Go Back   Pace and Cap - Sartin Methodology & The Match Up > General Discussion
Mark Forums Read
Google Site Search Get RDSS Sartin Library RDSS FAQs Conduct Register Site FAQ Members List Search Today's Posts

General Discussion General Horse Racing Discussion

Reply
 
Thread Tools Display Modes
Old 04-16-2019, 05:50 PM   #1
rdiam
Grade 1
 
Join Date: Sep 2016
Posts: 606
20 Race Cycles

Due to the variance in horse race betting, you are best off using a sample size of at least 50 races to analyze your betting proficiency at a statistically significant level. 20 races is too small to provide reliable results. The math don't lie.

As an aside, if you spend any time trying to create a betting odds line, the implied probabilities must add up to one. Reason: there can only be 1 winner in a race (let's assume no dead heats for this discussion). Any other method of creating a betting odds line is statistically incorrect.

Richard
rdiam is offline   Reply With Quote
Old 04-16-2019, 05:58 PM   #2
Jeebs
Grade 1
 
Join Date: Dec 2010
Posts: 909
Quote:
Originally Posted by rdiam View Post
Due to the variance in horse race betting, you are best off using a sample size of at least 50 races to analyze your betting proficiency at a statistically significant level. 20 races is too small to provide reliable results. The math don't lie.

As an aside, if you spend any time trying to create a betting odds line, the implied probabilities must add up to one. Reason: there can only be 1 winner in a race (let's assume no dead heats for this discussion). Any other method of creating a betting odds line is statistically incorrect.

Richard
What factors do you account for to get a good percentage read on contenders? I know that the primary battle is in finding the non-contenders. Dick Mitchell and Steve Fierro were always into the Pareto Principle (80% of the contenders win, so 20% lose). Would the 80/20 rule be statistically incorrect given that field sizes vary greatly from race to race?
Jeebs is offline   Reply With Quote
Old 04-16-2019, 11:48 PM   #3
Mitch44
Grade 1
 
Join Date: Nov 2014
Location: The Villages, Fl.
Posts: 3,705
This rule would work in racing as in any other endeavor.

Horseracing requires some good judgements be made by the user. Specifically in field size the user would have to round up or down and by a consistent application. Myself I would just always round up to the next highest number. I.e.6 horses would be 1.2 (20%) I would use 2 horses etc. The reason is that seldom does 1 horse completely dominant any race.

Sartin and Bradshaw always said "that seldom or rarely are there more than 3 real contenders in a race." 80-20 rule, 20% of a 12 horse fields is 3. I'm sure Sartin was well aware of this rule many years ago.(circa 1990 at least) I've been aware of it for some time but that long. Sartin's 3 contenders holds true for the majority of races. Like any racing rule there are exceptions just as in classes of races, strong and weak fields of the same class.How would one treat unknown factors? One could even after rounding up to always add 1 more contender to it.

If you can get the winner in your contenders at least 84% your doing good. Lt1 and I do better than that. Even with that if I'm investing more than a fun bet I still do a review of all the other horses in a race (non-contenders). I confirm their a toss or a keeper. An improving horse will not show up in most contenders selection methods because the horse hasn't been given the opportunity to establish its worth yet. I.e. a claim , equ. change and the horse improves greatly. The question becomes ;"where is its top"

Speaking of Sartin's 3 true contenders in a race, the average Sartin factor for a winner is 3 or less. How neat is that?

As far as a 20 race cycle Richard is correct in that it could be more accurate. Sartin said; "if it brothers you that much use a cycle of 21 " "beyond that is a complete waste of time." Those that need a data base of 1,000 races etc. are just wasting their time which could be used for better research of other ideas. Expecting exactness or perfection with all this is a fools errand. Concept is the key here not exactness. What happens in a 20 race cycle will happen in 100 or 1000. If it gives you confidence or you want a confirmation you can do a cycle of 100. I've never done more than a 100 cycle. If my idea doesn't pan out in 20 I'm out of there with a new idea or thought and new study.

Lastly for any cycle or collection the collector must do it in a consistent manner or the results will be greatly skewed.

Mitch44
Mitch44 is offline   Reply With Quote
Old 04-17-2019, 05:47 PM   #4
rdiam
Grade 1
 
Join Date: Sep 2016
Posts: 606
The nice thing about statistical inference is that it is objective -- no opinions needed. Sartin's opinion was to use 20 (or 21) race cycles: the math says 50 minimum.

As far as contenders and odds lines, I prefer to start with a random odds line and adjust it based on contenders. For those not familiar, if there are 8 horses in a race, the random probability would be 1/8 or 12.5%, and the random odds line would be 7-1 on each. Now if a horse is a win contender, it should have a win probability higher than random, and hence an odds line below 7-1. Similarly, non-contenders should have an odds line above 7-1. Once you have played around with this kind of separation, just remember to adjust your probabilities to add up to 1.0. I would agree with Mitch44/Sartin that you should only have 3-5 contenders in a race if your approach is going to produce +EV.

If you are really ambitious, you might also want to construct an odds line for how you expect the public will bet (see Sklansky, "Getting the Best of It"). The theory says to only bet on Expected Overlays to your betting line, and pass on Unexpected Overlays.

Richard
rdiam is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is On

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
20 Race Cycles - For Beginners and Old Alike Because I Can Jim Sartin Methodology Handicapping 101 (102 ...) 1 04-13-2017 03:56 PM
How to Make People Believe Because I Can Jim General Discussion 9 08-13-2013 05:43 PM
Track Your 20 Race Cycles Spreasheet / Video Bill V. Contest Admin Stuff 0 07-03-2012 12:59 PM
Official Contest Rules (please respond in the Poll to enter the Contest) Ted Craven Contest Admin Stuff 1 05-31-2012 12:48 PM
Wagercapping - Follow Up Articles Ted Craven Previous 'Handicapping Discussion' Forum 19 03-22-2010 06:06 PM


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.

All times are GMT -4. The time now is 02:15 PM.