 |
08-15-2018, 08:07 PM
|
#1
|
|
Registered User
Join Date: Jul 2002
Location: America
Posts: 6,955
|
Ratio of Odds
I was wondering if anybody has crunched the numbers and done a study that showed the ratio of odds to lengths behind in a race.
I used to have the data from a 1985 study, as well as one in the 90's, but was wondering if it has changed much since then. It may vary from track to track as well.
Ranges of odds would be OK.
Last edited by Secretariat; 08-15-2018 at 08:08 PM.
|
|
|
|
08-15-2018, 10:51 PM
|
#2
|
|
PA Steward
Join Date: Mar 2001
Location: Del Boca Vista
Posts: 88,447
|
Were you a fan of that book "More BASIC betting" that had those programs in BASIC code back in the late 80s? One of them was a program that spit out an odds line based on beaten lengths for horse racing....wondering if that's where you got the idea, as that is the only place I've ever seen that mentioned...
I actually used that code as the basis for my oddsline in the original PaceAdvantage software that I wrote and used way back in the day....
Last edited by PaceAdvantage; 08-15-2018 at 10:53 PM.
|
|
|
|
|
08-15-2018, 10:55 PM
|
#3
|
|
PA Steward
Join Date: Mar 2001
Location: Del Boca Vista
Posts: 88,447
|
BTW, I loved and STILL love that book...guy was preaching betting on overlays back in 1985 and using a computer to spit out odds lines...
I LOVED THAT BOOK!
It's a book literally NOBODY has heard of...and it was one of my favorites of all time, for it's day....
|
|
|
|
|
08-16-2018, 12:13 AM
|
#4
|
Join Date: Mar 2001
Location: Reno, NV
Posts: 16,900
|
Quote:
Originally Posted by PaceAdvantage
BTW, I loved and STILL love that book...guy was preaching betting on overlays back in 1985 and using a computer to spit out odds lines...
I LOVED THAT BOOK!
It's a book literally NOBODY has heard of...and it was one of my favorites of all time, for it's day....
|
I, too, kept my copy.
|
|
|
|
08-16-2018, 12:58 AM
|
#5
|
|
@TimeformUSfigs
Join Date: Jan 2002
Location: Moore, OK
Posts: 46,825
|
I have the book as well.
|
|
|
|
|
08-16-2018, 11:32 PM
|
#7
|
|
Registered User
Join Date: Jul 2002
Location: America
Posts: 6,955
|
Yeah, I used the book as well. Jaspar also had a second book - More Basic Betting. Still have both. With the advent of more Sports Betting they may be appealing to many.
Jaspar did put his code in and you can see the relationships he input for odds versus lengths and it was something that intrigued me.
I used it for awhile to help develop an odds line. However, I'm noticing differing values empirically. For example:
Jaspar lists horses that go off at 2/1 as essentially beaten two lengths, horses that win by a wire at 1/1, and horses who win by varying lengths even lower. Once you assign a value to each horse you can sum the probabilities and adjust to a 100% line.
However, Jaspar's program assumes a global oddsline and doesn't do much to adjust track to track. So was wondering if instead of looking globally, it may be more advantageous to look locally. My study is just beginning, but the month of August thus far at Penn has surprised me a bit just looking at winners. The median odds for winners is 2.1/1 and median win margins are 2.25 lengths.
Tried to do a linear equation to see if the std. deviation was low enough to make a solid comparision of odds to lengths just for winners. Best fit might be a curve or looking at ranges.
Was curious if anyone else was looking at this, and had seen differences from the 80-90's as now.
The one I got in the 90's was from Jim Bayle and was based more on Beyer ratings.
Last edited by Secretariat; 08-16-2018 at 11:37 PM.
|
|
|
|
|
08-17-2018, 12:29 PM
|
#8
|
|
The Voice of Reason!
Join Date: Mar 2001
Location: Canandaigua, New york
Posts: 112,744
|
Class, you want to see final odds of the winner plotted against the winning margin of the winner?
By track, distance, surface?
__________________
Who does the Racing Form Detective like in this one?
|
|
|
|
|
08-17-2018, 08:37 PM
|
#9
|
|
Registered User
Join Date: Jul 2002
Location: America
Posts: 6,955
|
Tom,
Ideally, my goal is to list a few things
1. Median Odds of Winners. per track (yes per distance or class would be nice, but samples sizes might be too small.) This lets you quickly analyze the type of races which offer the best opportunities.
2. My primary goal however is to create a comparision of odds to beaten lengths like Jaspar and Bayle did. I think there is variations among tracks as well which may help uncover odds line opportunities. Below is an approximation of Jaspar's figures. This isn't winners but just odds above 1.0. (1.0 was simply a BL - 0)
Odds - 1.01 To 1.19 BL - 0.25
Odds - 1.2 To 1.29 BL - 0.5
Odds - 1.3 To 1.39 BL - 0.75
Odds - 1.4 To 1.49 BL - 1
Odds - 1.5 To 1.59 BL - 1.25
Odds - 1.6 To 1.79 BL - 1.5
Odds - 1.8 To 1.89 BL - 1.75
Odds - 1.9 To 1.99 BL - 2
Odds - 2 To 2.49 BL - 2.25
Odds - 2.5 To 2.99 BL - 2.5
Odds - 3 To 3.49 BL - 3
Odds - 3.5 To 3.99 BL - 3.5
Odds - 4 To 4.99 BL - 4
Odds - 5 To 5.99 BL - 4.5
Odds - 6 To 6.49 BL - 6
Odds - 6.5 To 6.99 BL - 6.25
Odds - 7 To 7.99 BL - 6.5
Odds - 8 To 8.99 BL - 7
Odds - 9 To 9.99 BL - 7.5
Odds - 10 To 10.99 BL - 8
Odds - 11 To 11.99 BL - 8.25
Odds - 12 To 12.99 BL - 8.5
Odds - 13 To 14.99 BL - 8.75
Odds - 15 To 17.49 BL - 9
Odds - 17.5 To 19.99 BL - 10
Odds - 20 To 22.49 BL - 10.5
Odds - 22.5 To 24.99 BL - 11
Odds - 25 To 29.99 BL - 11.5
Odds - 30 To 34.9 BL - 11.75
Odds - 35 To 39.99 BL - 12
Odds - 40 To 49.99 BL - 12.5
Odds - 50 To 74.99 BL - 13.75
Odds - 75 To 99.99 BL - 15
Odds >=100 BL - 20
Always love your data mining equations and research into this stuff.
|
|
|
|
|
08-17-2018, 09:55 PM
|
#10
|
|
The Voice of Reason!
Join Date: Mar 2001
Location: Canandaigua, New york
Posts: 112,744
|
The data below is 3,002 horses at 6.0F at Aqu/Bel/Sar.
All classes, sexes, fast tracks. SB and non SB.
The beaten lengths were the ones from the paceline selected, not just the last race.
I re-did the graph to combine all the date after the 28.5 BL group, where it looked to "come apart" on the first graph.
Is this any help?
__________________
Who does the Racing Form Detective like in this one?
Last edited by Tom; 08-17-2018 at 10:10 PM.
|
|
|
|
|
08-17-2018, 10:15 PM
|
#11
|
|
The Voice of Reason!
Join Date: Mar 2001
Location: Canandaigua, New york
Posts: 112,744
|
Quote:
Originally Posted by Tom
The data below is 3,002 horses at 6.0F at Aqu/Bel/Sar.
All classes, sexes, fast tracks. SB and non SB.
The beaten lengths and odds are from the race just run.
I re-did the graph to combine all the date after the 28.5 BL group, where it looked to "come apart" on the first graph.
Is this any help?
|
CORRECTION ^
__________________
Who does the Racing Form Detective like in this one?
|
|
|
|
|
08-18-2018, 04:36 AM
|
#12
|
|
Registered User
Join Date: Jul 2002
Location: America
Posts: 6,955
|
Thanks Tom.
Yeah, what I thought. The odds/BL ratio has changed somewhat compared to Jaspar's data, and the intersection of odds/BL on your chart seems to occur around 10/1 whereas in his data it occurs around 5/2. May be modified in that he used results charts rather than pace line selected beaten lengths, but the info is useful and your line of best fit shows less deviation.
Much appreciated. Will do more research on this but this is good data to compare between a large versus small track to see variation.
|
|
|
|
|
08-18-2018, 11:35 AM
|
#13
|
|
The Voice of Reason!
Join Date: Mar 2001
Location: Canandaigua, New york
Posts: 112,744
|
Actually, the data came from the same line - odds and beaten lengths were from the race for every horse that finished within 40 lengths. After 40, the BL shows as "99."
I don't have the selected paceline data - I misread it.
__________________
Who does the Racing Form Detective like in this one?
|
|
|
|
|
| Thread Tools |
|
|
| Rate This Thread |
|
|
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
|
|
|
|
|
@
