Friday, October 30, 2009

The Spread, Arithmetic, and why OSU may have gone back to Multiple Formations

Has Ohio State abandoned an all 'spread' attack to go back to mixing in the I and 1-Back formations ? And if so, why?

Starting with the Illinois game, Ohio State effectively became an all-shotgun spread to run team.  According to Jeff Amey of the O-Zone, Ohio State lined up in the shotgun 87% of the time for the four games from Illinois to Purdue. (They had only done so 41% of the time in the first three games).  Then, against Minnesota, according to Amey, the Buckeyes went back to mixing looks, with 55% of their plays from Shotgun, and 45% from either the I formation or single back (Ace) formation.   

Perhaps, not coincidentally, the Ohio State offense's output improved against Minnesota.  OSU this year has  essentially functioned four games as an all spread team, and four as a multiple formation one.  Here are how the numbers compare:


All "Spread"  Production
"Multiple Formation" Production
Yards Per Game
292.8
387
No. of Plays
57.8
67.75
Yards Per Play
5.1
6.025



Moreover, Coach Tressel announced this week that he liked the formational balance from the Minnesota game, indicating that we will likely see multiple under-center formations going forward.  If this is true, the question becomes what happened to the short lived experiment of becoming an all spread team?

In a word, it's arithmetic.  Jeff Amey gives a valid answer that the spread basically made OSU too much of a Terrelle Pryor-centric offense and he wasn't able to sustain it.  But the deeper answer to this question is that the defense was able to gain a numbers advantage on OSU's spread that took away the Buckeyes' inside running game and forced Pryor to beat them both as the primary inside runner and throwing the ball down the field.  This the Buckeyes were unable to do.

To take a step back, the Defense always has a numbers advantage on the offense, because even if the offense blocks 10 men, the offensive ball carrier's counterpart is left free.  The 'spread's' true genius was that by adding the QB as a run threat, it occupied another defender, helping the offense with numbers.  As Rich Rodriguez describes here, when the QB can hold the backside defender that is one less man for the offense to block, creating better numbers. 

When OSU first unveiled an all-spread offense, they were able to take advantage of exactly this. fact  Illinois lined up on each receiver in a 2-deep shell and Pryor was able to hold the outside defenders, creating running lanes.  So, as you'll see in this clip, Illinois plays head up on every receiver with 2-deep safeties, giving OSU a 7 vs. 6 advantage in the box in addition to Pryor able to hold the front side nickel back with the run threat.



As teams have become more accustomed to defending the spread, however, they have developed a myriad of ways to slow teams down.  One is to cheat their outside linebacker/safety types into what Rodriguez calls the "grey area."  In essence, the alley players, rather than covering the inside receiver, cheat down to get involved in the run game.  Starting with the Wisconsin game, Ohio State started seeing this on steroids.  Here are two different clips.  The first is how Wisconsin defensed Ohio State's "trey strong" formation from which they like to run the speed option.  The second is how the defended the more balanced H-Back formation (basically they shifted their LBers over one gap to the strong side in the former).





As you'll see in both, it initially appears that Wisconsin, like Illinois only has six in the box.  But that is really an illusion, as their front and backside alley players are both playing in the 'grey area,' so it is really the equivalent to 8 in the box.  Wisconsin is effectively playing without a deep safety.

So now, rather than give Ohio State a numbers advantage, the spread has actually compounded the problem:  OSU is now playing 8 vs. 6 in the box (8 vs. 7 if you count the QB cancelling out a blocker), as opposed to the I where they would be 8 vs. 7, with a better chance of getting the playside defenders blocked, as described below.

Ohio State thus faced a numbers disadvantage starting with Wisconsin that limited their running game.  In fact, as hinted at earlier, the spread makes a numbers disadvantage in the box even more problematic then running out of the I does, for precisely the reason Homer Smith identifies,and its again because of numbers.  If you have no deep safety the defense will have one additional man to the front (playside) than the offense will have.  The only way to block him is to have a lead FB who can block through the hole.   Out of the spread with one running back there is no way to get to the unblocked playside defender.  As Smith says,

When there is no safety, you have the problem always presented by the man-for-man goalline-type defense: There is a frontal defender that you cannot release on, occupy with a fake, or block. In other words there is a free tackler for the ball-carrier... An I-formation will face two centerline, defenders or none. The center can have a chance to get a playside block, and a bootleg fake has a chance to hold pursuit from the backside. A handoff run has a chance.
So in sum, the spread is even less advantageous to run against stacked fronts then an I formation, and you can end up with situations like this where Purdue has four playside defenders and there are just not enough  people to block them.



So a spread team must find a way to address players cheating  play the grey area.  For years it was namely the bubble screen.  But teams have gotten adapt at defending the bubble screen this year, specifically by having their deep safety fly up at the first sign of a bubble screen, as Michigan, among others, are finding out.  (Not to dig too far into the weeds, but one effective counter I have seen to this this year is having the inside receiver fake the bubble screen and the outside receiver run a slant behind the safety, as we have seen both Ohio State, and teams facing Ohio State, do).

So if this play is taken away a spread team basically has two choices when teams want to stack the box a) use your QB as your runner so your halfback can block for him and essentially run I type plays, and gain the benefits Smith describes for having two backs or b) force them out of it by successfully moving the ball through the air.  Against Purdue, OSU tried both but neither were successful.

In the clip, OSU is essentially running their 'Dave' play with Pryor as the runner.  But the problem, as you see, is that Pryor is not a consistent inside runner.  He misses the hole and tries to string this outside...



Nor was Ohio State consistent enough in the pass game against either Wisconsin or Purdue to consistently move the football.

So in my estimation, this stark arithimetic drove Ohio State to abandon being an entirely 'spread' team to being in multiple looks.  What does mixing in the I formation and other pro-sets get you?  Homer Smith and Chris Brown have provided a series of benefits and drawbacks to each formation, but here are I think a few beneficial factors for this OSU team:

  •  As seen from these clips, teams cannot get many more guys in the box against OSU then they already were against the spread.  They have to still at least defend the outside receivers.  So at being in the I gives OSU a chance to get more hats on hats in the box
  • Relatedly, as discussed earlier, having a lead FB provides the opportunity to get more blockers playside then is available from the spread.
  • Play action from under center can create better dropback and bootleg faking off of a vertical run game.  To the extent that OSU wants to get Pryor on the edge, the bootleg and rollouts have been a great way to do so, as can be seen with the fake Dave play rollout.  Further, by using this type of play action, it gives more time for deep passing routes to develop, a problem OSU was having from the spread: 

  • It can enhance pass protection, particularly off of sprint draw action, by getting more blockers playside.
  • Preparation time.  Teams can no longer spend their entire week working on defending OSU's spread.   They instead have to split time between working against a spread and the I formation This makes them less able to unveil entire new defenses such as Wisconsin did against OSU.
  • At the same time, Ohio State runs largely the same plays from both formations, so for the line and WRs they do not need to practice double-it only changes the backs footwork.
  • Along these lines, by going back and forth, you can catch defenses in mismatches, such as having their nickel back in the game against the I.

  • Finally, it takes pressure off of Pryor.  By taking away the inside running game, teams were making the offense entirely Pryor dependent--by Jeff Amey's count, 90% of the plays against Purdue were put in his hands.  This hopefully provides Pryor a steadier run game and, perhaps more importantly, play action and bootleg and rollout game to put him in better opportunities to succeed.
Ohio State will not abandon the 'spread' game, nor should they.  It will remain a primary staple of their offense.  And they are not the only team to have problems out of the spread this year.  But mixing in the I and other pro style formations does make OSU more versatile and allow them to better counter defenses who want to stack the box against the spread.  That is why I expect to see Ohio State continue to mix and match between the two different styles for the remainder of the year. 








Wednesday, October 28, 2009

Tight Ends in the OSU Passing Game Part II

Since TE use (or lack thereof) forever remains a hot topic with OSU football I wanted to make one more point. But first one more note to show just how much the TE is involved in pass blocking. This is a pretty interesting (or revolting depending on your point of view) stat:

TIGHT END USE IN THE PASSING GAME vs. MINN 2009 (through 3 Quarters)
  • NUMBER of PASS PLAYS:  29
  • NUMBER of PASS PLAYS WITH AT LEAST ONE TE in the GAME:  21
  • NUMBER OF PASS PLAYS WHERE THE TE RELEASED INTO A ROUTE:  3

The TE releasing a total of 3 times (2 by Ballard, 1 by Stoneburner) is a pretty amazing stat.  The rest they were involved in slide protection as I discussed before.

In light of that, this second point may seem academic but I still think worthwhile, revolving the TEs role if they do in fact release on pass patterns. For the vast majority of teams, the TE is not the primary route runner on the vast majority of plays--unless you are in a situation where you have an Antonio Gates and he is your best threat. OSU is therefore by no means in the minority in not featuring the TE when they do release.

To take one example, here is a link to Norm Chow's dropback passing game. As many know, Chow is really one of the designers of the modern passing game.

http://bruceeien.com/offense/chow/CHOW.HTM

As you'll see, most patterns do not make the TE as the primary receiver--those that do generally being short, over the middle option and choice patterns. And, as several have noted, OSU has largely thrown the ball outside the seams this year, likely to try and reduce interceptions. As such, there are not many times when the TE is the primary threat. That the TE is not generally featured is even more true when you look at Chow's design for those plays for 3 WRs or more (this being a little dated when 2 x 2 formations were far more prevalent). The TE becomes even less of a primary threat as the slot receiver often takes his place running the primary inside receiver route.

"Smash" is an example of this. This is a basic route that every college team runs against cover-2 and its variants. Basically, the idea is to 'hi-lo' the cover 2 corner and make him commit to either the short or deep route, giving you a 2 vs 1 advantage. (For more on smash, check out here: http://smartfootball.blogspot.com/2008/01/divide-route-in-multiple-smash-concept.html). Take a look at Chow's various diagrams of the play.

As you can see, the play can be designed with the TE running the corner route. But, as you get to 3 and 4 WRs, it often becomes the inside receiver who runs it, and who generally offers a more vertical threat (or to put it another way, who would you rather have going deep, Devier Posey or Jake Ballard?).

And, likewise, OSU would generally rather go multiple WRs and run these route combinations with their 2 WRs to a side, rather than the TE. Here is OSU running a similar type of pattern against Minnesota (this is not quite a smash, as the receivers do a switch and it is the No. 2 receiver doing the short route, while the No. 1 receiver runs a post corner, but the same hi-lo principle applies).



The point of all this is to say that it is not unique for OSU to not feature the TE as the primary target on the majority of their pass plays. Instead, the TE is usually found by Quarterbacks who are reading through their progressions and check off to them. In this way, the TE's role in the passing game is somewhat equivalent to a counter play in the running game--if the secondary is going to cheat to take away my primary (outside targets) then I will go to the TE. For example, check out this discussion of USC's TE (McCoy's) day for ND:

http://www.trojanfootballanalysis.com/wp/wordpress/?p=1479

As they state, every time Barkley hit McCoy ND was on some type of blitz and Barkley found the mismatch with McCoy covered by a LBer.

So a lot of times, how involved the TE is in the passing game derives not from play design but rather the QB successfully reading his progressions. This year, however, Terrelle Pryor does not generally seem to be going through his entire progression. Instead, he generally locks onto one or two receivers (Dane or Posey) and if they are not there either forces the ball or runs. So the coaches may think the trade-off between sending him into patterns and keeping him in to block when he may not get the ball regardless favors the latter.

The other place the TE is a factor is play action and short yardage, as they have the advantage of being able to threaten to block and then release (again see McCoy versus ND). But even there, you have to find the TE and hit him (remember the goalline play action pass to Ballard against Purdue?). But more generally it is again just a choice by the OSU coaches to focus on throwing to the WRs outside on play action as well, as the clip above shows.

In sum, it comes down to primarily scheme, but also execution. The coaching staff has decided to focus on protection first, and worry about getting 4 or 5 wide receivers into pass patterns second. I offer a variety of reasons why they may try to do so in a previous post. Personally, I as a lot of people, would like to see the TE a bit more involved in the passing game, for two primary reasons:

1) because they are the best placed to immediately threaten safeties down the field and thus keep them from cheating up in run formations, and

2) because, as mentioned above, they have the threat of both blocking and releasing for a pass on play action, so they can occupy two defenders (the one the TE would be blocking and the one responsible for covering the TE).

But you have to walk before you can run and the coaches have clearly decided to first focus on a) pass protection and b) giving Pryor less to worry about so that he can make a couple of reads and then run the football if nothing is there. Eveything breaks down without decent pass protection (see Purdue) so it is hard to fault them for that choice.  My guess is that this will likely be the primary focus throughout this year. Though we have not seen much of Jake Stoneburner in the passing game thus far, his playing time is increasing and one hopes that, as a former Wide Receiver, he can offer more of a threat from the TE position heading forward.


UPDATE:

As I mentioned before, finding the TE often falls on the QB to make the right reads.  Here is a play action bootleg against Wisconsin where Stoneburner came wide open across the middle, but Pryor never got his eyes downfield as he tired to elude the DE-pause at the very end and you can see how open he was...

Tuesday, October 27, 2009

Tight Ends in the OSU Passing Game

 I initially had this as part of my in-game analysis but decided to separate it out as it has a more universal application.

 One quick word about the TEs and their use in the passing game. We are a 'zone blocking' or 'slide protection' passing team. That means that, rather than having a linemen responsible for a linemen, we slide left or right, and each linemen is responsible for their gap. The benefit of this is that it doesn't give a DLinemen a 2-way go on the center. The drawback is that you have to have your TE and/or RB in at least a check release, because someone needs to be responsible for the edge away from the slide. Occasionally, OSU will go slide protection on one side, and then man on the other to get a 5 receiver route. But most of the time, OSU likes to slide everyone and have the TE (and RB) responsible for the backside. Below is an example:



You'll see how the entire line slides right, and then Ballard and Hall are responsible for the left edge. Now, without getting into too much of a debate about the positives and negatives of doing so (for a fuller discussion that is well worth reading check out here: http://smartfootball.blogspot.com/2008/06/pass-protection-super-bowl-tom-brady.html), suffice it is to say that there are two different philosophies--one is that is that the passing game falls apart with pressure (see the Patriots in the Super Bowl) and it is better to first protect the passer. The other is to put 5 guys in the route, let the QB know he's going to get hit, and throw quickly. That is why most spread teams throw 'very short' quick routes. Note, that this doesn't have anything do with sophistication vs. lack thereof. Bill Walsh believed in the former, which is why they generally had check release patterns.

But the result for OSU is that the TE stays in to pass protect a good amount of the time. Generally, when the TE is in he on passig plays he pass protects. It's when we go 4 and 5 wide that we have 4 and 5 men pass patterns. I have advocated getting the TE more involved, espeically to threaten the deep seams quickly. But I think in OSU's defense they're working with the following factors:

1)  A young, inexperienced QB that is a much better passer when not pressured.
2)  A young, inexperienced QB that does not read progressions very well but instead locks on to particular receivers.
3)  A young, inexperienced OLine
4)  A starting TE, Ballard, who is not a big passing game threat.

As such, I think one can make a strong argument that this is the correct pass protection method for this team.  That is why, even with the 'fake Dave' playaction rollout play, you see us running it towards the 'twins' side so that we can throw the ball outside the hashes to our two best receivers, rather than using the TE coming across the middle on bootlegs.

All this being said, we see OSU use 'combo' protections to get 5 receivers into the route, so we know it can be done.  I would expect and hope that, as Pryor and Stoneburner get more experience, we can use the TE more vertically in the middle of the field and crossing on playaction...

Monday, October 26, 2009

OSU v Minny Game Analysis

So I'm trying this format so I can put some videos with my points and hopefully make things more clear-let me know what you think! BTW-like a kid with a new toy I got hooked on making video clips, so definitely a lot to go around!


Obviously a much better game in so many ways.


1. Have to start with TP. Many have obviously made theses points, but he just flat out made some huge strides forward in this game. When we were hoping that he would build on what he did in the 4th quarter against Purdue, it looks like he did so. He was so much more decisive. He stayed in the pocket and stepped into his throws, even with rushers on him. We did not see that at all against Purdue until the Fourth Quarter. Here, he did that nearly the entire game. I only saw maybe one throw off his back foot. He also for the most part did a nice job when he was rolling getting his shoulders turned to make the throw. I again only saw once where he did not get his shoulders turned. When he decided to run, he was decisive and tucked and ran. And he got upfield. That was almost as impressive. You can tell he made a conscious decision (and was coached!) to get upfield and not automatically run out of bounds. Some examples below.

Here is a great example of Pryor staying in the pocket, stepping into the throw, and delivering a strike to Posey on a curl route with an incoming rusher:



Pryor's reads were also better. Here is a good example. This is the first possession of the game and a good example of a coverage beater. We run a ‘spacing route’ to the right side, which is very good against cover 3, man, and double slants to the left side, very good against cover 2, man. Pryor sees middle of the field open (MOFO), meaning 2 deep, and goes to the double slant side. Hits the outside slant because he’s reading the no. 2 underneath defender who jumps the inside slant. Steps into his throw and puts it on target. Nice play design and execution.



Finally, his TD throw to Posey on the rollout was a great job by him. First he makes a great read. They jump the out route that we kept throwing to Dane on this play, leaving Posey open behind. Then he gets his shoulders square, sets his feet and throws a great deep ball.





One can claim that well, yea but it was Minnesota, but that I think ignores the larger picture. Pryor stepping into his throws, stepping into the pocket, being decisive has nothing to do with the opponent and everything to do with him and execution. If he continues to do this this offense will be a lot more successful. Granted, he still locks onto Posey way too much instead of going through his progressions, but how he played on Sat is a big step forward.


2. I really liked our formationing and play calling this game, in particular the mixing of the spread and I. In particular I like three things we did out of the I-

a) OSU did a nice job of breaking tendencies out of pro style formations. We passed 7 or 8 times out of either the I or the two TE bunch set we run, in 4 different ways-we dropped back, and we also ran 3 different play action passes-a sprint draw, a naked bootleg off of zone action, and then the Dave rollout pass I discuss below.

b) We went almost exclusively to a 'twins' formation in 'I'. Minnesota's reaction to this was to spread their Will LBer out to our inside receiver and bring their CB up over the TE. So essentially with our formation use we traded out a LBer for a CB in the box, which is a nice trade (as you'll see in the video below). I don't know if this was something they saw on film, but it will be interesting to see how teams defend us out of this set in the future.

c) As some may recall I was arguing in favor of more bootlegging and rollouts last week, and got my wish answered, LOL. But I will say we were hugely successful doing so. Just focusing on the 'fake Dave' play I discuss below, we ran this play four times. Pryor was 3-3 on those pass plays for about 60 yds and a TD, and also had one run for 12 yards. Minn bit hard on every play fake, and then, when they realized Pryor was rolling out, became more concerned with containing TP then their pass responsibilities, leaving huge holes in the intermediate zones that Dane was able to exploit.

Part of the success of the rollouts was the play design. The best bootleg and rollouts are those that look exactly like their run counterpart. Here, this is exactly what OSU does. As you can see below, we fake the Dave play, with the backside guard pulling and the line blocking down. Pryor then continues rolling out with the guard leading. We can then run any number of route combos off of this, though the favorite seems to be a smash variant with an out and post corner combo.




Then, later in the game after we had hit Dane several times underneath, you'll see the safety jumps the out route, Pryor makes the right read, and hits Posey deep.




OSU also did a nice job countering tendencies in the spread game, particularly the speed option. First we motioned the TB from the weak to the strong side to try and throw off the pre-snap reads for the speed option. Then we ran a counter play several times.



In sum, I like where the offense was this game. I personally am a fan of mixing the 'spread' and 'pro-style' formations and don't see any reason why it can't be done successfully. OSU did it well in '05 and '06, and other teams have had success with it, such as Oklahoma. The plays out of both are largely the same for the line and receivers, so it does not increase practice time, but it gives the defense two things to think about and takes advantage of the strengths of both formations.

More subtly, one can see an influx of new ideas that are making the offense a more integrated whole. We are starting to address some of the criticisms raised after the USC. We are not as formation predictable. We are doing a far better job with 'constraint' plays, such as counters, reverses, and bootlegs off of our primary plays. I like the increased amount of play action passing and boots. Right now Linebackers are biting big time on the play action. As they start to have to respect that, it should open up holes for the running game. Our plays, as described with the Dave pass play, are starting to look like each other. For example, I don't know if you would see the variety of counters we have shown to the speed option play in years past--we are doing a better job of not giving away tendencies. We are also seeing more integrated 'coverage beaters' type routes.

In sum, I think that the offense is growing conceptually and strategically from where it was last year as well, and hopefully look for this to continue.


3. The Offensive Line played far, far better. Before going into individual performances, you can tell that the line grew as a group and were really coached up this week. For the first time in a long time, we were really effective in short yardage. Granted, its Minnesota, but it was still a step forward when we run what everyone expects and still execute:



Look at the push Ballard, Cordle, and Boren get down blocking, and then the great job by Browning pulling...

Basically they functioned better as a unit this week. There were not the same number of blown assignments or misreads.

As to individuals:

--Cordle: He definitely deserved the offensive linemen of the week. To step into one of the toughest positions on the football field and perform like he did is very impressive. He had a very 'quiet' game, which as a linemen is a good thing. Nothing flashy, but executed his assignments and didn't have any breakdowns. He effectively controlled the edge on pass plays. He also played by far the most phsyical in the run game then I have ever seen him do.

--Ballard: Argh! Ballard makes me want to pull my hair out! The most inconsistent person imaginable, but when he plays well, he's really good. Well, the really good Jake showed up this week. In all seriousness, I have to give him props for stepping up after a terrible game at Purdue and playing really effectively. The best I have seen him done in pass protection by far. As I will discuss later, because OSU often uses a pure slide protection, the TE is often more important in pass blocking then the other linemen because they are sometimes out on an island with the DE. Ballard did an effective job for the most part doing so and deserves credit for it.

--Shugarts: When I say the line got a lot of coaching this week, you can tell J.B. got A LOT of coaching this week. Far better performance by him. He is obviously having trouble with speed rushers. So they had him cut block speed rushers quite a bit. This freed him to not overcommit to stopping a speed rusher and then getting beat inside. He also did a better job using his feet when he did pass block, and only got beat inside once. If he continues to work on his pass blocking he can be very good, because he's already a very good run blocker.

--Boren and Browning: I put them together because they are the two I don't worry about. Absolute road graders in the running game. The difference was they rebounded from their subpar performance against Purdue and got back to playing at, what I would say, is their all-Big Ten performance level. They are also really starting to play smart and understand thing conceptually. With Browning especially, you can tell that he is really starting to understand how to find the hole when he is pulling. And then, once either of them get through it, they blow up linebackers.



As you'll see on the first play, it's a trap in which they do a nice job dealing with a pinching 3 technique. Boren, rather than letting the DT pinching go free, stayed on him for a second to slow him down until he could pass him off to Browning. And then still got to the LBer. Browning did a nice job then controlling him. The on the second play, we run a zone play with a fold block by Browning. Look how Boren drives out the 3 technique and Browning drives back the LBer...

Brewster-I saved Brewster for last because I don't have much to say about him, and in this case that's a good thing. He is becoming more and more dependable and starting to reach the Boren and Browning level. He's playing with much better pad level and not getting driven back by NGs.

In sum, like Pryor, everyone linemen turned in a better performance. If this continues we have a chance to have a decent offense by year's end.


4. I was impressed with Hall and Martin. In particular, I really like how both run the 'Dave' play and think they do it better than their more senior counterparts. Dave is not a zone play-its one quick read and then go-you have to follow right behind your pulling guard. If you don't, its too slow to develop and gets stuffed. Look at how quickly Hall makes his decision and goes:




Similarly, here's Martin on his TD run:



5. And finally, OSU does in fact run the zone read, LOL, and has for some time This was a great run by Pryor. Minny actually had this defended well. They ran a 'scrape exchange' on the backside, meaning the Backside DE and the OLB trade responsibilities, in order to throw off the QBs read. The OLB was in a good position to make a tackle on Pryor, but then Pryor turned on the jets...



Pretty impressive, and the kind of thing that people don't even notice about him because he makes it look easy.

6. In sum, this was a step forward by everyone--the coaches in terms of scheme, conceptual playcalling, and breaking tendency, and the players in terms of execution. I again grant its Minnesota, but many of these things fall on the coaches and players, and not who they are playing. I consider this game a step forward and if we can continue to build on this, it could be a good November...


The Play-by-Play Summary (Through Pryor's Possessions) After the Jump...


Followers