Sports Line Auditing
Thread Rating:
My friend and I bet every NFL game using the point spreads on Bovada. He picks all games one week then I pick the next. All weeks picks are made at once, no later than Thursday. We’re both reasonably honest, but I do usually verify the lines he uses for his week.
I get bored checking lines sometimes. What would be a a good way for me to sample them instead? For instance, if I successfully verify 5 of 16 lines, how confident can I be that all 16 are correct?
Quote: Ace2This is more of a math question than a sports one.
My friend and I bet every NFL game using the point spreads on Bovada. He picks all games one week then I pick the next. All weeks picks are made at once, no later than Thursday. We’re both reasonably honest, but I do usually verify the lines he uses for his week.
I get bored checking lines sometimes. What would be a a good way for me to sample them instead? For instance, if I successfully verify 5 of 16 lines, how confident can I be that all 16 are correct?
If you don't trust him you shouldn't be betting at all.
Feels like you need an a priori and then apply Bayes.Quote: Ace2This is more of a math question than a sports one.
My friend and I bet every NFL game using the point spreads on Bovada. He picks all games one week then I pick the next. All weeks picks are made at once, no later than Thursday. We’re both reasonably honest, but I do usually verify the lines he uses for his week.
I get bored checking lines sometimes. What would be a a good way for me to sample them instead? For instance, if I successfully verify 5 of 16 lines, how confident can I be that all 16 are correct?
We’re only humans. Even I have made an occasional half-point error. In my favor of courseQuote: DRichIf you don't trust him you shouldn't be betting at all.
What if there isn’t a priori? Meaning each side will cheat as much as they think they can without getting caught.Quote: unJonFeels like you need an a priori and then apply Bayes.
If they know all games are being checked, they won’t cheat at all. Conversely, they’ll cheat on all 16 games if they feel nothing’s being checked. And anything in between.
Is that equivalent to a 50% chance of cheating on any single game?
Quote: Ace2What if there isn’t a priori? Meaning each side will cheat as much as they think they can without getting caught.
If they know all games are being checked, they won’t cheat at all. Conversely, they’ll cheat on all 16 games if they feel nothing’s being checked. And anything in between.
Is that equivalent to a 50% chance of cheating on any single game?
What’s the benefit of cheating and not getting caught? What’s cost of cheating and getting caught? If you model the payoffs you should be able to find a Nash equilibrium.
Cost: Damaged trust I suppose. The other side may become less willing to accept future wagers
Btw I’m referring to small “adjustments” like 1/2 point. In NFL, though, it’s amazing how often that hook makes a difference.
Quote: Ace2Benefit: More favorable line and increased probability of winning
Cost: Damaged trust I suppose. The other side may become less willing to accept future wagers
Btw I’m referring to small “adjustments” like 1/2 point. In NFL, though, it’s amazing how often that hook makes a difference.
why dont you just set a rule where you each send a screen shot of the account summary. I wouldnt say, hey I don't trust you but would just say since lines move so quickly it would be easier to just get a screen shot so I knew what I was following each game.
ZCore13
Though this scenario is loosely based on actual events, all I care about is the math involved. Also I love mathematical estimates/shortcutsQuote: Zcore13It takes like 3 minutes to look at 16 lines and verify. Do you not have that kind of time?
ZCore13
