SB Nation DC Editor's Pick
A Sabermetric Evaluation of Redskins Big 3 Running Backs
Sabermetrics is the art of using advanced, infrequently used statistics to evaluate players. They first became popular in baseball, but have gradually shifted to use in all of the Big 4 sports.
Baseball has WAR, hockey has Corsi, and basketball has Effective Field Goal %. But what about football? Christopher Harris over at ESPN recently wrote an article advocating for the usage of advanced statistics for football. He does a good analysis of the top guys, but I wanted to get a deeper look into our RB situation. So jump down to check out whether or not our aging group of running backs can carry the load.
A quick primer. The stats I am going to use are YCo, YCo/Att, YCo%, MT and Att/MT. YCo is how many yards the player gains after first contact. YCo%, arguable the more important statistic, is the percentage of a players total yards that he gained after first contact. MT is how many tackles the player "missed", which means either broken or avoided tackles.
|
Name |
Snaps |
Att |
Yards |
Y/Att |
YCo |
YCo/Att |
YCo% |
MT |
MT/Att |
|
273 |
124 |
502 |
4.1 |
357 |
2.9 |
71.12% |
6 |
.05 |
|
|
Larry Johnson (KC) |
328 |
132 |
380 |
2.9 |
294 |
2.2 |
77.37% |
5 |
.04 |
|
Larry Johnson (CIN) |
80 |
46 |
204 |
4.4 |
142 |
3.1 |
69.61% |
6 |
.13 |
|
207 |
98 |
389 |
4.0 |
242 |
2.5 |
62.21% |
7 |
.07 |
So what does this all mean? A few things, actually. Lets go player by player...
- Clinton Portis had a career low in yards, attempts and touchdowns, and tied his career low for games played. But it looks like a lot of this wasn't his fault. He had a fairly average 4.1 yards per attempt, but had an extremely high 2.9 yards gained after first contact. That means that he was on average touched by a defensive player after only 1.2 yards past the line of scrimmage. Portis, who thrives on beating out his opponents and avoiding contact, wasn't given any room to work.
- Once again, the stats show that Portis was hit early and often. He gained 71% of his yards after being hit. What this shows is promise. When he was healthy, Portis knew he was going to get hit hard, but kept pushing. Its a good sign from #26 and I expect it to continue this season.
- Remember when I said he thrived on avoiding contact? Yeah, I guess not. With only a measly 6 tackles avoided, you'd be lucky to see Portis avoid a tackle even once a game. He would only miss one every 21 attempts.
- Larry Johnson's 2009 was clearly a tale of two seasons. Playing in Kansas City, he was on the field for almost every snap. And, like Portis, he was hit. A lot. Johnson only managed to get 0.7 yards past the line of scrimmage before first contact. Because of this, he has the highest YCo/Att and YCo%. Johnson is clearly a between the tackles runner who pounds through the line to gain yards. But he needs more help than that to succeed. He also missed a ridiculously low 5 tackles, missing one every 27 rushes.
- When Johnson was cut and picked up by the Bengals, everything started to go right. Clearly Cincy is better than KC in every way possible, (as proof by the O-Line giving him an extra 0.6 yards before contact) but Johnson showed a huge spike. His Y/Att jumped 1.5 yards, and the % of yards gained after first contact dropped 8 percent. He also missed a tackle every 8 attempts, a really promising number.
- The key stat here: In 7 games with KC, 132 attempts. In 7 games with Cincy, 46 attempts. When he isn't getting pounded 20 times a game, he is successful. He is a super talented running back who has been run into the ground in KC. He needs less carries to succeed.
- Fast Willie Parker got some really good help from his O-Line, getting 1.5 yards past the line of scrimmage before first contact. But once he was hit, that was about it for him. He only gained 2.5 yards on average after that contact, so only 62% of his yards came after contact. He got his nickname for a reason. At 5'10, 210, Parker relies on beating out the defenders, not going through them. And for the most part, this worked. With one missed tackle every 14 snaps, Willie was pretty average at getting around the defenders.
What to Expect this Season
Basically, these stats are pretty positive. First and foremost, in order for the run game to succeed, our Offensive Line needs to play well. Getting a good push and making sure that our guys can gain speed before getting hit might be the most important part of our running game. Next, I am really excited about LJ. He isn't going to get the majority of the carries, but when he touches the football he should really help this team. Also, this just proves that we cant have one back taking all the carries. We need to be a RBBC team to succeed. Next, Parker can be a very important asset to this team if hes healthy. I think that we will keep all three guys and Shanahan will go into the season ready to play off each of their strengths. And last but not least, something that the stats cant show, is how hard these guys have worked this off-season and how important the sense of competition is. They are coming in fit and ready to fight, which always bring out the best in people. Do I think our rushing attack is going to dominate teams? No. But I do think it will keep defenses on their toes and give #5 the chance to succeed in this system and in DC.
48 comments
|
8 recs |
Do you like this story?
Comments
Where's Jefferson
He will love this post
Don't ever let failure get you down. Remember, a very successful major league hitter failes 70% of the time.
was thinking about Jefferson...
before I even finished the first sentence…..good post amkcaps, definitely an interesting look
Great analysis
Rec’d
I'm worried about the beer supply. After this case, and the other case, there's only one case left- Barney Gumble
by CptChaosSidekick on Jul 24, 2010 10:33 AM EDT reply actions
Don't forget Ryan Torain...
Ryan have always been Shanahan’s favorite RB. If he’s healthy, he will be number 1 or 2 for Shanahan’s offense. He is exactly the type of runner that is required for the zone-blocking that Shanahan loves.
I see Portis as a one cut and go type guy, similar to Torain, therefore it’s going to be up to Torain or Portis as the starter.
This is how I see it.
1. Portis – Every down guy
2. Torain – Come in for 12-15 carries
3. Johnson – For short yardage
4. Parker – Inactive
by Negative Nelly? on Jul 24, 2010 1:48 PM EDT reply actions
So if Torain comes in for 12-15 carries and LJ is used in short yardage….
How many carries are you expecting Portis to get as the (starter) every down back? There doesn’t seem to be many carries left. Unless you are expecting the ’Skins to rush for 40 times a game.
The Once and Future King
Rushing = 55%
Portis = 20 carries
Torain = 13 carries
LJ = 3 carries
by Negative Nelly? on Jul 26, 2010 8:49 AM EDT up reply actions
Thirty-six rushes per game would have been 2nd in the NFL last year for a team; 20 a game would have been 5th among individual players. Factor in Portis’ mileage and Shanahan’s history with running backs, and I can’t see it shaking out like that.
The Jets were a big running team.
The Jets ran 59% of the time out of 1270 plays for all games and ran 58% out of 1082 plays in the regular season.
They averaged 39 running plays per game in the regular season. Your analysis looks good, yet 40 running plays a game is not unreasonable.
by Jefferson1935 on Aug 3, 2010 1:06 PM EDT up reply actions
As far as I can tell
it’s all from last year.
Insert something witty here.
by Knubles and Bits on Jul 24, 2010 3:59 PM EDT up reply actions
Just checked.
From last year.
Seems like a bit of a small sample size to draw a lot of conclusions on. Especially considering the sad state of our offense last year.
I think Johnson will thrive in this offense
by only getting 7-12 carries. He will be the perfect change of pace and I expect him to be as successful as one could be with so few carries.
Insert something witty here.
by Knubles and Bits on Jul 24, 2010 3:59 PM EDT reply actions
Football Outsiders attempts to figure out the offensive line's contribution versus that of running backs.
http://www.footballoutsiders.com/stats/ol
You might try to make a comparison with your source which looks like something I’ve seen on FO.
I like Portis till he goes down with an injury
SBNation.com
watchkalibrun.com
Poyznus.com
Hoya Saxa
by David St. Martin on Jul 24, 2010 8:14 PM EDT reply actions
Sabermetrics
is sort of like some of the more arcane technical analysis in the stock market. And many of those indicators are almost meaningless. Remember all the numerous statistics that showed that Jason Campbell was a good quarterback. Of course, I’m just cynical about statistics.
But-but-but
Jason Campbell is a good quarterback ;_;
by BigOLinemen on Jul 24, 2010 10:01 PM EDT up reply actions
If you look at the average time the ball is in the air during his throws.
He ranks in the top 5 QBs in the league.
Average time...
Well, that says it all. Well, maybe not everything.
There's a bandwagon?
Sorry, didn’t mean to bash him. He probably qualifies as a “good” QB. But we’ll know a little more after this season. There are many on HH who really like him, and some use statistics to show how good he is, and I sometimes question those statistics.
I really disagree with the LJ analysis
If you look at the game logs your “he just has to have less carries to be productive” has no factual basis. You were just taking the average carries not looking at it game by game. He had 7 games with Cincy. His best game was against CLE. He had 22 of 46 attempts with the Bungles in that game. That skews his YPC average severely. That left 6 games with 9 carries or less. In those games he broke a 4.0 average twice(Against a horrible Chargers Run D and a Jets team resting for the playoffs). Those other 4 games he averaged 2.75 yards or less. The real analysis of LJ is he does well playing behind a great O-line against the Browns. He is more miss than hit in any other situation.
well, hopefully
If CP and/or WP come out and get it done 2.75 ypc will be enough as he will only be used in short yardage situations.
Donnio
A stat Professor told me this a long time ago and I’ve never forgotten it:
“Statistics don’t lie but statisticians work for somebody”
(Our politicians spin Doctors might be the best example of BS stats).
It is true that statisticians work for somebody (some are independent), but.
And that is a big but. For one, some times there are paradoxes in the numbers. An example of this was in wrist watches compared from 1954 to 1958 when the unit prices of all individual producers increased, but the average price of their sum totals decreased. What happened was the cheaper watches became a larger proportion of the total. As Bay Area Bullet mentioned in effect the “truth” of the details gets hidden in the totals in the case of LJ average yards per carry. The is not usually a conscious intent to deceive in something like football stats. The famous trick is to pick a favorable base year in the case of economic stats like gross national product to prove a favorable economic growth for some political purpose.
by Jefferson1935 on Jul 25, 2010 12:15 PM EDT up reply actions
Interesting
You mention other ways statistics can lie (besides cherry picking) – like picking your base number (often, year), and using misleading totals, when the devil is in the details. All these are reasons to be sceptical of statistics. Don’t dismiss them, but try to figure out what, if anything, they actually mean. By the way, there is a classic example on Yahoo right now – a big AP story purporting to show that the middle class in America is disappearing. A lot of disturbing statistics. I don’t dismiss it, but I think it is probably painting a misleading picture of the situation of the middle class (it seems to me there is still an awful lot of money around, but digging out the truth would be a lifetime project). It’s not clear if the study is from left or the right. It stresses leftist type arguments that the rich are getting richer (which I believe) and the poor, poorer (questionable, and also contrary to the main point of the article), but some of the alleged causes (globalization, companies moving overseas) seem more rightist type arguments. You can see some of the methods discussed here in that article – cherry picking, starting points, misleading totals, etc.
Jefferson1935 -
yes, I don’t think any of the football stats are purposely misleading. Most just need a careful reading, to see what they really mean.
Statistics
Yes – look at the so-called “scientists” peddling the global warming baloney. They see what they want to see, put out statistics that support their goofy case. But on your comment, it’s semantics, but statistics can and often do lie. Remember the old saying (Mark Twain?): there are “lies”, “damn lies” and “statistics”. One of the more common methods of making statistics lie is Cherry picking – just giving out the stats that support your case. That was what Al Gore did in his movie. Incidentally, I don’t think sabermetrics “lie”, and you might be able to glean some interesting insights from them, but I just wouldn’t exactly put too much weight on them.
Al Gore
He not only cherry picked, but also put out a bunch of alleged “truths” that were flat out wrong – like the polar bears in trouble, the ice caps and arctic ice disappearing, warmest this and warmest that (cherry picking time frames), etc. Illustrating another general way statistics can lie – omitting context and perspecive.
bite ur tongue. the man invented the internet for god’s sake. lololol
by les boulez bomber on Jul 25, 2010 11:42 PM EDT up reply actions
Just can't understand how Tipper could divorce the
man who invented the internet. I mean what else does the poor schmuck have to do, win the Presidency or something? :)
Time flies like an arrow. Fruit flies like a banana.
he actually never said he invented the internet
but don’t let that stand in the way of a funny joke.
"By far the worst performers on the team are in the front office." – Sally Jenkins
by smutsboy1 on Jul 26, 2010 9:46 AM EDT up reply actions 1 recs
I'm sure he just said something in a garbled fashion
or that it was taken out of context and abused by the media. But its been good for a laugh for at least a decade, if not 2.
Time flies like an arrow. Fruit flies like a banana.
But on your comment, it’s semantics, but statistics can and often do lie.
One of the more common methods of making statistics lie is Cherry picking – just giving out the stats that support your case.
That’s not an issue of statistics lying, that’s an issue of people using them in a manipulative fashion. To take the stance that statistics lie because people can manipulate them or use them in a misleading fashion is akin to not believing in photographic evidence because someone can doctor a picture.
The source of amkcaps information seems to have
been Pro Football Focus rather than Football Outsiders – http://profootballfocus.com/by_position.php?tab=by_position&season=2009&pos=HB&stype=r&runpass=run&teamid=-1&numsnaps=0&numgames=1.
It might surprise some of you, but I am fairly new at some of the stats besides the standard stats seen in the newspapers showing game results. Kevin introduced me to PFF with his article on the Redskins line in the 2009 season. Some of PFF’s numbers are perplexing such as those on safety play, i.e., A. Rolle who got the big contract from the Giants. I don’t see PFF as a mature system, but it has a good start.
“The Elusive Rating
In PFF’s ongoing quest to present the world with unique, insightful statistics, Sam Monson presents the Elusive Rating. The goal of this stat is to filter out the performance of back’s blockers and solely focus on a runner’s contribution”
http://profootballfocus.com/articles.php?tab=articles&arc=2010-05&id=162
Some surprising facts which explain why teams performances can radically shift. It is also why the NFL is a passing league.
“Walkthrough: The Teal Deal by Mike Tanier: The 750 Club
Twenty percent of the yardage in an NFL season is gained on two percent of the plays. … While those big plays aren’t divided equally, 750 plays come down to about 24 plays per team. That’s about 1.5 plays per team per week. The difference between a Super Bowl champion and a .500 team can come down to those rare plays.”
http://www.footballoutsiders.com/walkthrough/2010/walkthrough-teal-deal
can you please update this post
with League Averages for all the measureables.
and also maybe Chris Johnsons numbers.
So we can see what they shoudl be aiming for, and whether they are above or below league average in any of these measureables. otherwise they are just numbers on a page with no reference point
(yes it is the analyst in me talking again, I hear this same complaint from my boss about my reports from time to time, apologies)
Pommylee
I disagree with your Portis point too`
I have always seen Portis as being Peterson like in his desire to at times initiate contact, he literally uses defenders like a Pinball machine when he is at his best, bashing into them, knocking them off balance then pivoting away into space.
As much as I have some hopes for these 3, the stats on RB’s in the 30’s and with a large number of carries under the belt suggests to me that we need a UDFA RB to come out of the woodwork and offer us something if we have any desire to rely on the run this year
Pommylee
how do they compare to other teams RBs?
Given the limited amount of info (no stats for any other RBs much less something like a league average) I don’t have any reaction to this information.
after looking at the pro football focus web page
It does look like we have a good RB corps. I took their data and dropped the players who had fewer than 93 carries (to eliminate the guys who only had 5 carries). Then I sorted the data to see where the current Redskins RBs fell. When I sorted by YCo/Att Portis is 21st (Jamaal Chalres is #1 with a 3.6 so what does that say about LJ?), Parker is 37th and LJ was very low (I don’t recall the exact # but it wasn’t good). If you sort by YCo % then LJ is first, Portis is 6th, and Parker is 27th.
If the stats are really a good indicator (I can see the potential but I’m not completely sold on it yet) and out OL really is improved then we should have a very good running game.
Maybe I'm over simplifying, but
I see that any RB not named Parker can average 5 YPC as long as the line provides a push worth 2 – 3 yards every snap. Not bad if stats lead to reality, and if the line is as improved as I hope.
Parker shifts to the slot (in motion) and offers a screen threat to decoy Cooley or vice-versa, otherwise I don’t see much use in Parker. Portis can blow through big holes AND be effective after contact….
Good line = good running game
"By far the worst performers on the team are in the front office." – Sally Jenkins
Any RB can put up numbers late in the game
agains a poor, tired defense. Is there a stat that says when you really need 2 yds your back will get it? Heart, how good the line is, etc. all need to be weighed into the evaluation. If you’ve pissed off your OLine because you don’t practice expect numbers to dwindle. Whole lot more going on than just stats.
Time flies like an arrow. Fruit flies like a banana.
What I am hoping for...
I hope that any of our RB can get out of the backfield before have to use the highlight stick. I saw too many times last year that the RB had to make tacklers miss as soon as they touch the ball.

by 

























