Tweet
Finally, the model gets to shake off its 2019 dusty data, and is now ready and loaded with all current stats, so – look out! LOL. Well, the model could hardly do much worse in terms of its “picks”. But I’m still satisfied with its performance. Statistical projection models like Number Fire and TeamRankings haven’t done much better (NF has clocked at 41% ATS and TR, a woeful 33%) so I’m not down on my own model – yet. Remember, data projections are a guide, and projected “picks” are only a grading variable, and should never be tailed. Anyway, GLTA this week.
-
-
This season is slowly becoming a Covid 19 nightmare, just as it has been all year for everyone. Of course, it’s the players lives, and the lives of their families, which matters most, but just a few more teams with positive test results may wreck the schedule; and, unlike MLB, that kind of chaos is much harder to repair and could end the season or, at least, shorten it. Players need to take the Coronavirus reality more seriously, and the NFL needs to wield a heavier hammer on those who violate protocols (a moron like John Harbaugh should be suspended for the rest of the season). With every new positive test result, I’m beginning to despair for the NFL season.
-
Comment
-
For reasons known only to itself, the model has been refusing to sweep all closing lines each week until the last game of the week has been played. It takes too much effort to fix the problem, so I’ve been updating closing lines manually, with hit or miss results (my eyes and fingers are not what they were) – and I’ve decided to just stop, and wait until the model does it after the MNF game.
For Week 4 a couple of line changes flipped the final picks into wins for the model, but it’s usually the other way around. Still, ridding itself of 2019 data has, apparently, brought some respectability to the Limper’s MOV projections, so I won’t complain.
GLTA in Week 5.
Comment
Comment