Thanks for the posts. Maybe this is showing how little I know of betting, but - what do you mean when you say you have e.g. 30 cents on Dodgers/Padres?
From the table - AikidoML is producing odds, not win probabilities? Why? Why do you have two rows per game instead of just one with home/away team?
On the more general side - what numbers are you comparing on the table, exactly, to reach the conclusions you have? Which numbers/cells are added in after the program has run?
30 cents is a measure of edge - it means that I think I'm getting a similarly good price to if I got Under 7+130, so the program thinks it's got about 15% edge.
It's listed as two lines per page because that's how gambling lines are typically shown, and it is easier to work with that way when thinking about odds. In another context I might lay out the spreadsheet differently.
Odds are win probabilities - -209 means a 209 in 309 chance of winning, 141 means 100 out of 241, and all that. I chose to have it display the number in this form but there's no real difference. The Aikido numbers are what the program thinks are fair.
The columns Team through AikidoML are automatically generated, the rest I do manually (the ROT# grabber isn't working right). I choose manually which picks are good enough, and which to pass on. In an ideal world I wouldn't have to do that, but it should be clear why from the explanations.
So 30 cents is just the "conversion" between - and + odds? Is the edge in % always the cents / 2?
The picks/passes columns are odds as displayed on Pinnacle at the time you examined the model predictions? So we should compare the predictions to the value on pick/pass?
Very odd (ha) to me to think in terms of -/+ odds, I have such an easier time with direct win probabilities. Do you have any recommended reads for heuristics on thinking in American odds?
Again - sorry about the basic questions, I'm much more familiar with non-gambling sports analytics.
Yeah, those are Pinnacle odds. I reserve the right to shop elsewhere if Pinnacle is pricing me out, but I mostly treat this as a world with one book. I also can take runlines if they're sufficiently better than moneylines, and I want to bother.
For small amounts close to 50/50 yeah, 2 cents = 1% is fine. As you go farther along that breaks down (so after about -120 it starts to fray a bit, by -150 it's pretty wrong). I don't know of a read for it.
Thanks for the posts. Maybe this is showing how little I know of betting, but - what do you mean when you say you have e.g. 30 cents on Dodgers/Padres?
From the table - AikidoML is producing odds, not win probabilities? Why? Why do you have two rows per game instead of just one with home/away team?
On the more general side - what numbers are you comparing on the table, exactly, to reach the conclusions you have? Which numbers/cells are added in after the program has run?
30 cents is a measure of edge - it means that I think I'm getting a similarly good price to if I got Under 7+130, so the program thinks it's got about 15% edge.
It's listed as two lines per page because that's how gambling lines are typically shown, and it is easier to work with that way when thinking about odds. In another context I might lay out the spreadsheet differently.
Odds are win probabilities - -209 means a 209 in 309 chance of winning, 141 means 100 out of 241, and all that. I chose to have it display the number in this form but there's no real difference. The Aikido numbers are what the program thinks are fair.
The columns Team through AikidoML are automatically generated, the rest I do manually (the ROT# grabber isn't working right). I choose manually which picks are good enough, and which to pass on. In an ideal world I wouldn't have to do that, but it should be clear why from the explanations.
So 30 cents is just the "conversion" between - and + odds? Is the edge in % always the cents / 2?
The picks/passes columns are odds as displayed on Pinnacle at the time you examined the model predictions? So we should compare the predictions to the value on pick/pass?
Very odd (ha) to me to think in terms of -/+ odds, I have such an easier time with direct win probabilities. Do you have any recommended reads for heuristics on thinking in American odds?
Again - sorry about the basic questions, I'm much more familiar with non-gambling sports analytics.
Yeah, those are Pinnacle odds. I reserve the right to shop elsewhere if Pinnacle is pricing me out, but I mostly treat this as a world with one book. I also can take runlines if they're sufficiently better than moneylines, and I want to bother.
For small amounts close to 50/50 yeah, 2 cents = 1% is fine. As you go farther along that breaks down (so after about -120 it starts to fray a bit, by -150 it's pretty wrong). I don't know of a read for it.
Thanks for the responses!