UCLA Football: Summarizing Chip Kelly’s 2018 offensive scheme

PASADENA, CA - SEPTEMBER 15: Head coach Chip Kelly and Ethan Young watch warm ups before the game against the Fresno State Bulldogs at Rose Bowl on September 15, 2018 in Pasadena, California. (Photo by Harry How/Getty Images)
PASADENA, CA - SEPTEMBER 15: Head coach Chip Kelly and Ethan Young watch warm ups before the game against the Fresno State Bulldogs at Rose Bowl on September 15, 2018 in Pasadena, California. (Photo by Harry How/Getty Images) /
facebooktwitterreddit
Prev
4 of 9
Next
(Photo by Harry How/Getty Images)
(Photo by Harry How/Getty Images) /

RB alignment

This offense employs almost exclusively single back sets.  Under center plays always used a single deep set RB (with a single play exception against Oregon, where they lined up a TE as a lead blocker in an I formation-ish looking alignment).

The shotgun has three RB alignments relative to the QB, they can use “Pistol/Behind”, “Hybrid/Universal” (diagonally behind and next to the QB), or “Next”.

Hybrid was the bread and butter baseline. The pistol was the most obvious run tell and yielded terrible results (lowest SR/YPP). Pro (tightly correlated with under center) was for low upside reliable running and the previously discussed bad tendency breaking passing. Next was for passing 79% of the time and exceptionally good tendency breaking running at 7.3 YPP and 50% Success Rate. The distinction between Hybrid and Next alignment could be open to subjective interpretation. I have a theory that Next alignment was a lack of RB discipline more than a called formation distinction.

The theory goes that the RBs were creeping up when they knew they weren’t getting the ball and wanted to get into their pass protection or route running.  Irby and Olorunfunmi were the worst offenders here if my theory is true. Irby, for example, lined up Next 61 times in 156 snaps while Joshua Kelley only lined up Next 19 times out of 533 snaps.  They also occasionally used empty backfield (29 plays, 2.8 YPP and 31% successful) and a player on each side of the QB (4 plays, 10 YPP and 75% successful).

Runs from shotgun

The use of mesh hand-offs were >93% of the time with at least appearance of a read.  The 18 exceptions included WR sweeps, RB sweeps, the straddled triple option, and a conventional pitch option. Any play faking from shotgun looks like a zone read mesh.  Blur used lots of tosses and non-option/read hand-offs from the shotgun. Runs from under center have no post-snap option or read by design.

This table shows how the three main types of run plays complement each other and give play-calling options.  Non-Mesh (also correlated tightly with under center and Pro Ace RB) running is a low risk/low reward choice with the highest success rate, but lowest YPC.  Outside Zone Read (OZR) is the high risk/high reward call which was utilized the least (highest YPC, lowest success rate).  OZRs typically result in either a tackle at the line of scrimmage or a big positive play. The Inside Zone Read (IZR) is the high usage middle ground.  It’s important to note that I am using IZR and OZR as umbrella terms for option running inside or outside the tackles, but there are a lot of different ways they block them that aren’t even necessarily zone based.

Power vs. OZR

Oregon/Fishduck “Power” and “OZR” (Tutorials #2 and #3 above) are usually combined now by UCLA.  “Conventional” Blur OZRs (lineman all kick stepping together to zone block with no pulling) was used very rarely by UCLA. UCLA 2018 OZRs mostly relied on pulling lineman (so we technically should be calling it Outside Sweep Read, OSR or something similar but I’m trying to keep things simple, yeah I know, too late). OZRs were also often (but not always) combined with “Tackle-Over” (because they move a tack… actually never mind) unbalanced line formations that Justin Frye brought with him from his time at Boston College.  I covered Tackle-Over in depth after the USC game. OZR was only used 58 times all year (7% of plays) mostly against Cincinnati, USC, and Stanford.  It was not a UCLA staple.