Does anyone use longshots to identify track bias?

[Bill Cullen]

A friend of mine says he sometimes notices several longshots (he defines them as higher than 8/1) coming in the money in the first couple of races. He checks to see if they have the same general running style (on the pace/off the pace) and if they have the same running style, he designates the track bias accordingly (fast track/slow track). He then configures his betting for the rest of the races (dirt only) according to his determination of the track bias. He says he has had "some success" with this approach.

Any thoughts or comments?

Thanks,

Bil Cullen

[andicap]

Longshots to me are an indication of a bias, but not definitive proof. It sure is more of an indication than if a series of 2-1 shots finish first.
My barometer is if winning horses have %E levels that are consistently higher than they should be, e.g. for that track model or have top %E is a heavily pressured race. In normal races on normal tracks, the winning %E horse's will have a figure in the middle of the pack.
So if low %E's are winning when they shouldn't that indicates speed is dying.

The only problem I have is all this record keeping takes time -- and my damn job gets in the way. (How many of you KNOW you would win a lot more you had 25 hours a week more to keep records on what's going on and handicap more carefully? I know I would)

[Bill Cullen]

Originally posted by andicap


The only problem I have is all this record keeping takes time -- and my damn job gets in the way. (How many of you KNOW you would win a lot more you had 25 hours a week more to keep records on what's going on and handicap more carefully? I know I would)

You can frame that quote! That is the truth!

Bill C

[keilan]

Bill – I like the idea of cappers trying to determine the weight of the track early in the card. Your friend’s method of measuring track weight based on your description is faulted and will be wrong as many times as it will be right – thus little if no advantage.

Horses winning early in the card running gate-to-wire also don’t ensure that the track is typical or fast either. Low level claiming races fool many players in believing the track is playing fast when it really isn’t. If a player cannot accurately project the quality of each horse in the field, trying to assess the track weight is near impossible.

Andicap – You are getting closer to solving track weight with the use of energy. Listed below are some things you may want to consider.

1) learn to be flexible with %E (today’s field/winner may give you the predicted %E and it may not, know which type of races/fields when speed means more than %E and vice versa).

2) Estimate the pace of each race and learn to mentally adjust the %E figure. i.e.(how will horse A perform against a different pace scenario) Pace has more effect on the energy number than any other variable. Form, maturity, barn changes, etc need to be factored in also.

3) When a player fully understands energy you will no longer rely on track profiles with regard to %E. I know all the writers will tell you different, but they’re flat out wrong.

4) A few months ago Game Theory posted tracks surfaces rated from quickest to heaviest. This list will give some reference point to start from when playing/capping multiple tracks.

5) Okay now we get down to the most important aspect – Is your energy number reliable?

a) the variant calculated must be accurate
b) if the energy numbers vary in range significantly – I would check fractions and trip notes. Older male claiming horses are very reliable and tend to run very consistently.
c) If your software is adjusting for distance equalization this will skew your %E numbers considerably and make them un-reliable. I suggest you have your programmer provide ENERGY NUMBERS based solely on the actual distance run. You will learn to adjust the numbers in your head at a glance.

While I agree with most of your post, it is those generalities that you stated that separate the weekend warrior from the guy that plays daily. :)

[Bill Cullen]

Have any studies been done re later races on a card showing a disproportionate number of winners with a particular running style (on pace/off pace) as a function of the earlier races having longshots of a particular running style (on pace/off pace) coming in?

That would be VERY interesting.

Thanks,

Bill Cullen

[keilan]

Bill -- there probably is a correlation between running styles and long shots winning early, but that doesn’t assist you while your playing today’s card. The trick is to make the correct determination for the remaining races today. I fully understand what your friend is doing; I’m only giving you my opinion on its shortfalls. There are simply many other things to consider beside running style and horses that pay large.

[chickenhead]

Originally posted by keilan

3) When a player fully understands energy you will no longer rely on track profiles with regard to %E. I know all the writers will tell you different, but they’re flat out wrong.


would you be generous enough to give any further explanation as to what you mean by this?

I would think this might mean that the context (running styles of participants) of the race dictates the winning %E as much as the track profile, would like to know if this in any way what you mean.

[keilan]

Chickenhead – firstly track profiles can be misleading, it all depends on how the data is gathered. IMO all data should be filtered by class, distance, sex, surface, age etc.

Now back to your request: track profiles

Question --- can a horse that runs 6 furlongs whom typically possess an energy number of 53.50% win today’s race when the track profile shows that the range for winning horses is between 52.60% to 53.20%

Answer – maybe yes maybe no. It depends on strength of the field – pace/pressure – running styles of each horse – race shape – today’s track weight.

So if a player eliminates horses based solely on the track profile he has just missed a golden opportunity.

Trust the above example helps to clarify my comment regarding track profiles. Energy is a complex variable and should be used as a tool, if one makes absolute rules without multiple filters you will be left only spinning your wheels.

[kenwoodallpromos]

Use my track speed system; then use a stopwatch to time the first couple of races' fractions; then observe in the stretch who is, how many are, and what width of the track they are are fading; then use your longshot styles.

[Tadek]

Bill,

I am using three parameters.
One is calculated by mapping a running time of a race into a range of max – min values for that distance. The value computed is between 0 and 100. Obviously the first race has the value of 0 and following races have values aggregated from previous race values (fuzzy inference strategy).

Other parameter is calculated with payout of the winner. The same processing is done on this value – mapping of the value into a range of 0 – 100 and aggregation of values from preceding races.

Finally, the third parameter is a subjective value I enter to indicate if the speed holds or if it fades. This value is also aggregated for all consecutive races in the current day.

The resulting values are entered into the database and later can be used for profiling. You realize that this methodology allows for validation of a fairly wide range of possibilities relative to the current track condition. One can verify if a high value of speed correlates with short-odds-winners or large value of payouts indicates better long-shot opportunities and so on.

At this point however it is a project in progress. I have all the tools but I simply did not have enough time to enter the data and test it on a sufficient sample.

Regards

Tadek

[Fastracehorse]

Bill,

First off, assuming that a longshot runs well because of a track bias is quite an assumption.

There are many examples of horses that win at long prices because they deserved to - not because of a bias.

Having said the above, biases do occur and hence, are important - but assuming a horse ran well because of a bias, and not because you may have missed something, is a dangerous habit.

fffastt

[JimL]

keilan, I just printed this thread. Thank You! JimL

[keilan]

JimL -- your welcome

Time for a sidebar regarding the example of the 6-furlong winner.

Here’s the beauty in that scenario, the horse controls the pace and knocks off fractions that are comfortable for him, stays brave and runs to the wire at decent odds. This race is run every day.

Afterwards when the variant is calculated, the energy number mystically falls into the predicted range (52.60% to 53.20%) and is now part of the population for energy analysis.

Now that the trainer has figured the horse out (wink-wink) it’s time to move him up in class or race in the same or another condition. Now he gets horses that will make him run quicker internal fractions, maybe the track weight today is heavier, and that’s all it would take and this horse becomes a non-contender. A good horse to play against, again this race is run every day.

If you follow the track profiles they allow you to miss when he wins and the time you wager on him he is a false fav. For database jockeys it is imperative to understand what you’re mining. ;)

[Hosshead]

Originally posted by Fastracehorse@DRF

First off, assuming that a longshot runs well because of a track bias is quite an assumption.

There are many examples of horses that win at long prices because they deserved to - not because of a bias.

......assuming a horse ran well because of a bias, and not because you may have missed something, is a dangerous habit.

fffastt
So true.

[hurrikane]

Interesting thread. Not for the bias discussion though.

I'm of the belief that 99% of the handicappers wouldn't know a bias if it kicked them in the ...well anyway. Many longshots coming ITM by itself is meaningless.

I like Keilans thoughts, IMO..many data factors, especially those that use a variant, are meaningless if you do not take into account the strength of the field. Pars are meaningless for this and is the reason for the downfall of a lot of speed/pace systems.

Please, don't ask me if I have the answer. I don't. I can win at this game picking 20% winners and waiting for the odds. That means I have 80% losers. I can make good money..but I would love to have 100 more plays a week and 5% more winners.

[andicap]

Keilan,

I already do No. 2 below -- a central tenet of energy. High %E's win all the time when they can relax and pace themself so a 53.00 can become a 52. I love to bet against these types when they win big but come up against more pace and end up repeating their usual 53-54%E and finishing out of the money.

As for No. 1, I've found I can use %E in almost every race to at least eliminate a contender unless they are all bunched so closely that unless I have a form cycle or trainer angle insight I'll pass or see if some 8-1 shot or more can win. You're definitely right in that in many races %E will not get you all the answers or even an important one. But if it can point me to 1 contender out of my final four, who are severely disadvantaged, I'll take it.

I'm still not sure about track profiles. Yes, yes, I know they differ dpending on class, etc.so the usual models are too broad, but they arent as broad as one might think once you get past younger horses. I haven't found any difference in 14K and 50K claimers in NY.
I don't really depend on the profiles -- I compare the horses with each other, something the "experts" dont teach you, but which I feel is a much better way at looking at energy. Each race is its own puzzle and rather than compare the horses' to some artificial standard I want to see how the fit relative to their competition.

As for how they are computed, I have no idea how HTR does them. All I know is they consistenly work for me. How do they work for me: %E can generally explain to me what happened during a race even if I couldn't. I horse that looks on paper to be a contender but who'se %E is the lowest among the contenders is a doubtful prospect unless there is a complete pace meltdown. A lot of longshots come from that area because they often lack the glitzy speed figs of their "E" rivals and they plod along and occasionally pick up all the pieces. They are excellent horses to use for 3rd in trifectas tho.







Originally posted by keilan
1) learn to be flexible with %E (today’s field/winner may give you the predicted %E and it may not, know which type of races/fields when speed means more than %E and vice versa).

2) Estimate the pace of each race and learn to mentally adjust the %E figure. i.e.(how will horse A perform against a different pace scenario) Pace has more effect on the energy number than any other variable. Form, maturity, barn changes, etc need to be factored in also.

3) When a player fully understands energy you will no longer rely on track profiles with regard to %E. I know all the writers will tell you different, but they’re flat out wrong.

4) A few months ago Game Theory posted tracks surfaces rated from quickest to heaviest. This list will give some reference point to start from when playing/capping multiple tracks.

5) Okay now we get down to the most important aspect – Is your energy number reliable?

a) the variant calculated must be accurate
b) if the energy numbers vary in range significantly – I would check fractions and trip notes. Older male claiming horses are very reliable and tend to run very consistently.
c) If your software is adjusting for distance equalization this will skew your %E numbers considerably and make them un-reliable. I suggest you have your programmer provide ENERGY NUMBERS based solely on the actual distance run. You will learn to adjust the numbers in your head at a glance.

While I agree with most of your post, it is those generalities that you stated that separate the weekend warrior from the guy that plays daily. :)