 |
|
07-14-2015, 01:13 PM
|
#1
|
|
Registered User
Join Date: Feb 2002
Posts: 10,861
|
Par Times Available
I've been working on making par times for U.S. racing using many ideas from Bill Quirin's books, as well as ideas from the writings of James Quinn, Andrew Beyer, Charles Carroll and Nick Mordin. I've finished making pars for most of the tracks with dirt courses. (I'll make all-weather, inner track and turf pars as well as state bred pars next.)
I'd be interested in getting feedback about the usefulness and accuracy of these pars. Drop me a PM or post here and I can send you more for analysis.
Here is a sample for Belmont Claiming Sprints on the Dirt Course for Males 3 years old and up:
Females appear in a table below the males.
Adjustments for Restricted Claiming pars are given as well as for 3 year old only races.
All times are in seconds, including the table below of 3 year old adjustments.
I've also made pars for Allowance races 3 years and up and 2 year old claiming and allowance races.
The first two columns are the 2 furlong and 4 furlong pace times, hence the letter "p" after the furlongs. The rest of the columns are the distance of the race. 500 = 5 furlongs, 550 = 5.5 furlongs, etc.
Code:
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
BEL D M CLAIMING SPRINTS MALES 3UP AND OLDER
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
DISTANCE
Class 200P 400P 500 550 600 650 700 750
50000 22.49 45.62 57.62 63.68 70.08 76.48 82.94 89.54
40000 22.52 45.69 57.72 63.8 70.2 76.6 83.12 89.72
32000 22.56 45.76 57.84 63.95 70.35 76.75 83.31 89.91
30000 22.57 45.78 57.87 64 70.4 76.8 83.37 89.97
25000 22.6 45.85 57.99 64.14 70.54 76.94 83.53 90.13
20000 22.65 45.95 58.15 64.33 70.73 77.13 83.73 90.33
16000 22.7 46.05 58.32 64.53 70.93 77.33 83.94 90.54
12500 22.76 46.17 58.53 64.77 71.17 77.57 84.17 90.77
10000 22.82 46.28 58.72 64.99 71.39 77.79 84.39 90.99
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
BEL D M CLAIMING SPRINTS FEMALES 3UP AND OLDER
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
DISTANCE
Class 200P 400P 500 550 600 650 700 750
50000 22.59 45.82 58.02 64.08 70.48 76.88 83.34 89.94
40000 22.62 45.89 58.12 64.2 70.6 77 83.52 90.12
32000 22.66 45.96 58.24 64.35 70.75 77.15 83.71 90.31
30000 22.67 45.98 58.27 64.4 70.8 77.2 83.77 90.37
25000 22.7 46.05 58.39 64.54 70.94 77.34 83.93 90.53
20000 22.75 46.15 58.55 64.73 71.13 77.53 84.13 90.73
16000 22.8 46.25 58.72 64.93 71.33 77.73 84.34 90.94
12500 22.86 46.37 58.93 65.17 71.57 77.97 84.57 91.17
10000 22.92 46.48 59.12 65.39 71.79 78.19 84.79 91.39
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
RESTRICTED CLAIMING PACE AND FINAL TIME PARS CAN BE CREATED BY USING THE OPEN CLAIMING TIMES ABOVE AND THE SUGGESTED ADJUSTMENTS BELOW.
CLM NW3L Final Time is 0.2 seconds slower than the same open claiming level, 4 furlong pace is 0.1 slower and 2 furlong pace is 0.05 slower
CLM NW2L Final Time is 0.4 seconds slower than the same open claiming level, 4 furlong pace is 0.2 slower and 2 furlong pace is 0.1 slower
CLM NW1L Final Time is 0.6 seconds slower than the same open claiming level, 4 furlong pace is 0.3 slower and 2 furlong pace is 0.15 slower
MDN CLM Final Time is 1.0 seconds slower than the same open claiming level, 4 furlong pace is 0.5 slower and 2 furlong pace is 0.25 slower
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
THREE YEAR OLD ONLY RACES - FINAL TIME PAR TIME ADJUSTMENTS
600 650 700 800-818 832-850 900
----------------------------------------------------------------------------------------------------------------------------------------------------------------------
+1.8 Jan 1 +1.8
+1.6 Feb 1 +1.6
+1.4 Jan 1 Mar 15 +1.4
+1.2 Feb 15 May 1 +1.2
+1.0 Jan 1 Apr 15 Jun 1 +1.0
+0.8 Jan 1 Jan 1 Apr 15 Jun 1 Jul 1 +0.8
+0.6 Jan 1 Feb 1 Mar 15 Jun 1 Jul 1 Aug 1 +0.6
+0.4 Apr 15 Jun 1 Jun 15 Jul 15 Aug 15 Sep 15 +0.4
+0.2 Jul 1 Aug 1 Aug 15 Sep 15 Oct 15 Dec 1 +0.2
+0.0 Nov 1 Dec 1 Dec 15 ------ ------ Jan 1 +0.0
Use half of the adjustment for the 4 furlong pace par and one quarter for the 2 furlong pace par.1
|
|
|
|
07-14-2015, 11:19 PM
|
#2
|
|
Registered User
Join Date: Jan 2008
Location: pen
Posts: 4,581
|
that every final distance has identical pace pars doesn't seem realistic to me. 
you didn't find this for every track did you??
|
|
|
|
07-15-2015, 03:31 AM
|
#3
|
|
Registered User
Join Date: Feb 2002
Posts: 10,861
|
Quote:
|
Originally Posted by proximity
that every final distance has identical pace pars doesn't seem realistic to me.
you didn't find this for every track did you?? |
That's a good observation and is something I had to try to understand.
The short answer is that at some tracks and distances I made different pace pars for different distances, but at most tracks I only use one set of pace pars for sprints and another set for routes.
For some tracks at some distances the pace pars are very different -- Delta Downs 5 and 6.5 furlongs and Arlington 8 and 8.5 come to mind. There are others, but not many.
Bill Quirin's study showed that, in general, the pace pars step in an orderly fashion from class to class and can be used across multiple distances. So you can have one identical set of pace pars for all sprint distances and another set for routes.
He gives an explanation in his book of why it is ok to do it this way. However, when you come across a track with a lot of races at a particular class and several related distances you will often find the pars are quite orderly.
At Santa Anita the pace pars are nearly identical.
At Charles Town the 4.5 furlong pace pars are nothing like the 6.5 furlong pace pars.
Here is an example using CT in the following two tables :
Code:
----------------------------------------------------------------------------------
CT D M CLAIMING SPRINTS MALES 3UP AND OLDER
----------------------------------------------------------------------------------
DISTANCE
Class 200P 400P 400 450
25000 21.56 45.39 46.08 51.95
20000 21.6 45.46 46.2 52.09
15000 21.65 45.56 46.38 52.29
12500 21.68 45.63 46.49 52.42
10000 21.72 45.71 46.64 52.59
8000 21.76 45.79 46.79 52.76
7500 21.77 45.82 46.84 52.81
6250 21.81 45.89 46.96 52.95
5000 21.85 45.98 47.12 53.13
----------------------------------------------------------------------------------
Code:
-------------------------------------------------------------------------------
CT D M CLAIMING SPRINTS MALES 3UP AND OLDER
-------------------------------------------------------------------------------
DISTANCE
Class 200P 400P 650 700
25000 23.26 46.85 80.49 86.88
20000 23.3 46.95 80.68 87.08
15000 23.37 47.08 80.94 87.35
12500 23.41 47.17 81.12 87.52
10000 23.47 47.28 81.34 87.74
8000 23.53 47.39 81.57 87.96
7500 23.54 47.43 81.63 88.02
6250 23.59 47.52 81.82 88.2
5000 23.65 47.63 82.05 88.42
-------------------------------------------------------------------------------
The challenge of making a different set of pace pars for every distance is that the sample sizes are small and with so few races the data is pretty "noisy". When there is not enough data available to make reliable pars then you have to guess. I think it's a bad idea to guess what a par should be when you know that studies done on large datasets have shown that the one set of pace pars works well.
Even the data in the charts has a lot of noise -- especially when it comes to beaten lengths at the various calls. I think the more important thing is for the handicapper to figure out what the differences between pars are for various tracks. Once you make your variants and then use them to make speed figures you're going to want to know how the figures from "Track A" relate to "Track B".
|
|
|
|
|
07-15-2015, 03:56 AM
|
#4
|
|
Registered User
Join Date: Feb 2002
Posts: 10,861
|
Here is some more data:
The first table is raw data that was used to compute the pars that appear in the second table.
The last column of the first table is labeled "N" -- this is the sample size for that distance and class. Now, looking at the raw data it looks like the pace times get slower as the races get longer. The differences for 5.5, 6 and 6.5 are negligible. And given all the noise in racing data it is probably not a big deal.
Now the 7 furlong pace par looks like it should be slower. However, there are only 13 samples. You really need about 30 samples to be able to do analysis about the significance of the variables. So in this case I default to Quirin and use the same pace pars for all the distances.
Some par makers might use the raw data in the first table to make their pars. It looks reasonable. And to be honest, that probably works. However, I think it's better to use the distance and class with the biggest sample size and extrapolate from there. In fact, this is the only way to do it if you want to make pars for every track. There simply isn't enough data to make pars otherwise and if there was enough data it would be incredibly tedious to do by hand and it would probably never get done except for a small sample of tracks.
Code:
Track Dist Surf Course Class FT 2F 4F N
SA 550 D D Clm 10000 63.92 21.82 45.01 19
SA 600 D D Clm 10000 70.41 22 45.06 21
SA 650 D D Clm 10000 77.06 22.11 45.18 18
SA 700 D D Clm 10000(10-9) 83.88 22.71 45.53 13
---------------------------------------------------------------------------------------
-------------------------------------------------------------------------------------------------------------------------
SA D M CLAIMING SPRINTS MALES 3UP AND OLDER
-------------------------------------------------------------------------------------------------------------------------
DISTANCE
Class 200P 400P 450 500 550 600 650 700
75000 21.62 44.33 50.32 56.6 62.61 69.01 75.41 81.72
50000 21.66 44.4 50.5 56.7 62.76 69.16 75.56 82.02
40000 21.69 44.47 50.62 56.8 62.88 69.28 75.68 82.2
32000 21.73 44.54 50.74 56.92 63.03 69.43 75.83 82.39
25000 21.78 44.63 50.9 57.07 63.22 69.62 76.02 82.61
20000 21.83 44.73 51.04 57.23 63.41 69.81 76.21 82.81
16000 21.88 44.83 51.2 57.4 63.61 70.01 76.41 83.02
12500 21.94 44.95 51.37 57.61 63.85 70.25 76.65 83.25
10000 22 45.06 51.54 57.8 64.07 70.47 76.87 83.47
8000 22.06 45.17 51.71 58 64.3 70.7 77.1 83.69
-------------------------------------------------------------------------------------------------------------------------
|
|
|
|
|
07-15-2015, 04:59 AM
|
#5
|
|
Registered User
Join Date: Jan 2008
Location: pen
Posts: 4,581
|
what I would do for pace is to make the final time pars like you're doing and not try to make any specific (say $10,000 clm or whatever) pace pars but take all the final times that you used to make ALL of your pars for the track and distance and their accompanying pace times and analyze them.
i'd put all the times in a spreadsheet to get median (or average) times and standard deviations. I would then compare the standard deviation from the pace times to the standard deviation of the final times to get a ratio and (after setting the average pace and final times equal to each other) i'd use this ratio to create a pace par for each final time.
you mentioned the work of james Quinn. are your turf pars going to resemble the charts he made in figure handicapping? that would be interesting to see!!
|
|
|
|
|
07-15-2015, 12:37 PM
|
#6
|
|
Registered User
Join Date: Feb 2002
Posts: 10,861
|
Quote:
|
Originally Posted by proximity
what I would do for pace is to make the final time pars like you're doing and not try to make any specific (say $10,000 clm or whatever) pace pars but take all the final times that you used to make ALL of your pars for the track and distance and their accompanying pace times and analyze them.
i'd put all the times in a spreadsheet to get median (or average) times and standard deviations. I would then compare the standard deviation from the pace times to the standard deviation of the final times to get a ratio and (after setting the average pace and final times equal to each other) i'd use this ratio to create a pace par for each final time.
|
That's an interesting way to do it. I'll think about it. It sounds like it could work.
Quote:
|
you mentioned the work of james Quinn. are your turf pars going to resemble the charts he made in figure handicapping? that would be interesting to see!!
|
I haven't decided on how to do turf pars, yet. I'm still tweaking the dirt pars, but should be finished today. So I'll start digging into the turf later today.
Beyer writes a little about Quinn's turf pars and gives one example of a flaw in his system.
Cary Fotias made the best turf numbers I've ever used. So I might do something similar to what he did.
My hunch is that the turf speed figures aren't as important as the patterns that are formed by the figures, (i.e., Fotias' Turf Decline Line pattern).
Until I get a look at the data it's hard to say what form the pars will take. The only thing Quirin wrote about turf pars is that he doesn't make them because there are too few and the nature of turf racing doesn't lend itself to speed figures. Beyer found the same thing in turf racing here in the U.S. as well as in Australia and Europe.
However, Nick Mordin successfully uses speed figures and wrote about his method of making them.
David Edelmann, a statistics professor who has lived in Australia and Ireland, writes a bit about them in his book, too.
|
|
|
|
|
07-15-2015, 02:02 PM
|
#7
|
|
Registered User
Join Date: Jul 2013
Location: Seattle
Posts: 3,943
|
"At Santa Anita the pace pars are nearly identical."
It seems to me that 5 and 5.5 furlong races at Santa Anita are always run much closer to the track 3 year fastest time than the 6, 6.5 and 7 furlong races. Looking at 3 year fastest and track variants for the day, 5 and 5.5 races will almost always come out to 100+ while 7 furlong races will almost always be 100-.
This is for meat and potato claimers. I'm always very careful with any numbers at Santa Anita when applied to 5 or 5.5 furlongs.
|
|
|
|
|
07-15-2015, 03:15 PM
|
#8
|
|
Registered User
Join Date: Feb 2002
Posts: 10,861
|
Quote:
|
Originally Posted by ultracapper
"At Santa Anita the pace pars are nearly identical."
It seems to me that 5 and 5.5 furlong races at Santa Anita are always run much closer to the track 3 year fastest time than the 6, 6.5 and 7 furlong races. Looking at 3 year fastest and track variants for the day, 5 and 5.5 races will almost always come out to 100+ while 7 furlong races will almost always be 100-.
This is for meat and potato claimers. I'm always very careful with any numbers at Santa Anita when applied to 5 or 5.5 furlongs.
|
In my Santa Anita table of times for older claimers, I don't have any 5 furlong races. When I do maidens, allowance and younger horses maybe I will find some.
Here is a table of SA pace times for 5.5, 6 and 7:
Code:
SA 10k claiming
Dist 2F 4F N
5.5 21.82 45.01 19
6.0 22.00 45.06 21
7.0 22.11 45.18 18
SA 5k claiming
Dist 2F 4F N
5.5 22.07 45.55 11
6.0 21.80 44.74 18
7.0 22.71 45.53 10
As you can see the sample sizes are kind of small, but they do show that the times are pretty close at the 10k level. The 5k level has more spread, but that could be due to the small sample size.
The reason I think Quirin is probably correct about being able to use the same pace pars for sprints and another set for routes is because as the number of samples get larger they will tend to be identical.
Now, maybe in those 5k claiming races at 7 furlongs at Santa Anita the jockey is trying to save something for the finish. But I don't think so, otherwise you'd see it at 10k, too.
|
|
|
|
|
07-15-2015, 08:41 PM
|
#9
|
|
2 outta 3 aint bad
Join Date: Sep 2003
Location: Lehigh Valley,PA.
Posts: 2,217
|
Quote:
|
Originally Posted by highnote
The only thing Quirin wrote about turf pars is that he doesn't make them because there are too few and the nature of turf racing doesn't lend itself to speed figures. Beyer found the same thing in turf racing here in the U.S. as well as in Australia and Europe.
|
I find turf (internal) pars to be more consistant than dirt pars.
You will learn what I mean by experimenting with these.
As far as too few, you will easily find well over 20 turf races per meet/per distance...... at the common distances.
Of course not short meets like, ex. Keenland.
I use equibase historical charts for this (free) and update after each recent meet.
I strongly suggest you incorporate turf pars.. asap....=.....$$$.
PS;Great job on your charts,good post, 
Last edited by KingChas; 07-15-2015 at 08:49 PM.
|
|
|
|
|
07-15-2015, 09:22 PM
|
#10
|
|
Registered User
Join Date: Aug 2011
Posts: 28
|
Interesting,you seek to make par times for your own unique variants,
for your numbers.I looked at 3 of your 10,000 pars,final times.
And they all seem pretty tight to what I am using.within .7 tenths the
furthest,will look at more later.I use Gordon Pine and Dave Schwarts
Ideas when I need to put together a set for a track.but it slow work.
Pulling from charts or pp's.Looks good so far Highnote.
Gordon Pine link you may like,I like his par track model idea.
http://www.netcapper.com/TrackTracts...e/TT020906.htm
|
|
|
|
|
07-15-2015, 09:52 PM
|
#11
|
|
Registered User
Join Date: Feb 2002
Posts: 10,861
|
Quote:
|
Originally Posted by KingChas
I find turf (internal) pars to be more consistant than dirt pars.
|
Intuitively, this makes sense since horses tend to pace themselves in the early stages and then sprint to the wire, whereas, horses on the dirt tend to sprint early and then decelerate to the wire.
Quote:
You will learn what I mean by experimenting with these.
As far as too few, you will easily find well over 20 turf races per meet/per distance...... at the common distances.
Of course not short meets like, ex. Keenland.
|
I like to have at least 8 races per track, per class, per distance. That's a big limiting factor for turf.
Quote:
|
I strongly suggest you incorporate turf pars.. asap....=.....$$$.
|
AGREE!
Quote:
|
PS;Great job on your charts,good post,
|
Thanks! This is a fun topic. Nice to have people to bounce ideas off of!
|
|
|
|
|
07-16-2015, 01:08 AM
|
#12
|
|
Registered User
Join Date: Jan 2008
Location: Melbourne Australia
Posts: 916
|
Quote:
|
Originally Posted by highnote
I've been working on making par times for U.S. racing using many ideas from Bill Quirin's books, as well as ideas from the writings of James Quinn, Andrew Beyer, Charles Carroll and Nick Mordin. I've finished making pars for most of the tracks with dirt courses. (I'll make all-weather, inner track and turf pars as well as state bred pars next.)
I'd be interested in getting feedback about the usefulness and accuracy of these pars. Drop me a PM or post here and I can send you more for analysis.
Here is a sample for Belmont Claiming Sprints on the Dirt Course for Males 3 years old and up:
Females appear in a table below the males.
Adjustments for Restricted Claiming pars are given as well as for 3 year old only races.
All times are in seconds, including the table below of 3 year old adjustments.
I've also made pars for Allowance races 3 years and up and 2 year old claiming and allowance races.
The first two columns are the 2 furlong and 4 furlong pace times, hence the letter "p" after the furlongs. The rest of the columns are the distance of the race. 500 = 5 furlongs, 550 = 5.5 furlongs, etc.
Code:
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
BEL D M CLAIMING SPRINTS MALES 3UP AND OLDER
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
DISTANCE
Class 200P 400P 500 550 600 650 700 750
50000 22.49 45.62 57.62 63.68 70.08 76.48 82.94 89.54
40000 22.52 45.69 57.72 63.8 70.2 76.6 83.12 89.72
32000 22.56 45.76 57.84 63.95 70.35 76.75 83.31 89.91
30000 22.57 45.78 57.87 64 70.4 76.8 83.37 89.97
25000 22.6 45.85 57.99 64.14 70.54 76.94 83.53 90.13
20000 22.65 45.95 58.15 64.33 70.73 77.13 83.73 90.33
16000 22.7 46.05 58.32 64.53 70.93 77.33 83.94 90.54
12500 22.76 46.17 58.53 64.77 71.17 77.57 84.17 90.77
10000 22.82 46.28 58.72 64.99 71.39 77.79 84.39 90.99
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
BEL D M CLAIMING SPRINTS FEMALES 3UP AND OLDER
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
DISTANCE
Class 200P 400P 500 550 600 650 700 750
50000 22.59 45.82 58.02 64.08 70.48 76.88 83.34 89.94
40000 22.62 45.89 58.12 64.2 70.6 77 83.52 90.12
32000 22.66 45.96 58.24 64.35 70.75 77.15 83.71 90.31
30000 22.67 45.98 58.27 64.4 70.8 77.2 83.77 90.37
25000 22.7 46.05 58.39 64.54 70.94 77.34 83.93 90.53
20000 22.75 46.15 58.55 64.73 71.13 77.53 84.13 90.73
16000 22.8 46.25 58.72 64.93 71.33 77.73 84.34 90.94
12500 22.86 46.37 58.93 65.17 71.57 77.97 84.57 91.17
10000 22.92 46.48 59.12 65.39 71.79 78.19 84.79 91.39
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
RESTRICTED CLAIMING PACE AND FINAL TIME PARS CAN BE CREATED BY USING THE OPEN CLAIMING TIMES ABOVE AND THE SUGGESTED ADJUSTMENTS BELOW.
CLM NW3L Final Time is 0.2 seconds slower than the same open claiming level, 4 furlong pace is 0.1 slower and 2 furlong pace is 0.05 slower
CLM NW2L Final Time is 0.4 seconds slower than the same open claiming level, 4 furlong pace is 0.2 slower and 2 furlong pace is 0.1 slower
CLM NW1L Final Time is 0.6 seconds slower than the same open claiming level, 4 furlong pace is 0.3 slower and 2 furlong pace is 0.15 slower
MDN CLM Final Time is 1.0 seconds slower than the same open claiming level, 4 furlong pace is 0.5 slower and 2 furlong pace is 0.25 slower
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
THREE YEAR OLD ONLY RACES - FINAL TIME PAR TIME ADJUSTMENTS
600 650 700 800-818 832-850 900
----------------------------------------------------------------------------------------------------------------------------------------------------------------------
+1.8 Jan 1 +1.8
+1.6 Feb 1 +1.6
+1.4 Jan 1 Mar 15 +1.4
+1.2 Feb 15 May 1 +1.2
+1.0 Jan 1 Apr 15 Jun 1 +1.0
+0.8 Jan 1 Jan 1 Apr 15 Jun 1 Jul 1 +0.8
+0.6 Jan 1 Feb 1 Mar 15 Jun 1 Jul 1 Aug 1 +0.6
+0.4 Apr 15 Jun 1 Jun 15 Jul 15 Aug 15 Sep 15 +0.4
+0.2 Jul 1 Aug 1 Aug 15 Sep 15 Oct 15 Dec 1 +0.2
+0.0 Nov 1 Dec 1 Dec 15 ------ ------ Jan 1 +0.0
Use half of the adjustment for the 4 furlong pace par and one quarter for the 2 furlong pace par.1
|
a noble pursuit!
i think your relationships are astray.
for instance if you gave an arbitrary number of 100 for your 70.08 for 6 furlongs and 100 for your 82.94 for 7 furlongs.
then the number for 71.39 at 6 furlongs should be the same as the number for 84.39 at 7 furlongs, but they're not.
in other words your class relationships vary at different distances, where i would expect them to be the same.
|
|
|
|
|
07-16-2015, 03:25 AM
|
#13
|
|
Registered User
Join Date: Feb 2002
Posts: 10,861
|
Quote:
|
Originally Posted by steveb
a noble pursuit!
i think your relationships are astray.
for instance if you gave an arbitrary number of 100 for your 70.08 for 6 furlongs and 100 for your 82.94 for 7 furlongs.
then the number for 71.39 at 6 furlongs should be the same as the number for 84.39 at 7 furlongs, but they're not.
in other words your class relationships vary at different distances, where i would expect them to be the same.
|
Thanks for your feedback.
I don't understand what you're saying. If you assign a 100 to 70.08 at 6 furlongs and a 100 to 82.94 at 7 furlongs and each class step down the ladder is, say, worth 1 point, then 71.39 at 6 furlongs would be worth 92 and so 84.39 at 7 furlongs would also be equal to 92.
Or I am misunderstanding what you're saying?
|
|
|
|
|
07-16-2015, 03:55 AM
|
#14
|
|
Registered User
Join Date: Jan 2008
Location: Melbourne Australia
Posts: 916
|
Quote:
|
Originally Posted by highnote
Thanks for your feedback.
I don't understand what you're saying. If you assign a 100 to 70.08 at 6 furlongs and a 100 to 82.94 at 7 furlongs and each class step down the ladder is, say, worth 1 point, then 71.39 at 6 furlongs would be worth 92 and so 84.39 at 7 furlongs would also be equal to 92.
Or I am misunderstanding what you're saying?
|
if each time at each distance was worth 100(or whatever) at the 50000 level, then so too should each number be equal for any other class level.
i made those 5000 ones equal to 100.
so the 10000 ones should all line up(they should all be the same value whatever that value is)
but instead i get 550=80, 600=82,650=83,700=83 and 750=84.
thus the relationship is incorrect imo.
they should all be the same number,
then you could look at the gaps as that will give you an idea that something is wrong.
for instance the time gap between 50000 and 10000 is 1.31 , 1.31 , 1.31 , 1.45 and 1.45 respectively.
it is logical to expect the gaps to increase as the distances and classes get /longer/stronger.
but you have the 550,600,650 the same and then a big jump of .14 seconds to the 700 and 750.
the problem with most time people is that that they don't realise that pars(standards or whatever else you call them) are really percentage figures and nothing much else.
thus you should only need 1 time for each distance at each track.
so perhaps you may have 56 seconds for 1000 metres for IAMBEST class at WOOP WOOP, then you have figured that IAMSECONDBEST class is perhaps 99.5% of IAMBEST so you would expect the average IAMSECONDBEST to run 56.28
all that is important is the relationships between the classes and distances at the various tracks.
the times themselves are of minor importance.
|
|
|
|
|
07-16-2015, 12:30 PM
|
#15
|
|
Registered User
Join Date: Feb 2002
Posts: 10,861
|
I think we are approaching this from two different perspectives. It sounds like you advocate using only one par time for each distance. Nick Mordin calls these "Standard Times". His standard times represent what a theoretical Grade 1 or Group 1 horse could run at a given distance at a given track. Some tracks don't card Grade 1 level races so that is why his standard times are theoretical. But he can determine the theoretical standard times by estimating how various classes of horses run over both tracks.
My pars are based on the work of Dr. William Quirin from his book "Winning at the Races: Computer Discoveries in Thoroughbred Handicapping".
So I don't think your method applies to Quirin's method as I am trying to do it. That's not to say that your method is wrong or that it could not be adapted to Quirin's.
As far as the gaps between distances as the distances get longer; it makes perfect sense for this to happen because for dirt track racing the time it takes to cover a length increases as the distances get longer.
So from 10,000 claiming to 50,000 at 5.5 to 6.5 furlongs the gap is the same -- about 6 lengths. For 10k to 50k at 9 furlongs the gap is 10 lengths. There is a telescoping effect as distances get longer.
Also, the final time gaps between distances get larger by different rates at different tracks. For example, "Track A" might have a final time gap of 6.4 seconds from 6.5 furlongs to 7 furlongs and "Track B" might have a final time gap of 6.6 seconds between 6.5 and 7 furlongs.
I agree with you that "times" are not important and that the relationships between the "times" are more important. As long as the relationships are relatively correct you could use hours instead of minutes and get the same or similar results, relatively speaking.
Quote:
|
Originally Posted by steveb
if each time at each distance was worth 100(or whatever) at the 50000 level, then so too should each number be equal for any other class level.
i made those 5000 ones equal to 100.
so the 10000 ones should all line up(they should all be the same value whatever that value is)
but instead i get 550=80, 600=82,650=83,700=83 and 750=84.
thus the relationship is incorrect imo.
they should all be the same number,
then you could look at the gaps as that will give you an idea that something is wrong.
for instance the time gap between 50000 and 10000 is 1.31 , 1.31 , 1.31 , 1.45 and 1.45 respectively.
it is logical to expect the gaps to increase as the distances and classes get /longer/stronger.
but you have the 550,600,650 the same and then a big jump of .14 seconds to the 700 and 750.
the problem with most time people is that that they don't realise that pars(standards or whatever else you call them) are really percentage figures and nothing much else.
thus you should only need 1 time for each distance at each track.
so perhaps you may have 56 seconds for 1000 metres for IAMBEST class at WOOP WOOP, then you have figured that IAMSECONDBEST class is perhaps 99.5% of IAMBEST so you would expect the average IAMSECONDBEST to run 56.28
all that is important is the relationships between the classes and distances at the various tracks.
the times themselves are of minor importance.
|
Last edited by highnote; 07-16-2015 at 12:33 PM.
|
|
|
|
|
|
Posting Rules
|
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
HTML code is Off
|
|
|
|
|
