Carlos Hyde: Still Impressive When Using Advanced Metrics

By DJ Byrnes on March 5, 2014 at 1:04p

It didn't take a mathematician to spot Carlos Hyde's talent in his monster 2013 season, but Greg Peshek of crunched some numbers, and SURPRISE, the numbers back up what our eyes told us (for the most part). 

Here's how Carlos' numbers broke down compared to LSU's Jeremy Hill, Auburn's Tre Mason, Baylor's Lache Seastrunk and Washington's Bishop Sankey.

EXTRA YARDS. As Peshek explains, Extra Yd "represents an aggregate score combining yards after contact capped for 15 yards and broken tackles as a function of total carries. The score is capped at 100 and an average score is around a 65."

  Hyde Hill Mason Sankey Seastrunk
EXTRA YD 83.72 81.99 66.05 52.85 80.06
YARDS AFTER CONTACT 3.48 2.93 2.78 2.45 2.89
BROKEN TACKLE % 7.92% 8.59% 6.19% 4.56% 8.33%

Not bad numbers from Carlos, whose numbers back-up his status as a power-back.

DEFENSES FACED. Here's how Pashek broke down this category: "In charting each play I counted the number of blockers and personnel in the box at the beginning of each play. In doing so, we can create a ‘blocker differential’ or how many defenders there are compared to blockers. In the chart below, -1 represents one more defender in the box than blocker, 0 represents an even number of defenders and blockers, and so on. The values for each category are YPC. If there were not enough carries for the differential, they were left blank."

  Diff -2 Diff -1 Diff 0 Diff +1
Hyde 2.50 6.78 9.19 -
Hill - 7.27 7.16 4.44
Mason 3.27 6.49 4.99 7.87
Sankey 3.71 6.15 6.70 6.11
Seastrunk - 4.76 8.32 -

Interesting enough, the numbers say Carlos Hyde struggled against stacked boxes in short yardage situations. Peshek notes while Carlos is good in yards after contact, his struggles against stacked boxes "raises questions about which part of his game will translate to the NFL."

Carlos definitely saw more stacked boxes than Hill and Seastrunk, as there wasn't even enough data to plot their -2 differential. Interestingly, Sankey, whose yards after contact numbers were the worst of the bunch, posts good numbers here. So does Auburn's Tre Mason but he plays in an offense with a lot of window dressing/misdirection.

SHORT YARDAGE SITUATIONS. Peshek explains this category thusly: "3rd Down % is the rate at which the RB converted 3rd and short situations. Goal to go is the yards per carry in goal line situations."

  Hyde Hill  Mason Sankey Seastrunk
3rd Down % 75% 85.71% 92.31% 78.79% 70.00%
Goal-to-Go % 2.71 3.53 3.06 2.61 -

This set of numbers was the most shocking to me. I thought 75% on 3rd down was massive, until I saw Jeremy Hill posted a 85.71% clip and Mason posted a 92.31% clip.

This is where I would like to see the quantity involved in these numbers. How many times was Carlos Hyde ran in a 3rd down situation compared to Jeremy Hill and Tre Mason?

RUNS BY YARDS PER CARRY. Peshek says this category "shows the percentage of runs by yardage gained. It should give you a feel for whether a back was getting stopped short often or breaking off solid long runs regularly. It’s a bit less pronounced with this group, but in the second group we’ll see backs with much higher percentages of short runs."

  Hyde Hill Mason Sankey Seastrunk
< 4 Yards 36.14% 41.41% 41.69% 43.86% 40.38%
4-20 Yards 58.42% 53.54% 55.70% 48.77% 53.85%
20+ Yards 5.45% 5.05% 2.61% 7.37% 5.77%

Here, Carlos Hyde shows 1) he doesn't go down easily and 2) he has a surprising number of big yardage runs.

Now, these stats aren't perfect. One gripe I have is Carlos' three-game suspension. (He also only logged five carries against FAMU.)

Jeremy Hill was suspended for the LSU-TCU opener, but everyone else played a full 13 games. Carlos never got to play the likes of Wofford, Western Carolina, Idaho State or Kent State. He came in during conference play and was still the most consistent weapon the Buckeyes had all year.


Comments Show All Comments

Killer nuts's picture

My question is why was Jeremy Hill so much worse when his blockers outnumbered the defenders

Killer nuts's picture

And why Does Carlos never get to run with more blockers than defenders. I know the reason is that teams stack the box against us but it sure would be nice to have a threatening pass attack to get those guys out of the box

+2 HS
DJ Byrnes's picture

Peshek: "Hill was very consistent when faced with -1 and 0 differentials averaging around 7.2 YPC, but somehow dropped when he had an extra blocker. Based on the data, we’d expect it to go up but it’s likely an anomaly."

Can't read a block, perhaps? I thought that was pretty interesting as well.

Californian by birth, Marionaire by the Grace of President Warren G. Harding.

CALPOPPY's picture

Can't read a block, perhaps?

That was my thinking. I was wondering why Carlos was so good when the defenders were even. He obviously read the blockers well and hit the hole at the right time. So maybe Hill was not good in these situations.

I'm a hurtin' buckaroo.

d5k's picture

The sample size for the -2 and +1 categories must be tiny.  And the -2 category is likely dominated by goal line situations where your yardage is limited by the distance to the goal line.  +1 runs would only happen for these guys on weird situations like 3rd and 20.  Ohio State almost always faced 8 man boxes against their 11 personnel.

I_Run_The_Dave's picture

Yeah, I'd like to see these numbers alongside the defenses they were running against.  The B1G may have been bad last season, but they play defense.  The SEC did not play defense last year, and I'm not sure how the Big 12 compares to the B1G.  The ACC was a joke in that category.  The point about Carlos Hyde not getting to play against the cupcake part of the schedule is also valid -- lets redo these numbers with the 3 weakest defenses faced taken away for each (to make it 10 games each).

+1 HS
Buckabroad's picture

Great write-up. Damn am I going to miss El Guapo.

"The minute we stop expecting greatness, we become Wisconsin."

Buckeyeneer's picture

Is it because we run a spread that Hyde didn't have enough data for +1 blockers?

"Because the rules won't let you go for three." - Woody Hayes

THE Ohio State University

d5k's picture

+1 means that they are basically playing with 2 deep safeties without cheating off the slot receiver.  Against that type of front the read option would average 10 yards per play with Braxton/Hyde/Mewhort&Co.

+1 HS
TUNBUCK89's picture

You can crank the numbers in as many ways as you want. In the end, the numbers should NOT contradict what we saw with our own eyes. And what we saw from El guapo was simply.....GREAT!

Barnsey69's picture



Thank the Maker that I was born in Ohio, cradle of coaches, US Presidents, confederate-stomping Generals, and home of The Ohio State University Football Buckeyes- 2014 UNDISPUTED National Champions!

seafus26's picture

A stacked box in my opinion is tough to run against from the shotgun handing to a stationary back. Especially without a passing threat. I wanna see Carlos slamming into defenders once he gets to run downhill. The best of Carlos may still be to come

Go Bucks and michigan STILL SUCKS!

yrro's picture

I don't mind pulling out Hyde's stats, but reproducing the entire tables, even with cites, seems like we're treading the line on journalistic integrity a little bit. Everyone make sure you click on the rotoworld link. Dude charted a *lot* of games for that info.

+1 HS
Earle's picture

Would you say that he charted a plethora of games?

Valid point on clicking the rotoworld link.  I think I read somewhere that DJ does not claim to be a journalist.

Have you tried Not Your Father's Root Beer?  It tastes just like the real thing, but it packs a punch (5.9%ABV).  It's a little sweet for me though.  Two is my limit.

+2 HS
AndyVance's picture

As perhaps the only non-staff journalist on the board who isn't an anonymous name and icon, I'll play self-appointed ombudsman for a minute to address you point, which I think is a fair question.

In this case, DJ provided two separate links in the opening graph of the post, providing a reference back to the author and the source article. Furthermore, he correctly attributed quotes and paraphrased material from the source article, so there was no question about what was Peshek's work and what was 11w's take on the data and information presented.

Working at a trade journal that reports on A LOT of scientific literature, I'll say that this approach is not at all uncommon. An academic will write a paper, and a journalist will read it, analyze it, and present it in a way that their audience can understand and interpret. The reader can then go back to the source material for more information as they like. That's pretty much what DJ did here, and if I were his editor, I'd say he passed the proverbial "smell test" here.

What does drive me crazy, however, is when members of the community get sloppy with cutting and pasting large chunks of material from other sites into forum threads, or worse yet, attempting to cut and paste whole articles from paid sites. Most of the time the staff or mods put the kabosh on that, but I see it happen from time to time and it drives me nuts.

+7 HS
Go1Bucks's picture

Here Here!  Thank you AV for pointing out this extremely important truth.  While appreciating the good work done on 11W, one must always remember the rules to write by.  The Staff here knows it's business (as do you AV).

Go Bucks!

+1 HS
AndyVance's picture

Thanks! I think YRRO's question was a good one, and more than anything I appreciate that he had the right reason for asking the question: respecting the work of another author, who obviously put a ton of work into those figures.

+2 HS
yrro's picture

Thanks for the reply and analysis, Andy.

Yeah, I almost didn't didn't say anything. I think DJ did an excellent job of giving full credit to his source, and of making clear when he was using quotations and when he wasn't. The inclusion of pretty much every graph from the original article was what got me, as that was the meat of the effort involved in the original post. I certainly don't think it was to the point where anyone would feel they have a right to force the article off of the site or anything, though.

I will bow out to your greater experience in the area.

+1 HS
AndyVance's picture

I respect you all the more for asking the question. As a reporter, there's a fine line to walk in sourcing another writer's work, and then as a writer of my own original stuff, I've been frustrated before by less scrupulous writers lifting too liberally without sufficient attribution.

I doubt I'm anywhere near an authority on the subject, but I felt like your question deserved a thoughtful answer (before anyone decided to rain downvotes upon you for having the temerity to speak up... 'cause that happens, unfortunately).

+1 HS
Go1Bucks's picture

I neglected to thank YRRO earlier for his question as well.  Thanks YRRO.

Go Bucks!

2002osubuck's picture

How about all the times he converted 1st down on 2nd and 3rd downs with 7-10yds to go. Hyde definitely got a lot of yards on 1st and 2nd down carries. Also, Hyde didn't have a negative carry until the last few games. 

Seattle Linga's picture

He tires defenses out in the 4th quarter of every game he plays

+1 HS
Ericgobucks's picture

I am so happy for him and wish him the best, but it's going to be a little sad not to see him line up and blast guys in the second level. I hope you get paid well, Carlos. 

scrotieMcboogerballs's picture

Who will you miss more?:

A. El Guapo

B. Crafty

"Who do you think you are?!?!...I am!!"  -Pete Weber, bowling legend and dominator of the English language