Our defense is awful...

[tech_wreck47]

Yes, unless you are happy to be ranked 45th. But Dave Braine would say 45th is very good.

We are 45th without play a good offense, so yes it's pretty bad.

 

[33jacket]

We are 45th without play a good offense, so yes it's pretty bad.

We are 45th because of only 61 snaps a game defended. We are high 70s in yards per play. On top of playing bad offenes....its not good.

 

[YJMD]

Nuke. Yards a game is no good. Yards a play is closer. Ppd is best.

Yards a play:

2016 - 77
2015 - 92
2014 - 110
2013 - 78

Feel free to discuss if you think thats awful. I DO. No reason we cant be top 40 in this stat. We arent close. And this goes back to his loose scheme giving up too many yards a snap.

@AE 87 may have ppd numbers

This is where evaluating defenses gets very complicated. Some things to consider:
1. Is the avg yards per play between rank 78 & 110 statistically significantly different? (I assume not)
2. How do you adjust for quality of opposing offense?
3. Sacks & TFL are important and aren't fully captured by yards per play. Better to lose yardage 20% of the time and gain 6 y/play 80% of the time than to gain 4.8 yards/play 100% of the time.
4. Explosion plays are important and aren't fully captured by yards per play.
5. Turnovers have even more impact.
6. Forcing FG instead of TD is perhaps more important than not letting someone in FG range.
Etc.

But the real kicker here is that there is a whole lot of variance in these things, esp. with something like turnovers which has enormous impact but happens infrequently. Thus, a defense might be much more successful based on randomness (common cause variation).

The problem with approaching addressing these things is that our coaches aren't engineers. They are going to react to what they see, and if they fiddle too much they're causing problems rather than fixing them. We might be reacting to random variation rather than something real.

This is exceedingly difficult to measure as we are doing a different job every game (different opponent, different scheme, different conditions), so we'll have to be a little more qualitative. Focus:
1. Make sure we are doing standard work. More important that everyone knows what job to do and how to do it than that the job is being done right.
2. When looking at breakdowns, try to understand if we have special cause variation or common cause variation. In football defense, special cause variation might be things like (good - we figured out snap count, we recognized play so forced a pick) and (bad - our scheme doesn't account for the wheel route). We need to have a plan to address special causes we can identify.
3. We need to identify our causes of failure and cluster them and identify the process weaknesses we have (not structural - e.g. need someone bigger or faster, etc.)
4. We need to go away from the idea of solutions being finding the best player at some position/training our players to "be more consistent", etc. What we end up doing there is "chasing the tail". That is, our work is probably producing results (individually, like time to recognize play action) that are a predictable distribution, such as normal. By the first approach, we're trying to cut off the chances of having a slow time to react. That's great, but it only moves the median by a tiny bit.
5. Instead, we should focus on process of how we account for play action in a given formation, how we are being trained to recognize it, how effective is that training producing results (and why? -- unclear instructions, process takes too much time, relies on communication which is not reliable, outside of our player's capability so do something else, etc.)

 

[iceeater1969]

We are 45th because of only 61 snaps a game defended. We are high 70s in yards per play. On top of playing bad offenes....its not good.

Lets revist this after Miami game. By then we'll have played two good passing teams as well as three good running teams and that should give a pretty good sample size. If we are 45 @ that time I will be pleased because our younger guys will be maturing and the freauency of gap rush technique sill be increasing as we approach uga

If after miami we are > 75, then u were right and it is awful.

 

[33jacket]

This is where evaluating defenses gets very complicated. Some things to consider:
1. Is the avg yards per play between rank 78 & 110 statistically significantly different? (I assume not)
2. How do you adjust for quality of opposing offense?
3. Sacks & TFL are important and aren't fully captured by yards per play. Better to lose yardage 20% of the time and gain 6 y/play 80% of the time than to gain 4.8 yards/play 100% of the time.
4. Explosion plays are important and aren't fully captured by yards per play.
5. Turnovers have even more impact.
6. Forcing FG instead of TD is perhaps more important than not letting someone in FG range.
Etc.

But the real kicker here is that there is a whole lot of variance in these things, esp. with something like turnovers which has enormous impact but happens infrequently. Thus, a defense might be much more successful based on randomness (common cause variation).

The problem with approaching addressing these things is that our coaches aren't engineers. They are going to react to what they see, and if they fiddle too much they're causing problems rather than fixing them. We might be reacting to random variation rather than something real.

This is exceedingly difficult to measure as we are doing a different job every game (different opponent, different scheme, different conditions), so we'll have to be a little more qualitative. Focus:
1. Make sure we are doing standard work. More important that everyone knows what job to do and how to do it than that the job is being done right.
2. When looking at breakdowns, try to understand if we have special cause variation or common cause variation. In football defense, special cause variation might be things like (good - we figured out snap count, we recognized play so forced a pick) and (bad - our scheme doesn't account for the wheel route). We need to have a plan to address special causes we can identify.
3. We need to identify our causes of failure and cluster them and identify the process weaknesses we have (not structural - e.g. need someone bigger or faster, etc.)
4. We need to go away from the idea of solutions being finding the best player at some position/training our players to "be more consistent", etc. What we end up doing there is "chasing the tail". That is, our work is probably producing results (individually, like time to recognize play action) that are a predictable distribution, such as normal. By the first approach, we're trying to cut off the chances of having a slow time to react. That's great, but it only moves the median by a tiny bit.
5. Instead, we should focus on process of how we account for play action in a given formation, how we are being trained to recognize it, how effective is that training producing results (and why? -- unclear instructions, process takes too much time, relies on communication which is not reliable, outside of our player's capability so do something else, etc.)

This is why i have always said, and if you wish you can dig up my old posts, heck even in this thread, that what makes it awful for me is the game tape.

the stats just offer more proof to the eye test.

For me watching what I see, week in and week out, year over year, I say to myself this D is not good. And when i then see the stats....it validates what I see on film....so i statistically dont care what the difference is from 78 to 110. It doesnt matter. The tape suggests they are not playing well. Then the stats, whether 80 or 130 help proove it....

If we were good...over a full year, most of these statistical categories will trend to the top 40....or top 30.

 

[elwoodgt]

The "eye test" resonates with me. I loved watching our defense play under Tenuta. Every play was exciting, and not in the bad way. Remember the 7-3 win over Maryland in like 2003? It was like watching a no-hitter in baseball.

Under Tenuta, our defense looked eager and hungry. Sure, you might beat us here and there, but we're coming for you the next play. We won a lot of games with defense, and it was just fun to watch. Remember the 2006 Notre Dame game, when they described Philip Wheeler as the best player in Notre Dame's backfield?

 

[AE 87]

Obviously, you can't take away one play from one D and compare them to others. However, when assessing a D, you can take into account what they did apart from 1 play to get a sense of where they are. If you take away the 73 yard one, we held BC to 2.86 yds/carry. We held Mercer to 3.o4 yds/carry. Now, that's not great given their offensive history, but it's not chicken-liver either. It seems to me that we have a shot at being pretty stout against Vandy considering how much they like to run the ball, 63% of snaps.

We'll see.

 

[jeffgt14]

There are different ways to look at our D. While I keep defending our D I do agree that we don’t look very good on paper or on film. A lot of it has to do with Roof’s strategy but I agree with his strategy. He’s trying to not give up the big play and put the other team in a lot of 3rd down situations to try and get a stop there. He’s done a great job at that and it’s up to the players to get the stop on 3rd. I’ve seen a ton of times on 3rd down this year alone where the only reason we didn’t get a stop on 3rd down is because someone missed a tackle or didn’t make the play they should have.

I think our D compliments our O very well. We try and not put the game into the hands of our defense but that just means we have to execute on O and not turn the ball over. It’s basic but we let our offense win the game. I’d rather have that than an overly aggressive ballhawking D that’ll either look terrific or give up 50+ points.

 

[YlJacket]

From what I can see of Vandy's offense, I would equate them to BC in that I don't expect they can put together 10 plus play drives and score. Unless they are just blowing us off the ball, if we can keep them having to convert 3rd downs they will not execute. In that vein, will frustrating, I expect that the current bend but don't break approach will play well Saturday. It will frustrate our offensive co-ordinator who wants more touches for the offense, but will keep Vandy in the 10-14 point range.

 

[Captain killem]

Bend and don't break will work against lesser offensive coordinators. When you play the stronger more advanced coordinators it will become evident to what 33 is saying.

 

[Captain killem]

We have some good personnel on defense matter of fact 2 Gatorade players of the year so I'm not accepting the excuse of personnel.

 

[Supersizethatorder-mutt]

The "eye test" resonates with me. I loved watching our defense play under Tenuta. Every play was exciting, and not in the bad way. Remember the 7-3 win over Maryland in like 2003? It was like watching a no-hitter in baseball.

Under Tenuta, our defense looked eager and hungry. Sure, you might beat us here and there, but we're coming for you the next play. We won a lot of games with defense, and it was just fun to watch. Remember the 2006 Notre Dame game, when they described Philip Wheeler as the best player in Notre Dame's backfield?

And remember the 2003, 2004, 2005, and 2007 UVA games when his defenses got toasted by the Cavs?

 

[InsideLB]

On paper going into the season we had zero proven pass rushers and a brand new secondary. 2 new starters at backer.

Not surprised at the results so far. I agree it's frustrating to watch, particularly 3rd down.

I expect us to grow as a defense this year and be significantly better by the end of the year. Until then I support Ted and the guys as they work hard to get better.

In the meantime I am ignoring and tuning folks out so I can enjoy yhe season more.

 

[AlabamaBuzz]

Well, those of you that don't want to "accept the excuse of personnel" might want to look at where we are hurting the most in this area - DT and LB's....it ain't even close.

No D-coordinator in the country could make this D better than top 60 in my opinion with our inside DL's.

 

[jeffgt14]

I 100% disagree with that. We have some great talent at LB. We are a little weak primarily with depth at DT but they can play. I still think a Gamble/Adams duo starting would improve our defense greatly.

 

[AE 87]

Well, those of you that don't want to "accept the excuse of personnel" might want to look at where we are hurting the most in this area - DT and LB's....it ain't even close.

No D-coordinator in the country could make this D better than top 60 in my opinion with our inside DL's.

Pat Gamble is a 5th year Sr who was a high 3* DT out of high school with offers from Georgia, BC, Stanford among others.
Kallon was was a 4* with similar high offers.

I don't know how one concludes that we wouldn't get better results with different coaching.

 

[grandpa jacket]

Give me a name of a coach, one that will stay here and keep continuity from year to year.

 

[AlabamaBuzz]

Kallon was over ranked based on potential, and yes, I believe Gamble has lived up to his 3 star rating, but no more unfortunately. Even if you factor in a few of these guys playing up to potential, the potential is not high enough, and the depth is woeful. These are not guys that the factories would be starting, no doubt.

If we could get a few "big nasties" up front in the middle that could occupy 2 blockers each, then that would free up the LBs and ends to make more plays....that ain't what is happening, and it ain't happened very much in GT's history. And, problem is that the quantity of these "big nasties" in high school is very limited, and the number that could or would choose a school like GT is extremely limited.... (maybe looking at single digits every year out of high school for those that would even consider Tech, and that is looking nationally, not regionally) You guys that blame the coaches will continue to blame the coaches. When there is more talent, the coaches will be smarter.

 

[Techster]

I believe Gamble has lived up to his 3 star rating, but no more unfortunately. Even if you factor in a few of these guys playing up to potential, the potential is not high enough, and the depth is woeful. These are not guys that the factories would be starting, no doubt.

Gamble is a lot like Gotsis. He should probably be playing SDE rather than DT. The times he's played DE, he's actually looked pretty good. It's a same KCH, Glanton, and Adams are so young, otherwise we just might see Gamble at SDE.

Gamble was actually heavily recruited towards the end of his senior year. GT did well to hang on to him...there were quite a few factories trying to get him to flip.

 

[sidewalkGTfan]

Give me a name of a coach, one that will stay here and keep continuity from year to year.

What's continuity worth if you're only getting average (at best) results out of it?

I'm not saying Roof should be fired but I do think it's more than past time to see our D play at a higher level than what's it been so far with him as DC.