The Value of Hurries and Hits, Part Two

In part one***, I discussed the merits of the Redskins blitz packages on factors such as hits, hurries, tipped passes, and sacks. Today, we address a much bigger picture, and try to find out whether the pursuit of increased pressure and sacks are actually a good thing.***I've gone back and corrected some errors in the first part of this study. I used an attempts baseline of 904 when calculating my non-blitz percentages. That was the number of total defensive plays the Redskins did not blitz...not the number of passing plays they did not blitz. That, of course, includes the 480 some running plays that the Redskins face, which I had to throw out. The new numbers have not changed my conclusions really, but still, I implore readers to go back and re-read the new numbers for better (and less erroneous) context.We can break down sacks into the 5 main reasons they occur, and in doing so, the Redskins 24 sacks from last season break down thusly:Sacks earned by the defender/blocks blown -- 11Coverage sacks -- 7Sacks due to an overall collapse of pocket (i.e. ball held too long/more guys coming than can be picked up) -- 3Rushers untouched, via either a well designed or timed blitz, or protection breakdown -- 2Gift sacks (i.e. where the QB runs outside his protection) --1I ordered these from most well deserved to least deserved. At the top, a blown block indicates a defender overpowering an offensive player and rewarding himself by crushing the quarterback. Matchups come into play of course, but this is mostly a talent thing. In the middle, you have the kind of sacks where the playcall has a lot to do with it. There's little distinction between a coverage sack and a collapse of pocket sack, they both signify an complete defensive victory on the play. An untouched rusher or protection breakdown is less about a completely earned sack, and more about a failure to adjust by the offense. Finally, a gift sack is any sack where the QB trips or otherwise runs himself into a sack, and it's the equivalent of an unearned run in baseball. You can't do anything to earn more of these, they just have to come to you.Andre Carter was the only Redskin to force more than a single blown block sack last season. 7 other players, 6 of whom play on the defensive line, forced exactly one blown block sack. The Redskins did have three guys record multiple sacks that were charted as coverage: Anthony Montgomery, Jason Taylor, and Demetric Evans all earned two sacks with some help from their backfield mates. Note: two of those players are playing with teams not named the Redskins this upcoming season. The other has been relegated to the bench for most snaps.I want to focus on the sacks marked as blown blocks by the Football Outsiders' game charters for a second. These sacks were predominantly made in the first seven games of the season, in fact 7 of 11 occurred on or before the Cleveland game with 9 of 11 coming prior to the Redskins bye week. Only 4 of these sacks came on 4 man rushes, and none came on 4 man rushes following the Cleveland game. This means that the Redskins were disproportionately more likely to win their one on one matches with more rushers and more blockers than with a minimalist rushing scheme.This scatterplot shows how as the Redskins increased their pressure scheme frequencies, the sack production increased as well:Blitz vs Sacks.bmpSo, if we can establish that periodic blitzing helps the Redskins get sacks, and that other types of blitzing help the Redskins get better matchups to win and get to the quarterback with, should we just become a blitzing team?Perhaps. The yards per blitz type split suggest we should4 rushers = 6.07 yards per pass5 rushers = 5.25 yards per pass6 rushers = 3.65 yards per pass7 rushers = 7.04 yards per pass8 rushers = 9.67 yards per passWhat about the percentage of first downs/touchdowns per passing play?4 rushers = 26% conversion rate**5 rushers = 32% conversion rate6 rushers = 25% conversion rate**7 rushers = 45% conversion rate8 rushers = 60% conversion rateIn part one, it really looked like the Redskins should have rushed 4 and 5 rushers all the time, because that's how they generated those pressure plays, but since we looked at the actual production of this team based on these splits, we've actually seen that pressure hardly matters at all. There's another type of big play though, that we should consider:InterceptionsThe Redskins intercepted a mere 13 passes. 3 of these came on plays where the QB was hit, 2 came on passes tipped at the line, and 1 came on a play where the quarterback was hurried, but not hit. The larger point here is that more than half of the Redskins INTs last year came when an opposing quarterback threw into coverage without being under duress.We can also break down the interceptions by blitz split:Interception rate of 2.5% when rushing 4Interception rate of 2.3% when rushing 5Interception rate of 1.6% when rushing 6One interception when rushing 7+ConclusionPressure is probably the best indicator of pass rush because it increases your sample size, but you don't seem to get any sort of consolation prize from leading the league in QB hurries. You need to get the sacks, and to do that, it appears that you sometimes have to bring 6 guys.As far as rushing a certain number of guys, there doesn't appear to be a dominant strategy from the 2008 Redskins data. When the Redskins rushed anywhere from 3-6 guys, there was highs and lows, but no discernible advantage. Now, I can safely recommend against the 7 or 8 man blitzes, that they provide no advantage and we need to not have 25 of those plays again this year, even 10 is likely too many.I originally wanted to look at some week by week trends, and perhaps we will at a later date, but I'm going to conclude on the above note that pressure is a great indicator of good things, but meaningless unless you get the sacks.