The Cleveland Browns are surely hoping to pull off a '99 Colts or an '08 Dolphins.
Those are the only two teams since 1970 that have made a 10-win improvement from one year to the next. And while a 10-win jump from the hapless, zero-win 2017 campaign seems like a near impossibility, there is certainly some preseason buzz that this Browns team could at least win some games in 2018. And the "Hard Knocks" bump has barely even kicked in!
To a degree, ESPN's Football Power Index is joining the party, giving Cleveland a 55 percent chance to win at least six games this season. From a historical standpoint, that's more realistic, given that 68 teams since the merger made a six-win improvement from one year to the next.
By far the biggest reason the Browns have the best chance to improve by such a large amount is precisely because they won zero times in 2017. In a league with as much luck as the NFL, regression to the mean has a heavy influence on future seasons' results. Consider: Since the merger, 23 percent of teams that finished a season with 0-3 wins improved by six or more wins the following year. Teams with 4-9 wins accomplished the feat only 5 percent of the time.
So it stands to reason that since no team was farther from the mean than Cleveland, no team is more likely to make that kind of improvement. Which is exactly the case here, and the No. 1 reason why FPI gives the Browns the best chance in the league to make a dramatic improvement.
And what about the 10-win improvements of the '99 Colts and '08 Dolphins? For Indianapolis, that was Peyton Manning's second (and breakout) season in the NFL. In 2008, Miami brought in Chad Pennington at quarterback, and coach Tony Sparano and quarterbacks coach David Lee surprised the league with the Wildcat en route to an 11-win season.
Could Browns coach Hue Jackson actually pull off the 10-win improvement? Per FPI, there's a 2 percent shot of that happening.
So what else does FPI like about the Browns? Well, the anticipated positive regression isn't just about the Browns winning zero games a season ago. It's also that they didn't play like a zero-win team. Based on their end-of-season FPI rating from last year -- which is largely derived from their level of play throughout the year -- we would have expected Cleveland to win about 4.5 games last season, given its schedule. A large part of that is simply because it requires quite a bit of bad luck to win zero games.
Regardless, when FPI is evaluating the Browns and projecting their level of play in 2018, it thinks about them as coming off a season of bad play -- not all-time bad play.
Presumably, Vegas also sees last year's Cleveland team along these lines (their Pythagorean record based on points scored and points allowed was 3.3-12.7, per pro-football-reference.com), as the Browns' win total earlier this week was 5.5 at the Westgate Las Vegas Superbook. Vegas win totals are also an input into FPI.
Beyond that, there's one other major factor propping up Cleveland's projected win total: Tyrod Taylor. The former Bills quarterback is a QBR darling because he possesses two skills that are undervalued by traditional statistics: ball security and running ability. In other words, he's better than most people think.
Even more important, he's a heck of a lot better than DeShone Kizer. Taylor has a career Total QBR of 60.6, far above Kizer's 29.4 from last season. Taylor's expected QBR is factored into FPI.
Of course, as much stock as we put into our model, it isn't perfect. So the question is, what possible factors might FPI be missing, and do those make the Browns likely to be better or worse than it projects?
1. Baker Mayfield's impact
The first overall pick in this year's draft is the heir apparent to the ominous title of Browns Franchise Quarterback. FPI treats all freshman quarterbacks the same ... effectively as a replacement-level player. It will quickly adjust should Mayfield play and play well. But until then, it won't think much of him.
In general, this is a fair method -- plenty of players flame out, after all -- but it's not unreasonable to assume Mayfield has a better chance to become a good quarterback than the average rookie, given he was the first overall pick. Higher draft picks do tend to outperform lower draft picks (duh), even if it isn't to the degree the draft pick trading market believes. While Mayfield probably will not be on the field at the start of the season, FPI does consider backups and might be slightly underrating the former Oklahoma quarterback's potential contribution.
2. Taylor's value
While QBR attempts to quantify a quarterback's value, it is influenced by the skills of teammates. There might be a case that this has depressed Taylor's numbers, because it's quite clear that oft-injured Sammy Watkins (along with a handful of games with Kelvin Benjamin last season) was the most receiving help Taylor got in his time in Buffalo.
The wide receiver with the most receiving yards last season for the Bills was ... Deonte Thompson, with 430. And, Benjamin aside (who played in only five Bills games and four Taylor starts), their highest-graded receiver by Pro Football Focus last season was Andre Holmes, who ranked 74th in the league. The year prior, Watkins was the highest-graded receiver, at No. 47.
It's quite possible that this has held back Taylor's statistics, but this is a dangerous game to play without considering every quarterback's situation. There are certainly other quarterbacks who have played with worse offensive lines than Taylor, and almost everyone in the league has played with a worse running back (and LeSean McCoy is a big factor in the pass game, as well). As a result, I'm hesitant to assign much more credit to Taylor than he is already given by QBR.
3. New additions
While none of these additions are explicitly considered by FPI, they are implicitly considered by the team's season win total -- a big component in our preseason number. In other words, no further effect here; we can move along.
4. Is their win total correct?
If there's a flaw in relying on Vegas, it's that the betting markets can be influenced by bettors. Most of the time, the market is quite efficient. But there are some situations in which lines may not reflect reality. One theory I thought possible was that the stench of the Browns would drive bettors away from Cleveland and that their season win total would be lower than it ought to be as a result. But I was wrong!
Per Jay Kornegay of Westgate Las Vegas Superbook, the sharp bettors love the Browns because of how competitive they were last season (despite winning zero games), and as a result, if anything, the number and vig (or "juice") are a shade high. So there goes that theory.
So where does that leave us?
If FPI is missing anything on the Browns, it probably isn't missing by much. So while there's no doubt that Cleveland will find its way to the win column, our model isn't quite as bullish as ESPN's Mike Clay, who declared Cleveland a sleeper candidate to reach the playoffs. But it is perhaps just a tad more bullish than ESPN's Bill Barnwell, who wrote that he thought they were unlikely to be much more than a four-win team.
Ultimately, FPI gives Cleveland an 83 percent chance to finish with between four and eight wins, so set your expectations in that range.
Evan Kaplan and David Bearman contributed to this article.