Hey folks!
I'm trying to decide what percentage of my bankroll should I bet on each game. Here's some data
-I can consistently predict who wins 68.6800895% of the time. This is over a span of 1340 games.
-I predict correctly 79.6442688% when betting on the favourite, and when betting for the underdog I'm correct 34.9544073% of the time.
-I have also logged the odds and how my win percentage for each odd. For example at 1.3 decimal odds moneylines, I'm correct 93% of the time. I've managed to come up with a pretty strongly correlated linear equation for this data.
So should I use my overall percentage when calculating Kelly, a weighted one depending on if I'm betting on a favourite or not, or use my equation to estimate my win percentage based on the decimal odds?
Also should I use some sort of fractional Kelly? Also since games can be simultaneous, how should I deal with that?
I'm trying to decide what percentage of my bankroll should I bet on each game. Here's some data
-I can consistently predict who wins 68.6800895% of the time. This is over a span of 1340 games.
-I predict correctly 79.6442688% when betting on the favourite, and when betting for the underdog I'm correct 34.9544073% of the time.
-I have also logged the odds and how my win percentage for each odd. For example at 1.3 decimal odds moneylines, I'm correct 93% of the time. I've managed to come up with a pretty strongly correlated linear equation for this data.
So should I use my overall percentage when calculating Kelly, a weighted one depending on if I'm betting on a favourite or not, or use my equation to estimate my win percentage based on the decimal odds?
Also should I use some sort of fractional Kelly? Also since games can be simultaneous, how should I deal with that?