UCLA Football: Summarizing Chip Kelly’s 2018 offensive scheme

Pre-snap Motion vs. Play Faking

There was very little reliance on pre-snap motion or play faking; this was not a misdirection based offense. There were only 96 play fakes on 468 called pass plays (21% of the time).  There was a pre-snap motion on 154 plays out of 846 total (18% of the time).  These two concepts resulted in 6%-7% bumps in success rates, and play fakes boosted passing YPP by 2.3 YPP.

Watch any Scott Frost offense from UCF or Nebraska, and you’ll see much more motion and play fake frequency. This is one of my biggest head-scratchers about Chip’s 2018 scheme.  I would really love to hear Chip’s take on why they don’t use these two more.  You could suggest that play faking wouldn’t be as effective with UCLA being a pass-first offense,

but this twitter thread and article disagrees using data.

While the Blur featured a lot of WR and RB (2nd RB in game lined up wide) screen passes (using bunch and stack receiver formations) to keep defenses spread wide, this offense seems to have much more downfield route emphasis. There were probably less than five screens all season, and never to an RB. Bubble screens seemed like an obvious remedy to  DTR’s early-season struggles with progressions and an overpowering arm. Screens basically eliminate QB progressions and thrive on passes with zip. The only conclusion I can draw is that Chip doesn’t intend to use quick wide screens long term, and chose to not band-aid it in to get through this year.

RPOs

Run-Pass Options (RPOs) exist in the UCLA playbook, but don’t appear to be common or a staple. While I’m sure that I have no standing to judge, I think that some of the sports media are quick to incorrectly label a play fake with the appearance of a zone read mesh as an RPO. During the early nonconference games, I noted a handful of plays I thought were probable RPOs by UCLA.  Here’s an RPO that went run, where you can tell half of the line is run blocking while the other half is pass protecting.  Here’s an example of an RPO that went pass, where the line is run blocking, but DTR kept the ball for a quick throw.

As the season wore on, I stopped noting in the play log that any plays were potentially RPOs.  That could have been the fact that they stopped getting used, or it could just have easily been a lapse in my own play logging awareness of them.  If the former is true, this would be an example of Chip overloading the players running a new system early, where he backed off and decided to simplify things for the players once they got into conference play. RPOs are a topic where I feel especially ignorant and will be looking for tips on diagnosing and tracking them in 2019.