So after writing roughly eleventy billion film reviews here on Arrowhead Pride, it occurred to me that using 500 words to describe my grading process each article is a waste of time. I vented my frustration on Twitter, and had about 8 quadrillion (we're big into numbers today) people point out the obvious solution: write an article which describes the methodology of the film reviews, then link to it every time I write one.
So that's what this will be. I'm going to cover
1) The methodology I use for offensive linemen, cornerbacks, and OLB/DL in film reviews;
2) What numbers I track; and
2) Why I use that particular methodology.
I've talked about this stuff before, so some of this will be a bit cut-and-paste. But I think it'll be good to get everything together (at least with positions I consistently review). Keep in mind, each film review will generally consist of two parts: the numbers and the actual film review.
Let's do this.
Offensive Line Review Methodology and Reasoning
With offensive linemen, I review every snap on all-22 (primarily for the "Madden" view), specifically watching the player I'm grading. While I pay some attention to the formation, etc., the overall success of the play doesn't matter. All that matters is what the offensive lineman does.
While watching the film, I compile stats of the offensive lineman's wins, losses, and neutral plays.
A win is where a lineman stops a pass rusher or eliminates a run defender (by walling him off or moving him out of the play, depending on what's required of him). A neutral play is where a lineman doesn't really win or lose clearly, where a lineman is part of a double team (with rare occasions where it's done so well it counts as a win or so poorly it counts as a loss), or where the lineman doesn't do much (say, a quick WR screen where the OL doesn't have a chance to block anyone before it's over).
I then do a little math (yuck) and calculate what the lineman's win percentage and loss percentage is compared to his total snaps. Loss percentage is significantly more important than win percentage in my opinion. An offensive lineman losing can completely derail a play. They are essentially paid to not lose. Fighting a defensive lineman to a draw will generally give a runner or QB enough room to work with. A good "win" percentage can push a good performance to a great one, though.
Generally speaking, the magic number is 10 percent. I want my offensive lineman to lose on less than 10 percent before I'm comfortable with them as a starter. Obviously, the lower the better.
Why do I do it this way? Why not track sacks allowed or pressures allowed? For the simple reason that sacks and pressures are often dependent on what the QB and the rest of the offensive line is doing. It doesn't always add much value when you already know how often an offensive lineman lost to know what the results of those losses were. In fact, it can sort of muddy the waters. Often a lineman might "lose" only to have the QB scramble and make a nice play with his feet. That doesn't make the play by the lineman any better, but avoids the "sack allowed" stat.
In other words, I'm trying to leave out all stats that don't isolate the lineman's performance and focus ONLY on what he's doing. That's the numbers part.
The film review part is the real meat, where I try and talk about a lineman's footwork, athleticism, lateral movement, hand fighting, punch, strength, and overall technique. It's basically the why portion of the article, and is far and away more important than the numbers.
I believe combining the numbers I track with the film review leads to the absolute best we can do to really determine how well a lineman played.
Cornerbacks Review Methodology and Reasoning
For corners, I do a lot of the same thing I do for offensive linemen. I review every snap on All-22 and track successful coverage and failed coverage on individual snaps. I then calculate success percentage. I also track the number of times a corner is targeted, the number of catches he gave up (and the yards allowed), and the number of passes he defensed (both knocked down and picked).
It should be noted I only track those numbers when a corner is in man coverage (because doing it with zone coverage would be ridiculous) and when the corner actually has to cover (so not on a WR screen, for example).
(An eternal H/T needs to go to Cian Fahey, the first writer I saw do things this way with corners)
With corners, there are a few ways they are judged by the masses. The first (and worst) is by their tackle and interception numbers. That's a really, really unreliable way to try and figure out how well a corner can play. For starters, a corner could well have a bunch of tackles because he's getting worked over and bringing down receivers after a zillion catches. That's, you know, bad. And interceptions provide a glimpse into what a player did about 1% of their snaps. It doesn't tell you anything about their consistency in coverage.
The more modern way people judge corners (largely due to PFF's work) is how many catches they gave up versus how many times they were targeted. The problem with this methodology is two-fold. First, we're still dealing with a relatively small percentage of a snaps a corner takes. Even if a guy got targeted 15 times, he was in man coverage another 20-30 times in that game (at least that, in Bob Sutton's system). So we're still leaving out the majority of snaps.
Additionally, catches allowed doesn't account for great quarterback and wide receiver play. There is no such thing as perfect coverage in the NFL. There's always a chance that the QB is going to fire in a rocket into a six inch window that you can't possibly defend (you know, getting Rodgers'd), or that the receiver is going to do something ridiculous (getting Hopkins'd). But when you're just tracking targets and catches allowed on those targets then this coverage...
:format(webp):no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/6707257/Chiefs_2.0.png)
... where Marcus Peters had great position on DeAndre Hopkins and did a good job contesting the ball (in other words, did exactly what he's supposed to do), is somehow considered WORSE coverage than this play...
:format(webp):no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/6707261/Chiefs_2.0.png)
... where Peters got his hips turned around at the line of scrimmage and Calvin Johnson blew right past him to get plenty of separation. The only reason this wasn't a huge gain? Stafford sailed the ball over Megatron's head.
What I'm demonstrating here is that tracking catches allowed brings things into the analysis that are beyond the corner's control. By focusing on whether the coverage is successful or a failure (regardless of the result) narrows the analysis to ONLY what the corner did.
Successful coverage is where a CB sticks with his man closely enough to force a great throw or great catch to get a completion. Failed coverage is where separation is allowed to the point that an NFL QB and WR should have an easy pitch and catch if the throw is made. By looking at this every single snap, we can figure out how a corner was playing and isolate that from the opposing QB's talent (though not the opposing WR's talent, as receivers like Jeremy Maclin are a lot tougher to stick with than, say, almost everyone else).
Generally speaking my cutoff for corners is a 70 percent success percentage. Corner is an incredibly difficult job, and they're going to get beat multiple times every game. Additionally, it's very important to pay attention to the way a corner loses (which goes to the film review portion). One corner who is successful 70 percent of the time might be a fair sight better than another corner of a similar number due to the fact that the second guy is getting beat on more harmful routes (like posts and double moves), while the first guy is able to contain his losses to less damaging plays (like, say, quick outs or something like that).
Man, we gotta shorten these up. Now that you've got an idea as to what I'm doing it should go faster.
Wide Receivers Methodology and Reasoning
With wide receivers, I essentially do the reverse of what I do with corners. I still track each snap on All-22, but a win for a wide receiver is gaining separation in the route while a loss is... you know, not.
With receivers, I also try and track drops, as well as types of routes run (which allows you to go back and figure out which routes receivers have the most success with). Two other things I intend on tracking (even though it's pretty easy to find on various sites at this point) is YAC and saves. A save will be any time a WR bails out a bad throw by making a great catch. Think of Maclin's sliding reception on a deep bomb against the Bills. Smith badly underthrew it, but Maclin adjusted so well to the ball's path that he wound up with a big catch.
Again, the idea here is to see what the receiver is doing even if he's not getting targeted, which allows for a much more complete picture as to the type of player he is (and allows us to take the QB out of the equation to an extent). For example, if a player doesn't get targeted on a play like this...
Pressing Jeremy Maclin isn't really a great plan. He could teach a class on getting free releases at the LOS. pic.twitter.com/CSzuy9mtu7
— Seth Keysor (@RealMNchiefsfan) June 14, 2016
... that doesn't change the fact that he beat the CB like a drum, right?
See, these are already getting faster!
Defensive Linemen/OLB Methodology and Reasoning
Much like with wide receivers and corners, defensive linemen and outside linebackers (not inside linebackers, which I'll explain in a moment) are essentially the reverse (or inverse?) of offensive linemen.
I watch (usually on the "Madden" camera angle) every snap, then track the number of times a DL/OLB is able to beat the offensive lineman in front of him. If he does so, it's a win. If he's completely stonewalled as a rusher or moved aside (or walled off) as a run defender and makes no impact on the play, it's a loss. Anything in between is a neutral. I also track pressures and hits, functional double teams (h/t to Craig "Kalo" for that one), where a defender forces a double team and occupies it enough to free up another player, and "stuffs" (runs contained by individual effort for a relatively short gain).
Generally speaking, you're going to see a lot of neutrals and losses for defensive linemen. It's a tough position to play. I don't have a set percentage I look for (because I've done fewer of them), but for defensive players it's important to not lose often (just like with OL). However, wins are MUCH more valuable for defenders than offensive lineman. I win means a QB pressure or a sack, or perhaps a run stuffed. A lot of teams will put up with guys who lose a little too often but make up for it by winning a bunch too (rather than racking up a bunch of neutral plays).
What does a win look like? Hey, Chris Jones, help me out bro.
I keep saying "this is the last one for now," then Jones destroys another play and I have to post it. pic.twitter.com/rcoExmwCjz
— Seth Keysor (@RealMNchiefsfan) April 30, 2016
When outside linebackers go out in coverage I use the same methodology I use for corners, IF I'm tracking that side of things for that particular player.
Why am I not including inside linebackers here? Quite simply, because inside linebackers have a very, very different job than OLB's and defensive linemen. There's quite a bit less taking on of offensive linemen one on one. For inside linebackers, I haven't come up with a good system yet (other than stuffs and missed gaps), so any review I do on an inside linebacker is going to be more "this is what his film looks like" oriented than numbers oriented. If you have suggestions for numbers to track for ILB's, I'm all ears.
Quarterback Methodology and Reasoning
I did a giant film review on Alex Smith last offseason (and will do another one soon), as well a a piece on Chase Daniel against the Chargers. I'm planning on changing up a few of the things I track this season (no more "nice throws" stat tracking, sorry guys), but will continue to look at the following:
Missed shots- Open wide receiver down the field 15 yards or more on snaps the QB should have seen him (ie didn't get killed the second the ball was snapped) and just flat-out missed him (ie doesn't throw him the ball, without the legitimate excuse of finding an open guy elsewhere).
Happy feet- Clean pockets bailed out on by the QB.
Drops- Catchable throws dropped by WR's (genuinely catchable, not "if it were Jerry Rice he would've caught that) and yards lost (which includes reasonable YAC).
Saves- Plays wherein the QB faced heavy pressure or blanketed receivers but made something happen with his feet or a great throw.
Flushes- Plays wherein the QB gets chased out of the pocket before anyone is open.
Inaccurate passes- I mean, this is pretty self-explanatory.
Potential picks- throws that should have been intercepted by the opposing defense.
I also plan on tracking the depth of each throw made by a QB, to get a better idea of how often the QB is testing the intermediate to deep zones.
This one is still a work in progress, so I may track a few more things as we go along.
I think that's a good place to call it a day here. Most other positions I might observe will use some combination of the methodologies I've talked about here (tight ends will use a lot of what you see with WR's and OL, for example), so there's no need to get in-depth on them. I'm attempting to think of a formula for safeties at the moment, as well as running backs. Any suggestions are welcome on those two positions as to what I should track (or any other position. I'm always willing to hear ideas).
So I hope all this helps explain the process I use as I review film, and will serve as a beacon we can point back to any time I'm reviewing film (I made add on as I develop my system for other positions). If not, well, at least we passed a little time together, right?