Situational Run vs. Pass

I love this stuff, and agree with the general idea that teams should be passing more on first down. Running isn't dead, but it is definitely more situational. I've seen some good analysis lately that shows that on first and 10 anywhere outside of the redzone an offense needs to gain 5 yards in order for them to increase their odds of scoring on that drive. In other words, 2 and 6 is a win, albeit a small one, for the defense. Given that it's safe to assume that most first down runs aren't going to pick up 5 yards, it's a safe assumption that most of the time you run on first down you're putting yourself behind schedule.

Now you might say "but if we throw on first down and get an incomplete then we're behind schedule and have to throw on second down" and to that I say "yes, you're right, but the same is true on 2nd and 8 as well".

Zim is just too "old school" unfortunately.  Remember last year?  For the first 5 games teams made our defense look like swiss cheese simply by passing on first down and it was like a mind blowing concept to Zimmer. 

Looks like we have a 48% success rate on first down passes for 9.1 yards while we have a 43% success rate on first down runs for 4.0 yards.  Right there the analytics of it tell you we should be passing more on first down.

It's really interesting because our first down run success rate is pretty bad, even when you compare it to a team like the Texans who run at a similar rate and aren't known for their running game at all, even they have a 55% success rate for 5.2 yards.  That probably has to do with the fear of Watson making them pay if they get too aggressive with the run on first down, right now the good teams clearly don't fear cousins.  

The other interesting thing is this is an average which means our first down run success rate probably goes astronomically low against a team like the Bears.  

Not to mention there's a run/pass double standard that I'm not comfortable with lol. Why is gaining 9.1 yards only a 48% success rate, while gaining 4.0 is a 43% success rate just because it's a different play? Last time I checked, a team has to gain 10 yards no matter what they call. The fact that we have to treat runs differently destroys the entire argument. Not that I think a team should never run. I do think it's important to keep the defense off balance because I imagine if teams passed 100% of the time, those passing averages would come down.

Yeah I was going to say, how much of that running success is due to Watson scrambles? The data probably should have filtered out QB scrambles since most of those occur after a play breaks down.

                  ATT-YDS   AVG
Carlos Hyde        51-199  3.90
Duke Johnson       19-147  7.74
DeShaun Watson      8-54   6.75

Looks like Watson doesn't account for too much of the success, but Duke Johnson certainly does. Makes you wonder if teams even look at data when you see something like this lol. Why is Hyde still getting the bulk of the carries on 1st down? Duke has 1st runs of 40, 32, and 24 yards. Hyde's longest 1st down run is 18. And 19 (37.3%) of those Hyde rushes have gone for 1 or fewer yards. Duke has had 8 (42.1%).

Meanwhile the Vikings have a similar problem.

                  ATT-YDS   AVG
Dalvin Cook        51-205  4.02
Alexander Mattison 16-109  6.81
Ameer Abdullah      3-16   5.33
Kirk Cousins        6-4    0.67
Mike Boone          1-2    2.00

Mattison has 4 of our top 5 longest runs on 1st down and has only 5 (31.3%) of 1 or fewer yards, while Cook has 13 (25.5%). Only 3 against the Bears however, 3 against the Packers, 3 Raiders, 2 Falcons, 2 Giants.

Backup running back stats can get inflated coming in fresh after the starter has worn the defense down. It is how Gerhart fooled the jags into giving him a big contract.

True, but shouldn't teams take advantage of this then? Why not spell Cook with Mattison a little more to try and get some more of those explosive plays?

Did anyone see any first down passing numbers from this week? I feel like we were more aggressive passing and passed to set up the run so to speak, but I would be curious if the numbers corroborate or not.

Wasn't quite sure where to put this: