We're now 3 winning weeks from 4, and our season is getting back on track

As I detailed last week, this season has been tough. No season has had fewer underdog winners (48) since 2013 (45), and we're ten down on the better seasons. No excuses, of course, and I'm happy that of the 17 underdogs who've won over the last 3 weeks, we've picked 8 of them.

Year	Upsets	Weeks with 3 or fewer underdog wins
2013	45	2
2014	48	2
2015	57	1
2016	58	2
2017	50	2
2018	48	4
2019	55	2
2020	55	1
2021	58	0
2022	58	0
2023	50	2
2024	48	4

Last week we nailed the Panthers (+240), Steelers (+130) and Dolphins (+110) in what was already a profitable week prior to the start of the late window on Sunday. Unfortunately, we did miss one incredible lean pick, which was the Broncos, on their last second blocked Field Goal. That one is a little difficult to digest, given how on the money the pick was, and how unexpected the loss. Had it won, we'd have had our biggest underdog winner of the season.

I was also unfortunate with the Texans, who narrowly lost in any event. It's a lesson learned to keep an eye on those last minute injury reports, as my pick was almost entirely based on Nico Collins playing. By the time I was aware he was missing the game, it was already kickoff. Doh! You can't win 'em all, but that is on me.

Oh, and one final point - the Patriots and Saints, our two next-best underdogs who didn't quite make the cut, won their games, while our top 5 favorites all won, so the picks, as you'd expect, are getting better as the season wears on.

Top picks

A good week saw us win 3 and lose only 1 (the Jets, less said the better) and generate a profit of +19 units, our second-highest of the year. The Panthers pick was also the second-biggest upset we've predicted on the year at +240, beaten only by the Bucs upsetting the Lions at +295 in week 2.

All underdog picks

Sadly our leans were much less successful. The Broncos will remain the one that got away, while the Texans was based entirely around Nico Collins playing, and he then... didn't. Unfortunately, too late to change that pick. The only genuine whiff of the three was the Titans +7.5. Ironically, had either of the other two picks been covers, they'd have won.

Points total picks

Our Discord users are aware, but our points total picks are killing it. We changed some of the algorithm criteria in week 6 to update a previously successful model, and it has yielded a 58.3% hit rate (42-31) and 71.4% (10-4) last week. Win total picks are at the bottom of each analysis page, and in our Discord, I also post the model's results - unvarnished and unaltered - on Tuesdays, albeit they change as the week draws on and some factors like injuries come into play.