Which team will be the biggest surprise in 2014? Last year, the Houston Texans shocked Vegas and analytics fiends alike. Before the season, the Texans’ over/under win total was 10.5. Football Outsiders Almanac projected them to have 9.3 wins and gave them a 67% chance of making the playoffs. Basically nobody saw the 2013 Texans’ implosion to 30th in DVOA coming. Interestingly, though, the Texans are part of a larger trend in the kinds of teams that have been having enormous drop-offs in performance.

Consider the graph below. It looks at the change in DVOA for good-but-not-great teams, those that ranked between 6th and 15th in the previous year. ^[1]Before 1989, I use Andreas Shepherd’s estimated DVOA. I thank him for sharing his data.

Historically, the good-but-not-great teams have regressed a little bit. From 1985 to 2010, those teams dropped on average between two and four points of DVOA. The trend was relatively stable for each five year period. While we would expect some regression from good teams, the size of that regression has changed since 2010. Over the last four years, the good-but-not-great teams have dropped an average of ten points of DVOA, the biggest regression by far since the merger. Note that if we drop 1983 to account for regression coming out of the strike-shortened 1982 season, we get a DVOA change for 1980-1984 of about four points of DVOA, making 2010-2013 even more clearly on its own island. This idea leads into my first prediction for the season.

Prediction #1: There will be a falling-off-the-cliff team. At least one team in the top half of the league by DVOA in 2013 will be in the bottom three by DVOA in 2014.

There has been a falling-off-the-cliff team just 15 times in the 44 seasons since the merger, but it has happened each of the last four years. Since 2010, 6 of the 12 teams that finished in the bottom three for DVOA actually finished in the top half in the previous year. While the last four years may just be further evidence that sometimes, splits happen, my personal feeling is that there’s a good chance that this year will be the fifth year in a row. And if there is a falling-off-the-cliff team, previous examples can help to identify the best candidates for that ignominious distinction. Going back to the merger, here are the teams that transitioned from the top half of the league into the bottom three in DVOA the following season.

Since 2010, each of the six falling-off-a-cliff teams had a huge drop-off at quarterback. It’s hard to learn too much from the two biggest nosedives at that position. The 2010 Cardinals and 2011 Colts replaced Hall of Fame-level quarterback play with, well, somewhat less awesome quarterback play. ^[2]The pattern in Figure 1 still holds if we drop the 2010 Cardinals and 2011 Colts. For 2010-2013, the average DVOA drop for good-but-not-great teams falls from 10.3 to 8.7 points. The 2010 Panthers went from very bad numbers the previous year from Jake Delhomme (but good ones over five starts from Matt Moore) to atrocious with Jimmy Clausen and Matt Moore. The 2011 Buccaneers went from good Josh Freeman to bad Josh Freeman. The 2012 Titans went from solid Matt Hasselbeck to not-so-good-and-not-always-healthy Jake Locker. The 2013 Texans traded in OK Matt Schaub for bad Matt Schaub and slightly worse Case Keenum. ^[3]Writing this sentence made me think of the 1962 Mets and Harry Chiti, the only player in baseball and probably sports history to be traded for himself.

Bearing this in mind, here are our candidates for the 2014 falling-off-a-cliff team, the teams that ranked between 6th and 15th in DVOA last year:

To find the best candidate for the team to fall apart, I just want a team with the potential to fall apart at the game’s most important position. That leaves me with a pretty good set of options: Kansas City, Arizona, Chicago, and St. Louis. I almost included Cincinnati, but an implosion from Andy Dalton does not really fit his trajectory of solid but not spectacular play. Chicago makes the list because Jay Cutler gets hurt a lot and Jimmy Clausen already was responsible for torpedoing one recent team into the NFL depths. St. Louis got bad quarterback play already last year, and Shaun Hill may not be any worse than what they have. Given his age, though, it is certainly in play for him to be terrible even though that seems not to fit. Kansas City has some merit and is Chases’s pick, since their line may take a step back and Alex Smith was only OK even last year.

Arizona seems to be in a similar situation to St. Louis. Carson Palmer is almost exactly the same age as Hill, and despite having navigated very different career paths, there may not be much of a difference in ability as of 2014. Palmer has better weapons, but since St. Louis has a head coach who’s done pretty well even with some iffy quarterback play. Arizona also has the toughest schedule in the league, just another reason I like Arizona as being the leading candidate to finish as the falling-off-a-cliff team.

Projection #2: The falling-off-a-cliff team will be Arizona, Kansas City, or St. Louis. Chicago enters the mix with a long-term Cutler injury. Arizona is the leader at the starting gate.

In contrast to the good-but-not-great teams, the great teams have shown a recent tendency to stay great. In fact, from 2010-2013, the best teams have retained their value at a rate just a little below 1970-1989, when teams could keep their core intact much more easily.

The real outlier in the graph is 2005-2009, when the best teams lost on average almost 20 points of DVOA in the following year. Since 2010, the best teams have only lost about half of that value. Over the last four years, four teams account for more than half of the spots in the top five by DVOA: Denver (2), Green Bay (3), New England (4), and New Orleans (2). Not coincidentally, those teams have great quarterbacks. ^[4]Yes, this sentence is only slightly more informative than “The sky is blue.”

So things have been relatively stable at the top recently. The 2009-2013 Patriots are the first team since the 1991-1998 49ers―who should have won more Super Bowls― to rank in the top five in DVOA for five consecutive years. Expect more of this consistency from the best teams.

Prediction #3: The Seahawks, Broncos, Saints, Packers, and Patriots will all be in the top eight teams by DVOA in 2014, barring a quarterback injury.

When you start multiplying probabilities, the odds on that get pretty long. If you put together the odds on these teams winning their respective divisions, you get about a 14% chance of all five bringing home the bacon. Since Vegas likes to make money, the odds are actually a little longer than that. Still, it’s hard to see how these five teams end up outside the top eight without losing their quarterback.

Finally, how about the teams at the bottom? There has been an unusually large amount of upward mobility lately. Looking at the bottom ten teams by DVOA, here is how they perform the year after they scraped the bottom of the NFL barrel.

Since 2010, teams in the bottom ten of the league have gained an average of 14.8 points of DVOA the following year. That increase is about five points more than the previous and relatively stable trend. Getting off the mat has never been easier.

But it does help to not be starting from the absolute bottom. Going back to 2004, here are the teams that finished in the bottom ten in DVOA one the previous year and then surprised the next year with a DVOA of at least +10.0%.

The 2013 Chiefs are clearly the anomaly. The other ten teams since 2004 that went from bad to pretty good ranked no lower than 28th the previous year. So if we want to identify the best chance for a team to graduate to pretty good, we probably want to focus on the teams that finished higher than 28th last year. To make this a little harder, I’ll focus on the teams that finished between 25th and 28th: the Falcons, Vikings, Giants, and Browns.

Prediction #4: At least one of the Falcons, Vikings, Giants, and Browns will finish with a DVOA over 10.0 and rank in the top ten teams in the league.

All of this leads up to my season projections. After all that careful reasoning based on historical data and recent trends, now I lay all that to waste and put an exact number of wins on each team. I also put each team’s projected DVOA ranking for 2014 in parentheses. And even though picking the playoffs is completely insane at this point, I do it anyway.

So I am looking for recent trends to mostly continue this season. The biggest dogs will remain very large. At least one strong team from last year will fall to the very bottom of the league. And one team from close to the bottom will rise from the ashes to at least temporarily join the NFL elite.

I look forward to being horrendously wrong about all of this, of course. ^[5]In the spirit of getting things totally wrong, here are my picks for week 1. First, back in May I bet on the Buccaneers getting 2.5 at home against Carolina. The line on that has moved to Tampa … Continue reading Can’t wait to see Paul Ryan’s twin on TV’s greatest channel in just two days.