Inside the Playbook: Michigan's Bounce Draw
Jim Harbaugh and Tim Drevno must have been watching their
Barry Sanders highlights when they decided to add the bounce draw to the playbook
it to the playbook. Or, at least, they were catching up on their Run and Shoot Mouse
Davis offense. It’s commonly stated that, at heart, Harbaugh runs a west-coast
offense (WCO). More or less that is true, but he has also – like most great
offensive minds – borrowed pieces from a variety of other offenses. In that
past we’ve seen forays into the spread attack with Colin Kaepernick and some
T-offense randomly against Maryland that one time. In this instance, he
borrowed from a Run and Shoot offense and integrated it relatively seamlessly into
how he utilization trips formations to the field within his nominal WCO (a
short, quick, spacing passing attack to the trips side and G Lead sweeps run
plays to the trips side) and downfield attack. In this article, we are going to look at how this
play functions and why it works within the Harbaugh framework.
Run and Shoot Origins
If you thought this was an excuse to watch Barry Sanders
highlights… yup.
(None of the links allowed me to embed)
As a run and shoot play, Mouse Davis called it an 80/81 Draw. Run and Shoot utilized a lot of 4 WR, 1 RB formations, incorporated pre-snap motion, and included a lot of route options post-snap for the WRs to choose from. It was initially invented by Middletown (OH) High School football coach Tiger Ellison, but received national acclaim when Mouse Davis refined and used it at Portland State. Later adapted with great success in the late 80's by the University of Houston, eventual UH Heisman Trophy winner Andre Ware was then drafted by the Lions, who were one of quite a few NFL teams to begin incorporating the offense in the early 90s.
The run and shoot at heart is a fairly balanced offense. However, because of the emphasis on the spread out passing game and the limited numbers of blockers in the box, defenses generally were not very creative in how they had to defend the run. But because of that, they often quickly bailed out of the box to drop back into their zone coverage as they feared the passing attack so much, and the passing attack could be quick because of the number of WRs and how spread out they were. This lead to defenders having their backs to the football or gaining a lot of depth, which set up nicely for a draw play. Force the defense to commit to the field, block the first level and seal the 2nd level inside the box, and then you have a lot of open space near the boundary sideline to run. That's the theory behind the "Bounce Draw".
Bounce Draw Rules
There are two different ways I have seen this run: the way Michigan runs it and the way the old run and shoot ran it. Because of the strength of Michigan's center in 2016 (Mason Cole) in space, they utilized him at the second level, and limited those that blocked LBs to the LT and the C, by design. So the rules had to change a little bit. Against most formations, this means the playside tackle (PST, always the LT in the case of Michigan) blocks the playside LB, the Center blocks the next LB to the backside, and the playside guard (PSG) blocks the first DL.
The second way allows for a bit more flexibility because it allows the OGs to release to get blocks. However, you have to be confident that the communication will hold up on twists and stunts and that the releasers will do a good enough job at the first level before releasing, which may be a reason Harbaugh limited who was releasing.
Note:
- F1 is the first frontside box defender counted from the ball outward; it could be a DT, DE, or LB.
- F2 is the second frontside box defender; it could be a DT, DE, or LB
- 0 is lined up on the nose of the football and is not counted
- B1 is the first and second backside defenders
Here are some diagrams.
Here we are against an Even 4-2 front. The LT uses the DE's momentum to wash him into the backfield before releasing to the playside LB and sealing him inside. The LG will seal the DT inside. This DT is the primary read for the RB. The center is releasing to the backside LB.
Now against a 3-3 Stack. Again, the DE attacks outside, although he starts from a head up position. Here is a key difference in how Harbaugh will run this. Because that DE is the 1st DL playside, the OG will follow the LT to the DE and maintain that block. The Center and RG will combo to the MIKE.
Looking at it a different way to block this 3-3 Stack defense, the MIKE could be seen as F1 and the WILL F2. That means the LG would release to the second level as soon as the DE committed outside. The center can clean up the 0 tech. In this case, post snap, the DE has made himself F3, while the LBs will have playside gap responsibilities. You could also make a rule so that the LG pinned down on the Nose Tackle and allowed the Center to release, but you run the risk of giving the Center a very difficult path to the MIKE, and he's unlikely to be able to make that block.
Now what if the DE pinches inside from an Even 4-2 front? In this case, because the LG has responsibility on the DT, the LT has to stick with his man and pin him inside. This does leave the WILL free , but the hope is, with the defense otherwise pinned inside, that the RB has the WILL in a lot of space and can make a man blitz. This is not a great defense for this play, but it can still work because it becomes a difficult play for the WILL, particularly if the LT can really wash the DE down. Note that the RB first has to read the DT, and his next read will be the DE sealed inside.
Going back to a 3-3 Stack defense, the RB's read is the DE. Here, the LT can wash the DE into the LG, who can overtake that block. The RB will read that as having to go inside the LG. The LT will then release to the WILL, who will likely have responsibility for the C gap (outside the tackle) and so will be easier to seal outside. This means this play won't "bounce" until after the LB level. Another thing to note on this play is the combo between the Center and RG. Another option here is to fold block with the RG and pin the NT to the backside with the Center. This ensures a clean gap for the RB to crease the LOS, and the draw aspect hopefully buys enough time for the fold blocker to get to the second level.
Note that in all examples above, the OL is not actually attacking forward in their blocks. They will pass set initially at 0-1 yards depth and then begin manipulating the defenders or working to the second level.
Michigan Examples
Now let's look at some examples. First I'll show the defenders as they will be called out (or blocking calls will be made based on their alignment), then the blocks, and then some cut ups, and finally the video.
Here, PSU is in a base 4-2 Nickel defense, so labeling is pretty simple.
Here is an interesting one. The numbering correctly indicates that the frontside LB is going to retreat, so he will become the LT block once he releases the DE. However, had that LB rushed, he would have effectively become another DL. This means that the OL would have had to adjust to this (assuming the backside LB also rushed), the LT would still shove the DE upfield and release. The LG would take DL1. But the Center would now have to attempt to seal the frontside LB to the backside of the play, and the RG and RT would have to pinch inside to seal off the backside LB and the backside DL. This means the backside DE is left unblocked, but most likely he won't take an angle that will allow him to make a play on the ball anyway.
Here, PSU is in a base 4-2 Nickel defense, so labeling is pretty simple.
Here's what the blocking scheme should look like (and it plays out essentially exactly like this).
At the snap you see the pocket starting to form, just as the receivers release and the RB works laterally as if he is going to assist the LT with a chip.
A half beat later, you see the OL start to engage in the blocks and manipulate where defenders are going. You also see the LT begin to push the DE down and release, and the C release. Notice the depth of the LBs. One has stayed flat footed, the other moved up about a yard but is relatively flat footed as he starts to understand what is going on in front of his eyes.
Neither LB is attacking the LOS yet, with the backside LB just now redirecting from retreating to coming down hill (he had to gain depth as he had inside coverage responsibility on the TE working the seam). The frontside LB has barely moved up and is now flat footed looking at two OL releasing down field. The LT is taking a wide angle and has plenty of time to get outside to seal the LB inside. Meanwhile, the RB gets small so as to duck behind the LG and find the hole that is provided by the LT shoving the DE up field.
And now perfect angles to make the blocks and spring a big gain.
A few more examples. Rutgers in the same general alignment leads to the same general outcome.
How It Works Within
Michigan Offense Framework
As I previously explained, there are two main ways that this integrates specifically well for Michigan's offense. The first is the sweep play, which Michigan often utilizes by running from a 3x1 formation. Below is an instance against FSU in the bowl game.
The second is that Michigan uses 3x1 formations quite commonly in the pass game. In fact, this may be the more important point. They will use it for deeper concepts (Post-Scissor (scissor concept with two post routes) in back-to-back clips, inside post-wheel or vertical switch) which help sell the down field routes
They also used it in the short and intermediate passing attack (mesh-spot concept (read more here), again, Snag-7 (two man concept with a bench route) with backside speed out, NCAA Route (learn more here))
So I want to give a feel for what all of this means. It means within the common framework of Michigan's offense, they utilize the same or similar formations to attack downfield, to attack the short and intermediate areas of a defense, to attack the wide side of the field with the run, and to isolate defensive backs to the boundaries. They utilize this same look to do a lot of things, including things the draw initially looks like immediately post snap. The formation, while an alert to the defense to maybe think "draw", does not mean one thing.
Now lets step back a little bit to get a bigger grasp of this play. 3x1 formations often simplify defensive schemes. From a high level, there are really only four ways to defend trips from a coverage standpoint:
- Rotate your zone to numbers
- Play straight man with field DBs (meaning no over the top help)
- Rotate/bracket with the boundary safety
- Bracket with a LB
Option 1 leaves you weak against many passing attacks out of a trips, because trips can inherently flood the short or deep area of the field on one side. Over rotating often effectively leaves the boundary WR in man coverage without help or takes the defenses eyes away from run fits or the backfield as they look for receivers coming from the far side of the field. Because of the quick threat, often times LBs are forced to bail as soon as they read pass sets from the OL, which they do on a draw play.
Option 2 leaves you in man coverage against a formation that allows for a lot of rubs. While you can run combo coverages to mitigate that, it still requires you to defend without help in a lot of space. The man coverage also turns the DB's backs from the ball. But at least you will keep two LBs in the Box and often times a deep center or a deep half safety to mitigate the damage.
Option 3 takes your safeties eyes away from the backfield. It does allow him to play over the top of #1 to the boundary if no works vertical, but as we've seen, Michigan often runs vertical concepts out of this look. In fact, the pass example linked against Illinois above shows the boundary safety with his eyes on the far side of the field in Illinois's two high defense, and the run goes a long ways because of it.
And Option 4 completely takes a LB out of the box as soon as he sees a pass set. Again, it's a draw play, so that will be the read. That gives a numbers advantage to the offense, and possibly an additional number as the DE is pushed past the play.
What all this means isn't that it's an impossible play to defend. In fact, the if play side DE is disciplined in his keys and really presses the OT instead of flying up field, he can squeeze the play and cut it down in the backfield or prevent the OT release, making the play itself much more difficult; but that also slows his pass rush. But playing smart defense, even with all the disadvantages listed above, can mitigate this play, but that's easier said than done because it makes it more difficult to do many of the other things that are asked of you within the frame of the defense.
Great article. You explain concepts really well, I also like the section on the advantages of 3x1 formations.
ReplyDeleteI have a request. Can you do an article on playcalling? How and why plays are being called, like down and distance vs complementary plays.
Maybe the difference in playcalling for a "prostyle" team vs Air Raid, or Flexbone Option Team.
Really enjoy your blog.
Thanks! Really good ideas for some posts down the road. May have to limit it a bit but I like the concept
DeleteWhat's the outlook for this play's role in 2017?
ReplyDeleteMason Cole moves from center to left tackle, presumably they like what he can offer on something like this. What about new center Kugler?
Cole should be good at LT for this play. He's always moved well and blocked the second level well, so no concerns with that. Kugler is a bit more of an unknown. Could see them open to releasing a guard in some instances, though not sure they really want to change the rules too much. I think Kugler should be fine, usually blocking the backside LB isn't terribly difficult. But OSU snuffed this play out when Cole couldn't release from his center position to get to the MIKE last year, so it isn't trivial.
Delete