INDIANAPOLIS, IN - AUGUST 19: John Beck #12 of the Washington Redskins is tackled by Gary Brackett #58 of the Indianapolis Colts during the game at Lucas Oil Stadium on August 19, 2011 in Indianapolis, Indiana. (Photo by Andy Lyons/Getty Images)
As the Redskins preseason kicks-off this week I thought it would be good to highlight some stats to avoid, and those to focus on this preseason. On some level all stats can be misleading especially in preseason due to small sample size and context. Even still they can be proper evaluation tools if used appropriately.
Context concern: Context in any stat is extremely important, but it is even more important during the preseason because the differences can be so drastic. Any time their is a play the first thing that should be done is an evaluation of the circumstances surrounding that play. If it was a 50 yard TD pass, how many of those yards were from the receiver, how many from the QB? Did the receiver make a nice catch, or was it a better throw by the QB? Was their a busted coverage or did the defensive back make a bad read on the ball? These are just a few of the questions that should be looked at on any play, and in preseason when you have such differences in talent context is even more important. Was this a catch by a number 3 receiver versus a corner who won't even make the practice squad? etc. There are two examples I want to highlight about how context matters:
Brandon Banks Punt Return versus the Buccaneers: Last preseason Brandon Banks looked to be in a roster crunch and had yet to do much to help his case on special teams. In the final preseason game last year Banks took a 95 yard return to the end zone and all but sealed his spot on the roster. What that play doesn't tell you, is that the punter who made the kick, was an undrafted rookie free agent who had no shot of making the team. The kick was a little low, and primed for a big return. Giving Banks too much credit for this return is why context matters.
John Beck Quarterback Battle: Last preseason John Beck and Rex Grossman were locked into a training camp battle, and it seemed as though their numbers were fairly similar, Grossman got the nod. Though their numbers were similar, Grossman's were clearly better and he went up against more 1st and 2nd string units. That alone though isn't why Grossman was the starter. Looking back at the preseason you could clearly point to times where Beck didn't handle pressure well (4 sacks to 2). And when he was under pressure Beck would either scramble for minimum gain or dump the ball off (the Colts game highlighted a couple plays where he did this). Though many fans didn't realize it at the time, this foreshadowed just how bad he'd be as a starter, despite a stronger arm and more mobility than Grossman.
Some stats I don't like:
Quarterback Yards and yards per attempt: this is predicated on so many factors, that given the weakened competition and the small sample size can be very misleading
Quarterback TD to interception ratio: The sample size here is way too misleading, unless there is a large number of interceptions, this stat really can't be trusted.
Kick-return average: this stat at any time is flawed because it counts yards gained in the end zone as positive yards, but is very troubling with a small sample size.
Running back yards per carry: Given the small sample a few big runs can greatly skew the stats. If a running back has 20 carries and breaks off one 40 yard run and one 30 yard run, he would only need to average 1 yard per carry on his remaining 18 runs to have a 4.4 ypc mark, which looks very impressive.
Catches, yards and yards per reception for Receivers/Tight ends: Similar to running backs the ypc can get skewed pretty easily with the small sample size. And measuring number of catches is hard because the team tries to get so many receivers and tight ends involved.
Defensive sacks: Sacks are just part of the pass rush equation and given some weak back-up O-lines, QB's who don't know how to get rid of the ball, and running backs who don't know how to pass protect, these numbers can get out of whack quickly.
Stats that I do like:
Quarterback completion percentage: Now this stat can still be off (as evidenced by both Beck and Grossman over 62% last year), but it can still be a solid barometer. It's even better if you grade each pass and give more credit for completions versus NFL caliber defenders. Also looking at ball placement.
Quarterback time in the pocket: This one is something you need to do on your own, but even if you don't use the stop watch you can see the difference when a QB is holding on the ball too long. If a quarterback is holding on to the ball for longer than 3 seconds consistently he's going to be in trouble. Even if it's not hurting his play in preseason it could make a difference once defenses bring more creative blitzes and you face starters more.
Running back rushing chart: Now this is not to say one should ignore 30 and 40 yard runs like my example above, but first you have to look at the context of who is he going against. Not just from a talent perspective, but is he rushing against just 6 guys in the box, etc. Second, I want to see how many positive runs he's getting. On top of that I'm looking at how many good runs (i.e. over 4 yards), though again context matters b/c a 3rd and inches 1 yard run is still pretty good. Also I like to eliminate a rushers top 10% runs and bottom 10% runs, to see where his average is.
Defensive players pass rush: Sacks are only one part of the pass rush equation so I want to look for everything, pressures, hits, even just getting penetration.