Are pass rush and interceptions really related?

Discussion in 'Texans Talk' started by gtexan02, Feb 5, 2009.

  1. gtexan02

    gtexan02 Working?

    Joined:
    Mar 24, 2005
    Messages:
    15,785
    Likes Received:
    1,243
    Location:
    Boston
    My main wish for the offseason is a ball-hawking safety. Frequently my posts are countered with the argument that average safeties can be made to look great with an a good pass rush.

    So to find out if interceptions really are the result of a good pash rush, I looked at the stats from last year:

    The top 5 teams last year for intereceptions were:
    1. Baltimore (11th in sacks)
    2. Cleveland (30th in sacks)
    3. Chicago (22nd in sacks)
    4. Greenbay (25th in sacks)
    5. Tampa Bay (20th in sacks)


    The bottom 5 teams last year were:
    28. Buffalo (28th in sacks)
    29. Seattle (10th in sacks)
    30. Dallas (1st in sacks)
    31. Denver (26th in sacks)
    32. Detroit (16th in sacks)

    The average sack ranking for a top 5 INT team was 21.6
    The average sack ranking for a bottom 5 INT team was 16.2



    Surprised? It makes logical sense that a team with a good pass rush would force a QB to make more poor decisions. But teams often plan accordingly, and as it turns out, teams with good pass rush actually get fewer interceptions. This is probably because teams playing against a team with a good pass rush use short drop backs, quick outs, and more runs to counteract the good pass rush.

    A lot of interception takeaways is more the result of a good secondary. It does not depend entirely on the pass rush, as people seem to believe.
     
  2. Wolf

    Wolf 100% Texan

    Joined:
    May 1, 2004
    Messages:
    37,900
    Likes Received:
    1,699
    Location:
    Kerrville
    to add to your list(nice job by the way) .. I wonder what those teams did in the form of knockdowns and hurries?

    I know sacks are nice and a good measuring stick,but sometimes it is more about disruption


    Mario would be a good example on our team, he might not have as many sacks as the year before (not much but 2), but if I recall he created a lot of havok
     
  3. gtexan02

    gtexan02 Working?

    Joined:
    Mar 24, 2005
    Messages:
    15,785
    Likes Received:
    1,243
    Location:
    Boston
    Yeah, I really wish those stats were easier to find. All I have is NFL.com and ESPN. Anyone have a better stat site?

    I think Dallas is the most interesting example. They have an incredible pass rush, but last season had a horrible secondary. And the result is that even a 60 sack season cannot move them past third from last.

    Baltimore, on the other hand, has a ball-hawking secondary. They did not have to get many sacks to get their INTs
     
  4. dalemurphy

    dalemurphy Hall of Fame

    Joined:
    May 3, 2004
    Messages:
    6,124
    Likes Received:
    662
    Location:
    Austin

    THe first thing I noticed is that all top 5 teams run fairly unique defenses that put players in position for INTS:

    Baltimore and Cleveland- both run a 3-4.. So, though Cleveland had a poor pass rush, I'd guess a number of INTs happened as a direct result of the zone blitzing they do.

    Chicago and Tampa Bay- both run the Cover 2... So, that system allows a lot of gambling by the CBs. Also, both teams were good against the run which probably means they got thrown on a lot.

    Green Bay- they have special talent at CB and also play a more agressive Man coverage defense than anyone in the NFL.


    Also, I don't think sack totals are the best indicator of a good pass rush. Clearly, though, scheme adoes have a significant impact on how many INTs a team comes away with. As a Texan fan, that's what is so exciting! Our scheme on defense has been SH*T!! It, the talent, and the aggressiveness will all be better next year. Yeah!

    Interesting that Dallas was so bad with the scheme and pass rush they have. First, that does indicate their horrific play at safety. Also, perhaps they shouldn't have been so anxious to let JReeves go. They certainly could have used him last year. When Newman missed time, he would have been the best CB on their team.
     
  5. TexansSeminole

    TexansSeminole Hall of Fame

    Joined:
    Apr 7, 2006
    Messages:
    8,794
    Likes Received:
    404
    Location:
    Orlando, FL
    It may be more complete to incorporate the QBs quarterback rating and completion percentage against those defenses. Sometimes QBs don't throw INTs in the face of pressure but throw the ball away.
     
  6. gtexan02

    gtexan02 Working?

    Joined:
    Mar 24, 2005
    Messages:
    15,785
    Likes Received:
    1,243
    Location:
    Boston
    That would give you a better idea of how much a good pass rush hurts a passing offense, but I was specifically interested in finding out if Interceptions were the result of a pass rush.

    In addition to having very low sack totals every year, the Texans never seem to get many takeaways through the air. I always wondered if the two were related, but it seems like even the #1 pass rush doesn't guarantee you a lot of INTs. It takes a ballhawk in the secondary to pull those down

    But just for fun:

    Top 5 Teams in QB rating allowed (low number)
    1. Baltimore (11th in sacks)
    2. Pittsburgh (2nd in sacks)
    3. Tennessee (5th in sacks)
    4. Green Bay (25th in sacks)
    5. Philadelphia (3rd in sacks)

    Worst 5 Teams in QB rating allowed (high number)
    28. Jacksonville (20th in sacks)
    29. Seattle (10th in sacks)
    30. Arizona (14th in sacks)
    31. Denver (26th in sacks)
    32. Detroit (16th in sacks)

    The average sack ranking for a top QB rating team was 9.2
    The average sack ranking for a bottom 5 QB rating team was 17.2

    So in this, have a good pass rush certainly helps your pass defense. But it does not result in more turnovers. This makes sense, too, I think

    One other interesting thing to note is that getting a lot of INTs does not necessarily equate to being a good pass defense team
     
    Last edited: Feb 5, 2009
    keyser likes this.
  7. TexansSeminole

    TexansSeminole Hall of Fame

    Joined:
    Apr 7, 2006
    Messages:
    8,794
    Likes Received:
    404
    Location:
    Orlando, FL
    Yea, I was just getting ahead of myself I guess, haha.

    I think your right. A pass rush may give those guys more opportunities but you still need a ballhawk to actually make the interception.
     
  8. IlliniJen

    IlliniJen Hall of Fame

    Joined:
    Sep 17, 2007
    Messages:
    9,083
    Likes Received:
    1,214
    Location:
    Midtown, Houston
    Do you think that analyzing one season worth of stats will give you enough information to make a reliable assumption off of?

    I would say you'd need at least 5 years of data to look at, in case this year was an anomoly in any way. It was indeed a strange year.
     


  9. gtexan02

    gtexan02 Working?

    Joined:
    Mar 24, 2005
    Messages:
    15,785
    Likes Received:
    1,243
    Location:
    Boston
    lol, are you trying to get me fired! :) 5 years :pirate:

    I don't have time to do the full on stats, but I will say that looking back just by glancing at the stats seems to indicate that you may be right.

    There are always a couple of odd balls (a team with a ton of sacks that didn't snag many INTs, or a team with tons of INTs that generated a weak pass rush), but the trend doesn't seem to support last years numbers if you glance back at the last few years.

    That being said, its hard to account for teams like Cincinnatti(2007) who was 6th in INTs, 32nd in sacks or Miami (2006) who was 3rd in sacks, 31st in INTs.

    To do a real analysis, we'd need to graph number of sacks by number of INTs over the last 5 years for every NFL team and see if there any sort of trend. Maybe I'll do that tonight
     
    Last edited: Feb 5, 2009
  10. Bulluck53

    Bulluck53 All Pro

    Joined:
    Aug 30, 2007
    Messages:
    804
    Likes Received:
    81
    We can have fun with this.

    Looking at Net Yards per Passing Attempt Allowed [(Passing Yards-Sack Yards)/(Passing Attempts-Times Sacked)] compared to league leaders in sacks, the top five went like this...

    NY/A
    Steelers.....Ravens
    Eagles.....Titans
    Cowboys


    Sacks
    Cowboys.....Steelers
    Eagles.....Vikings
    Titans


    and then Interceptions

    Ravens.....Browns
    Packers.....Bears
    Bucs
     
  11. Bulluck53

    Bulluck53 All Pro

    Joined:
    Aug 30, 2007
    Messages:
    804
    Likes Received:
    81
    And after looking at those statistics as far back as '04 there is no consistent correlation between the three. We need to find QB pressures in order to get the result we're looking for.
     
  12. gg no re

    gg no re iggnorent

    Joined:
    Jan 9, 2005
    Messages:
    4,901
    Likes Received:
    522
    Has the off-season gotten this boring already?
     
  13. Bulluck53

    Bulluck53 All Pro

    Joined:
    Aug 30, 2007
    Messages:
    804
    Likes Received:
    81
    Haha so true, but I'm still getting rid of the bad taste form last year (its in the past now)

    And besides, what better way to pass the time than talking in-depth statistics from years passed?
     
  14. b0ng

    b0ng Ooops

    Joined:
    Jul 21, 2007
    Messages:
    7,889
    Likes Received:
    613
    That's a hard stat to find since I don't think that the NFL keeps track of hurries, knockdowns, etc. I think Football Outsiders is the place that might have that.
     
  15. gtexan02

    gtexan02 Working?

    Joined:
    Mar 24, 2005
    Messages:
    15,785
    Likes Received:
    1,243
    Location:
    Boston
    Alright Jen! Ask and ye shall receive :)

    So I tabulated all the results for every team between 2004 and 2008 (5 years)
    I plotted number of sacks vs. number of interceptions, and then looked for a correlation. Here is the resulting data plot:

    [​IMG]

    Sacks are plotted on the x-axis, INTs on the y
    As you can probably tell from the graph, there is a slight positive trend, implying that more sacks = more interceptions.

    However, this trend was not statistically significant.
    (R = 0.14, N = 160, P > 0.05)

    What this means is that we cannot conclude that more sacks results in more interceptions. And as we go further back into the data, we would expect to corroborate this result.

    If you asked me what I believe, I personally think this says that more sacks does slightly increase your odds for more turnovers through the air. Our p value is pretty close to the significance level cut off, so with more numbers we may actually result in a real trend.

    The important piece of information is that there is not a clear, steep trend between the two. There are way too many points on the far left (low sack numbers) with huge INTs and vice versa for me to believe the two are strongly linked.

    While a good pass rush probably helps INTs, this says to me that a good secondary makes its opportunities regardless of who the front four (or three) are.

    This is surprising, because I think a lot of people just sort of assume that a strong pass rush results in tons of INT opportunities.... and this just isn't the case
     
    Last edited: Feb 5, 2009
    JayCee and TexansSeminole like this.
  16. TimeKiller

    TimeKiller Guest

    Joined:
    Nov 3, 2008
    Messages:
    7,648
    Likes Received:
    548
    Location:
    This giant tree, standing 10,000 ft. high but not
    If hurries and D-line deflections were available I'd much like to see a graph (;)) correlating hurries to sacks or deflections to ints.

    Logic would say a Dline with it's hands up causes tip drills and more ints. Certainly it counts as an INT, but it's not exactly like a corner jumping a route and may be inflating a normal secondary into a good one.

    I think a large portion of hurries to go with a large portion of sacks is the true measure of a pass rush instead of sacks alone. Lots of hurries means the Dline was getting after folks, not just "Well, we had to get back there eventually" type of sacks.

    I nominate this the most interesting thread of the offseason. I know it's early but this is just interesting. Good job gtex, gotta spread the rep though...
     
  17. infantrycak

    infantrycak Mod. Staff Member

    Joined:
    Apr 29, 2004
    Messages:
    54,920
    Likes Received:
    2,605
    Location:
    Austin, TX
    Stats, Inc. keeps them but only I believe for their reporter and team clientele. Gotta give the media something to be "special" with.
     
  18. brakos82

    brakos82 Hall of Fame

    Joined:
    Oct 28, 2006
    Messages:
    19,847
    Likes Received:
    978
    Location:
    LOHAR (Land of Hipsters and Rain)
    The scary thing is... I know what all of that means... :hide:
     
  19. Thorn

    Thorn Dirty Old Man

    Joined:
    Apr 14, 2007
    Messages:
    24,541
    Likes Received:
    1,892
    Location:
    Houston
    gtexan02 if I could give you rep for the work you did on that post, I would. :texflag:
     
  20. Ole Miss Texan

    Ole Miss Texan Hall of Fame

    Joined:
    May 8, 2006
    Messages:
    8,693
    Likes Received:
    685
    Dude... that is awesome. You should totally go get a Phd or something. ;)
    Someone give the man some Rep cuz I can't!

    That's really interesting stuff. And I can now see why/how that's the case.

    However, think about this! If a team has a lot of sacks, that means they tackle the QB in the backfield while he has the ball, thus giving his team a 0% chance of making an interception. The more sacks you get, that's the less throws a QB is making and less chance for a Defense to actually make an interception.

    Har har!!:pirate:

    That's gotta be why the Texans told Mario to go get some sacks. I think Richard Smith was saying to Weaver "Don't sack him, let him get rid of the ball, that way we MIGHT get an interception... better yet Anthony... drop into coverage."
     

Share This Page