« The Falcons Had Too Many Tackles | Game 8 Preview: Jaguars »

Dr. Z Doesn't Like The Passer Rating System, Young Punks


If you've been reading for a while, you know that SI's Dr. Z annoys me sometimes with his "Things were so much better before fire" attitude towards football. There's a lot to be said for the old days. The game was more brutal and therefore more fun to watch, there was less player douchiness and diva-ism, "playing both ways" had a much cooler meaning, and nicknames were earned instead of self-imposed. But science along with an ever-increasing public interest of the game has made today's football players better than those from the previous eras. The players from the 1920s to the 1960s didn't have the training and steroids and nutrition and steroids and constant coaching and steroids that modern players all have access to. Oh, and we have steroids now.

This time around, Z goes off on a rant about the passer rating system and how mysterious and spooky it is and how it needs to be adjusted so players from past eras can look better.

Achievements have gone way above the old standards, but Elias has maintained that same system for 35 years, with the same benchmarks and the same schedule of rewards. The passing game has changed dramatically, but Elias plods on, stuck in its standards of 1973, when its system came in.

I know what Z is looking for. He wants a system that properly acknowledges the quarterbacks of the past and yet still gives Tom Brady and Peyton Manning their due. But what the current rating system measures are still the most important tangibles to a passer: completions, yards, touchdowns and interceptions. Remember, this is not a quarterback rating system. It does not measure leadership, grittiness, the ability to audible out of a play at the line, victories or who dates the hottest models. It measures the ability to pass the ball efficiently. That may be why it is called the "passer efficiency rating".

Z's inner-codger was never more obvious than when he spent half of his first page illustrating how complicated the system is and how no one understands it.

... it's a prehistoric monster that no one understands, an illogical piece of antiquity that influences so much of the game when it shouldn't.

Well, folks, I'm about to blow the doors off this motherfucker. I'm going to explain the passer efficiency rating to you. You heard me. I'm going expose the formula to the blinding white light of public scrutiny and peel away the mystery that has shrouded passer ratings for decades. This is like when that mystery dude in the mask revealed how certain magic tricks were done, except with less pyro and chicks in sparkly bikinis and more numbers. I'm cool like that. Ready?

The passer efficiency rating measures exactly four things: completion percentage, yards per attempt, touchdowns per attempt and interceptions per attempt. Each of those things is just as important as the other; they are all weighed equally in the formula. The standards are as follows:

  • 30% - 77.5% in completion percentage
  • 3 - 12.5 in yards per attempt
  • 0% - 11.875% in touchdowns per attempt
  • 0% - 9.5% in interceptions per attempt

If you rate lower than the lowest numbers in each category, it's the same as a zero. If you rate higher than the highest of those numbers, it's considered perfect. Anything in between is measured proportionally. Then the results are added together (interceptions are subtracted, though, since the lower that number is, the better.) That's really the heart of it.

There is some simple arithmetic used to bring all those measures down to a single standard, and that's the part that usually makes people's eyes glaze over. But none of that stuff affects the rating itself, just how it is represented. The way it is done now, a perfect score is 158.3 (39.58 for each category) and the worst possible score is 0. I don't know why they use that particular scale, but it doesn't matter. All that matters is the four bullet points above.

Now, how useful is it at all? I don't know. A lot of things out of the passer's control can affect the rating. A receiver can catch a 6 yard slant and then run 80 yards for the touchdown and the passer gets credit for the score. He didn't really have much to do with the touchdown, though. And the receiver can also have a perfectly thrown pass deflect off his hands and into a defender's and the passer would be penalized for the interception. You'd like to think those things balance each other out and you're left with at least a decent idea of how well the quarterback is performing as a passer. But, again, that doesn't tell the entire story.

Let's use a Buccaneer example. In 1979, Doug Williams had a passer rating of 51.2 which by today's standards sucks donkey balls. But Williams took few sacks because he was always throwing the ball away when his line couldn't protect him. Williams got the hell beaten out of him on a regular basis but still started every game that year. He was a natural leader in the huddle and his teammates trusted his skills and decision making. The team went to the NFC Championship game and no player on that team would argue that Doug Williams wasn't a major reason why. It's just not evident in his passer rating.

So, Z, just leave the rating alone. For what it's good for, it's fine. And you'll beat your bald head against a wall trying to quantify everything that makes a good quarterback. Do you really think Sammy Baugh or Sid Luckman or Dutch Clark give half a shit what their passer efficiency ratings were? Hell no. They were too busy with booze and hookers to care about anything so trivial.

Chances are very good that this particular pass fell incomplete, but the Bucs were lost without him for 15 years after he left.


Comments (6)

To quote LT in Waterboy, "this brings me to my second point, don't smoke crack."

Are you actually putting yourself in the impossible position of defending a passer rating system so severly flawed it makes the BCS look good?

GQ (of all magazines) had an excellent article on both how the passer rating scale came into existence and the multiple flaws in the system.

But most importantly, you COMPLETELY missed the boat on Dr. Z's point. He was arguing that we should use a NEW scale since what used to be excellent passing is now average. The introduction of the west coast offense and rules changes favoring passers have made quarterbacks more efficient, AND that a system created in the 70's no longer accurately measures qb performance.


I will admit that my overall premise that Z is defending the quarterbacks of the past was not explicit in his article. But you know the only reason this came up with him at all is because he has been bombarded with predictions that Tom Brady will have the best season a quarterback has ever had, and that eats him up. The best anything of all time can't occur after 1970 for him, so he needed to bring attention to the fact that the passer rating system is flawed.

But you, along with everyone else who calls the system into question, don't mention exactly what new metrics you want to include so that modern passers can be fairly graded while still remaining relevant to all historical players. Should all attempted yardage be included so we can give less weight to West Coast guys who just check the ball down to backs? What about down and distance? Maybe the score at the time of the pass? Certainly a team that is far behind is going to pass more... should that be counted more or less?

And you think the current system is complicated?

Or are you just arguing that the scales that I listed in the entry should be adjusted? Maybe 77.5% shouldn't be perfect anymore. Maybe that should be bumped up to 82% or 85%. I could get behind that, I guess. The math would change slightly, but the overall formula for measuring passers would stay the same. The article you cite points out problems in the current system, but it does end on this note: "It has flaws, yeah, but the passer rating system achieves its goal: it establishes a standard, so we can compare today's quarterbacks with one another easily, and with yesterday's to see how the game has changed." I think that was the goal of the system all along.


Here is what I recommend:

(Yards per touchdown + game minutes played) multiplied by (the dollar value of all endorsement contracts signed during the year) divided by (the team's winning percentage on grass after appearing on Monday Night Football), multiplied by (an arbitrary amount to represent the player's stripper girlfriend).

That ought to settle this.


Seriously though, I think this whole passer rating thing is garbage. It's fantasy football run amok. A great quarterback is a great quarterback because he leads the team where it needs to go.

Think back on the greats. Except for Joe Montana, we don't remember them for being "efficient passers" or having good ratings. We remember them for the qualities that can't be measured. Roger Stabach defined professional. Namath was arrogant and cool. Elway refused to lose. Mario's had perfect touch (at times). Farve was the gambler.

None of these guys had the numbers of the modern players, but I would take them over the modern guys any day.

I think the passer rating is another stupid attempt to give braindead announcers more things to talk about. They use statistics as a crutch for not understanding the game. This is just one more number that they can ooh and ahh about in the booth.

I have long ago stopped paying attention to stats. All that matters to me is whether the guy can play and whether the team wins.


changes? easy, i would change the damn scale (although don't ask me how, I never did well in math). what kind of sense does it make to have a max of 158.3?

create something that even dopes like me can understand. otherwise the qb rating will eventually be relegated to the dustbin of history anyway because no one understands it.


That's it? You're easy. Just multiply the current passer rating by .632.



Post a comment

(Because of trolling spambots and other insidious technology, I have to approve comments before they're posted. It shouldn't take long. Thanks for your patience.)

Name:
Email:
URL:
Remember personal info?
Comments:
(you may use HTML tags for style)