forums

Upon Further Review: Manny Diaz as DC

View as article
JohnZarrilli said:
Lance, I'm interested to hear more of your thoughts on why you think YPP is a useful metric to measure Miami's defensive performance under Manny. I know sports stats guy think YPP, as you said, does a good job at capturing overall performance and success, but I question its applicability to Manny's style of defense.

My eyes tell me that Manny does a better than average job at fielding a D that creates a lot of negative plays, which can definitely skew the YPP stat. My eyes are also telling me that our D played a lot of off-ball zone coverage (or gave significant cushion in man) and allowed QBs to dink and dunk down the field. Both would tend to make our YPP appear better than our actual defensive performance.

Just so other posters can follow along. A tfl on first down for -2 yards followed by a 6 yard gain and another 6 yard gain would average out to around 3.33 ypp, which looks great on paper, but ultimately we still gave up the first down. Iterate that over a whole season and YPP maybe doesn't do a great job at capturing whether the opposing offense is sustaining drives and scoring points.

You're someone who's much more versed in the world of sports stats, but I would speculate that a stat that takes into account scoring defense (defined here as points per drive) and adjusts for a whole host of variables like level of competition and field position might not be as favorable to Manny as the YPP stat. Of course, as you pointed out in the original post, that requires a lot more work and is not as easily calculable.

Thoughts?
Click to expand...
You bring up some great points. Yards per possession would seem to be a more useful stat. Yards per down set also.
 
Advertisement
Lance Roffers said:
Why? They’re a subset of the same thing.
Click to expand...
they’re not. you can have different numbers of plays per possession. same for down set.

also - bad third downs extend possessions and add yards. you can look great on first and second but if you don’t get it done on third, you’re going to struggle as a D. YPP on third isn’t the right Q because distance can vary - third down % is already tracked and obviously matters.
 
Last edited:
Ethnicsands said:
they’re not. you can have different numbers of plays per possession. same for down set.

also - bad third downs extend possessions and add yards. you can look great on first and second but if you don’t get it done on third, you’re going to struggle as a D. YPP on third isn’t the right Q because distance can vary - third down % is already tracked and obviously matters.
Click to expand...

If you want an all encompassing look, check out SP+. It takes pretty much everything into account.

Miami was 11th in 2016
16th in 2017
14th in 2018
 
Advertisement
Lance Roffers said:
It lies dramatically. The best third down for an offense? The one it doesn’t face.

Is a defense good because it gives up a bunch of first downs before it even reaches third down? Of course not. Is a defense great if it allows 0-3 on third down in a game? What if it was giving up giant chunks of yardage on 1st or 2nd down and never even getting to a 3rd down? What if it was giving up consistent 3rd & 1’s? You aren’t stopping those often.

This fascination with third down statistics is baffling to me. You’re reducing your sample. You’re increasing your variance. You’re decreasing your stickiness. You’re actually introducing the need for more context.

Yards Per Play does a far better job of profiling a defense than 3rd down data does.
Click to expand...

Success rate pretty much nails it all.
 
HighSeas said:
When Diaz was DC there was one game where his scheme and gameplan was a huge letdown. 2017 ACCCG I posted a preview breakdown in advance of the game and I legitimately thought the defense would have a lot of success. Clemson offense had a journeyman OLine, Kelly Bryant at QB, Fr Etienne and a JAG upperclassman splitting the backfield, and at WR they had all gadget and slot types. Clemson had a very predictable, limited offense with short passing and the same 2-3 QB run schemes on short yardage and red zone. Diaz called a very passive, soft spot drop zone kind of game, Clemson had some early success, got some lucky breaks and snowballed.

Since Baker arrived as DC pretty much every game has followed that blueprint (passive, soft, tons of zone).
Click to expand...
BC '18.
First 21 minutes of the game BC ran 41 plays for 244 yards, 14 first down, and scored 17 points.
 
Advertisement
NJshoreCane said:
BC '18.
First 21 minutes of the game BC ran 41 plays for 244 yards, 14 first down, and scored 17 points.
Click to expand...
boston college football GIF


I mean...what can you do?
 
Ethnicsands said:
they’re not. you can have different numbers of plays per possession. same for down set.

also - bad third downs extend possessions and add yards. you can look great on first and second but if you don’t get it done on third, you’re going to struggle as a D. YPP on third isn’t the right Q because distance can vary - third down % is already tracked and obviously matters.
Click to expand...
OrangeBowlMagic said:
Success rate pretty much nails it all.
Click to expand...
But it doesn’t because it weights success the same.
 
Ethnicsands said:
they’re not. you can have different numbers of plays per possession. same for down set.

also - bad third downs extend possessions and add yards. you can look great on first and second but if you don’t get it done on third, you’re going to struggle as a D. YPP on third isn’t the right Q because distance can vary - third down % is already tracked and obviously matters.
Click to expand...
They are entwined with one another and YPP > YPD without question. What you’re really looking for here is percentage of possible yards achieved per drive.

Other than to follow-up on some data questions asked, this will be my last reply on this topic.
 
Rellyrell said:
And this is where the nail meets the hammer.

So I was told what 2018 showed was that Manny’s defense made bad offenses ever worst b/c when facing us, they were even below their season avg. lol.

OK, so let’s put it like this;
LBJ has a pick up basketball game against me. Sure, I’m decent in basketball, but I’m no where near his height, athleticism, or professional skill set. If he went on Twitter & bragged about holding me to 1 point, what purpose does that serve? Clearly he’s a superior athlete than I am. (Although I think I could get at least 4 points, lol)

The point? If Miami is trotting out 5 & 4 star players against teams featuring 3 star & UR filled teams, then what’s the point of bragging or pumping out ur chest? What we need to see is how does Diaz’s defense do against competent teams. In 2018, we played a whole slew of incompetent teams, which is why at the end of the season, we had a bottom third S.O.S out of all P5 teams.

Stats can be skewed anyway u want. Ppl say LBJ is the greatest of all time b/c of what he’s accomplished, while others say well let’s look at context, the era he’s playing in, the rule changes, how many games it took him to accomplish this vs. player X, etc. etc.

It leaves it open for interpretation & debate, but using one metrics & saying this is y Diaz is borderline elite is b.s, imo. Again, per “stats”, We were a top 15 defense in the nation, but anyone w/ eyes could see we weren’t, and when u peel back the layers of that ‘14 season, and look at other metrics where we performed poor in, it was clear our eyes were right.

I like to take a gander at both. Again, I think Diaz is a good DC; in no way shape or form and I’m saying he’s trash, but his philosophy, once caught on to, dips in efficiency. The reason why we can only use UT as a gage is b/c he bounces around after a yr at every stop (which, if imma be honest, is suspect af).
Click to expand...

In '18 vs the best teams we played:
Only MSU and Bama held LSU to fewer yards. They played UGA, uf, and Auburn
No one held UVA to fewer yards
Pitt played ND, Penn State, UCF and Clemson that year. Only Clemson ( by 1 yard) did better.
Wisconsin was right in the middle.

The only way to compare is to look how other teams did against our opponents.
The same way scores are looked at.

I do have my concerns with Manny's Defenses.
I can also say the '18 defense was good.
 
Last edited:
Advertisement
Lance Roffers said:
But it doesn’t because it weights success the same.
Click to expand...
Football doesn’t have a control variable other than W/L, and that itself is a function of D, O, team, situation, opponents. So all these metrics give you some insight but they do not answer the question and folks who want them to are just staring too hard into the crystal ball. There are different things a DC is responsible for, and some of it changes based on O and opponent (and obviously game situation and personnel).

The most meaningful stat is obviously PPG. But opponents vary, rules change, talent match-ups change. So you back off that and you’re trying to look at what went into it. How’d we do against the pass vs run? How’d we do on third downs? How’d we do in the redzone? Turnovers. Negative plays. Possession time. All somewhat relevant. YPP? Meh. It’s a smoothie. If you want to sum it all up, why not the one that matters - ppg?

If you’re trying to assess a DC for real, not just read stats, you’d presumably focus on game planning and adjustments as important tells (in addition to the base metrics). You can think about some different ways to measure them. First quarters and third quarters, perhaps. Also, focus on the 4 tough games - the rest shouldn’t matter.
 
Lance Roffers said:
They are entwined with one another and YPP > YPD without question. What you’re really looking for here is percentage of possible yards achieved per drive.

Other than to follow-up on some data questions asked, this will be my last reply on this topic.
Click to expand...
That cannot be right ‘without question,’ because it’s incredibly easy to show how you can have a lower yards per play and get gashed and someone else has a higher yards per play and stops drives. Third downs mean a lot.

Set 1: 15 yard first down, 3 yard run, 5 yard pass completion, 1 yard run, punt. 4 plays, 24 yards, 6 YPP, 1 first down, punt. Yards per possession: 24.

Set 2: 4 yard run, 4 yard run, 4 yard run.... and on and on ... TD. Lots of plays, 4 YPP, lots of first downs, TD. Yards per possession: all of them.
 
Advertisement
OrangeBowlMagic said:
What do you mean by that? Like it’s not opponent adjusted?
Click to expand...
I had asked about it because of the outliers that result from having one long run figured into YPP.

Lance pointed out that success rate has a similar problem because it weighs a 3 yard gain on 3rd and 2 the same as a 75 yard run on 3rd and 2.

The only issue I can see with it is that a 4 yard gain on 1st down followed by a 3 yard gain on 2nd down followed by a 4 yard gain on 3rd down only has a success rate of 33%, but that fails to show the impact of gaining a fresh set of downs.

www.canesinsight.com

Upon Further Review: Manny Diaz as DC

I judge Manny by the eye test. How does Miami look against good offenses? Not very good. On the other side Miami usually looks good defensively against the weaker teams (Duke last year) padding statistics making the defense look statistically better. The attacking style works against less...
www.canesinsight.com www.canesinsight.com
 
We'll see what Manny is about this year for sure because the DEs and LBs are average at best. Where will the havoc plays be coming from?
 
Advertisement
Back
Top