Calgarypuck Forums - The Unofficial Calgary Flames Fan Community - View Single Post - Rube's 2019 NFL Gambling Odyssey (and NFL Preview Thread)

rubecube · 09-24-2019, 04:38 PM

I guess I can sort of go into a bit of detail on what I'm doing now. Basically what I'm doing is creating my own point spreads for games based on some of the stats I used to project win totals for the season and then looking for discrepancies between the spreads I've set vs. the spreads the sportsbooks have set.

So for example, I had last night's game as CHI -10.5, whereas the books had it at CHI -4, representing a 6.5 point advantage/value for me, which was one of the highest advantages I had this week, the others being CAR +2.5 (my line was CAR -6.5) and PIT +6.5 (my line was PIT -1.5). Obviously the point isn't for my lines to be correct. They're just there to alert me to where the books are most likely to be wrong.

I'm not still using the 2018 stats at this point, because that'd be stupid, but I used them as baselines and then adjust things every week after the games have been played.

I think where I went wrong in the first couple of weeks was not considering which of my lines were extreme outliers and just going based off of what I saw as the biggest advantage. To correct that, I tracked the median, mean, and mode point advantages of the winning games the spreadsheet predicted as well as the losing games to give myself a range to stay within. I don't know if that will still be necessary as the season progresses and I can set my spreads based more on current data but it's probably a decent strategy to have for the first few weeks of a new season when the sample size is too small and still largely relying on the previous season's data.

Or it could all be junk and I end up tossing it out in another couple of weeks.

09-24-2019, 04:38 PM	#70
rubecube Franchise Player Join Date: Mar 2006 Location: Victoria Exp:	I guess I can sort of go into a bit of detail on what I'm doing now. Basically what I'm doing is creating my own point spreads for games based on some of the stats I used to project win totals for the season and then looking for discrepancies between the spreads I've set vs. the spreads the sportsbooks have set. So for example, I had last night's game as CHI -10.5, whereas the books had it at CHI -4, representing a 6.5 point advantage/value for me, which was one of the highest advantages I had this week, the others being CAR +2.5 (my line was CAR -6.5) and PIT +6.5 (my line was PIT -1.5). Obviously the point isn't for my lines to be correct. They're just there to alert me to where the books are most likely to be wrong. I'm not still using the 2018 stats at this point, because that'd be stupid, but I used them as baselines and then adjust things every week after the games have been played. I think where I went wrong in the first couple of weeks was not considering which of my lines were extreme outliers and just going based off of what I saw as the biggest advantage. To correct that, I tracked the median, mean, and mode point advantages of the winning games the spreadsheet predicted as well as the losing games to give myself a range to stay within. I don't know if that will still be necessary as the season progresses and I can set my spreads based more on current data but it's probably a decent strategy to have for the first few weeks of a new season when the sample size is too small and still largely relying on the previous season's data. Or it could all be junk and I end up tossing it out in another couple of weeks.