View Full Version : Keeping pace shape records
andicap
08-08-2003, 08:51 AM
Does anyone have success keeping extensive records of pace shapes (various ways to define them, such as E-E race with lots of E types, etc.)?
That way you know a horse who just emerged from an E-E race for example and likes to race near the front was at a disadvantage and if today's shape is say a E-P race (1 E and several pressers) he will have it easier.
Larry Hamilton
08-08-2003, 10:13 AM
This is another facet of handicapping that at first blush looks simple--it isn't even possible. Here's why:
If you use Running Style or Quirins Speed numbers is not relevant. The key is the math. With running styles, there are E, EP, P, S and NA. The question is how does a race's shape affect an outcome, so the first task is to determine the shape. There are an average of 8 horses in all races across the country. There are 5 variables possible for each horse. Therefore there is a total possible count of shape of races of 8^5 or >30,000 possible shapes. So, if everything was perfect, you could expect to have a shapecount of 1 for 30,000 races. 30,000 races takes about a 3 months to run. So you should have enough in your data base in about 10 years. Worse, if you did, how effective do you thing 10 year old data would be?
andicap
08-08-2003, 10:27 AM
Larry,
Here is how Frandsen Publishing defined looking for pace shapes
Projected ESP Race Shape
Every race sets up a little differently based on the running styles of the horses in the race.
ESP
RACE
SHAPE _ _ ESP RUNNING STYLES
_ _ _ _
EEE _ _ 3 or more “E” horses
EE _ _ 2 “E” horses
E _ _ 1 “E” horse; no “EP” horse
E-EP _ _ 1 “E”, 1 or more “EP” horses
EP-EP _ _ 2 or more “EP” horses
EP _ _ 1 “EP”, no “P” horses
EP-P _ _ 1 “EP”, 1 or more “P” horses
P-P _ _ 2 or more “P” horses
P _ _ all “P” horses
P-S _ _ 1 “P”, 1 or more “S” horses
S _ _ all “S” horses
_ _ _ _
If an E horse comes out of an EEE race and finishes out of the money and now finds himself in a "E" race today, isn't his chances better than it would
look on paper???
Larry Hamilton
08-08-2003, 10:48 AM
a number of comments:
1. Their list of those shapes considered omits tens of thousands of possiblities. Without a db which considers all the individual shapes, how could they know what wins when?
2. They have attemtped to circumvemt the data collecting problem with groupings. i.e. 3 or more E horses instead of 3, 4, 5, etc. I must admit the first time I tackled this paradigm, I did the same thing. I did the same thing with Quirins numbers. Look at this problem backwards--if you say that races with 3 E's in it yield the winner 30% of the next race if they are alone. But that says that ONE OF THE GROUP has a 30% shot. The individual has much less, probably a third of the total. I am aware that I have made the math overly simplistic and maybe even incorrect--but the fact remains, I have tested this as well as I could and was extremely disappointed when my one-of-three didnt win 30%.
The point is Groupings don't work--dont believe me? Test it.
Larry Hamilton
08-08-2003, 11:26 AM
I didnt say that correctly.
Groupings dont work for picking individuals, they do work for picking groups!
andicap
08-08-2003, 12:24 PM
Larry,
If I'm not interested in using Pace shapes to actually select horses, but just to give a grade to a race in a horse's PP as "fast," "Honest" and "slow." Would it have more validity there.
In other words, the EEE and EE would be classified as "fast," so horses exiting those races would be marked with a "FAST" in their PPs to denote the excess of early types.
Today, the horse faces no other early types and it might move way up.
your comments?
Larry Hamilton
08-08-2003, 12:46 PM
My initial knee jerk to what you said comes by way of explaining a visual matrix.
Down one axis, "Slow", "Honest", "Fast" to represent a score for the race exited. Across the other axis, this horses current style. Each cell in the matrix will have been a measure of how that style does out of that kind of race. This, I am sure, is what the guys you quoted attempt to do.
Further, if one cannot measure the value, whatever form that value takes, that goes in the cell, what is the purpose of the matrix?
Will have to think more about it. It appears that you have reduced the groups to 3, but that doesn't make the grouping problem go away, unless all you want to do is hang a lable on the race. Why would you do that as it appears that the natural next question is, "How will my E horse do today, given he came from a "Slow" race?"
Valuist
08-08-2003, 01:10 PM
I say keep it simple. In my Weekly Charts, I'll denote "fast early, slow late" or whatever when there's an unusually paced event. Yeah, its not scientific, but I can remember it easier when the horses run back as opposed to making various numerical ratings.
Fastracehorse
08-08-2003, 08:40 PM
<Keeping pace shape records
Does anyone have success keeping extensive records of pace shapes (various ways to define them, such as E-E race with lots of E types, etc.)?
That way you know a horse who just emerged from an E-E race for example and likes to race near the front was at a disadvantage and if today's shape is say a E-P race (1 E and several pressers) he will have it easier.
+++++++++++++++++++++++++++++++
+++++++++++++++++++++++++++++++
I just look at the PP's and determine how the race may unfold.
It's not hard to do once you have seen thousands of races and have learned by being beaten thousands of times.
And it is reasonably accurate - you need more than one pace-scenario though. This is true because many times the race-shape looks like there will be a melt-down on the front end - but isn't when 1 or 2 jockeys pull hard on their charges. Somebody who was supposedly compromised now has an uncontested lead.
The above happens in alot of grass races.
fffastt
VetScratch
08-08-2003, 08:59 PM
Andicap,
ESP
RACE
SHAPE _ _ ESP RUNNING STYLES
_ _ _ _
EEE _ _ 3 or more "E" horses
EE _ _ 2 "E" horses
E _ _ 1 "E" horse; no "EP" horse
E-EP _ _ 1 "E", 1 or more "EP" horses
EP-EP _ _ 2 or more "EP" horses
EP _ _ 1 "EP", no "P" horses
EP-P _ _ 1 "EP", 1 or more "P" horses
P-P _ _ 2 or more "P" horses
P _ _ all "P" horses
P-S _ _ 1 "P", 1 or more "S" horses
S _ _ all "S" horses
I think quite a few database enthusiasts extract from BRIS downloads (ALW and others) and have taken pace shape analysis several steps further than what has been documented by Fransden (and they have undoubtedly gone further than what they document). When NA types are considered, one can expand the number of shape categories, compile distribution stats about each style in each shape category, and break out statistics by track/surface/distance/race-type-or-condition. This can give you dozens or hundreds of samples to study, depending on how you approach it. However, as Larry said, the possible number of discrete shapes is far to large to yield credible samples for each of more than 30,000 possibilities.
A few useful generalities emerge from such studies, but I think most of them are "common sense angles" for expert players (although what makes sense to experts often contradicts what the general public may believe).
Some nuggets exist, but looked at from a "value" perspective, most of the good nuggets of information correspond to rarely found ESP distributions within the major shape categories that you listed.
Others may disagree, but I think this sort of analysis pays fewer dividends than "advertised." Mostly because there are few big surprises.
Fast,
I'm glad you pointed out the fact that speed duels that you'd expect often don't take shape on turf races. This used to get me sssoooo frustrated because I thought I had the perfect bet on an off the pace horse only to see one of the front runners win on fairly easy fractions. I now take pace into account but don't use it to determine my bet. I'm not sure but I get the feeling the trainers and jockeys read the form and sometimes out-smart themselves. I've also seen this happen when there's very little speed and you see 3 horses suddenly fighting for the lead thinking they can steal it. It's frustrating and I haven't figured out what to do about it except to bet the horse I like best given it gives me good value and hope for the best.
I also pay attention to fast race shapes (EEE) but will still bet my E horse if it looks good for the same reason I mentioned in turf races. In some cases like this I'm actually dumbing down my handicapping and it hasn't affected my bottom line yet.
Have you figured out a way to use it to your advantage?
VetScratch
08-09-2003, 11:18 AM
Niko,
In line with your post, one of the general observations that can be derived from race shape studies is that E-types fare better in North American turf races than many handicappers suspect.
I get an IV of 1.28 for E-types in Maiden Turf Routes when the shape is EEE (3 or more E-types). Some players would look for ways to throw out all such E-types on the belief that none will survive to score at the wire.
Derek2U
08-09-2003, 11:53 AM
In NY Racing, the only distance that's favorable to E's is the flat
1 mile. A very good factor for longer Turf races is when a horse,
at an earlier age, was an E in dirt races, yet matured into a non-E
& now tries the turf. Yesterday, in Saratoga's 9th, Kid Kailin was
such a horse.
andicap
08-09-2003, 03:18 PM
VS and others,
none of your responses surprise me. I just wanted to know if anyone had success doing it -- and no one has said yes. (Maybe they are and don't want to let on and spoil the prices.)
I've long been suspicious of this conventional type of pace handicapping -- partly because the cat's out of the box now. It's like the absolute wrong time to buy a stock is once its profiled in Business Week.
I have seen some huge longshots with poor speed figures (or poor recent figures and good back figs) win when 6 Es were in a race and the pace just collapsed. So I think those things do happen with some consistency, but more often then not, they don't.
I guess the best thing to do is study 50 races at your track and see what happens.
As always predicting it beforehand is the hard part. Trainer intentions, jockey intentions, the way the earlier races have been run (if jockeys and trainers see all early horses or a lot of late horses winning recently, the condition of the horse and what it will or won't do today all affect the race. I have become less confident in who's going to win what race but more confident in my overall bottom line recently if that makes sense. If the horse is good and the odds are right bet, if not don't is about what I'm down too (a little more complicated than that). In looking at race shapes I'd have to guess 95% of the time I can't gain an advantage from it but as you mentioned Andicap there are situations that you can take advantage of if you wait long enough.
Vet Scratch, I know you got ripped on a little bit in an earlier post but I've immensly enjoyed your contributions to the board and would like to see more people of your caliber join. It's hard to tell when people are serious or just giving someone a hard time. I'm not sure why people have to attack others on the board to protect their own egos. Discuss or ignore are two good options. Sorry about adding the side note here.
vBulletin® v3.8.9, Copyright ©2000-2024, vBulletin Solutions, Inc.