Outside the Box: Knile Davis

Knile Davis' day against the Dolphins and his PFF grade is explored in more detail.

| 3 years ago

Outside the Box: Knile Davis

OTB-2014-davisThere are certain common statistics that have been drilled in to the brains of NFL fans as benchmarks for defining standout performances; the 100-yard rushing day among them. It’s easy to have a peek at the box score and run with the usual storyline, chalking up a strong day for a running back based on him hitting triple digits on the ground, but there’s often more to be told.

One such case came to the forefront this week as Kansas City’s Knile Davis, while getting the start in place of injured Jamaal Charles, eclipsed the 100-yard mark for the first time in his young career. Many Chiefs fans, riding the high of a much-needed win in Miami, were no doubt scratching their heads when they saw Davis’ overall PFF grade for the game come in at a significantly low mark of -6.6.

So how can the grade be so far off from general perception? Sure, he scored fantasy points but did he really turn in “the best game of his career”? It’s largely a product of taking in the whole picture, or not taking it in, as the case may be.

Davis finished with 132 yards and a touchdown on 32 carries. The touchdown run was impressive and is an easy one to lodge in memory as he made multiple moves – and made multiple defenders miss – on his way to the end zone from 21 yards out.

All told, 64 of Davis’ rushing yards (48%) came on that run and three others, and each of those – and a few more — received positive marks, but that’s about where the feel-good portion of his outing ended. His other 28 carries amounted to an average of 2.4 yards per and were largely pedestrian until you get to his two fumbles.

davisThose two significant downgrades along with the normalization process that is applied to all players per positional expectations, meant that even the good vibes from his TD run and his yardage total weren’t enough to avoid a negative in the rushing category.

But, as is the case with all positions on the field, there’s more than one facet of play to be judged. Davis was on the field for 54 snaps and 20 of those were pass plays. Targeted only once on the various short backfield routes he ran, Davis dropped the lone ball that came his way, sending his receiving grade into the red as well.

And, when asked to stay in and pass block – an underappreciated but vitally important aspect of running back play — on a small handful of snaps, Davis struggled, as he was charged with a sack and a hurry (another sack initially on his book was removed during our regular review process), resulting in another negative mark tugging his overall down.

Taken together, negatives for rushing, receiving, and pass blocking added up to Davis’ disappointing overall grade as presented in our Premium Stats section and reflect his whole body of work on the day, not just the eye-catching, but overly simplistic, box score stat line.


Follow Rick on Twitter

  • Tim Edell

    I love PFF and i thouroughly enjoy reading everything this site offers. I would find it interesting based on how PFF grades players to see how Barry Sanders would of graded back then. Here was a player who could have 15 carries for 20 yards and the next thing you know he has 16 carries for 100 yards. Would the 15 unproductive carries leave him with a negative grade on a weekly basis??

    • chickenhed

      It would likely be around a neutral grade unless he never fumbled, pass blocked like a boss, and he was excellent in the receiving game (then his overall would likely be slightly positive).

      So many people see PFF as a site that dictates whether a player is “good” or not. PFF goes MUCH deeper than that. You come here if you want to know if a player is good, CONSISTENTLY and in ALL aspects of their game.

      If you want your player to ONLY be a great runner and do everything else poorly (pass block, receive, hang on to the ball at all times, etc) then PFF is actually not the best place to come (unless you focus on their run grade and not their overall of the player). That’s why so many people freak out when they see a player that got 100 yards get a negative grade.

      What does one value in a player? If you wanted a player that sparks legendary ability once every 20 plays, but is terrible for 19, then you can just ignore the scores here. That one play (and I’m NOT being sarcastic here) may really be worth the 19 bad ones.

      Apologies for the rant. Your question is perfectly valid. My honest answer is that, in the case you gave, Barry would receive a neutral grade. Does that mean Barry is a bad running back? No. It means that Barry was an inconsistent running back during those 31 plays. There are many players that are so insanely good that it is worth keeping them for the plays they are below average in because the return on the great plays is so spectacular. Cordarrelle Patterson is a good example of this.

      No site can give you every detail of every player. I use PFF as one of many sources to get an idea of a player’s play by play ability.

      • eYeDEF

        I don’t that’s fair to Sanders to call him an ‘inconsistent’ running back though. That would defy the eye test of what we saw on tape. It’s not like he played poorly on the plays where he was stuffed for a few yards. He was a remarkable back with breakaway talent that defenses would often overload the box to stop. Is it really accurate to call him ‘inconsistent’ when there are 8 men in the box ready to gang tackle him and he’s got nowhere to run?

      • skwirrl

        PFF is a steaming pile of dogshit is what it is. One of the biggest steaming piles on the entire internet.

      • LightsOut85

        The one more-reasonable (IMO, compared to what you’re describing (which I agree with)) critique I’ve read – is that the grades don’t (seemingly, as we don’t know the exact method) give enough “credit” to the aforementioned big-plays. (Perhaps desiring some sort of “win probability” aspect? Something to measure how it changed the game, in the context of that game). That is, when players are “close” in consistency (not 1 who’s wildly inconsistent), but the less consistent one makes more “impact” plays.

        I don’t know who they have in mind when they think of this, but my guess would be a situation like 2 DTs: 1 is graded better vs the run but doesn’t have great “hard stats” (because he’s getting graded well for things like blowing up a block, holding up 2 blockers, etc), and 1 who is graded worse because he is not as consistent (but not wildly so), but also makes “game changing plays” (run-stops, etc). They usually add that a team “knows what they’re getting” with the latter-player & value the ceiling/potential (that is, he’ll play aggressively, shooting gaps to get into the backfield – and sure, he’ll miss sometimes, but that has to come with also making plays the former-player isn’t capable of making, etc).

        While I can sort-of understand that viewpoint, I also might say…..they also give us a variety of hard-stats (…though not as many as I know they have :p). If one wants to put a premium on “play-making” then sort the DTs by run-stop%, etc. PFF doesn’t have any influence in giving out actual player awards (although that’d be better than the actual methods in place >_>)…no need for people to get their undies in a bunch.

        (I will say, I think it’d be more prudent to break up the grades in their articles (for those w/o a sub), so people CAN see where they did well/poorly. Same with forming their pro-bowl/all-pro teams solely by overall. Navarro Bowman made one last season despite 49% of his overall being pass-rush, where only 7.3% of his snaps came (ie: he was just very EFFICIENT, but didn’t bring much overall value there)).

  • Kevin Bohorquez

    I don’t find it accurate that you guys grade a lost fumble and a fumble that was not lost as the same thing. It was a fumble, but he managed to recover it, the other one, he lost it, why are they the same?? they don’t have the same result. So, maybe the grade system you guys use is not that accurate. Same thing about the long runs, it seems that it is bad for a RB to have a long run here in PFF.

    • tttrickyyy

      Kevin – I think the idea is that despite the fact that he recovered it he still fumbled it. Who recovers the fumble does not cover up the fact that the fumble still occurred.

      • Kevin Bohorquez

        Then we are playing with the “he coulda lost it”…. those are not facts and are not an accurate way to meassure his production at that specific game. Sure he should get blame for fumble it, but he managed to recover it, that’s why I’m asking if fumbling it and losing it are the same as fumbling it and recovering it.

        • TFootball

          The act of fumbling isn’t a “luck” thing, skill and practice can prevent it from happening (or, a skillful player can create them defensively). Recovering a fumble is much more a thing of chance. Calling all fumbles, recovered or not, a negative thing is fair because the chance of a positive outcome on a fumble is near zero.

    • Brian

      I’m actually OK with that, if he fumbles he fumbles…and I like PFF since they analyze the actual performance of a player. How a great pass that is dropped is not a bad QB play, for example. I could see your point, but I like they take the view of fumble is a fumble for that player.Besides, it is regardless a “bad” play. I guess there could be one of those freak plays where it gains you yards, but THAT is pure luck lol

    • Chris

      There’s data out there that show that fumble recovery evens out to 50% chance in the long run.

      The fact that he had 4 long runs doesn’t outweigh the fact that he had 28 not too great runs. Overall, his running was average (~4 YPC – this is the league average) and on top of that he had 2 fumbles and 1 dropped pass. I’d say his day should be below average overall.

  • Anthony

    Looks like a lot of rationalization in order to defend the indefensible. I love how PFF always uses the “if you take away all of his good runs he really didn’t do anything” argument. They used the same argument to discredit Charles’ 2012 season. It’s clearly nonsense logic.
    Why not then drop the runs where he lost yardage as outliers as well? There was at least one run where he lost ~5 yards before he could even taken 2 steps.
    This is a dumb grade, pure and simple.

    • Andreas Haakshorst

      I used to have some beef with that point as well, but you got to realize that nobody here is saying: “ignore the good runs”. The point PFF tries to make is that those few good runs get overshadowed by a lot of bad runs (or plays in general). An overall grade – being cumulative – therefore will be negative.
      In case the relation between good and bad runs was exactly the other way around, in fact the bad runs would be ‘dropped’ and overall you’d get a positive grade.
      @PFF: I think this specific argumentation you’re using can just too easily be misunderstood. I know you are just trying to bring the base statistics more in accordance with the grade. But if the argument is that those stats just don’t reflect a players performance, maybe trying to make them do just that is not such a good idea.

      • [email protected]

        This is one of the many major flaws in their grading system. They reward consistently mediocre performance and downgrade game altering plays.

    • JJ

      Splash plays are worthless without consistency, and consistency is where the law of averages comes into effect and players like Davis watch the Super Bowl from their armchair wondering “What if?…”

      • [email protected]

        Consistency is useless without splash plays. Splash plays are good without consistency. That team that won the Superbowl was #2 in explosive plays and #1 in forced turnovers and turnover differential. Making 10 above average plays is not as good as making one huge play.

        • JJ

          The Seahawks and Broncos were the most consistent teams in football last season, so nice try no dice.

      • Anthony

        Look, it’s clear you didn’t watch the game. He consistently picked up 4+ yards, broke tackles! and moved the chains. Davis was consistent- he averaged over 4ypc and had a long run of 20 yards.

    • Ken

      Yup, let’s just ignore that he dropped a pass and also did a bad job in pass protection. Those are very insignificant duties as a running back. All that matters is, he had a 100+ yard game and a TD. I’m sure the coaches don’t care about the other things he did a poor job of.

      • JJ

        Ken if the stat sheet is all that matters to you then why are you on this website discussing the finer details? not taking a dig at you just curious.

        • Ken

          Dude I was being sarcastic……

      • Anthony

        Did you even read what I said? Did I even mention the 100 yard game or the TD? No.

  • Fintasy

    Opposite to the Tannehill case, where he keeps receiving glowing PFF scores, but he’s on the verge of being benched this week.

  • skwirrl

    Except KC’s line is probably the worst in the league, (Much like the dogshit stats that PFF produces.). Which makes Knile’s performance extraordinary.

    • Graniterlm

      You mean the same dog shit stats that you are over here dining on? I guess you must have had a bad plateful

    • [email protected]

      All stats are dogshit. Statistics are the most easily manipulated thing in the world, you can make a statistic to back up any argument no matter how wrong it is.

  • Chris

    It’s funny hearing all the whiny Chiefs fans. Y’all sound a lot like Seattle fans whining about Russell Wilson’s grades.

    If a guy has 2 great runs and 28 average ones, his run grade will come out positive. Unless he fumbles twice, then it’s gonna be neutral or negative. Then he drops his only pass so his passing grade is negative. And he allows several pressures including a sack in the few times he’s asked to pass block.

    As told in the article, the one great TD run where he broke several tackles sticks in people’s heads, as does the “100 yard game”. But that isn’t judging how he played all day, and all day he was average to below average in all facets of being a RB. From that point it’s simple math. Negative plus negative plus negative equals really negative.

    If he eliminates the fumbles, catches his only target, and doesn’t allow a sack (just a hurry or two) then he would be average to good in all phases with nearly the exact same statline and highlight run, and he’d finish with a positive overall grade.

    Summary: Don’t fumble, don’t drop passes, and don’t allow sacks. I don’t see how anyone can argue that.

    • LightsOut85

      “But but, he’s on my favorite team! And just because I don’t understand statistics doesn’t mean I’m bad at watching football – I’m great!”

      • [email protected]

        You know there is one NFL team who for years has been all-in on statistics and has a devoted statistic department and has used advanced statistics in their player evaluations. They had a QB who they were thinking about and the dug deep into the advanced statistics and decided to give him another chance.

        The team is the Jacksonville Jaguards, and the QB is Blaine Gabbert.

        Statistics is not a tool for understand anything, least of all sports.

  • pit

    I love when you guys want to skew stats or make someone look bad, all you do is pluck the big runs out of the equation. Those are kind of important.

    • Chris

      I love when you want to make someone seem like they had a good game by focusing only on the splash plays and ignoring the other 30 carries where he was below average.

      PFF grades their whole game, every facet of being a running back. And the truth is Knile Davis was below average in everything he did outside of those few splash plays.

      Why is that so hard to understand?

      • LightsOut85

        To add on:
        Let’s not use grades, but actual run distances. If you look at the play-by-play from this game, you’ll see Davis had 32 runs at an average of 4.125 YPC.* What the PBP allows us to see is the down/distance situation. *(Which, isn’t even that great. The only really impressive box-score stat is the TOTAL yards, but every knows (or should know) that’s a counting-stat & it influenced by how many times you did the action…it doesn’t tell you anything worthwhile).

        If you find Davis’ success-rate for this game (where, to be a successful run you must gain 40% of the yards needed for a 1st on 1st down, 60% on 2nd down, and convert the 3rd/4th down), it’s 37.5% (Meaning 62.5% of his runs failed to set up his team in a favorable position the next play). That is VERY poor. Granted, some of this is on the OL, but PFF’s grades account for that (with things like yards-before/after contact, etc).

        This is just 1 type of look, but it shows you how breaking it down play-by-play can give you a much clearer picture that looking at “counting stats”.

        • Chris

          Great stuff man

        • Gèr Steffens

          Very good illustration of how raw numbers are meaningless. Like you said 132 yards looks great, but when there’s 32 carries it’s a lot less impressive. And since in many cases the run game is based on staying ahead of down and distance having a few big plays and a lot of touches doesn’t mean you did your job.

          If you remove the total yards you have a running back who, like you said, kept his team on track 37.5% of the time, had 4.125 ypc, 1 touchdown, 2 fumbles, a dropped pass and gave up a hurry and a sack. Not exactly a great game.

          It reminds me of one of my pet peeves: If running back A gets 20+ touches team B wins. Generally when running back A gets 20+ touches it’s because his team was ahead in the 4th quarter so he got a lot of touches late. Numbers are meaningless without context.

          • [email protected]

            Numbers are meaningless, you have that part right. Last night was the first time i saw Knile Davis play and he looked pretty good to me.

  • Izach

    The overall problem with the PFF grading system is that they consider all phases of the game to be equal if he dropped 1 pass even if it was his only pass that isn’t worth the negative grade he recivied for the whole category, it was 1 play of the 54, same with the pass protection, there’s a reason he didn’t block or was targeted 20 times of the 20pass plays he was in. Now I agree he failed in those aspects of the game but those 3plays in the passing phase of the game don’t count as equally as the other 51 plays they just don’t. His two fumbles are a much bigger problem but he only lost 1 which means it shouldn’t be graded as negatively. All in all Kile didnt have as good a game as FF players think he did but was no where near as bad his -6.6 PFF grade.

    • [email protected]

      If we were going to adequately discuss all the problems of the PFF grading system we’d die of old age first. Danny freaking Woodhead was their 7th ranked RB last year. Antoine Winfield was their #1 ranked DB and then the next year wasn’t even on an NFL roster.

  • [email protected]

    So you were saying? That Knile Davis guy is no good! PFF is the worst. Seriously.