KFAN Rube Chat

The biggest transaction this off-season may have been the Jared Allen trade from Kansas City to Minnesota. Allen led the league in sacks last year with 15.5. He'll be going to a Viking defensive line that was impenetrable against the run but featured an average pass rush. The Vikings defense was so good against the run, opponents passed nearly 70% of the time compared to a league average of about 55%. Allen's ability as a pass rusher could have a very large impact. We know the Vikings pass rush will probably improve, but by how much? And how will the improvement affect the Vikes' bottom line?

Importance of Sacks

First, let's look at the general importance of sacks in terms of team wins. Normally when I model team wins, I combine sack stats with passing stats to calculate net passing yards. Along with turnover rates, running efficiency, and penalty rates, net passing efficiency is used in a regression to estimate total season wins for any given team. This method is useful for estimating the overall importance of the passing game compared to other aspects of football. But in this case, we want to specifically know the effect sacks have on total team wins.

By using basic yards per attempt (YPA) and sack rate instead of net passing efficiency, we can isolate the effect sacks have on winning. Using sack rate (sacks per drop back) instead of total sacks is useful for two reasons. First, it is far less susceptible to problems caused by the direction of causation. That is, defenses that face lots of passes would tend to have higher sack totals because they have more opportunities. Secondly, it allows us to compare teams and players that faced vastly different numbers of pass attempts (such as the Chiefs and Vikings).

Jared Allen's Potential Effect

Allen's 15.5 sacks were remarkable because he played in only 14 of 17 games last season. He also played on a 5-11 team that more often than not faced offenses with safe leads more interested in running out the clock than passing. His sack rate was an amazing was 3.8%. That's as high as the entire entire Bengals' defense. Had he played in all 17 games and faced the very high number of passes the Vikings defense faced, Allen would have over 26 sacks! But it's unlikely the Vikings' opponents will pass nearly as often in 2008, especially with Allen leading an improved pass rush. If they face the average number of drop-backs, Allen would have about 20 sacks, still a near-record-high amount.

In estimating the number of added sacks Jared Allen could bring to the Vikings, we can't just add 20 to their previous total of 36. The Vikings defense already had 9.5 sacks from the right DE position last year, 5 from Ray Edwards and 4.5 from Brian Robinson. A reasonable estimate would be that Allen could add a net of 10 additional sacks, taking them from an exactly average 36 sacks to a solidly above-average 46. This assumes he performs nearly as well as last year and he remains healthy--both fairly big assumptions.

Note that the additional sacks Allen may bring need not be his sacks. By chasing a QB out of the pocket or by occupying multiple pass blockers Allen could enable teammates to get the sack, but the credit would still belong to him.

Sack Rate's Contribution to Winning

Based on a regression of team win totals using sack rates, running and passing efficiencies, turnover rates, and penalty rates, sack rates are fairly important. For every standard deviation improvement in sack rate, a team can expect about an additional half win. That's comparable in importance to offensive running efficiency. The table below lists other important stats in comparison to sack rate and how they tend to impact team win totals.

.nobrtable br {display: none} table {border-collapse: collapse; border-width: 1px 1px 1px 1px; border-style: solid; } th {padding: 3px} td {text-align: center; padding: 3px;} #colorcol {background-color:#ffffe0}

Stated differently, for every percentage point increase in sack rate, a team can expect an extra 0.36 wins. Jared Allen's additional 10 sacks would improve the Vikings' sack rate from 5.0% to 6.4%. That additional 1.4% roughly translates into 0.50 additional wins for Minnesota.

Sacks correlate with other defensive stats, which complicates the analysis. Sack statistics are representative of a defense's overall pass rush. There are hurries and QB hits that are interrelated with other stats such as defensive passing efficiency, interceptions, and fumble rate. (I'll have to make one more assumption--that a pass rush causes the effect on passing efficiency, fumbles, and interceptions instead of the other way around. I realize this is not absolutely correct. A good secondary can cause the QB to hold the ball longer, for example. But it appears to me that the direction of causation primarily begins with the pass rush.)

The fact that sacks correlate with defensive passing efficiency, interceptions, and fumbles means that the benefit of adding a solid pass rusher is stronger than it first appears. Unfortunately, it also means the regression model is less than perfect. Ideally all of the independent variables in a regression are independent of each other as well as independent of the dependent variable. (This is why my normal model does not break out sacks and uses net pass efficiency instead.) However, no model is ever perfect, and I could only hope for a solid but rough 'order-of-magnitude' estimate anyway.

I estimated the effect of sack rate on defensive passing efficiency, interception rates, and fumble rates. I won't bore you any further with regression results, but the correlations were:

.nobrtable br {display: none} table {border-collapse: collapse; border-width: 1px 1px 1px 1px; border-style: solid; } th {padding: 3px} td {text-align: center; padding: 3px;} #colorcol {background-color:#ffffe0}

Defensive Stat Correlations with Sack Rate

Statistic	Correlation with Sack Rate
D Pass Efficiency	0.21
D Int Rate	0.17
Fumble Rate	0.43

In short, the cumulative effect of sacks on these other stats contributes to an additional 0.41 wins. In total the Vikings could expect about 0.9 additional wins from an extra 10 sacks. Realistically, this estimate should be taken as very rough, but it does provide an idea of the magnitude of impact a premier pass rusher could provide his team.