Ian Allan
Wednesday night I posted the quickie note about fumbles — pointing out that there was no strong numbers showing that defenses would tend to recover lots of fumbles just because they had done so in the past. The very best defenses (in terms of fumbles) over the last 18 years have on average recovered only a half fumble more per season than the very worst defense.
But upon further examination, let me slightly alter those remarks. I spent another couple of hours toying with the numbers on Thursday night, and there is some connection between sacks and fumble recoveries. And sacks are less of a luck-driven stat – they can be somewhat accurately forecast.
I plugged in the info for every team since 1990 – 545 teams – into the Super Computer Deluxe (well … a Microsoft Excel spreadsheet) and found that the teams with the best pass rushes, on average, can count on recovering 1 more fumble per year and 2 more interceptions per year. The teams with the weak pass rushes tend to be the opposite – 1 fewer fumble recovery and 2 fewer interceptions than the NFL averages.
Of the 545 teams I looked at, 111 of them had 45-plus sacks (7.5 over the league average). They averaged 13.7 fumble recoveries per season and 18.9 interceptions per season. Of that group of 111, 57 percent finished above-average in fumble recoveries, and 66 percent finished above-average in interceptions.
At the other end of the scale, 117 teams finished with 30 or fewer sacks. This group averaged 11.8 fumble recoveries and 15.4 interceptions. Only 34 percent finished above-average in fumble recoveries, and only 33 percent finished above-average in interceptions.
So if you can latch onto a team that generates more sacks – a stronger pass rush – that at least tilts the odds in your favor in terms of fumbles and interceptions. There’s still plenty of luck involved, of course, but things are more likely to go your way.
The question then, of course, turns into whether sacks can be accurately forecast. I think that to a degree, they can. They’re certainly more chartable than fumble or interception numbers from the previous year.
In trying to find a defense that will finish with lots of sacks, one system is to select teams that had plenty of sacks the previous year. So let’s go back to our study group of 545 teams. Of that group, 107 defenses had 45-plus sacks. The following year, 62 percent of those teams finished above-average in sacks, and 40 percent finished with at least 4 more sacks than the NFL average. That’s not a terrific correlation; 38 percent of that group finished below average in sacks, and the group, on average, finished with 9 fewer sacks. But it tended to do better at least than the rest of the NFL.
Look at the 108 teams that finished with no more than 30 sacks. Only 34 percent of that group finished above-average in sacks the following year, and only 14 percent finished at 4 sacks above the NFL average. So basically, if you select one of those hapless defenses that finished with about 28 sacks last year, you have only a one in seven chance of that defense finishing up in the 40-plus sack range where it starts to become more likely that they’ll generate more fumbles and interceptions.
Pure numbers, of course, are only part of the analysis. When the Vikings traded for Jared Allen, that made it far more likely that they might be one of those teams finishing in the 42-plus sack range – and made it far more likely that Kansas City would finish below average in that statistical category.
Bottom line: When building your draft board, my suggestion is that you calculate which teams will finish with the best and worst pass rushes, using a combination of the sack totals from the last few years, plus factoring in the personnel and scheme changes for each team. Then, those numbers should be blended against the historical data that we’ve found. Estimate the majority of defenses to finish average in fumble recoveries (13) and interceptions (17). For the teams you think will finish with the strongest pass rushes, estimate 1 more fumble recovery and 2 more interceptions. For the teams with the weakest pass rushes, estimate 1 fumble and 2 interceptions below the league averages.
—Ian Allan
- Comments [0]
Readers' Comments
Add a Comment
Already a registered user? Please sign in to add comments.
To add comments, you must become a registered user of our site. To register, please click here.
