Welcome to the first week of Upset Watch 2021. In this column, I break down the week’s entire slate and pick the underdogs who are most likely to beat the odds in week 1.

Over the last few years, Upset Watch has accumulated over $5,000 in profit, with last year accounting for over half of that figure - finishing with an excellent pick of the Buccaneers as underdogs in the Super Bowl, when the media consensus was 76% in favor of the Chiefs.


How do we do it? Read on to find out our methods, but first, let’s talk about an exciting addition coming today.



Our improved favorites feature is now live! Last year’s system was frankly, terrible. I had originally designed the favorites feature in 2018, and we’d even launched it at the back end of that year, only for a slightly rogue decision by a former member of our tech team to delete it when migrating to our new technology stack in the summer of 2019.


Well, they’re back. You can now click the favorites icon in the top right of the table (click the three little dots on mobile) and you’ll see stars appear next to each expert. Clicking those stars and turning them yellow will add them to your favorites. They'll automatically save, and you can of course search for experts and fans while the favorites selector is open.


Once you're done, you can select your favorites from the same drop down menu that has ‘experts only’, ‘fans only’ and ‘combined’ in it. You can also remove experts that way too.



A big improvement with this feature is that the consensus will now update with only these experts, and you can also add users to the list alongside the pros.


Once we have a second pass of this, we’ll add multiple lists that you can name accordingly, and the ability to integrate pro stats into the lists, but for now, the favorites panel should allow you to very simply see your favorite experts, and that’s exactly what I believe most people want!



If you’re a first time user of Pickwatch or first time reader of this column, let me talk you through the column. Our aim is not to predict the winners of every game. There are many tools on the site that will bump your overall percentage up and help you win pick’em leagues or other competitions.

Upset Watch is different. We’re lasered in on underdogs, which pay out higher rewards from sports betting. For example, a $100 bet on the Cowboys at +340 in week 1 will yield $340 profit if they win. Conversely, picking the Bucs will only give you a $22.50 profit from that same $100 stake.

Around 32% of NFL games end in an outright upset. The entire point of it being an upset is that it will be a surprise. Upset Watch exists to find out where those 32% occur statistically, and help you match up the historical trends with the upcoming slate of games.

Part 1: Number of upsets and underdog covers
First, we need a big picture. To do this, we take every game played over the last 8 years, and we break it down by week to find the average number of straight up wins by underdogs, and the number of underdog covers.

Why do we do this? Because our public data shows that fans do not like picking underdogs, and they gravitate towards favorites to cover the spread, even though the spread is roughly a 50-50 proposition in most years.

Here’s what I mean. In the table from 2020 below, you can see that the average fan will pick 4 underdogs less than they need to each week, whereas Upset Watch picks just half a game less on average.

Of note, you can see that in many weeks, the number of fan picks was wildly incorrect. In week 13, there were 9 underdog covers. We predicted 7, the fan consensus predicted just a single cover.


So we understand that in some weeks there will be more, in others less. Check out this chart, that shows the variation from week to week.



From the highs of 9 games in week 2, to the lows of just 5.5 games in week 9, our principle is simple: Always pick the right number of upsets.

Part 2: Where do Upsets occur?
So now we know that we need to find 7 teams who will cover, and in week 1 of this season, we also need 5.3 (rounded down to 5) outright winners from among the underdogs.

But where do upsets occur more often?

Again, we look at previous seasons, and specific weeks as a guide. Overall, underdogs in week 1 are just 43 and 83 straight up, so your chances of a victory are just 35% if the spread favors the opponent.

However, there are sweet spots to be found. For example, when the spread is under +2.5, teams are 16-9, so the underdog is actually more likely to win than the favorite in that circumstance. That applies to two teams this week.

We also want to know where they don’t occur, and that’s generally in big spread games. At most, you should be picking 1 team with a spread over +7 points this week, as those teams have managed just 6 victories between them since 2013 from 27 attempts.

When the spread comes out to +3, roughly 40% of those games end in an underdog victory, so of the 6 teams on this week’s slate currently at +3, we could expect at least two of them to triumph.

Finally - covers. Those same big underdogs who we wouldn’t pick to win, fare a lot better against the spread. Teams over +7 are 18-8-1 since 2013, and have a 66% chance of beating the spread.

All told, we’re picking:

  • Two teams facing low spreads
  • Two teams facing +3 spreads
  • One team facing a big spread
  • And two more teams facing big spreads to cover, but not win.

PICKS (Pro Users Only)

This article is now part of our Pro service. To read the picks, you’ll need to be subscribed to Pickwatch Pro. It’s just $30 for an entire year ($2.50 a month!) or $7.99 for a monthly subscription.




For a full list of features, click here, and if you’re in Arizona, Colorado, Indiana, Iowa, Michigan, New Jersey or Pennsylvania, you can get a year free by placing a $10 bet with our friends at BetMGM.