What’s wrong with Georgia Tech’s run game? (Is anything?)

dwyer1Fits and starts. Georgia Tech’s offense is, by most statistical measures, beating its marks from last season. In ’08 the Jackets averaged roughly 370 yards of offense, while this year they are second in the ACC with over 400 total yards per game. Scoring is up by over six points a game too, up from 24 to roughly 30. But the perception is that Paul Johnson’s vaunted flexbone offense is not doing so hot. Indeed, the perception vs. reality debate centers on Jonathan Dwyer, who, if you ask most fans (or see the emails I get), is having a disappointing year despite being third in the ACC in rushing yards per game.

There’s definitely some truth to the idea that Johnson’s offense has not been crisp. Exhibit A were the nationally televised games against Miami, where the Jackets fell behind and could not get the offense going, and Clemson, where a strong first quarter and gutsy fourth bracketed two quarters of very little production. And Dwyer’s 400+ yards rushing this year are muddied by his 66 yards against Clemson and seven against Miami. So what is the prognosis?

I talked to a few flexbone experts and the thoughts were these. The first let’s just get out of the way: Johnson is still using Chan Gailey’s players, and doesn’t yet have its own. I don’t want to harp on this but I am sure there is at least some truth to it. The line in particular was disappointing against Clemson and Miami, and Johnson will ultimately be judged when he gets a full roster there. But that doesn’t much help us in the here and now.

Second, and most interesting, is that there is sentiment that Johnson doesn’t totally trust his quarterback and is predetermining more reads than we might think. I don’t think that is as surprising as it sounds. Johnson said in clinic talks over the summer that he predetermined a lot of the reads last year, and there is some precedent for this: Tom Osborne said in the Nebraska heyday up to 70-80% of the reads in a given game might be predetermined by the call. I’m an option purist, and moreover from a viewing perspective I can’t always tell if a play was a bad read or a predetermined one. But that would help explain some of the stunted dives to Dwyer that have not seemed to go anywhere at times this year. (But don’t ask Mississippi State or North Carolina.)

Finally, there have been some changes in defensive tactics. Most notably almost everyone is playing Georgia Tech with a nose-guard to help stop the midline option and to either stop Dwyer up the middle or at least muddy that read. Miami in particularly played their base defense but simply moved one defensive tackle over to nose guard. This isn’t an end-all be-all scheme, as it has opened up some outside lanes and various counter plays, but GT has not always executed those plays well. Rewatching the Clemson and Miami games in particular one is struck by the fact that there are big plays to the outside to be had, but the Jackets just keep missing key blocks. Now credit both Clemson and Miami for fending off the myriad chop cut blocks and making tackles, but if Johnson wants to continue having success they will have to make teams pay for crowding the middle, and the passing game can only go so far.

The demise of Johnson’s scheme has been premature, though, like any other squad, it comes down to execution and GT will have to prove that its success will continue. The Jackets face a reeling Florida State squad this weekend, followed by a streaking Virginia Tech team. We’ll learn a lot about PJ’s boys in the next two weeks.

Paul Johnson bonus. Below are highlights from the 1992 Holiday Bowl, where Hawai’i, with Johnson as offensive coordinator, defeated Illinois.

  • Tom

    “the myriad chop blocks”

    CUT blocks, not CHOP blocks.

  • Jonathan B

    “Now credit both Clemson and Miami for fending off the myriad chop blocks and making tackles…”

    No! Et Tu, Chris Brown? Please don’t contribute to the confusion about PJ’s completely LEGAL blocking scheme and the patently ILLEGAL chop blocks. One is a one-on-one block, the second is a dangerous block in which the defender is engaged and another offensive player blocks him below the waist. Two different things.

  • Trey

    Part of the perceived problem is one of exposure. The offense was pretty mediocre in the early high-exposure Thursday night games, but looked very good in the last two games that nobody saw (vs. UNC and @MSU, both regional broadcasts). The offense was very good in the last two games.

    If teams are better at defending it, they’ve primarily become better at not allowing big play home runs in the running game. That’s a big reason why GT does look much worse in one stat, yards per carry. Currently GT is at 4.83 YPC which is shockingly low for a CPJ coached team.

    But what doesn’t show up on the stat sheets, and maybe you can’t really see on a TV broadcast, is that GT is doing better than in 2008 on the routine dive plays and midline keepers that don’t get broken for big yardage. Last year those plays tended to go for 1-3 yards while this year they are frequently gaining 3-5 yards. You can see this in the time of possession numbers — often a meaningless stat, but telling in this case. In the last two games GT had the ball 42 and 36 minutes. That didn’t happen last year.

    I hadn’t heard your theory about Johnson predetermining reads, but it would explain something that looked strange: in the Miami game there were a few plays in which Nesbitt would pull the ball only to get destroyed on the edge while Dwyer pounded 6-8 yards without the ball. I wondered at the time if Miami was doing something strange to make Nesbitt miss his reads.

    Finally, Johnson’s offenses have always been stronger after Nov 1st. This has happened every year since 2003, and usually by a huge margin; and it was also true at GSU. Navy tended to have front-loaded schedules and initially I thought that had a lot to do with it. But Navy played GREAT in all but one of the 5 bowl games CPJ coached them in; the GSU offense played better in the playoffs than the regular season; and GT’s offense was strongest in November despite a very back-loaded schedule.

    That being the case, I think it’s too early in the season to judge the offense. I’d wager that GT ends the season with over 5 yards per carry.

    It’s also worth noting that passing yardage is up 50% without very many more attempts, because Tech is leading the nation in yards/attempt by a 15% margin.

  • Brent

    The nice thing about Paul Johnson is he does not mince words. He readily admits that problems in the Miami and Clemson game stem from poor execution. Missed blocks were everywhere in the Miami game and kept GT from hitting the long run payoffs that makes the system so attractive and Dwyer’s stats potentially so good. The offensive line was porous in that game and gave Nesbitt no time to make his reads. I had not picked up on the predetermined calls, so perhaps that was also an issue. However, defenses have been committing to take Dwyer out of the game, which creates opportunities to pass to a talented set of receivers in Thomas and Hill. Stretching the defense to help in the secondary will naturally open up the run game and improve Dwyer’s stats. I would not want to be FSU this week trying to decide where to put their chips. The beauty of PJ’s system is that it is built around efficiency and taking advantage of what the defense gives you. If they take care of business against FSU, the VT match-up will be one to watch. I would be most worried about that kicking game, which puts GT at a disadvantage in both field position and scoring opportunities.

  • Tim

    I remember hearing somewhere, possibly from an announcer during one of the Thursday night games, that Paul Johnson was the kind of coach that would come right back with a counter to something the other team did on the field. That’s why I thought it was strange that they waited until the 4th quarter of the Clemson game to try some of those counter plays and QB keepers that helped them win in the end. Is that personnel and practice time again, or maybe a little redneck cussedness that he’s going to call the same plays over and over until his guys start making the right blocks he wants?

  • Craig

    It seemed to me that in this last game Mississippi State was loading the line to stop the run, which left Thomas one on one with a guy several inches shorter. Johnson saw this several times had Nesbitt just throw the ball straight over to Bay Bay.

    What impressed me most about this game is that Nesbitt is throwing more accurately and with more confidence. This will, I hope, keep other teams from crowding the line.

    I’m interested to see what Bobby Bowden has in mind for the “daggum wishbone” Saturday. If Nesbitt keeps his cool and the receivers don’t drop the ball, he could break 300 yards passing.

    I’d still rather watch them run the option than throw, but that’s my preference.

  • Ben

    A lot of it has to do with schedule. The Jackets have already faced the best run defenses they will face this season. FSU and VPI’s run defenses have been sieve like this season, and the next 4 after those 2 games are UVA, Vanderbilt, Wake Forest and Duke. In 2008, the schedule was not as front loaded with strong defensive teams as the 2009 edition is. It takes 4-5 games typically before for the timing to come around. The offense is starting to get cranked up now, these questions will cease in the coming weeks IMO.

  • Tim James

    Oh yeah, I almost forgot about FSU’s secondary, Craig. Hopefully Nesbitt can shred them.

  • Scott

    I believe GT led the nation last year in plays over 20 yards. They’ve had a few long runs, but I’ll bet if you broke it down, you’d find fewer “long” plays in the running games, bringing their avg/carry down. Combination of teams playing a little better defense but also our blockers missing a few key blocks that would have resulted in more big plays (as mentioned).

    They’re more than making up for it in the passing game, which is one of the most efficient in the nation (#1 in yards/attempt, #1 in yards/completion).

  • Rob

    The ypa passing numbers that Scott mentioned arent even close. GT is #1 at 11.4 ypa, Notre Dame is #2 at 9.9.

    Florida St is the only defense giving up more than 10 ypa. Saturday could be interesting.

    RE: long plays vs long drives. Last year GT couldnt manage long drives. Eventually they would stall and/or end with a fumble. This year, fumbling is way down and getting the 3-5 yards on “failed” dive plays is keeping drives going. The offense this year looks like what people often think of with a run first offense – grinding out long drives – instead of what I have been telling people GT’s offense is, a big play offense.

    Also, in the last 2 games, there have been very few triple option plays run. A lot of midline, toss sweeps, and other plays, but very little straight TO plays.

  • Dan

    Tech is simply not an option team right now…at least not a triple option team. The interior of the line has a long way to go. I would love to see a break down of the passing game and routes ran in the MSU game.

  • Jon

    Demataryius Thomas is a beast. If Nesbitt has any resemblance of accuracy, Thomas would be putting bigger numbers than it already is, not that his production is subpar. His production is outstanding and he would be a more well known WR if he had played in a passing offense. He’s that good.

    With that being said, he doesn’t have much experience in running the full route tree and he has a tendency to run sloppy routes.

  • Elliott

    In your Hawai’i video the QB is #4 Ivin Jasper.

    Who is OC at Navy now!

  • Elliott

    Oh, and you also see HC Ken Niumatalolo next to PJ in the box.

  • Terrance

    3 Things come to mind.

    1) We played Miami as the 3rd game in 12 days. I want to see any other college team do that.

    2) CPJ doesn’t have his players? CPJ will not be able to find better Skill guys than we have now. O-Line yes, Skill no.

    3) The passing game stats are good. AS with any TO team, if teams go man you throw for 250 yards i.e., Miss State. Teams are keying JD which means his stats should be down and every game Thomas should catch for 150.

  • bj

    Have you checked out Tech’s Yards Per Pass Attempt and Yards Per Completion numbers yet? Tech leads the nation in both, by a wiiiiide margin.

    People are down on Tech by and large because Tech’s rushing numbers are sagging a little, but that’s in part because defenses are committing both safeties to option assignments, and Tech is passing more as an adjustment. Tech didn’t pass particularly well vs Clemson, but started putting it together vs Miami, and was lights out vs Mississippi State.

    Breaking down what Tech is doing in the passing game might be worth a later blog post.

  • bulldogoption

    I think it really boils down to DCs having better schemes. It had likely been a long time since the DCs that GT faced last year had defended true triple option with a dive back. You could see in some of those games last year (the Georgia game comes to mind) that there were some major breakdowns on defense……who’s got force/pitch. With each passing year, those DCs and defenses will get better at defending the option.

    Doesn’t have the players….eh? I agree with a poster above that its going to be hard to get better skill kids than they have now.

    Predetermining reads…I think that this is fairly common among some option coaches. They don’t want a 19 year old determining their livelihood. It has its advantages and disadvantages. The disadvantage is that you can miss out on big opportunities….the dive would have busted for a TD but the pull sign was given. The advantage is that it takes some pressure (mental) off of that 19 year old kid.

  • Joe

    It’s Demaryius Thomas (nickname “Bay Bay”), hence the “B. Thomas” on his jersey. He is also an excellent blocker.

    Very true about there being few triple option plays run the past two games with more midline, toss sweeps, counter traps, draws, etc. Perhaps that, coupled with the predetermined option reads, contributes to the fewer fumbles and more “grind it out” style mentioned earlier vs. the higher-risk pure triple option and more big plays/long runs.

    Love your website!

    PS: Did I see the Fumblerooski in that Holiday Bowl video clip?!

  • Dallas

    I think another major note on your second point would be that, if Johnson doesn’t trust his QB to even read the option plays most of the time, it’s likely he doesn’t have much freedom to check at the line either. Those shifted fronts (like you were talking about Miami using) are vulnerable if the QB can check and call an option play that hits at the bubble (i.e., if they shift a DT over to a three tech to try to muddy the read on inside veer, the QB can check to midline to take advantage). I don’t know that this is necessarily the case, of course, but I would have to imagine that Nesbitt probably isn’t allowed a lot of liberty at the line.

  • MTK

    Sorry for the following dorky question Chris, but how technically do you keep up with all the games? Cable/satellite provider’s DVR? Tivo? Other? Do you pick a few to record each week? Do you live in front of the tv on weekends? Curious. Really enjoy your site.

  • MTK: I don’t watch every game, though it might seem like I do. I watch plenty; use DVR; and espn360 comes in handy. Plus I’ve watched a lot of football in great detail so it doesn’t take me too long to size up what I’m looking for.

    Joe: The fumblerooskie was pretty awesome.

  • MikeLew

    Johnathan B-

    As a former receiver, I certainly know the difference between the cut block and the chop block, and often had to cut safeties and OLBs(being the biggest receiver on a (bad D3) college team at 6′, 175 will cause that), but I know too many former D-linemen that had their knees and ankles screwed up by completely legal cut blocks to agree with your implication that the cut block isn’t more dangerous than a regular block.

  • HAAL Fan

    It is fairly common practice to run in plays with set reads. Navy was known for “burning” plays to set up later plays based on player reactions. They’d adjust blocking schemes so that one player was allowed to make a big tackle for no gain or a 1 yard gain only to jump on the same ball carrier the very next play and have the QB “replace” his previous position for a 15 yard gain. I’m not surprised he’s doing it, but it is a huge detriment at that level, seeing as how you lose the “Check with me” system previously mentioned.

  • Golden Hand

    Tech has one, and only one problem (it ain’t DC adjustments). Their offensive guards are really, really subpar players, and their offensive tackles are too small. The left tackle weighs 255! The two starting guards are slow, and kinda clumsy, especially the right guard (watching him try to pull and lead is like watching an elephant waltz). Tech struggled a lot last year in running the offense efficiently, and it’s just more of the same. Expect improvement through November, though. Johson is that good.

    Johnson has been recruiting offensive linemen that fit his scheme (middle-sized, but quick and aggressive). They’ll be a LOT better in a couple years, assuming he keeps recruiting skill players as well as Tech has in the past.

  • Brad in KY

    GT just rushed over, through, and around FSU for 400 yards. FSU’s defense isn’t very good, but 400 yards is still a huge number.

    I guess the question we should be asking is, “What’s wrong with Georgia Tech’s defense? (definitely something!)”

  • Mark

    I’m far from an expert on this, but I remember reading, before the season, several predictions that Dwyer wouldn’t have the kind of year he had last year simply because he’s moved from the “A-back” position to “B-back”. By design, he’s the one pounding it up the middle while Roddy Jones and the other guys get the flashy plays to the outside the majority of the time.

    I would love to hear what Chris thought of that FSU game, though.

  • Simpsons Kidrobot

    I significantly adore the vinyl Kidrobot Simpson’s mini information – As with several Kidrobot sets, The Simpson’s arranged is sold in blind boxes only. The fixed is comprised of 24 figures, including three chase statistics. Each figure is 3 inches in height.

  • Valeria Aamodt

    There’s noticeably a bundle to learn about this. I assume you made sure good factors in options also.