Not for the faint of Math
Good Day All
some musings on a major change to Mullumby Defender, not to the calculation of the predicted scoreline and margin, but rather to the confidence in it from game to game.
Currently, the model assumes the true "home margin" result will lie in a normal distribution around the predicted value with a standard deviation equal to the MAE of the season. Reasonable as a overall tactic, but what about doing better for individual games?
For the Mullumby Defender selected games, the success rate for line bets (50% bookie rated) prior to reliability is around 60-65%. After adding the "reliability" criterion which relates the reliability of each team in the past to be successful when selected in a game, the success rate.... well we are on 11 wins from 13 bets in 2018, and the historical modelling puts it around 75-85%.
This means apples are not apples, some predictions are better than others.
Further evidence is found in the Mean Absolute Error for the reliability selected bets is currently 19, and the overall MAE for the model is 30.
So, using past successful bet rates incorporating the selected bets that we don't end up using because reliability is less than 0.35 does seem to be working.
I would like to change the overall model method for predicting the probability of success to an individual team based methodology.
The models heart is a "team strength" in points and a "team defence" as a %
If I use the standard deviation associated with both of the above and use Physics error handling all the way to the home team margin, then I will end with a different "error" for each game to use as the standard deviation of the normal distribution when calculating the probability of a line bet success.
I will be careful NOT to modify the "original" Defender which is going so well, but rather to make another in a different place.
Any thoughts are welcome :)
some musings on a major change to Mullumby Defender, not to the calculation of the predicted scoreline and margin, but rather to the confidence in it from game to game.
Currently, the model assumes the true "home margin" result will lie in a normal distribution around the predicted value with a standard deviation equal to the MAE of the season. Reasonable as a overall tactic, but what about doing better for individual games?
For the Mullumby Defender selected games, the success rate for line bets (50% bookie rated) prior to reliability is around 60-65%. After adding the "reliability" criterion which relates the reliability of each team in the past to be successful when selected in a game, the success rate.... well we are on 11 wins from 13 bets in 2018, and the historical modelling puts it around 75-85%.
This means apples are not apples, some predictions are better than others.
Further evidence is found in the Mean Absolute Error for the reliability selected bets is currently 19, and the overall MAE for the model is 30.
So, using past successful bet rates incorporating the selected bets that we don't end up using because reliability is less than 0.35 does seem to be working.
I would like to change the overall model method for predicting the probability of success to an individual team based methodology.
The models heart is a "team strength" in points and a "team defence" as a %
If I use the standard deviation associated with both of the above and use Physics error handling all the way to the home team margin, then I will end with a different "error" for each game to use as the standard deviation of the normal distribution when calculating the probability of a line bet success.
I will be careful NOT to modify the "original" Defender which is going so well, but rather to make another in a different place.
Any thoughts are welcome :)
Comments
Post a Comment