Recruiting for the 3-3-5 (a test case)
I've read here and elsewhere that going to the 3-3-5 will hurt recruiting, especially D-Line. The reasons given vary, but center on the assumption that D-Line players won't want to play the 3-3-5 since they are not "showcased" (or words to that effect). I decided to take a look at the one program I know runs the 3-3-5 consistently (WVU) and see if their defensive recruiting (and especially D-Line recruiting) is consistent with this claim.
**NOTE: ALL rankings according to Scout**
Last year, their defensive line recruiting was pretty good (2 3-star DTs, a 4-star DE and a 3-star DE to go along with two more 2-star linemen). That is a pretty good haul for WVU. For comparison, they only pulled in 6 3-star or above players on offense, compared to 4 on the D-Line alone (along with 8 more LBs or secondary players).
In 2007, they got a 4-star DT, a 3-star DE and one additional 2-star DT. Defensive recruiting again was a bit better than offensive, with only 2 of 16 defensive recruits below 3-star status (compared with 2 out of 11 on the offensive side).
In 2006 they had a crappy class in general, but note that 4 of their 5 3-star or higher rated players are on the defensive side of the ball.
In summary, while the recruiting classes at WVU are generally lower quality than UM classes, within those classes, RR doesn't seem to have had a problem recruiting for the 3-3-5, especially the D-Line. If fact, it seems he had an easier time recruiting defensive players than offensive players.
Side note: It does appear true that the supers WVU (rarely) got tended to be on the offensive side of the ball--Devine, Gwaltney, Jenkins--but those were few and far between.
December 18th, 2008 at 11:26 AM ^
this is a poor test case.
December 18th, 2008 at 11:34 AM ^
Because...
December 18th, 2008 at 12:40 PM ^
hurt the arguments of people who hate Rich Rodriguez
December 18th, 2008 at 12:54 PM ^
it has nothing to do with "hatred" or rich. if he goes with an unproven dc with better ones out there, im gonna question it. everyone should be questioning it.
theres a difference between supporting the team and questioning decisions. everyone here has too much of a hard on for rich to realize that he makes mistakes, so what if someone points them out
December 18th, 2008 at 1:04 PM ^
So, essentially, any coach you haven't heard of right now or who you don't like is unproven, and the 2 or 3 you want hired are "better", right?
This is my problem with you. Not the fact you are questioning. The fact you are questioning, but the only possible answer is the answer you approve of. That's not questioning, it's being a baby.
December 18th, 2008 at 1:07 PM ^
did i say coach or scheme?
this is what i dont like about you. you cant have an adult discourse about this without calling me a baby.
December 18th, 2008 at 1:10 PM ^
"if he goes with an unproven dc with better ones out there"
Uh, yeah, you did say coach. That's what a DC is.
December 18th, 2008 at 1:13 PM ^
sorry...
a coach whose going to run an unproven system
December 18th, 2008 at 12:50 PM ^
the argument is something like this
wvu does not recruit well. they recruited slightly above average on the defense line for the past three years (except one year, they recruited poorly all around).
THEREFORE...
the defensive line was slightly above average compared to the rest of the team = all their defensive recruits wanted to play in 3-3-5 = (and this is where the line starts getting fuzzy) recruiting at michigan will not be hurt.
there is no causation in this logic. it is way to attenuated to draw a comparison between 3 years at wvu and what will happen at michigan.
i get the argument for change on the offense side of the ball. its one thing to get a recruit to come to michigan in the spread because every single team runs a version of the spread. NO ONE runs the 3-3-5. wvu. air force. anyone else name a d-1 school? you think if this thing worked more places would have picked up on it. we all agree that the coaches know more. well, the ones who are in the know aren't using this scheme so what do you think it says about this scheme.
okay, then you say well, nick saban at Alabama runs a 3-4. thats the same thing. but its not. because the 3-4 is not the stack, if it was we wouldnt be having this discussion. it might have the same amount of people on the line, but that doesnt matter. every defense has the same amount of people on the field, but they are all different.
the problem is, and i think everyone who is honest with themselves could foresee this happening, a recruit having offers form all the "top" schools that run 3-4 or 4-3. something that is known and well established. and then there is michigan's 3-3-5. which is unknown and unproven. you're a 5 star recruit with nfl aspirations. you tell me where you're going....
December 18th, 2008 at 12:53 PM ^
You say "if it works, other would run it" - how do explain the fact that WVU has, by and large, been a better defensive team than Michigan since Rodriguez arrived (as measured by yards-per-game)?
December 18th, 2008 at 1:05 PM ^
id say...
show me those stats. and while you're at it give me defensive rankings of the two, rushing yard rankings, passing yard rankings, how many points each team gave up. etc. one category "by and large" does not mean a better defnse across the board.
then you will tell me it doesnt matter because wvu doesnt get the talent we have.
then i will tell you it doesnt matter beacuse the big east does not compare to the big ten.
then everyone else will chime in and call me a fairweatherd fan, in not so nice terms.
this happens every day. why dont you, and everyone else defending the 3-3-5, answer this honestly. if michigan gets hopson, or casteel, or someone else we've never heard of to run the 3-3-5....would you, as a michigan fan, be excited for the defense? would you be pleased?
or would you be more excited if they got chavis or rhoades or someone else established?
December 18th, 2008 at 1:11 PM ^
I'll be excited. Just like I would have been excited if Michigan hired Urban Meyer when the spread was "exotic" and nobody "good" was running it.
December 18th, 2008 at 1:18 PM ^
and ill be running my head into the wall when we have to hear all the same excuses:
the defense is learning a new scheme it takes time to learn a new scheme they're terrible now but they'll be better next year. besides that terrible scum bag of a coach lloyd carr left the cupboard bare.
December 18th, 2008 at 1:31 PM ^
Regardless of whether a 3-3-5 guy, a 4-3 guy, a 4-2-5 guy, or a 6-1 guy comes in, our players will be learning a new scheme.
December 18th, 2008 at 3:58 PM ^
You forgot "Stevie Brown is doing this to hurt me"
December 18th, 2008 at 1:27 PM ^
dex, baby, i wish you would come back online. i miss you already
December 18th, 2008 at 1:39 PM ^
The stats, for 5 years:
2008:
Points per game - WVU 9th, UM 71st
YPG - WVU 35th, UM 58th
Pass YPG - WVU 36th, UM 78th
Rush YPG - WVU 45th, UM 46th
2007:
PPG - WVU 7th, UM 23rd
YPG - WVU 4th, UM 25th
Pass YPG - WVU 11th, UM 7th
Rush YPG - WVU 15th, UM 57th
2006:
PPG - WVU 48th, UM 8th
YPG - WVU 53rd, UM 6th
Pass YPG - WVU 100th, UM 74th
Russ YPG - WVU 9th, UM 1st
2005:
PPG: WVU 10th, UM 22nd
YPG: WVU 8th, UM 43rd
Pass YPG: WVU 27th, UM, 52nd
Rush YPG: WVU 11th, UM 43rd
2004:
PPG: WVU 26th, UM 38th
YPG: WVU 31st, UM 29th
Pass YPG: WVU 39th, UM 42nd
Rush YPG: WVU 34th, UM 27th
So, in the past 5 years, UM has been better once, and roughly equivalent once. WVU has had a better defense 3 times, and has 3 top 5 finishes in PPG in the past 5 years. To say "it doesn't work" is sticking your head in the sand.
December 18th, 2008 at 11:31 AM ^
Please give examples of an opposite scenario. Where the school runs a 3 3 5 and a history of poor recruiting.
December 18th, 2008 at 11:35 AM ^
drewsharp says this is a poor test case. I would like to know why that is.
December 18th, 2008 at 11:57 AM ^
The idea a d-lineman needs to be left free to roam, get sacks, and basically not be double teamed in order to be showcased is absolutely ridiculous (and I realize you are NOT putting this point forward in your post, just to be clear).
If you go into the 3-3-5 as a D-Line player, and you command the double team you need to command to make the defense run like a well oiled machine, you are being showcased. This is what the NFL wants. If someone doesn't realize that a DT who can force a double team at will is one of the highest prized commodities in the league right now then I question if they actually watch the NFL or just read through stat lines to see who is "good".
December 18th, 2008 at 12:01 PM ^
There is a reason why Albert Haynesworth and Tommy Harris get paid like QB's.
December 18th, 2008 at 12:28 PM ^
Mr. Ngata would be a prime example of this - he is the star nose tackle for the Ravens and all he does is command double teams, which allows Ray Lewis and company to roam free. He definitely played at a Pro Bowl level this year and Lewis knows this.
I'm not sure about Haynesworth - I just think he's so dominating as a defensive tackle that he's commanding double teams, but not necessarily as a nose tackle. A better example would be Kris Jenkins of the NY Jets. He started out as a 4-3 defensive tackle and has moved to the nose tackle for the 3-4 and is playing quite well. Shaun Rogers is another example.
December 18th, 2008 at 12:31 PM ^
Haynesworth went from a 3-4 in college to a 4-3 in the pros. Most think he's the defensive player of the year, though he has almost no stats - he just clogs shit up.
December 18th, 2008 at 1:29 PM ^
I could be wrong, but I don't think Haynesworth was a 3-4 tackle in college. He and John Henderson were both DT's drafted in the first round of the same year, so I'm thinking they ran a 4-3 unless one of those monstrous guys was playing DE, which I doubt.
December 18th, 2008 at 3:30 PM ^
50 tackles and 8.5 sacks is a lot for any linemen, especially an interior linemen.
December 18th, 2008 at 1:06 PM ^
Finally found an article I read a few years ago, about the Ravens' nose tackle before Ngata, and why he was voted the team's best lineman despite his lack of stats: http://sports.espn.go.com/espnmag/story?section=magazine&id=3617911
The author interviews several people who just gush about Kelly Gregg's ability to take on multiple blockers and leave Ray Lewis free to roam. And also mentions the big pay day he got for doing it so well.
(Is he still starting? Or did Ngata completely replace him, or what? The article listed him as Ngata's teammate, and they are both still on the roster. I don't ever watch Ravens' games, so I don't really know who starts where on the defense this year.)
December 18th, 2008 at 1:09 PM ^
Gregg is on IR right now - he signed an extension in 07 though, so Ngata hasn't completely supplanted him yet. Ngata is a freaaaaakkkkk.
December 18th, 2008 at 12:25 PM ^
I'm not real familiar with Tennessee. Did Haynesworth play in a 3 man front?
December 18th, 2008 at 12:40 PM ^
PA Blue, thank you for showing me that DT's can be recruited for a 3 3 5 defense. Even if it was only one example, at least you brought statistical information to back your point of view.
Thank you Chitownblue. More examples of a stud DT who played a 3 man front in college are encouraging.
December 18th, 2008 at 12:43 PM ^
Something tells me this guy will get drafted somewhat highly as well:
http://alabama.rivals.com/cviewplayer.asp?Player=402298
December 18th, 2008 at 12:43 PM ^
it's a poor test case because WV doesn't recruit well period and certainly not at Michigan's level. We recruit different people at different levels.
Variations in a WV-level class would not necessarily be the same as variations in a Florida, LSU, or USC class.
That said, there are no good test cases that I can think of.
December 18th, 2008 at 1:03 PM ^
During the years I referenced, UMs D-Line classes were not much better than WVU:
Last year we, of course, had only Martin, a 4-star
2007 we had a 4-star DE and a 2-star DE
2006 we had 5-star DT (Slocum), a 4-star and a 3-star, to go along with a 3-star and a 2-star DE.
That is a 5-star, 3 4-stars, 2 3-stars, and 2 2-stars in the last three years for UM compared with 2 4-stars, 8 3-stars and 2 2-stars fro WVU. That is a 3.00 star average for WVU vs. a 3.37 star average for UM. Given what you say, I'd expect it to be much different, but it isn't.
In fact, you have to give consideration to the conclusion that WVU OVER-ACHIEVED in recruiting on the D-Line given the general liabilities they have in attracting top talent:
UMs D-Line classes tend to be LOWER in average star rankings than their classes in general, while WVU's D-Line classes tend to be HIGHER in average star rankings than their classes in general. UM classes generally tend to have average star ranking of 3.5-3.7 but our D-line recruits over that 3 year period have an average of 3.37 stars. WVU classes in general range from 2.4-3.1 stars on average but their D-Line recruits average 3.0 stars.
December 18th, 2008 at 12:54 PM ^
I understand the apples to oranges argument. But if WVU can recruit 3 and 4 star DT's and DE's then we should probably be able to do the same here at Michigan. Recruiting 3 and 4 star kids would not be taking a step back for Michigan by any means. Since we do not have a two deep roster of 5 star kids playing the D-line as it is.
December 18th, 2008 at 1:16 PM ^
I'm not saying the 3-3-5 stack is good... or bad... or anything. All I'm suggesting is that people who say quality D-Linemen won't come to UM if we run this system are purely speculating, and any evidence that does exist (albeit limited) contradicts that notion.
December 18th, 2008 at 1:30 PM ^
I thought you brought up some useful stats, but you're just not going to convince some people.
This discussion reminds me of a quote from that Clay Shirky video Brian put up earlier this week: "They didn't care that it worked in practice because they had already determined that it couldn't work in theory."
December 18th, 2008 at 1:39 PM ^
I am not afraid of innovation. Rodriguez fathered the modern spread. He may have a plan to revolutionize the defensive side of the ball as well. The mistake I think RR made was choosing a DC with philosophical differences to start the 2008 season.
December 18th, 2008 at 1:48 PM ^
Any defense - if it's run well - can be successful.
Any offense - if it's run well - can be successful.
This is why Paul Johnson can run the triple option to the tune of 282 yards a game and Mike Leach can throw for 10,000 yards a season.
It's also why the Ravens can run a 4-3 and win a Super Bowl while the Steelers run a 3-4 and win a Super Bowl.
We shouldn't be worried about what defense we run. We should be worried about the quality and cohesion of the coaches who are teaching it and calling the plays.
December 18th, 2008 at 2:01 PM ^
Well said.
December 18th, 2008 at 2:22 PM ^
the 1-1-9 can be successful too. But it's going to be harder than with a 3-4 or a 4-3 purely via schematic reasons. There's a reason nobody runs the 5-2 anymore.
Furthermore, part of whether a 3-3-5 is successful is the players you can recruit for the system. I think the WV example isn't totally irrelevant, but it's not close enough for my taste either.
And exactly as you say, the quality of coaches matter. If we get the 3-3-5, I want Casteel, Mr. 3-3-5 himself. Not a bunch of RR's cronies who don't have the same coaching ability.
December 18th, 2008 at 2:28 PM ^
"If we get the 3-3-5, I want Casteel, Mr. 3-3-5 himself. Not a bunch of RR's cronies who don't have the same coaching ability"
You might get your wish. I have a feeling the longer the DC job is without a replacement, the more it means RR is waiting for Casteel,
December 18th, 2008 at 3:38 PM ^
Borrowing from the database I used in my last post, it seems Michigan's defensive line recruiting, which has been perhaps our worst point in that time, was still a lot better than WVU's. Attrition, however, made these much more even.
Michigan's average Rivals ranking for the 13 ranked Defensive Line recruits from 2005 to 2008 was 5.7 (low 4-star). The average rank of each recruit was 31st.
West Virginia brought in 9 ranked recruits, averaging 5.6 (3.5 stars) and average ranking of 46.
Michigan recruits:
Terrance Taylor - 5.9 (4 stars) - 9th DT 2005
Adam Patterson - 5.9 (4 stars) - 10th DT 2006
James McKinney - 5.9 (4 stars) - 10th DT 2005
Mike Martin - 5.8 (4 stars) - 20th DT 2008
Ryan Van Bergen - 5.8 (4 stars) - 20th LE 2007
Eugene Germany - 5.8 (4 stars) - 22nd LE 2005
Jason Kates - 5.8 (4 stars) - 23rd DT, 2006
Chris McLaurin - 5.5 (3 stars) - 20th RE 2008
Renaldo Sagesse - 5.7 (3.5 stars) - 43rd DT, 2007
Greg Banks - 5.6 (3.5 stars) - 51st LE, 2006
Quinton Woods - 5.6 (3.5 stars) - 53rd LE, 2006
Carson Butler - 5.5 (3 stars) - 55th LE, 2005
John Ferrara - 5.6 (3.5 stars) - 55th LE, 2006
West Virginia Recruits:
Tevita Finau - 4 stars - 15th LE 2008
Larry Ford - 3.5 stars - 21st LE 2008
Christopher Palmer - 3.5 stars 39th DT 2008
Julian Miller - 3 stars - 39th RE 2007
Rashad Roussell - 3.5 stars - 43rd DT 2005
Asa Champman - 3.5 stars - 46th DT 2007
Jeff Braun - 3.5 stars - 58th DT 2008
Tyler Urban - 3 stars - 69th LE 2008
Junius Lewis - 3 stars - 82nd DT 2007
(LE/RE stands for left end and right end. Basically, LE is your strongside, or BIG DE's, while RE or weakside end is your pass rusher-type).
December 18th, 2008 at 3:48 PM ^
You make my point for me pretty well. I'm not saying UMs D-Line recruiting was the same as WVU's. I understand that UM attracts better recruits, on average, across positions. But your data show that the gap is actually pretty SMALL for D-Linemen:
1) 3.5 star average for WVU vs. low-4-star average for UM
2) 5.6 rating average for WVU VS. 5.7 FOR UM
3) 46 average position ranking for WVU vs. 31 for UM
These are NOT big differences, and are smaller than we'd expect given all we hear about how hard it is to recruit to WVU.
The biggest difference is the sheer number (13 for UM and 9 for WVU) but that is to be expected (playing 3 rather than 4 down linemen at a time means you need fewer).
December 18th, 2008 at 3:52 PM ^
Really, I don't see a whole lot of difference between what we were doing before, and what WVU was recruiting, except what you'd expect between the top program in college football history and the Pride of Appalachia. If a 3-man line set does anything, it should hurt recruiting of rush ends, but Michigan has typically filled that position by recruiting guys projected at inside linebacker (Woodley, B. Graham, Crable, et al.), as evidenced by the fact that the only rushing end we've brought in these past four years was benchwarmin Chris McLaurin.
That said, you eliminate a defensive tackle spot, but we've been hideously thin at that position since, well, 1997. This year I thought it was a position of strength next to what the rest of the defense looked like, but with Taylor and Will Johnson departing, that may not be so true in 2009.
Of course, there's always the recruiting message. Our best shot for a 5-star this season is Will Campbell, who essentially would be demoted by a scheme swap from 1st sub under two rotating positions, to 3-deep on one position. That may not really be true -- lots of time before all positions are set and he could out-complete Martin, Sagesse or Van Bergen. But it could be a line.
December 18th, 2008 at 4:05 PM ^
There has been a lot of talk about D line recruitment, but one of the factors that makes running a 3-4 hard to do in college is finding big enough outside linebackers who can engage and separate with offensive tackles. 3-4s are highly successful in the pros because there are only 32 teams, so it makes it easier for the 10 teams who run a 3-4 to find 260lb defensive linemen who can stand up and play outside. Competing against a 100 colleges for 260lb defensive linemen who might be able to stand up and play outside is incredibly difficult. Obviously Alabama proved that a 3-4 can be run, but notice that 3 out of their 4 starting linebackers are in excess of 250lbs and they have arguably the best NT in the game. It's the perfect storm for the 3-4 to be successful. But as of right now, we don't have that size on the edge or in the middle to run a 3-4. The closest we came to being able to run a 3-4 was 3 years ago when we had Woodley and Crable on the outside and Branch and Taylor on the line and even then it made more sense to keep Woodley in the 3 point stance.
I will admit though that with the exception of Roh, this year's class seems geared towards the hybrid DE/DT that play inside and outside on a 3-4 line. Dequinta Jones was originally a DE commit, Campbell and Lalota can play either inside or outside, and Pearlie Graves has the measurables (6'3'' 270 4.7 40) that would convince you he could play outside or inside as well.
December 18th, 2008 at 4:43 PM ^
That's a fantastic point.
Really, to my understanding, the difference between 3-3-5 stack and 3-4 is whether you employ a bandit FS/OLB hybrid, or that guy is a defensive lineman standing up.
The 3/4 is my favorite defense for a running spread, since you can designate one of four guys to be the spy on each play, and there's still three potential blitzers coming from anywhere.
The 4/3, rather, seems to me to be an optimal starting point for facing a stand-up QB, since your four guys can often get pressure against a 5-man front, leaving you all your linebackers to help in coverage and cover anything coming out of the backfield.
The 5-2 or 5-3 is your best bet for facing a Georgia Tech or Bo-era option offense. You put a nose in every hole and can either take the pitch man out early, or force an early decision.
If you're going against a spread passing game, a blitz-happy nickel package seems to work the best (think of those 38-0 drubbings of ND and Purdue in the mid-'00s.)
Each of these offensive sets are designed to give the offense more talent in their preferred zone of action than the defense. The Option wants to break it down to one tackler versus the ballcarrier and a pitch-man. The Spread wants to move defenders all over the field then exploit 1-on-1 matchups in space. The West Coast uses timing to exploit the momentary advantage of a receiver who knows his route over a defender's reaction time.
To do this, each offense, and variation on such themes, gives up someplace else. To approach this scientifically, then, for a defense in 2009, what kind of offenses will we be facing?
The 3-3-5 is a reactionary defense. It sees what the offense is doing, then adjusts mid-play, usually with the designated bandit filling a hole, and the linebackers then deciding which is the spy and which is the late attacker.
What kind of offense does this stop? I don't know. I haven't seen it enough to say.
If I'm picking defensive schemes, personally, I'd watch Ohio State's offense again and again, and go with whatever screws them up the most. You can build any of these sets into a good defense against whatever you see.
December 18th, 2008 at 6:10 PM ^
Caveat - I am no GSimmons, nor even close to understanding the intricacies of the 3-3-5
Maybe I'm being naive, but I just don't see the major stumbling blocks on defense like we saw on offense. I always figured that with offense, if you are a 6-4, slow QB asked to run the spread, you are in trouble, but you might be great for a pro-style; on defense, there seemed like less striation, less opportunities for a player to be limited by a particular scheme. Simply put, you hear all the time about "system" players on offense; rarely do you hear about a "system" player on defense except in certain occasions with LBs.
Yes, a 3-3-5 is different than a 4-3-4 or 3-4-4, but they are not so fundamentally different that players well-equipped for one won't gravitate toward a school because the school runs another. If you are a monster DT, you are going to get doubled in the 3-4 and 4-3 just as much as you would in the 3-3-5, but that is your job and if (when) you get to the NFL scouts know that your stats don't tell half the story.
Now, I can see the argument that DE recruiting may suffer a bit because, obviously, a 4-man front is going to create far more 1-on-1 matchups than a 3-man front, and DEs generally get measured by sacks/pressure and little else. Luckily, though, most NFL teams are looking harder at those hybrid DE/LB types than in years past, meaning a guy who can move from the DL but can still cover (like you need in the 3-3-5) would be at a premium. So yeah, there may be some DEs who prefer a 4-3 or a 3-4 because that was what they ran in HS or they think they translate better to the NFL, but if they are athletic enough I'm not sure they'll lose any effectiveness in the stack.
December 18th, 2008 at 8:07 PM ^
fair enough....
but if michigan is running a 3-3-5 come september and sucks, i dont want to hear any excuses that its the first year running the scheme, they dont have the right players etc etc.
December 18th, 2008 at 8:47 PM ^
And if they run it and succeed, you'll be back to say you were wrong? I mean it has to go both ways, right? You don't want to hear excuses if they suck; well I don't want silence from you if they are good, agreed?
December 18th, 2008 at 10:25 PM ^
haha. yes if the 3-3-5 turns out to be the best thing to happen to michigan football ill be back to say how much of a moron i was for questioning the infinte wisdom of rich
December 18th, 2008 at 10:40 PM ^
ill even one up that...ive been saying how 335 is going to hurt recruiting...well before shaffer quit everyone assumed will was coming to michigan. they all said it was a lock. well, if he does still end up coming, despite the fact hes probably gonna play in the 335 ill call myself a moron then too.
Comments