One of the better statistical predictors of success is turnover differential. While it would be tough to argue there’s a one-to-one correlation, the odds of success are low with a negative turnover differential. The 2018 Buffalo Bills had a -5 for their differential, good for 23rd in the NFL (so pretty bad). What might surprise some fans is that it was only -5. Why would -5 be shockingly low to some readers? You’ll see soon enough.
A year ago we made some best guesses on what would happen to the Buffalo Bills takeaways in 2018. A positive turnover differential in the drought-breaking season had many thinking that the 2017 season had a remarkably high number of takeaways. This naturally led to an exploration of this idea and a best guess on how they’d fare in 2018. I had this to say:
The stability of major players involved...indicates there’s no guarantee of a drop off. More importantly, the Bills weren’t really all that great at takeaways last year... There’s nothing that suggests we should anticipate such a drastic change. It’s a good guess that the Bills will remain in striking distance of this year’s  results with plenty of room to actually improve in this important metric.
I’m really glad I added that sentence about the possibility of improvement, because that’s exactly what they did. To translate the graph, the blue line is the actual number of takeaways by the Buffalo Bills each year. The red line is the league average, and the black line is the trend line for the league. The trend line is extended to 2019 so we get a handle on what the league might expect.
The Bills were above average in 2017 and still managed to increase the number of takeaways. This is despite a league-wide trend going back two decades of lower and lower turnovers. As an aside, I’ve heard the idea that the 2018 takeaways were more boom or bust than 2017, which was discussed in my 2018 season wrap up you can find here.
Now if you’re taking the time to make such fancy charts and data dumps on takeaways, you may as well go ahead and do one on giveaways too. With Tyrod Taylor, the Bills were incredibly good at protecting the ball, which made guessing what 2018 would bring a problematic endeavor. I went on record with:
It takes a special kind of bad in today’s NFL to break the 30 barrier.
Preparing fans for the worst-case scenario then, growing pains might lead to a 25-27 sort of season.
I did try to warn everyone there would be a major increase, but it would appear even I underestimated the size of the jump. And here’s the reason fans might be shocked by a “mere” -5 differential. That differential was despite an abysmal 32 giveaways. Only one team was worse (Tampa Bay Buccaneers) and only one team tied Buffalo (San Francisco 49ers). Respectively, those teams had -18 and -25 differentials. Yikes!
For takeaways, there’s some evidence to suggest that coaching and player stability decrease the volatility of this metric. For Buffalo that means nothing but positives. Leslie Frazier returns as defensive coordinator and Sean McDermott hasn’t gone anywhere. Kyle Williams is the only significant change to the starting defensive roster. The mid-20s is a very reasonable guess. With a projected league rate of about 21 takeaways per team, that would make for a third year over the average.
Could they improve? It’s possible but I personally wouldn’t count on a major swing. The Chicago Bears ended the regular season with 36 takeaways to lead the league. To do that, it would likely mean that Ed Oliver has a phenomenal rookie season as he’s the biggest change on that side of the ball. You can’t ever rule out good ol’ fashioned luck either. That could play into a large increase or drop-off. But until further notice, this is a team that takes advantage of opportunities.
This year’s projection has the benefit of a Josh Allen baseline to work with. In 2018, Allen’s baseline was one turnover (interception or fumble lost) for every 51 plays. Projected over an entire season around their 2017 mark of 1,059 offensive snaps, that comes out to 21 turnovers that Allen would be responsible for. Other players are likely to contribute around 5-7 turnovers, putting the Bills in the mid, or even high twenties.
The largest potential difference in positively impacting the turnover differential lies on the arm of the quarterback. It’s a good thing we’ve already looked at the average amount of improvement quarterbacks in Josh Allen’s shoes have experienced in the last decade. If you read that article, you might recall that most quarterbacks showed improvement in the key statistics that were looked at.
For interceptions, modest improvement is likely to occur. This would lower the giveaways by one or two over the course of a season. Based on the decade of data, it’s not unthinkable that Allen can cut his interception rate in half. That would lead to a drastic change in giveaways, possibly all the way into the teens.
There’s no perfect model to extrapolate quarterback growth from rookie to sophomore year. If the best guesses are an indication, Buffalo should be near a 0 when it comes to turnover differential in 2019. A stable and opportunistic defense will do it’s part, while a steadily improving offense starts heading in the right direction.