How many races do you need to feel confident?

[BGboothA]

Ok, I have 26,992 races in my database since April. Looking at different angles for how many qualified races would it take for you to feel confident in an angle.

For instance, I have an angle that has occured 81 times in that 27000 or 0.3% of the time, but has a positive return of $1.60 per race. This is an extreme example but the question remains.

How many races should an angle have to be a true quality angle?

Others are obvious, some have 2,500 or so races and show positive ROI but how small of a sample is too small?

[brucedenton]

Can you share this angle??

[str]

Ok, I have 26,992 races in my database since April. Looking at different angles for how many qualified races would it take for you to feel confident in an angle.

For instance, I have an angle that has occured 81 times in that 27000 or 0.3% of the time, but has a positive return of $1.60 per race. This is an extreme example but the question remains.

How many races should an angle have to be a true quality angle?

Others are obvious, some have 2,500 or so races and show positive ROI but how small of a sample is too small?

I think it depends on the actual angle.

Some are off the wall while others make very solid sense.

Can you give an example of some of the smaller size angles in question?

[BGboothA]

Sure as many of you know I produce my own pace figures and handicap in a very mathematical way. I am a data geek and I am finally ok admitting it.

Angle #1 - When our top pick has an HBD Early Pace figure that is 80+ higher than the 2nd highest HBD Early Pace figure that horse is winning at a 33.7% clip with an average payout of $11.52 over the course of 101 races. *This is after taking out the statistical outliers meaning any payout that is more than double the average payout.

[str]

Sure as many of you know I produce my own pace figures and handicap in a very mathematical way. I am a data geek and I am finally ok admitting it.

Angle #1 - When our top pick has an HBD Early Pace figure that is 80+ higher than the 2nd highest HBD Early Pace figure that horse is winning at a 33.7% clip with an average payout of $11.52 over the course of 101 races. *This is after taking out the statistical outliers meaning any payout that is more than double the average payout.

I guess I should leave this to the handicappers but IMO , for this type of angle, you would need many more races to feel ok with taking it forward. 101 races seems light.

[BGboothA]

I guess I should leave this to the handicappers but IMO , for this type of angle, you would need many more races to feel ok with taking it forward. 101 races seems light.

I completely agree with you here, which is what led me to the question...I don't feel confident at 101 and feel like I need more, but I always feel like I need more data...at what point should I feel good.

[JBEX]

I remember reading years ago Bill Quirin who was one of the pioneers of pace handicapping did a study with a database of tens of thousands of races. If I remember correctly horses who had the lead at the pace call were a positive ROI. also
believe that as the odds went up the ROI's got higher.. This goes against the normal odds being blindly bet as the ROI's go up as the odds get lower. Agree that your sample isn't large enough but this would lend credence to the idea you might have something

[JBEX]

Sure as many of you know I produce my own pace figures and handicap in a very mathematical way. I am a data geek and I am finally ok admitting it.

Angle #1 - When our top pick has an HBD Early Pace figure that is 80+ higher than the 2nd highest HBD Early Pace figure that horse is winning at a 33.7% clip with an average payout of $11.52 over the course of 101 races. *This is after taking out the statistical outliers meaning any payout that is more than double the average payout.

btw factoring in that you eliminated the extreme payoffs that is an over 90% ROI which is quite remarkable. According to the Kelly Criterion you could bet 20% of your bankroll on every race although I would cut that in half to factor in the bombs. $1000 bank would have generated $10000 in volume at a 90% ROI.. Do the math that's pretty exciting stuff!!