Is using poisson distribution to predict o/u pitcher strikeouts OK? The 4 criterion listed in the book are over 20 trials, odds off occurrence are less than 10%, go up in increments of 1 and past distribution is the same. The only criteria which is not met is odds off occurrence, the average strikeout rate is around 18%. Does that rule out any way of successfully using poisson to price the prop or could it still be close enough that it would still work?
Our head moderator, author, and professional sports bettor Justin7 once made a video explaining why poisson isnīt a good idea for strikeout props. It was pretty worthless, but you may be able to find it.
If it's not a rare event, which in this case it is not, then try using a binomial distribution instead of a Poisson distribution to determine the probability. No guarantee it will work, but it's worth a shot testing it I would think.
The way I have the model set up using binomial distributions won't work with out making a ton of calculations. I am going to take a sample of starting pitchers strikeout in each game started to find a standard deviation and then use the normal distribution in excel and see how that works.