Flat betting versus the Kelly Criterion - a simulation

[specialronnie29]

dark horse if you think feel like ev betting opportunities are getting smaller and smaller and squeezing you out then you should listen to these guys

believe me they have no problem finding good bets

neither do i

intuition lol

[Romanov]

This was an epic read

[danshan11]

Exploring ideas is one thing, but there's a limit to how many times people can discuss "how should I size my bets? What if I size my bets based on the shoe size of all the players on the team I think will win?".

these are the true problems with Kelly,
1. We assume the market is fairly effecient so that in turn means the closing line is the most effecient and is closest to the "implied probability" and to use Kelly correctly we need this number when we bet and it is not available until the event closes.
2. using our "edge" I consider myself very good at "guessing" the closing line and any game that has enough edge to bet is usually a game where my "guess" is larger different than normal and this creates the edge but in this instance it is already sort of saying I am wrong and I am going to use my "wrong" guess to determine my bet size using kelly assuming an edge based on a wrong guess to begin with. Even Pinnacle lines move 20-40 cents each way normally every game!
example I cap 5 games
game my guess current line
1 -122 -117
2 -166 -167
3 -145 -136
4 -144 -144
5 -175 -135 here is the edge game
so I want to bet this one right so I use my -175 and the current -135 to determine my bet size and in reality there are a couple problems here
1 why am I 40 cents off ? am I smarter than pinny? probably not
2 In most cases in my experience that line will never make it to -175 usually -145 -150 maybe -160 this means my bet size was based on bad info already
3 I am basing my win prob and bet size on a number -175 that is flat ass wrong! We are the small group of smart genius bettors but come on dude we are not 40 cents smarter than anyone!
you can run sims all day of Kelly verse anything but the reality is nobody knows their edge or probability until the line closes and that can create huge problems.
Ok so we say ok we are so accurate that we know that when this scenario appears we know it will average -145 so we use that number for our kelly calc and that is probably smarter but still not accurate.
and most of us cant even get this far on every bet, this to me says MOST people should not use Kelly period. 1 to 2% of Bankroll is probably the absolute smartest way for anyone except some super elite super monster that has his edge down to a very close to exact number period! Kelly is not for the average bear for sure!