the Pace Day

Sorted by Turn Time Alone - I use this to find pace attackers.

Sorted by Final Fraction - Come in handy on the turf.

Sorted by Speed Fig

Sorted by My Terminal Velocity Fig - Turn time + Speed

As I see it, most races are a game of follow the leader or early speed. I use the Turn Time and Final Fraction (probably more of a deceleration number as stated in the above posting) to try and run down the leaders. I look mostly for exactas.

Here is an example of a report. It was sorted on my TV Fig.

GRRRRR - Sorry I gotta learn how to post tables in these forums.....

[michiken]

Ok, I got frustrated with trying to make pretty tables in this forum so I have attached a pdf explaination.

[garyoz]

In my program, I have found it best for myself to search only the last (2) pacelines.


Basicly then you use best of last two, or average of last two? Paceline selection is the critical component of almost all software approaches. It is something that you should give alot of thought to. By automatically limiting it to the last two races I would suggest you are limiting your upside. There has been a ton written about paceline selection on this board and other place. It was also the basis of the Sartin/PIRCO workshops for many years.

I've come to the hypothesis that if you pick the same paceline for each horse, regardless of the program or data source you will come up with pretty much the same selections. I haven't run the data to back this up, but I have "collected" and used many software programs over the past decade often using a few programs simulatenously.

[Zaf]

Michiken,

Thanks for sharing what you do. I always like to see what other punters are doing. Do you use excel to crunch the numbers? How do you get the race conditions on the header of each race ?

ZAFONIC

[michiken]

Depending on the amount of time I have, I generally look at the last 2 races only.

The reason I do this is because I search for internal fractions or positive indicators that the public may miss. Sometimes I find a horse that has a lousy speed fig but a monster set of internal fracs (acceleration).

The 2nd reason that I do this is because of the way many horses run at different distances. An example could be when a horse raced in a route tightener in his previous race and cutting back in to say 5.5 or 6 furlongs today. If I only chose a paceline at the distance, perhaps I would miss the route tightner?

Also depending on how mentally drained I am from work, my program is flexible about how many pacelines I can go back - I can use 1 to 10. If I have some free time, I may look deep into a race.

Sometimes I just want to keep it simple because looking at numbers all day and night makes my head spin. In this case, I take the best paceline of the last 2 only with this *Caveat*

I watch the races and try to determine what types of horses are winning. If I see something I like, I may re-run my report and sort the data differently.

I agree with your point that paceline choice is subjective. I will add this very good suggestion to my 'things to research' in the future.


Basicly then you use best of last two, or average of last two? Paceline selection is the critical component of almost all software approaches. It is something that you should give alot of thought to.

[michiken]

My warhorse database of choice is an old DOS based program called Paradox. I learned how to program in its lanquage in the 1980's. Due to its column limitation, I have to use the multi file data from bris. The program I wrote imports the comma delimited text into 3 separate databases. I then wrote a query to extract the pacelines, M/L odds, race conditions, etc.

I don't like Excel and I am not a big fan of Microsoft products (because I am forced to use them at work). I have been using computers since 1979 and sometimes can be stubborn and old fashioned when it comes to computers.

I am in the process of converting my computer over to Linux (operating system) and open source software because its is Free as in Beer. Even this website is using php scripting language and mysql database to store data for this forum.

Michiken,

Thanks for sharing what you do. I always like to see what other punters are doing. Do you use excel to crunch the numbers? How do you get the race conditions on the header of each race ?

ZAFONIC

[whobet]

Hey michiken, does your computer have a "mouse"

[Zaf]

You sorted your figs by your TV number. In what instances do you find that number useful ?

ZAFONIC

[michiken]

Hey.........!

- of course my computer has a mouse LOL! Just cause I like old software doesn't mean that I like old Hardware......... and my avatar is not reminiscent of the announcer at Mountaineer despite what someone inferred.

Zafonic,

To explain my fig TV fig affectionately dubbed 'Terminal Velocity':

In every pace handicapping book I have ever read, it seems that it concetrates on who will hit each call first. To me this seems to rank horses on average pace or a linear fashion. These kind of methods seem to point only to one type of horse [E] or [E/P]. I wanted a way to find pace attackers.

Many years ago, I watched a video seminar where Dick Mitchell (i think) explained how he ranked contenders with his 'ability times'. This rating consisted of adding a horses turn time to the Pace call. This intrigued me because it added an element of acceleration to the formula.

I found that using the ability time method is often good at identifying who can lead a race but I still found it a little lacking. While it identifies front speed, it did not explain the closers or stalkers that fly down the lane or why a high ranked abilty time horse died in the stretch. That is why I added a decel or FF rating.

To create the TV Fig, I tweaked the 'ability ranking' for my use. Instead of adding the turn time to the Pace Call, I add it to the speed figure. This gave me more of a 'sustained ability' rating. It also helped me find what I call 'flat liners' - horse that seem to have the highest figs at every call but call it a day in the stretch.

I have about 5 years of data on my machine and spend time redboarding races. Using this Fig did help explain a few things like 'why did a horse that ran a speed fig of 95 get beat by one who ran an 84'. I am sure it has to do with the pace mix of the race as the 95 horse prolly got softened up in the early stages and the 84 rated horse had enough in the tank (as indicated by his internal fractions) to run him down in the stretch.

I find this rating useful at most distances and cheaper horse tracks except when the track is acting like a conveyor belt to front speed. It does not always seem to hold for the higher class tracks like GP, AQ and SA even though I do examine it. In my opinion, I think it more effective at routes.

As we all know, there is no 'black box' that can do it all. This is just one simple tool in my arsenal. I can say that I have had days where this rating shows me high priced winners, exactas and tri's.

One of my goals is to add a morning line to my sofware so that I can create one using this ranking. Normally when I make my selections, I will scan down the top 4 or 5 ranks and throw out the decelerators. Then I will take the highest odds horse and key them over the others. The unfortunate part is that I do not have an accurate formula to base an odds line on.

[Zaf]

Thanks Matt & Michiken,

I enjoyed reading this thread !

ZAFONIC

[DJofSD]

My warhorse database of choice is an old DOS based program called Paradox. I learned how to program in its lanquage in the 1980's. Due to its column limitation, I have to use the multi file data from bris. The program I wrote imports the comma delimited text into 3 separate databases. I then wrote a query to extract the pacelines, M/L odds, race conditions, etc.


Nothing wrong with Paradox. Yes, it may be long in the tooth and Borland is no longer supporting BDE but being X-based you can still use it and code software to use it.

DJofSD

[Big Bill]

michicken,

You stated in your earlier post that "to create the TV Fig, I tweaked the 'ability ranking' for my use. Instead of adding the turn time to the Pace Call, I add it to the speed figure."

I plan on doing some checking of my own with your TV Fig, but before I do I need to make sure that I understand what you add the turn time to. I assume that you add it to the speed figure of the horse and not the speed figure of the race, i.e., the winner's speed figure.

Big Bill

[michiken]

Yes I add the turn time to the horses speed fig.

I will post all of my formulas at a later date when I have more time.

Long Odds to All!