par times

[aaron]

I realize that each race class has a par time.I also realize that there are races which are run faster and slower than par.The question I have is if we are comparing horses by speed ratings and patterns where does par time come into being? A horse may have run a slow race according to the par of the race,but if he is against horses who are also slow,what is the advantage of knowing par times?I know a horse may have won in a race that comes up below par and is now in against horses who ran above par,but wouldn't his slow race show up in the speed ratings versus those of the other horses?

[cj]

Aaron,

I can't tell you why, but races where no horse has recently ran to the par for the class do not usually handicap well with speed figures. I have found figures to be useless in these situations.

CJ

[Dick Schmidt]

Aaron,

In the situation you describe, pars are indeed useless. But throw in a shipper from a track you never follow that has run a full second faster than any of your nags and then try to decide if it is for real or just ran on a faster track. For that, you need pars.

Dick

[Lindsay]

CJ wrote:

"I can't tell you why, but races where no horse has recently ran to the par for the class do not usually handicap well with speed figures. I have found figures to be useless in these situations."

This is an interesting point, CJ. Many people agree with you that figures are less effective in these situations. Why are they less effective? Here's a guess: If horses are entered for, say, 25,000, in all likelihood, at some point in their careers they have given evidence that they belong at or near that level, probably by running figures that are competitive at that level. But they aren't running them now. This is a situation that seems ripe for horses to make seemingly random jumps back to their best form, hence the unpredictability of such races.

[Tom]

I agree with CJ-if nothing has run close to the par, look for other things beside speed. My favorite angle here is a dominant pace horse that quits-today, he could-go-all-the- way!
Or a layoff horse with back class is sometimes a good price.
My rule of thumb is I want a horse to have run withing a couple lengths of par. Beyer-wise, at 6 furlongs, this about 5-7 Beyer points.
There will be a lot of opportunites on the inner track - a lot of races in the $25,000 - $35,000 races will be either very good or very bad for the class. You can catch some nice prices in this class range.

[LOU M.]

Try to read Figure Handicapping by James Quinn,I know I suggested this book in another thread for a different reason and no I'm not related to him but the book has a wealth of info and will point you in the right direction about when and how to use pars.

[jotb]

Hello Aaron:

"I realize that each race class has a par time".


Each race condition has a par time but keep in mind that pars are man made and if one is relying on par times to calculate a track variant, it's significant that the par times are accurate. Keep in mind that you are working with 9 races on average per day to calculate your variant. If your desire is to work with the "A" racing circuits, then the job beomes more difficult to calculate the true speed of the racing surface. The reason for this is because the major racing circuits will card several racing distances from 1 turn to 2 turns, including various race conditions, gender, and age groups. What happens to most variant makers in this type of situation? The variant maker, with not much to work with in terms of races, for the day will now wind up with 1, 2, or 3 races that are extremly strong or weak. The question you are now confronted with is, are these races as strong or weaker than they should be? Your adjusted variant brought you to do this situation but it originally started with your condition pars.
Now for me, the whole idea of speed figures is not so much comparing 1 speed figure to another but more in the way of finding stronger than normal races per class level. I want these type of races and follow-up on the horses within this race. Racing secretaries in NY will write as many as 6 races a month for NY Bred MSW colts at a specific distance. Same class, same purse but many times different horses. Some MSW's will be stronger than others in terms of pace and final times and eventually most of these horses will compete with one another on a specific day. This is where a variant maker should be able to identify which horses can substain a certain expected pace or final time from their previous MSW races prior. If you make an error from these previous MSW races then the whole process is dead. If I attached a fig that is much higher than normal, I will follow-up on the winner from that MSW race and check the next race that he runs in ALW N1X to see if that race fig was true or false. Of course the horse is now running in tougher condition and normally will take a couple of races to get through the next condition unless he is just a freak. If the horse runs close or slightly above the last fig in the new condition the following race, I am now confident that the race was strong and now wait to pounce on those MSW's that were in the strong event. Tons of work and a pain in the neck but it need's to be done in order to achieve success. Like I said earlier, when you are not working with the same conditions and distances on a specific day the variant calculated might be distorted somewhat. Of course we have a choice of incorporating 2 variants, one for sprints and one for routes but once again we are not working with enough of the same type of races each day. My theory is to have each race condition par done seperately at the racing circuit and there are plenty and take it a step further by incorporating several groups of conditions into one group. This procedure tightens things together somewhat. Believe it or not, you can incorporate certain race conditions for males and females into one specific group. For instance, in NY and like most other racing circuits normally colts are generally faster in terms of final times but some of the higher class filly races will have the same average final time as compared to lower class conditions for colts. The accuracy becomes greater in terms of pars if for example, you incorporate the filly NW1X with MSW's for colts. Each might have a same average final time of 111.00. My point here, is if you condense race conditions into groups the chance's are greater for an accurate variant than just working off the single pars per condition. If you use both systems, it should help establish an even greater accuracy from pace to final times. This should help the process when developing your track variant and at the same time identify or not make the mistake of making a race stronger or weaker than what it should be in reality. I'm not sure if this makes sense to all of you but I tried the best I could. I'm also sorry that this was long and drawn out. I have a hard time explaining these thoughts and ideas in writing. Appreciate any comments on this.

Thank you and happy holidays to all,
Joe

[aaron]

Joe- Thanks for your reply.The amount of work that goes into making a correct set of pars seems unbelieveable.Do the results compensate the amount of work required?Would just looking for patterns and keeping track of key races yeild similar results?I don't have the answers to these questions,but the topic is fascinating.Its been my experience that sheet players do not use pars,but are heavy on patterns that horses develop.What would be the best use of pars? From what I could gather if a player becomes as skillful as Joe is with his pars some good bets could be made.
I believe each race is different and what we use as a key ingredient in one race may not factor in at all in another race.
An example of this is trainer patterns,which I may use as 80% of my handicapping in one race,but might not have any effect on my handicapping of another race.

[Fastracehorse]

After reading your post it is obvious to me that U re trying to find overlays based on identifying stronger than expected races for a certain class.

The method U choose to do this by is calcualting subjective pars for these classes. If the winner of a race ran higher than par then U have identifieda # of possible overlays.

Did U know that U are doing the same means as chart work analysis??

Except chart work is easier and is more objective. That is, if a winner of a MSWt event comes back to beat N1X foes then U circle that winner of the MSWt event. Therefore, U again have a # of possibilites for overlays, based on exactly what U are doing with pars.

Where the charting method is superior to yours is in time-effectiveness and less subjectivity. And, that is why I suggest U take a look at doing chart work with DRF charts.

U're method has the advantage of knowing immediately that a horse has run better than par, assuming of course that your pars are correct. Having said that, visually impressive winners also lead to key races and again, is much easier than creating par times.

So to sum up, while creating pars is admirable, it is this handicappers position that chart work does exactly the same thing but with much less time involvement.

Joe, can U think of another reason why someone would want to do pars other than a need to work harder than usual??

fffastt

[OnTheQT]

Pars are of some advantage, if accurate. They are supposed to refect the average time ran by the winner at a given class and distance according to sex and age on a fast track.

Raw time without modification is not a very good method for determining advantage or disavantage.

[andicap]

Fast,
The problem with the circle method is that often times you don't know it's a key race until it's too late. One circle doesn't denote a key race and lots of times all you get are two circles.

I also found when doing the key race circles that it paid to denote when a horse came back to finish 2nd at long odds (say, 8-1).

I stopped doing it because there were so few times when I was able to capitalize on a key race because of that.

the best use of chart work I would think is finding horses who raced well against a bias (a real one) and are coming right back, especially if they finished poorly and are dropping in class. or getting a positive jockey change, etc.

[Fastracehorse]

True Andi,

Not every key race has a next out winner.

But that is not the point. The point is one is trying to find evidence leading to overlays and key races are an excellent tool for that.

I don't know what method U use but whatever it is, it doesn't find U an overlay every time either. Such is the nature of handicapping horse races.

Finding overlays will not be found under every turned stone. However if U turn enough stones with the right tool, there are enough nuggets to keep U interested.

Chart work is important.

fffastt

[jotb]

Fast:

I am not sure if I understand your reply. I have devised "condition par charts" and do not see a relationship between "condition pars" and "result charts". The condition pars are used as a foundation for "speed rating figures". In order to have accurate condition pars a numerous amount of work is needed. The main ingredient that is needed to obtain accurate pars are "past races". The actual amount needed for this is around 10 to 15 thousand races which is about 5 years worth. You might think this is ludicrious but to find an accurate average per condition, per distance, I feel this amount is necessary. There are race conditions for distances that are written less than a handful of times each year. How can you obtain a true average of this type if you are working with only 5 to 10 races over 3 years. If you incorporate these type of conditions into other race conditions or just leave them out because of the lack of races, I can assure you obstacles will comes about. For instance, if you leave out some conditions because of the lack of races and several of these races are now carded for a given day, you will then be forced to calculate the speed of the racing strip on this particuliar day with just a handful of races. What happens when the races that you are left with on this day are run at different distances. For instance, 3 races were used that you did not have a condition par for which leaves you 6 races to work with. Now the 6 remaining races are run at distances as follows: 5.0, 6.0, 6.5, 7.0, 8.0 and 9.0. If you make 2 variants for the day per sprint and route. I can assure you that each of these races once adjusted will have major differences. Some races will look much stronger than what they really are and vice a versa. This is not enough races to retrieve an accurate variant. It is difficult to obtain an accurate variant with 9 races nevertheless using only 6 races. For these reasons and then some, I have come to the conclusion why many handicappers that devise their own condition pars become discourage and proceed to create or find other avenues. I have found ways around the above obstacles and feel quite confortable with my condition pars as the stepping stone to accurate speed figures but the use of result chart analysis is only necessary when I feel that my speed figure rating for a particuliar race might have been off. I reinforce this by keeping a close watch on the horse that came from this event to only see if my number was a proper fit. The bottom line is, speed figure's are used as a guide to eliminate the pretenders from the contenders, to be able to identify horse's upward and downward racing cycles and lastly, to be able to identify slowly maturing horse's that are on the improve. There are other reasons why speed figures are used but for me, these are the top three on my list. My desire is to have speed ratings per horse that are better than what is out on the market today. This goal helps to have an edge over the general public. Lastly, no matter how good ones speed ratings might be, there will always be the need to use other handicapping factors before final decisions are made for investment and even then you need some racing luck on your side.

Joe

[Dave Schwartz]

Jotb,

I think I will weigh in here with an opinion.

First, there are different types of pars and you must match the par type with what you are trying to accomplish.

We make 2 types of pars:

A. Track-to-Track - Perhaps a better name would be "distance-to-distance" because the purpose of these pars is to address the question, "How fast is a particualr distance?" These pars are only used to adjust from one track/distance to another and are made from the theoretical $10k par at each track.

B. Class Level - These are the pars which, theoretically, represent what it takes to win at different levels.

In our approach to making pars, we concentrate on "A" to get the track right then we go back after to look at the Class Levels.

We begin by trying to understand the class relationships at the most common distance. Specifically, we ask the question, "What was the time difference between each class level and the track par?"

Once we have nailed down the time differences for each class, we etch them in stone and attempt to duplicate those relationships at the other distances.

One cannot simply say, "What was the average time for NW3 at 7 furlongs? because, they may only have run once or twice at this distance all year. And maybe not at all.

Most people are surprised to find that most tracks do not run all class levels at all distances in a given year. But, once you think about it you realize there aren't that many races.

Most tracks run 10 or more distances (including turf). They will also run 12-14 class levels, and don't forget that females are different than males. The most common distances (6f, 8.5f) at a given track will represent probably 1/4 of all races and the most common classes (3 core claiming classes, 1 maiden claiming, msw, and NW1) will represent probably 2/3 of all races. (I don't claim these ratios to be accurate but you get the idea.)

For example, how many male, NW1 races have been run this year at Santa Anita at 6f? Answer: 6.

How about NW3 at 7fur? Answer: 1.

And what about 1 1/4 miles? As they say in New York, "Fugettaboutit."

And remember that Santa Anita runs quite a few days. What about the tracks like Fairplex or Timonium that only run for a couple of weeks or less?

In spite of all this, you do not, in my opinion, need 5 years of data to make good pars (although there are some tracks where you MIGHT use that much data for turf distances). You typically need 2 years of data to make "track-to-track" pars.

The issue of making good pars lies in the continuity from year-to-yea in the class levels. When you understand the class levels, you wind up with good pars.

Just my opinion.

Regards,
Dave Schwartz

PS: Our 2003 pars will ship by January 15th this year. Stay tuned. <G>

[jotb]

Hello Dave:

I figured someone would feel that 10k races are just too much of a sample. I'm not so crazy about the short-cut idea of projecting what times should be for certain conditions because the sample is not large enough.

I feel even with the large amount I suggested earlier, it would wise to use another format to tighten up the pars somewhat. Then again depending on the racing circuit that you choose to build the condition pars from is another story. I believe that if racing strips are not changed every few years as far as the base, banks, and compostition, then why not have a large enough database of races to find the true average for all conditions and distances.

The same holds true for purse structure. Just because a racetrack has slots and the purses rise somewhat does not necessarily mean that a MSW is different than another MSW that was run before the slots were put into the racetrack. Then again the quality of horses could be different but this would depend on the racing outfits. For instance, NJ racing does not have the slots but I feel that there is quite a difference in the quality of horses between Monmouth and the Meadowlands. Monmouth will have horses that are of greater quality than the Meadowlands because of purse structure and racing outfits that will disappear after the Monmouth meet is over. The funny thing here is, that final times at the Meadowlands are faster than the final times at the Monmouth. The same conditions but the final times are faster over the Meadowlands strip. Are the horses that were running at the Monmouth meet all of a sudden rapidly improving in terms of final times. I think not! I noticed that Beyer #'s were higher for horses at the Meadowlands than Monmouth on average per horse. Why is that?

The adjustments from racetrack to racetrack within a racing circuit is without a doubt a must. The problem here is how much of an adjustment and what factors are used leading up to the adjustment. This is why I feel that devising speed ratings for every track in the country can be a difficult task especially when it comes to accurate speed figures. This makes for a tremendous amount of work in terms of understanding racing strips within racing circuits and the knowledge to be able to identify the quality of horses within the racing circuit at different times of the year.

Dave, I feel you are an extremly knowledgable handicapper and respect your opinions on this subject because of the experience and hard work you have put into this but I firmly believe there probably is no exact format that can be used to devise speed figures. The accurate part can be achieved but this all depends on how much weight you place on each factor pertaining to this end of handicapping. JMHO

Best regards,
Joe

[Dave Schwartz]

Jotb,

I agree with what you are saying... You cannot simply build a set of averages and call them "par."

And for some tracks multi-year is critical. There are some tracks where you would simply be guessing without multi-years. Especially on the turf. Sometimes on the turf I have had to consider as many as 4 years of races to get a handle.

And, as you said of me, I must have the same respect for you or anyone else that attempts to make pars for every track. If they work for you, that is all that matters.

Best Regards,
Dave Schwartz

[kitts]

Dave Schwartz is right on. Those pars and/or track to track numbers take humongous efforts to say the least. I was at Cynthia Publishing where Gordon Pine first created his pars. He scheduled his vacation for the day after they were done every year. I was with HDW when Jim Cramer started his worthy numbers and saw the effort he put into track to track figures. He did one circuit at a time and it took him hours each day to keep numbers current for each track. One circuit after the others. Comparing horses that won at this surface/distance/class against other horses that won. And then place horses. And on and on. I can only admire the efforts of Dave.