Tackling Technique: UConn, ND, MSU, Iowa

Submitted by maizedandconfused on

 

This season has had moments of extremely heated debate, regarding the performance of the defense. The schism in views is based on two views:

1. GERG is a bad coach 

2. Our talent level is severely depleted

I decided to look at tackling technique of our defense, more specifically on the bad technique aspect of it. This particular point of emphasis is almost consistently attributed to coaching, as good technique usually only transfers to the game field when it is instilled upon repeated coaching and re-coaching in situations where poor technique is used. 

This season, time and again, we have the ball player wrapped up, and then 3 yards later they have picked up the 1st down on 3rd and forever. I thought it was about time to look at tackling, both who is making/missing them and why.

About the analysis:

1. I regarded made tackles as anytime a player made contact with the ballcarrier and that ballcarrier ended up down by contact using solid tackling technique. I did not look at technique of made tackles, as after 1/2 of tape review most of the tackles that were made were of good technique. Any tackle that forced the ball-carrier out of bounds was counted as a made tackle.

2. I designated missed tackles as failing to make a tackle in space, taking an extremely bad angle on a tackle that should have been made, or simply just getting the hit but not bringing down the ballcarrier.

4. Bad Form takes into account any missed tackle that used any of the following:

  •        Head on the upfield side
  •        Arm/Jersey tackling
  •        Any hit at or above the numbers
  •        Getting "shook" in open field due to not breaking down or overpursuit.

The difference here is missed tackles in my mind sometimes come from being literally overpowered or stiff-armed, not a technique avenue. If the UM defender made the hit with the head on the right side and attempted to wrap up but the ballcarrier just slipped through, I counted this as a missed tackle only. If, however, the same play happened with the UM defender coming in high and behind, this was counted as both a missed and bad form tackle.

To derive the "Tackling Efficiency" I used the following formulas as necessary (feel free to critique or ask me to add things, as I am not a math whiz)

  •  Player Tackling Technique: (Made Tackles - Missed Tackles)/(Bad Form Missed Tackles+1)
  • Team Tackling Technique: (Made tackles - (Missed Tackles+ TDs))/(Bad Form Missed Tackles)

Results:

Individual Players Tackling Efficiency:

  UConn    Notre Dame MSU    IOWA    Season Ave
Roh 0.66 1 0.5 0.5 0.66
Banks 1 -1.5 -0.33 3.5 0.66
Sageese 2 0 -0.5 0 0.375
Martin 0.25 2 2.5 N 1.58
Van Bergen 2.5 -0.66 1 3 1.45
Demens N N 1 6 3.5
Mouton 6.5 0.5 2.25 0.13 2.35
Gordon #4 0.25 6 -0.1 0.33 1.62
Gordon #15 0 2.5 1.75 1.5 1.91
Rogers 0 2.5 -0.33 0 0.54
Kovacs 1.25 4 3.5 6 3.70
Ezeh 0.84 -0.5 0.5 -0.25 0.15
Floyd 3 0 2.33 0.63 1.49
Talbot N 1 -0.17 1 0.61
Black N N -0.5 -1 -0.75
Johnson 3.5 N N 1 2.25
Patterson N N N -0.5 -0.5
Herron 0 1 N N 0.5
Christian N 1 N N 1

* N denotes games where player did not participate or had no made or missed tackles

A few notes from the individual analysis:

  • Kovacs is by far our best tackler, with Mouton, Johnson and surprisingly Gordon #15 as the 2,3,4 respectively. 
  • Of the guys on this list, the biggest suprise was Martin, however I think it is important to remeber that this tackling eff. calculation puts weight on total tackles made, and Martin gets doubled. A lot. (in review of my stats, Martin only missed 3 tackles total, with 2 by bad form)
  • In the two games Demens played, he was an absolute tackling machine.
  • Ezeh really cannot tackle.

Team Tackling Technique:

  UConn Notre Dame MSU     Iowa Season Ave.
UM 2.78 1.28 1.14 1.83 1.75
Opp. 3.21 2.17 7.4 5.7 4.62


We were much better at tackling UConn in space than anyone else, not suprisingly this was the team with the least athletic roster.Notes from the team portion:

  • MSU was an absolutely terrible day, with almost a missed tackle per made tackle, and lots of bad form tackles all over
  • We are consistently worse at tackling with good technique than all 4 of the opponents analyzed, two B10 "good" defenses and 2 middle of the pack to bad defenses
  • On the season for these 4 games we are averaging about 2 missed tackles with 1 missed tackle by bad form every 3 tackles made

Conclusions:

These statistics correlate with what I have been seeing every week. Mouton is great at times and inconsistent at times, and Kovacs is a stud at bringing down the ball carrier. Ezeh was a sub-par MLB and Martin gets doubled. I have high hopes for Demens, Johnson and Gordon (both of them).

Surprises for me came with the discovery of Gordon (not Cam), that dude was all over the place when he was in/not getting blown off the ball by a OT. Secondly, I think Floyd has the potential (if he develops his man coverage) to be an absolute terror on the corner. He can and will tackle in space. 

As a team, I think the debate as to what this attributes to is still open for debate. Does this attribute to talent? Possibly. Is this more attributable to the coaching/GERG? Yes. However, with the limited bodies he has on his side of the ball, it might be due to the fact that live tackling is just not an option with this lineup. 
Regardless of the cause, the tackling is terrible, and is apparently getting worse through the season. This attributes in my mind to the coaching, specifically the D-Coordinator. 

I can only hope the tackling technique is going to get better, because as it stands through these 4 games it cannot get much worse. I will do my best to analyze Indiana, PSU and Illinois when I get some more time, to give a more all encompassing picture of our D (takes a significantly longer time because I have to watch the whole game). 

Comments

profitgoblue

November 10th, 2010 at 4:09 PM ^

The most interesting/surprising piece of the analysis is Roh.  I wonder if his weak numbers have been caused by switching between DE and LB or if his form needs some work . . .

QVIST

November 10th, 2010 at 4:19 PM ^

Very informative. I do recommend we get J.T. Floyd in the lineup ASAP as I do not think Flyod can keep the position locked down for much longer.

caup

November 10th, 2010 at 4:19 PM ^

You're not using the correct input parameters.  Sure, Kovacs and Demens are usually sound tacklers.  But you are punishing the DLs unfairly.

maizedandconfused

November 10th, 2010 at 4:26 PM ^

analysis was strictly based on tackling. I'm not punishing the DL unfairly, they don't make a lot of tackles. That being said, Banks and Sageese missed A LOT of tackles in the four games (9 total with 7 with bad form) and Roh, surprisingly, missed more than a few (6/2). 

So while I understand your consternation about the DL, they have missed their fair share of opportunities to tackle.

And, I already explained the Martin thing (he gets doubled.. a lot). 

However, if you have some suggestions for altering to the correct input parameters I am all ears. 

qed

November 10th, 2010 at 4:36 PM ^

It would be nice for there to be a maximum value like 1 or 100.  I am not sure how much better Kovacs can be ... but I also might be misreading your information.

maizedandconfused

November 10th, 2010 at 4:39 PM ^

The NCAA record for tackles in a game is 34 (Mike Singletary), so potentially it could be 34 as the high.

But in all seriousness, as a player 

+5 is Ray Lewis re-incarnate

+3 is a very solid tackler

+1.5 is a good to decent tackler 

<1 is sub-par with more missed than made. 

Blue in Seattle

November 10th, 2010 at 4:40 PM ^

Is it possible that the time that would normally be devoted to tackling technique is being used up with trying to train a player on their new position in the defense?

It would be interesting to see how the scores came out based on the class experience of the players in question, e.g. how to the freshman compare to the seniors?

also as a way to analyze GERG (although it seems like this kind of analysis is way too time consuming) can we compare last year to this year?

A final factor that I would say indicates GERG is a good teacher or at least sway to the talent is all the young guys that seem to be at the top.  Demens, Kovacs, Gordan #15.  And especially Kovacs, who I've never heard connected with any comment as to his athletic ability being an asset.  So who is teaching Kovacs?  How is a "walk on" ending up being a tackling stud?

To me it seems like Obi just doesn't have the instinct/intellect to be trainable, Mouton has ability but also untrainable in that his bad habits never go away, thus half the time his athletic ability is enough, and half the time it doesn't fix his bad decision (like the sack on the last illinois play, Mouton definitely turned on the jets to get to the QB before Roh, and I think Cam Gordon, but why the hell didn't he keep the right outer edge contain on the QB when he had two teammates to his left???

I think the overall poor play of Obi and Mouton have lead to a lot of other decisions the coaches have had to make that were not optimal under normal conditions. 

Do we honestly think that the coaches were surprised when Roh begged them to stop being an LB once Demens showed he could play?

And how do you end up teaching Demens but not Obi?  how could that be on the coaches?

I think it all comes down to a lack of upper classman starters.  All this inexperience would be covered up on the 2 deep, and in fact is on every other Big Ten team.

 

Undefeated dre…

November 10th, 2010 at 4:56 PM ^

Wow, this is great. Clear you put a lot of time into this, coding every tackle, etc. Nice work!

What are your thoughts on these metrics? For this, I'm assuming that Bad Form Tackles are a subset of all Missed Tackles.

Missed Tackle % = Missed Tackles  / (Made Tackles + Missed Tackles)

Bad Form Tackle % = Bad Form Tackles / (Made Tackles + Missed Tackles)

Bad Form-Induced Missed Tackles = Bad Form Tackles / Missed Tackles

The only weird thing is that we'd want smaller, not bigger, numbers here. But I think this does a better job of normalizing for the number of tackle attempts.

We could try to add in the TD's, but if we have no one in position to tackle on a TD, or it's a pass in the end zone, we shouldn't hurt our tackle percentage. So we should only count TD's if there was a missed tackle, and the formulas above already do that.

Again, great work!

ESNY

November 10th, 2010 at 5:24 PM ^

Exactly.  Should be a ratio or a percentage so the defenders can be compared across positions.  

Your goal is how often a player who is in position makes a good tackle vs. bad form vs. missed tackle. 

The denominator should be Tackle Opportunities and the numerator will be either Made Tackles or Missed Tackles.  And then you can dive further into the cause for the Missed Tackles (form, overpowered, angles, etc.)

ESNY

November 10th, 2010 at 5:40 PM ^

I think its clear, perhaps its my post that isn't.

I think the two key stats are:

Good Tackle % = Made Tackles / Tackle Opportunities (tackles + missed tackles (where ball carrier was touched) + missed tackles (total whiffs or overplayed the ball, etc.)

Missed Tackle % = Missed Tackles / Tackle Opportunties

And then further delve into the cause of the missed tackles:

Bad Form

Poor Angles

Overpowered

Out of position

 

maizedandconfused

November 10th, 2010 at 7:46 PM ^

The problem with this particular deal...

Is that if fails to accentuate BAD FORM missed tackles..

I put the formula together as you stated, and all this did was basically rank the players by made tackles.. the formula I used compounded all the input you want, and penalized players in a 1:1 ratio for missed tackles, and further penalized them in non-linear fashion for bad form missed tackles. 

Undefeated dre…

November 12th, 2010 at 11:23 AM ^

I care about bad form missed tackles. But I also care about missed tackles, straight up. If you want to know how many missed tackles are due to bad form missed tackles, then we just go with (bad form misses) / (total misses).

I know you want to create a single metric to rule them all. You'd like to emphasize bad form missed tackles. I think there's a general preference that doesn't penalize/reward based on tackle attempts, so that every perfect tackler, whether they had 1 attempt or 10 attempts, would have the same grade. So how about something like this:

[(total missed tackles) / (total made + total missed)]*(bad form missed / total missed)? But then that just simplifies to the second suggested formula, which is:
(bad form missed) / (total made + total missed). If you're trying to get a sense of what percent of tackle attempts are bad form misses, this is straightforward.

If the data's in Excel, I'd be happy to play around with it. I'd never do anything with it publicly; I'd just PM you what I found and you can post as you see fit.

jim48315

November 10th, 2010 at 6:35 PM ^

Nice post.  Too many kids got away with bad form in high school against inferior athletes.  Too many of them also like to do what they feel are highlight reel moves.  Good tackling requires contact with the chest or shoulders and driving with the legs.  Too many kids leave their feet, either to "launch" themselves or drop to their knees in an effort to get low and get leverage.  Too many times coaches assume the kids can tackle.  And in 2007 Lloyd assumed Mallet could take a snap from under center.  And in the Indiana game, RR assumed Denard could take a snap from under center.

But with low scholarship limits, it is hard to get practice hitting, and it is very hard to find out who the tacklers are when there is no hitting.

greenphoenix

November 10th, 2010 at 7:57 PM ^

Regardless of the cause, the tackling is terrible, and is apparently getting worse through the season. This attributes in my mind to the coaching, specifically the D-Coordinator.

So, the two big ten teams here, Iowa had better tackling than MSU, so isn't that a potential improvement? Agreed the tackling against PSU was awful, but it looked pretty good against Illinois.

I think I'd look at the tackling in the big ten separately from tackling against other opponents. This is a good, good conference.

It's still bad, but my hunch is that the team efficiency against illinois will be the best of any B10 game so far.

Search4Meaning

November 10th, 2010 at 8:03 PM ^

It is not perfect, but then no one really expected it to be.  I understand your points on the DL, but I still believe that your system is slanted away from them - but that's OK.

As a starting point to confirm suspicions, this rocks.

maizedandconfused

November 11th, 2010 at 1:07 AM ^

I tossed in a +1 for every tackle for loss, that could offset the slant away from the DL.

 

Alternatively, DL very rarely make more than 5 to 6 tackles a game. Just being located where they are, a tackle from then is almost always a tackle for loss or small gain. If your DL is leading your team in tackling, then your defense is ridiculous.

Taking a look at NFL leaders in tackling, there isn't one DL in the top 40.. all S and LBs there..

greenphoenix

November 11th, 2010 at 6:31 AM ^

I think your formula is looking at the right things; I wouldn't include TFL becaause it brings a lot of other factors in (scheme, blown up plays, etc.), and I think your point about who "gets" tackles on most teams is dead on. I do think, however, that we can create a formula that tries to get to pure tackling technique that includes even low frequency tacklers like the line.

Here's the thing: The total number of made tackles is in the numerator, so that means that even a poorly performing tackler may have a chance to get a bigger number just from the merit of how many total tackles they make. So why not make that your denominator baseline? Then your formula gets super easy:

Made tackles / (total tackle attempts-missed tackles).

"missed" tackles in this case would only refer to those tackles that didn't include bad form.

A ninja tackler in this model gets a high score of 1, regardless of whether they made twenty tackles or two; a completely hapless tackler ges a score of 0. The "poor technique" tackles don't even have to be counted, because they are implicit in the lower/higher number of made tackles.

So, a DL who has this kind of day:

4 attempts

2 missed tackles

1 made tackle

gets a 0.5. (1/(4-2))

An Ezeh-like linebacker who has this kind of day:

17 attempts

5 missed tackles

4 made tacles

gets a 0.33  (4/(17-5)).

So, in this case a DL player with a small number of tackles would have a better day than a busy linebacker with poor form.

What do you think?

maizedandconfused

November 11th, 2010 at 9:37 AM ^

But i feel like it doesnt address the underlying issue. The total attempts is always equal to made tackles + missed tackles, and bad form is just a percentage of missed tackles (every missed tackle is just a missed or a missed with bad form). Think I could change it to get down to the 0-1 scale, I like that

maybe 

Made tackles/ (total attemps x bad forms)

For yuour Ezeh youd get something around 0.22

and for your DL you would get your 0.5 (assuming 1 bad form for every 2 missed tackles which is about right for most players)

greenphoenix

November 11th, 2010 at 9:48 AM ^

Yeah I totally see your point about how a missed tackle can be good or bad. One way to address this is to explicitly separate "good" missed tackles from "bad" ones, and in my described formula change it to:

Made tackles / (total tackle attempts-missed "good" tackles)

This formula you propose also gets at this:

Made tackles/ (total attemps x bad forms)

But the multiplier in the denominator amplifies the error rate, which you're probably not trying to do.

What about switching it around a little bit:

(Made tackles - bad forms) / total attempts

A guy with a 1.0 in that model would have had a remarkable day, an astonishing day, because he would have nailed every open field tackle that was missable as well as having perfect technique. It would be hard to get a 1.0; even guys with a really good game would not get there if they missed any open field tackles.

Which is okay, right? But it does then combine technique issues with things like scheme, speed, size differences, etc.

I want to make sure I understand your underlying reasoning. I just love the analysis you've done here and would love to help nail down some of the position related "slips".

Not sure if this violates protocol, but maybe we have an IM discussion about it or a phone call? feel free to email me at [email protected]

bluenyc

November 11th, 2010 at 9:35 AM ^

Thanks for the info.  I thought I remember it said, not sure by whom, that GERG was a great teacher.  I would think that great teacher meant that he was good at teaching tackling.  No idea, just wondering what others heard or thought. 

As a poster said above, maybe alot of time is used to teach players who are moving around and tackling may get less time.  I have an indirect question in relation to tackling, with the time we are going to lose because of the violations, will tackling technique get worse?  How much will the time lost effect our defense in terms of fundamentals?