One.

A single team who was an underdog won in week 9. That's never going to help us, and it caps a run that we have never seen before.

Since we started tracking dogs back in 2013, last week was only the second time in 108 weeks that only one underdog triumphed in the first 9 weeks of a season. The average was almost 5 underdogs in previous years.

Beyond that, there's another thing that's hurt us a lot this year. If you have a bad week or two, that's to be expected, but we're also seeing fewer underdogs winning than most other seasons.

Year	Upsets	Weeks with 3 or fewer underdog wins
2013	40	2
2014	44	2
2015	46	1
2016	50	2
2017	48	2
2018	42	4
2019	47	2
2020	49	1
2021	53	0
2022	51	0
2023	45	2
2024	43	4

Translation? It's statistically pretty unusual to have so few underdog winners, and more importantly, those losses are not spread evenly over the season. If you're down 1 game every week, that's still plenty of opportunity. If 5 teams win in week 2, for example, and the average is 6, your chances of picking a winner are still good. Unfortunately, we've now had 3 weeks in 2024 where we've had 3 fewer underdogs than we would expect to win, and another with 2 less. 

All of those weeks correspond with our worst weeks, so while I'm not overjoyed by these terrible weeks, I do think that at some stage we'll begin to see a regression towards the mean, and more underdog winners over the back end of the year.

Top picks

Unfortunately, the Panthers weren't in our 4 top picks (see, I didn't bet the farm!) and that meant a -20 unit week.

All underdog picks

We did have one winner, the Bucs to cover (but not win) against the Chiefs on Monday night. That meant that for all picks, we were down 25.5 units.