Very fair question. To some degree I felt bad posting all these but at the same time I backtested the last few years and had similar volume with the results being good. I actually didn't post two late plays last week because of what you're saying, I had a bunch of plays at that point (of course they both won). So I just decided I'm going to post whatever the model spits out.
Oh duh and I should of mentioned, my model relies on in-season data and I loosen my requirements for plays starting in Week 14 (My model starts in Week 8).
Overall my results last year were: 82 Plays 54.7% And they beat the close 68.2%. This is strictly from backtesting based on opening lines (I didn't play or post these anywhere) but previous years results were similar. I also tested 7 years of playoffs and the results were actually better than regular season both in terms of closing value and w/l record but a much smaller sample.
And actually now that I'm looking at my spreadsheet for last year if I filter only for plays from Week 14-Week 16 they won at 50% and beat the close at 63% over 31 plays. The plays from Week 8-Week 13 beat the close 76% and at won at 64% over 51 plays. Neither is a huge sample but I think there may be something there. Either I'm loosening my requirements too much or my model doesn't work as well or some combination of both. Of course variance could be in play too but that's a fairly significant difference. I need to look at some earlier years now to be sure.
This is unrelated (I think) but I should mention that I take zero account of injuries into my plays. Playing them early in the week probably favors me in this regard as I'll be on the right/wrong side equally of a lot of the late injury scratches. But I probably should have said that at the beginning of this thread.
Thanks for asking though, and giving me another 20 hours of work
Not sure if I wasn't supposed to post past records without them being verifiable. If that's a no-no, just delete. Thanks.
It's technically against the rules, but the spirit of the rule is to prevent new posters from saying "Hi guys, I hit 58% last year, check me out!" When a respectable poster mentions backtesting results for the sake of discussion, it's not really the same thing and not something I'd feel the need to delete, so no worries.
Ok cool thanks. I didn't even think about it at the time I was responding and then it popped into my head that I had read that a while back. Good to know either way.
Updated Record: 25-19-2 +4 Units
Closing Live Value: 26/45 57.8%
Goats posted a good question a few posts back and unfortunately I haven't had the time to research the answer to it yet. I do think there's a chance my plays don't have value from Week 14- Week 16 so until I can say for sure or with reasonable certainty, there won't be any plays until the playoffs (I have tested 7 years of playoff results and the plays have been the strongest during this period).
Goats if you're reading, I have been super busy lately. I'm actually working on another sports betting project that's taking up a ton of my time. That along with tutoring a poker player, a wife, and Christmas around the corner, I don't know how quickly I'm going to be able to figure out the answer to your question. But I will get to it.
I doubt anybody's following these but if you are I'd wait a bit on this one - just moved to Indy -2.5 -113 on Pinny. Worst case it moves back to -110 or so and can still get 2.5. But if it keeps moving, it's possible some 3's show up.
Updated Record (Regular+Post Season): 26-19-2 57.8% +5 Units
Closing Line Value 27/46 59%
Postseason: 1-0 , Beat Close 1/1
Tracking postseason in addition to regular season because results have been substantially better than regular season in the back-testing I did. If that's a problem, please delete.
Updated Record (Regular+Post Season): 26-19-2 57.8% +5 Units
Closing Line Value 27/46 59%
Postseason: 1-0 , Beat Close 1/1
Tracking postseason in addition to regular season because results have been substantially better than regular season in the back-testing I did. If that's a problem, please delete.
You can break your record down any way you want, just make sure to give a complete record including everything at some point.
Honestly I don't know too much about NFL line movement. I'd guess the public is on Denver but besides my system liking Seattle, there are quite a few other things working against Denver that a sharp friend of mine pointed out. But these are things the public is very unlikely to know about or even care about.
I do see some 2.5's out there now. So yeah wouldn't surprise me to see some juiced 3's.
Final Record (Regular+Postsesason): 27-19-2 58.7%
Closing Live Value: 27.5/47 58.5% (Counted Seattle as a push. Pinny closed 1.5, BM 2)
Postseason Record: 3-0, Beat close 2.5/3 83%
I will be looking at the reason for so many plays in the latter weeks when I get home this fall. I'm fairly sure I know what it is, just haven't had the time to look at it.
It's that time again. I'll be posting some system generated plays in this thread as I did last year.
There was an issue Goats pointed out towards the end of last year where it was spitting out way too many plays and it's because I reduced the threshold on what makes something "playable". So that should be fixed this year and hopefully things go as well as last season.
I'm going to track CLV in the same simple way. If it moves off my # in my favor, that's a win. If it moves off my # against me, that's a loss. It's not perfect by any stretch but I'm not looking to prove anything here, just to hopefully share some winners. FWIW, I'm playing all these plays myself.
Comments
#301 San Diego +10 -110
#306 Atlanta -5.5 -110
#308 Tampa +5 -110
#310 Tennessee +2.5 100
#317 Houston +6.5 -120
#319 Buffalo -2 -110
#322 Miami +2.5 -110
#328 Oakland +3.5 -105
#332 Pittsburgh +3 -120
Adding: #326 Carolina -11 -110
Come on, his username isn't "ALittleBitTilt".
(I think he knows that I'm joking around)
I had the following # of plays by week last year:
Week 8: 8
Week 9: 9
Week 10: 7
Week 11: 8
Week 12: 10
Week 13: 9
Week 14: 9
Week 15: 10
Week 16: 12
Oh duh and I should of mentioned, my model relies on in-season data and I loosen my requirements for plays starting in Week 14 (My model starts in Week 8).
Overall my results last year were: 82 Plays 54.7% And they beat the close 68.2%. This is strictly from backtesting based on opening lines (I didn't play or post these anywhere) but previous years results were similar. I also tested 7 years of playoffs and the results were actually better than regular season both in terms of closing value and w/l record but a much smaller sample.
And actually now that I'm looking at my spreadsheet for last year if I filter only for plays from Week 14-Week 16 they won at 50% and beat the close at 63% over 31 plays. The plays from Week 8-Week 13 beat the close 76% and at won at 64% over 51 plays. Neither is a huge sample but I think there may be something there. Either I'm loosening my requirements too much or my model doesn't work as well or some combination of both. Of course variance could be in play too but that's a fairly significant difference. I need to look at some earlier years now to be sure.
This is unrelated (I think) but I should mention that I take zero account of injuries into my plays. Playing them early in the week probably favors me in this regard as I'll be on the right/wrong side equally of a lot of the late injury scratches. But I probably should have said that at the beginning of this thread.
Thanks for asking though, and giving me another 20 hours of work
I'll post here though with what I find...
Also, I'm going to try to look at this by early next week so expect a post/response then..
It's technically against the rules, but the spirit of the rule is to prevent new posters from saying "Hi guys, I hit 58% last year, check me out!" When a respectable poster mentions backtesting results for the sake of discussion, it's not really the same thing and not something I'd feel the need to delete, so no worries.
Closing Line Value: 20.5/34 60%
Pending:
#306 Atlanta -5.5 -110
#308 Tampa +5 -110
#310 Tennessee +2.5 100
#312 St Louis 5.5 -110
#317 Houston +6.5 -120
#319 Buffalo -2 -110
#322 Miami +2.5 -110
#324 Minnesota +4.5
#326 Carolina -11 -110
#328 Oakland +3.5 -105
#332 Pittsburgh +3 -120
Closing Live Value: 26/45 57.8%
Goats posted a good question a few posts back and unfortunately I haven't had the time to research the answer to it yet. I do think there's a chance my plays don't have value from Week 14- Week 16 so until I can say for sure or with reasonable certainty, there won't be any plays until the playoffs (I have tested 7 years of playoff results and the plays have been the strongest during this period).
Goats if you're reading, I have been super busy lately. I'm actually working on another sports betting project that's taking up a ton of my time. That along with tutoring a poker player, a wife, and Christmas around the corner, I don't know how quickly I'm going to be able to figure out the answer to your question. But I will get to it.
Happy Holidays all.
KC +2.5 -110
I doubt anybody's following these but if you are I'd wait a bit on this one - just moved to Indy -2.5 -113 on Pinny. Worst case it moves back to -110 or so and can still get 2.5. But if it keeps moving, it's possible some 3's show up.
Closing Line Value 27/46 59%
Postseason: 1-0 , Beat Close 1/1
Tracking postseason in addition to regular season because results have been substantially better than regular season in the back-testing I did. If that's a problem, please delete.
You can break your record down any way you want, just make sure to give a complete record including everything at some point.
San Diego +10 -112 (Bookmaker)
Closing Line Value: 28/47: 59.6%
PostSeason: 2-0 Beat Close: 2/2
I'm with ya. You think we'll see 3?
I do see some 2.5's out there now. So yeah wouldn't surprise me to see some juiced 3's.
BoDog's got 3 -120 up
Closing Live Value: 27.5/47 58.5% (Counted Seattle as a push. Pinny closed 1.5, BM 2)
Postseason Record: 3-0, Beat close 2.5/3 83%
I will be looking at the reason for so many plays in the latter weeks when I get home this fall. I'm fairly sure I know what it is, just haven't had the time to look at it.
Till next year...
There was an issue Goats pointed out towards the end of last year where it was spitting out way too many plays and it's because I reduced the threshold on what makes something "playable". So that should be fixed this year and hopefully things go as well as last season.
I'm going to track CLV in the same simple way. If it moves off my # in my favor, that's a win. If it moves off my # against me, that's a loss. It's not perfect by any stretch but I'm not looking to prove anything here, just to hopefully share some winners. FWIW, I'm playing all these plays myself.
Week 9 (All lines taken from Pinny)
Carolina +2
Jacksonville +11
Houston +2.5
Baltimore PK
W/L: 1-3 25%
ClV: 2-2 50%
Cleveland +6
Miami +2.5
KC -1.5
NO -4.5
Baltimore -9.5
YTD Results: 4-5 44%
Beat Close: 4-4-1 50%