How could a website with the name "Advanced NFL Stats" let the occasion of Calvin Johnson's record breaking season go by without making note? Saturday night Johnson broke Jerry Rice's record of most total receiving yards in a single season with one game still to play.
But the truth is that the record for most receiving yards (or rushing yards or whatever yards) is not "a statistic." It's trivia. It's the answer to a question on the ESPN sports-trivia-a-day calendar you just bought for your nephew for Christmas. It's the kind of thing that gives stats a bad name, or at least a misleading one.
Don't get me wrong. I think Calvin Johnson is awesome, and all the numbers agree. Fans and media are right to make a big deal out of such an accomplishment. It's just that records like this are what most people think of when they hear the word statistics. It's understandable then, that when coaches or players think of stats, they dismiss them as pointless or "for losers." After all, how could knowing who holds the record for most receiving yards in a season, or memorizing how many yards that is, possibly help a team win?
I sometimes find myself in a situation like this. I'm in an airport or on a plane working on the site from my laptop, and a fellow traveler discovers I'm that stats guy he read about. "Oh, you're the guy that does all those football statistics?"..."Yup, that's me."..."Oh, neat. So tell me, is it true that Mark Brunell has the record for the most interceptions by a left-handed quarterback?" It's at this point I realize I have no hope of explaining what I really do.
I'm under no delusion it's possible to change the common perception of statistics as little more than giant catalogs of trivial facts. I just called this stuff stats because that's what we called the classes in grad school that covered things like regression and other similar techniques. At first I resisted the term analytics because it sounded like one of those made-up corporate-speak words that made something that had been around for generations sound novel and unnecessarily complex. My word processor doesn't even think analytics is a word....add to dictionary...ok, now it does. But now I'm fond of the term because it avoids the stigma of trivia. It also rolls off the tongue a little smoother than operations research. At least football is free of a term like sabermetrics, which has a nerd-quotient barely above that of Dungeons & Dragons.
No matter what we call this stuff, it won't matter to my chatty friend in seat 14B. "No. It's actually Ken Stabler," I say. Boy, is he impressed.
- Home
- Unlabelled
- Calvin Johnson's Receiving Record and "Stats"
Subscribe to:
Post Comments (Atom)
Cybermetrics.
The singular is "trivium".
Thank youum.
'arithmetic' is probably the best word.
:)
happy holidays everyone.
Good post.
counting. Just call it counting.
NFL data analysis consulting
Probabilistic analysis or probabilistic statistics
I think where most "anti-stats" folks go wrong is in the misreading of stats. Jon Gruden noted on the broadcast last night that Johnson's record featured lots of "meaningless" yardage--as if to say that only yards gained in the commission of a win, or at least a closely contested loss, should count--and a possible record number of pass attempts by Matt Stafford. But Johnson's record says nothing about whether it was the "best" season by a wide receiver, nor if it was a winning season; just that he had the most yards by a player at his position. Full stop. That means SOMETHING, it just doesn't mean what Gruden wants it to mean.
Stats, analytics, trivia, whatever you want to call them, they only mean what they mean.
Funny, during the Lions game Gruden said something like "that's great, the Lions gain a lot of yards, but the stat that matters is wins. There's no stat that tells you how those yards turn to wins..." I remember thinking "yes there is..." Now I get why WPA is cool.
"Green Bay has the 32nd ranked defense statistically" - Joe Buck, 2011
I thought you were going to do an advanced analysis of Calvin Johnson's season.
Good post, but I can't agree with the implication that a high nerd-quotient is a bad thing.
The problem is, the average person knows almost nothing about probability. I think it should be integrated into the curriculum in elementary school, and have its own required course in high school and college.
The problem is, the average scientist, engineer, or mathematician knows almost nothing about probability.
I would say roughly 80-90% of "technical" people, and probably 95-98% of the population in general (the "average person") thinks we live in a deterministic world, not a stochastic one. Football is a great example. "He went for it on 4th and short, and didn't make it, so he should have punted". Not "He went for it on 4th and short and there was a 90% chance that that would have been better than punting, but he got unlucky and hit the 10%". I once demonstrated the power of probability to my (non-technical) sister by beating her repeatedly in a Monte-Hall type game, and her response was that I was winning due to "mathematical chicanery".
Agree that probability should be integrated into curriculums earlier, but at least making mathematically and science focused study it as part of their curriculum would be a start.
Nate Silver abhors a trivium
Statistics has a second dictionary definition that is "a collection of quantitative data". I wonder what the stats are on the percentage of time "statistics" is meant to be the first versus the second definition - or in your case, how often it is used to mean the first definition but taken by the Howdydoodats to mean the second?
While it's fun to play with words, I think your skills call for you to leave semantics aside and spend some time coming up with an analysis to capture just how exceptional Charles Tillman's performance was against Johnson this year. He allowed four (4) catches per game and 53 ypg for Johnson on 12.5 targets per game, in a season where Johnson's average in other games were 8/128/12 in this historic season. That's pretty epic.