Well hello there, welcome to upset watch week 7, our weekly look at the stats that matter when picking underdogs in the NFL.

First order of business is to talk about an exciting change that;s heading to Pickwatch over the next few weeks:

Matchup Hubs

I first conceived matchup hubs back in 2015 as a way of narrowing our data to be relevant only to a specific game, rather than the entire NFL. For example, if you were picking this week’s game with the Lions and Dolphins, rather than seeing someone’s record overall, you’d see their picks for those two teams in isolation.

Taking this one step further, I decided that it’d also be useful to see those two teams in the context of their upcoming matchups, so how well do the experts pick the Lions on the road this season and the Dolphins at home? By narrowing our data in this fashion, we could easily assess who has the most skill when picking the two teams, and who their pick was.

Despite various attempts to collate the stats necessary to run it in blog posts, my ultimate aim was that it would be an automated service that updated to reflect the picks that were live in the database. It was catch 22 trying to do it by hand, in that the time necessary to collate the expert data that was changing right up until game-time meant that it was impossible to then write 16 articles a week in the hour or so before kickoff.

In 2016, for the Super Bowl, I changed that tack a little and moved away from just the narrow ‘expert data’ and towards the entirety of the stats available. I began tracking things like points per game of both the team and their opponents, defensive points allowed, and the size of spreads, whether they were favored etc. Essentially, a bunch of stats that was very pick’em orientated.

This was the genesis for what will be happening in Matchup Hubs. Now that our team are settled in, we’re starting work this week on not only doing the above automatically for each game, but becoming a ‘one stop shop’ for each game.

On every matchup hub page, you will find:

Game info (time, date, venue, weather) Game stats (points per game, yards allowed, passing yards allowed etc) Injury reports Spread info (spread history over the week) Power rankings charts over the season Form (straight up, vs the spread, home and road) Expert data (narrowed only to the two teams as above) Betting trends (eg: public money %)

You’ll also be able to make your picks, of course.

Our aim with this feature is that you should be able to find out everything about a game (and any game over the previous five years) without having to trawl through lots of different sources. The biggest challenge we have is automating the data via an API (I won’t get too technical, don’t worry...) which means we don’t need to do any manual input each week. We think we’ve got that sorted, but of course, the next challenge is making it all work for you guys.

This is, of course, merely the start, but we think you’ll be really excited by matchup hubs. Anyone who finds our basic data useful will be even happier knowing they can see it in a wider context for each game, with multiple sources of data and information with which to make your picks.

Speaking of which, on with the upsets…

Last week

We went 3-4 on the upsets last week, mainly due to a flipped pick in the Chicago @ Miami game and getting shot of our Indy/NYJ pick in light of the injury report. Should have stuck to our guns on the former. That’s still left us with a 53.8% rating over the season, but we are in need of a big week, and I’m hopeful that this week is the one.

Methodology

This week we’re again looking at how many underdogs to pick based on the previous 5 years of week 7 matchups. Remember last week how we told you that people irrationally pick the favorite vs the spread? Well again, despite just 3 underdogs being backed vs the spread by the public in Vegas, 8 of them covered, just as we predicted. Let’s check out this week:

2013: 7 2014: 8 2015: 7 2016: 9 2017: 4

The average is a straight 7, which means we need to pick 50% of the underdogs from this week’s slate, but I think we need to break it down a little further in light of our recent mediocrity.

Let’s look a little closer at who those underdog wins belong to. There were 74 games, of which 35 resulted in ATS victories for the underdog and 3 were pushes (49.3%). Where it gets more interesting, is in the percentages of wins in certain scenarios, namely, home and road underdogs.

Road underdogs: 21-28 (42.9%) Home underdogs: 14-11 (56%)

Now we’re cooking. Home dogs are more likely than road dogs to cover, but how does the size of the spread change things? After all, if a team is an underdog by a lot at home, maybe that indicates they’re really bad, whereas if they’re marginal ‘dogs, it could be that Vegas is less sure.

Home underdogs where spread is less than or equal to 3.5: 9-4 (69.2%) Home underdogs where spread is greater than 3.5: 5-7 (41.7%)

Bingo. Almost 70% of home underdogs of less than 3.5 points cover the spread. In fact, in terms of straight up upsets, all of those ‘covers’ also resulted in an upset straight up.

What does this mean? It means we need to select 3 of the 4 home underdogs with a low spread, so that’s Miami (+3), Arizona (+2), Chicago (+3) and the Jets (+3). With just one game with a bigger spread (49ers are +10 at home to the Rams) we’ll pass on it and move on to rthe road dogs.

There are a whopping 9 road underdogs this week. How do they fare? Well...

Road underdogs where spread is less than or equal to 3.5: 11-22 (50%) Road underdogs where spread is greater than 3.5: 10-17 (37%)

Again, we favor the underdogs with lower spreads. Dallas (+1.5), New Orleans (+2.5) and Cleveland (+3). We’ll plump for two of those. We’ll also take 2 of the 6 road dogs with bigger spreads, meeting the 37% quota above.

So that’s my reasoning for the makeup of these picks, but who have I got this week?

The Picks

Upset of the week: Detroit @ Miami (+3)

My upset of the week performance this season has been terrible. You should just fade this pick, ideally, which is why I’ve decided to make this play the weirdest game of week 7, Brock Osweiler’s Miami Dolphins against the Detroit Lions.

Let’s be honest, Brock is very fortunate that he had the real hero of Sunday’s performance on his side, Albert Wilson. The two short passes that Osweiler sent less than 10 yards in the air were subsequently taken to the house via 43yd and 75yd TD runs, stats that significantly changed Osweiler’s stats with minimal impact from the man himself. That said, there’s no doubt that Osweiler has some talent in there somewhere. There’s a reason that Adam Gase likes him, and that, to me, is the key to this game.

Gase is nothing if not a stellar offensive mind. I wouldn’t say he’s done the right thing tying his cart to Ryan Tannehill (I’ll never let it go, Ryan), and I think there’s plenty of reason to think he’ll not be too disappointed to move on from Captain Turnover, The fumble in the jungle, erm… In(ter)ception…

Look, I think the Lions D is bad, not just ‘oh that’s quite bad’, but ‘wow, really?’ bad. They’re fortunate that the Broncos and Cardinals are the only two worse teams in the league vs the run, but this is the thing, the Dolphins are a mid-range (16th) rushing team, and the Lions are a poor rushing team. This is one of those games where someone like Wilson, who will slip tackles with ease, will cause havoc in the secondary against a porous defense, and couple that with Frank Gore rushing like it’s 2007 again, and you’ve got a good combination.

Also, the Lions are playing in this game and we haven’t got a clue who they are yet. Maybe they’re unfortunate with their 2-3 record, but they seem to be really odd on the road, and this feels like exactly that type of losable game…

2: New Orleans (+2.5) @ Baltimore

Of the three road dogs with spreads under 3.5, the Saints are the one that comes to mind as a team that can do some serious damage to their opponents. The Ravens are an incredibly inconsistent team. It’s like the mid-00s again, with a defense that is leading the league and an offense treading water at times.

My worry for Baltimore would simply be that they have only played one truly great offense this season (Cinci) and they lost in that game. I’m not actually sure any of their opponents can match up with Brees and the Saints, however, and the problem is that they have the second highest time of possession in the league, on average, yet a turnover differential of -2 and are ranked 28th in yards per attempt.

Essentially, what I’m saying is this: If the Saints go 2 scores up, it’s pretty reasonable to assume that the Ravens will struggle to stay in this game, and that two scores scenario is never far away with the ‘aints.

3: Denver @ Arizona (+2)

The Cardinals are the worst team in the league, but the Broncos are also terrible, and right now, there’s an increasing sense of inevitability about Denver’s trajectory. As mentioned above, these two teams are the two worst in the league vs the run, and while I fancy that the Broncos have a better weapon in Philip Lindsay to take advantage of that, I’m more sure that the Denver O-line is a disaster area. Garrett Bolles is a walking penalty, flagged more than any other player in the league over the last 2 seasons, and it speaks of a level of incompetence at front office and coaching levels, that someone can do that and remain pretty much an unquestioned starter.

Case Keenum… I like Keenum, but behind that line he’s a dead man walking. Also, he looks increasingly like Adam Sandler in The Waterboy. The Broncos pass rush is good, which may cause Josh Rosen problems, but I see this being one of those big days for David Johnson, and keep your eyes downfield for Christian Kirk after DJ sets up the play action big play…

4: Cincinnati (+6) @ Kansas City

My love for the Mahomes-Hill combination and KC in general is well documented on these pages, but the Bengals are the kind of team that will cause the Chiefs D some serious problems. I’m not sure about the win, but a +6 spread is the best value this week of the big-time road dogs.

The Bengals are not playing well on defense (ranked 30th) but the equalizer may be that the Chiefs D is just that bad, it may allow them enough ways to stay close. Indeed, the Bengals average 29ppg, which is enough to be 6th in the league. What happens when two bad defenses play two good offenses? I’m hoping that the answer is exactly what we think, but either way, I like the chances of one team or the other having a chance to win the game late on.

5: Carolina (+4.5) @ Philadelphia

Our second ‘big spread road dog’ pick, is the maddening Panthers. I’m sticking with them until they win, and this may be their best spot. First off, they love being underdogs under Ron Rivera. They’re 11-5 as ‘dogs since 2015, and that’s not a coincidence. Rivera trains his teams to fight adversity, and the Panthers are at their worst when they’re big-time favorites.

The Eagles were flattered by one of the worst QB performances I’ve ever seen from Eli Manning and the Giants last week. A half-way competent quarterbacking day could have been a serious challenge for the Eagles, who were mediocre in their early season outings under Carson Wentz. The Panthers have the 4th best defense in the league and the Eagles give up a hell of a lot of sacks… it’s not a good combination, especially with a generous 4.5pt spread.

6: Dallas (+1.5) @ Washington

These two teams are incredibly well matched, in fact so much so, that if you account for the 1.5pt advantage that Vegas confers to home teams, the difference is negligible.

The only category that truly separates these two teams is in their rushing attack. The Cowboys outstrip the ‘skins by a huge distance and have the second ranked rushing attack in the league. That won’t be a welcome proposition for the Redskins who have serious problems stopping teams using the run in situational football. It’s not that they’re giving up a huge number of yards, it’s that when they do, it’s usually for a first down. That ‘wear them down’ approach will be tough for Washington on Sunday, and might yield touchdown paydirt for Ezekiel Elliott.

7: New England Patriots @ Chicago Bears (+3)

Ah, my beloved Bears. Big Mitch Trubisky (don’t even know if he’s even that big) had one of the best games of his young career last week, taking on a meaningful opponent and gashing them while not making many mistakes. The defense really lost the game on two big plays (see also; Wilson, A) that should never have been as bad as they were.

The Patriots were, of course, on solid ground in Foxboro last week and despite the Chiefs running them close (and giving us an ATS win) they seemed to hit a good groove. They’re a different beast to the traditional Brady/Belichick offenses, but they’re grinding out results.

I’ve just got a feeling that despite what Belichick is inevitably telling his team, they may just look past the Bears after such a hugely climactic battle with the Chiefs, and see this as the beginning of a more ‘easy’ run that also sees them take on the Jets, Packers, Bills and Titans. Sure, it’s a favorable run, but those are the ones that catch teams out.

The rest:

8: Cleveland @ Tampa Bay (-3)

9: Minnesota (-3) @ NY Jets

10: Buffalo @ Indianapolis (-7.5)

11: LA Rams (-9.5) @ San Francisco

12: NY Giants @ Atlanta (-5.5)

13: Houston @ Jacksonville (-4.5)

14 Tennessee @ LA Chargers (-7)