PDA

View Full Version : Please explain the QB rating


Moriarty
10-17-2007, 08:31 AM
I'm not a football (American version, for our non-US friends) neophyte. And I understand the basic concept of the QB rating: the higher the number, the better; 135 or so is perfect (strangely, it seems that a college QB has a higher possible rating than an NFL QB); Rex Grossman scoring a 0, not so much.

But can anybody explain to me how the QB rating is calculated? What variables go into it (I seem to recall something about yards per attempt being a factor)? Moreover, is there another methodology that might produce a better indicator of a QB's performance?

Hung Mung
10-17-2007, 09:01 AM
Firstly, the NFL and Canadian Football League use the same rating system which is different from the college system. That's why college quarterbacks can have a higher passer rating than pros.

Here's the Wikipedia breakdown (http://en.wikipedia.org/wiki/Quarterback_rating#NFL_and_CFL) of how the number is calculated and an explanation of why the perfect rating is 158.3.

Basically there are four factors: Completion percentage, yards per attempt, touchdowns per attempt and interceptions per attempt. There is a maximum value for each component in the equation. Therefore, if you throw 5 passes for 5 touchdowns, 200 yards and no INTs, because each component's value only goes so high, you only end up with 158.3.

There's another section in the link about the NCAA's passer rating system and a discussion of the good and bad of each. Check it out.

garygnu
10-17-2007, 09:35 AM
Note that it's a passer rating, not a "QB" rating, as it does not measure things like footwork, release time, fumble tendancy, sakcs, etc. A throwaway pass to avoid a sack counts against the rating, while taking a sack does not.

Wee Bairn
10-17-2007, 09:35 AM
I always wondered why the 50th best passer in college had a better rating than the best NFL QB.

And Peyton has 4 of the 17 perfect games in the NFL, and Marino, Favre, Elway and Brady combnied have none? Interesting.

Wee Bairn
10-17-2007, 09:41 AM
And the all time rankings show that John Elway is one overrated dude. Ditto Bradshaw.

garygnu
10-17-2007, 09:52 AM
And the all time rankings show that John Elway is one overrated dude. Ditto Bradshaw.
Yeah, and the six rings between them?

Kid_A
10-17-2007, 10:02 AM
And the all time rankings show that John Elway is one overrated dude. Ditto Bradshaw.
There isn't a more useless stat in professional sports than QB rating.

Wee Bairn
10-17-2007, 10:11 AM
Bradshaw has four rings, but also barely completed 50% of his career passes, and only had a few more TD's than interceptions- these stats would get him a backup job at best in 2007. I know the rating isn't perfect, and I know eras are different, but I think it helps separate the true greats (Marino, Manning, Montana) from the dubious ones- (Bradshaw, Namath, Elway, etc.). Bradshaw was one solid component of a great team, not a great player, like Steve Kerr in the NBA has five rings, but no one thinks he's a hall of famer. QB's get too much credit for successes and failures both.

As for the rings, Bradshaw was on a team with like 10 HOF'ers for ten years. If the Ravens had kept their team intact for a decade, Trent Dilfer may well have four rings instead of one, same for Brad Johnson. And if you look at Elway's stats, excluding passer rating, you'll see an uncanny increase in his quality around 1995- around the time Terrell Davis joined the team. He was the real MVP of those teams, not Elway.

kidchameleon
10-17-2007, 10:11 AM
There isn't a more useless stat in professional sports than QB rating.

I dunno, Shaq has a pretty good 'field goal' %....

Wee Bairn
10-17-2007, 10:34 AM
Bradshaws average game- 12-23 166 yards, 1 TD, I INT. These are HOF numbers?
Elway- 18-31, 220, 1.25 TD 1 INT, a bit better, but no cigar.

Rough averages for some current average or worse QB's, some with rings:

Trent Dilfer- 13-23 155 1 TD 1 INT
Brad Johnson- 17-28 189 1 TD 0.75 INT
Rex Grossman 15-28, 185 1TD 1 INT
Vinny T. 16-29 200, 1TD 1 INT
Eli Manning 17-31 201, 1.25 TD 1 INT

If you're going to judge talent by rings, then every 70's Steeler at every position is the best ever at his position. Obviously some voters think like this, as Lynn Swann being in the HOF with an average of three catches per game is another friggin joke.

CJJ*
10-17-2007, 10:40 AM
I don't disagree with the formula, but a little math reveals the reasoning behind it.

The rating is basically an enhanced "yards per attempt" statistic. It's somewhat intuitive that a passer who can acquire more yards each throw is "better", however you'd also like to reward a QB with a better completion percentage, since a passer who hits for 50 yards on one of five attempts is not as helpful to your team as a consistently perfect passer who gets 10 yards every time he throws the ball. So the yardage total is enhanced by adding 20 yards for every completion. You'd also like to account for touchdowns and interceptions, so add another 80 yards for each TD (really 100, since each touchdown also necessarily includes a pass completion), and subtract 100 yards for every interception.

This "enhanced" yardage total, divided by pass attempts, is fairly close to passer rating (it is directly proportional to ithe NFL number). The NFL mucky-mucks wanted to put this on a scale similar to how most folks were graded in school (100 is an "excellent" rating, ~95% is an A, ~87% is a "B", etc.) so you need to divide this total by six and treat the result as a percentage. Finally, you need to add 6.25% (call it "extra credit" :) ) to get the final NFL figure.

zamboniracer
10-17-2007, 10:46 AM
The NFL QB rating seems to put too much weight on TD pass percentage. Yards per attempt is a better measure, adjusted for interception by deducting 40 yards per interception, as per Allen Barra and George Ignatin's book "Football by the Numbers 1986".

Philster
10-17-2007, 10:52 AM
Bradshaw and Elway didn't exactly play in domes or a sunny paradise.

A QB is severely hampered by weather. If you play in the NFC East, you get such paradises as Giants Stadium, or the old Veterans Stadium in Philly.

Brady is enjoying an usually warm autumn -- probably record warmth and calm. He is on a record pace...no coincidence. Watch his numbers flatten as the season gets more nasty.

When its 26f or -8 celsius, and the winds are swirling around, a good QB will post average ratings, keep the team mistake free and make decisions as the environs dictate. You can watch a winner, like Elway, during these times, or you can watch a stat machine (Manning) flaunt his numbers in Domesville.

John DiFool
10-17-2007, 11:29 AM
And Bradshaw spent a goodly portion of his peak years before the 1978 liberalization of the passing game. He was admittedly horrible before 1975 (which includes an injury-plagued 1974 title year), and what's been forgotten since then is all the nasty articles written about him in this time period by the press, but from 1975 on (age 27+) he was pretty damned good. Averaged close to 8 yards per pass, which is outstanding (Marino and Montana exceeded 8.0 only twice each in their careers), TD/Int was 164/129, which you can probably live with.

I've made no bones about my dislike for Elway elsewhere/when, so I'll just say his peak also occured kind of late, starting in 1993 (age 33), when his TD's shot up and INT's went down; before then he was pretty mediocre, but came up big in a few big games and got the rep he needed to breeze into the Hall.

Wee Bairn
10-17-2007, 12:41 PM
In fairness to pre 80's QB's, I believe that it used to be the defense was allowed to hit the receiver as soon as the play started and then as often as he wanted- there was no five yard rule for contact like there is now, IIRC, which may partially account for the reason modern QB's, even those thought of as average, dominate the career QB rating rankings, and have much better completion percentages.

Moriarty
10-17-2007, 06:40 PM
Thanks for the input.

I think that football, as the ultimate "team game", doesn't mesh well with individual stats. That's why, when you see the mediocre numbers of some early QBs, you are really missing the point: this is a game about intangibles, and ultimately about wins. Flashy numbers are simply overrated (I say this having been a longtime Dolphins fans, who used to argue vehemently about the greatness of Dan Marino).

If you want to judge players, we need to think about their actual performances, and not just the stats that try and represent those performances. Watch some highlights of Lynn Swann, for example, and it's easy to realize he's a HOF player, his 3 catch per game average notwithstanding. And how, exactly, should we decide a lineman is great, since there are few stats (pancake blocks, anyone?) to measure them by?

That's why I was interested in the QB rating. It's the most stat-intensive measure in football, a game really best judged by the final score and how we got there. Any measures of a QB's performance that don't include the greats from yesteryear (including Bradshaw) among the best is a flawed analysis.

Public Animal No. 9
10-17-2007, 08:10 PM
...like Steve Kerr in the NBA has five rings, but no one thinks he's a hall of famer. ..
Aw, don't rag on my buddy Steve. He was clearly the force behind all those championships - he won with two different teams behind him, and let those supporting "no-names" take all the credit! :p

Marley23
10-17-2007, 08:26 PM
In fairness to pre 80's QB's, I believe that it used to be the defense was allowed to hit the receiver as soon as the play started and then as often as he wanted- there was no five yard rule for contact like there is now, IIRC, which may partially account for the reason modern QB's, even those thought of as average, dominate the career QB rating rankings, and have much better completion percentages.
There's also the proliferation of the West Coast offense.

astorian
10-17-2007, 08:45 PM
Has the formula for quarterback ratings been changed at al lsince the mid-Seventies?

I ask because, today at least, the rankings place most quarterbacks almost exactly where they belong. But in the Seventies, I think the ratings tended to give too much credit to dink passers with high completion percentages.

BobT
10-17-2007, 08:53 PM
I'm pretty sure the passer rating formula has been constant.

If it's been changed, you would have to retroactively refigure it for everyone or else the numbers wouldn't be comparable.

RealityChuck
10-17-2007, 09:30 PM
Has the formula for quarterback ratings been changed at al lsince the mid-Seventies?

I ask because, today at least, the rankings place most quarterbacks almost exactly where they belong. But in the Seventies, I think the ratings tended to give too much credit to dink passers with high completion percentages.I think QBs were ranked by things like total yards and yards per pass before the rankings were set up. That actually made some sense -- it's similar to the way baseball ranks batters (there are leaders in different offensive categories, but no one tries to put it all together for one number that defines the best hitter).

Thus looking at the 1968 AFL, John Hadl was probably considered the top QB, since he led in yardage, completions, and TDs.

BobT
10-17-2007, 10:00 PM
If Major League Baseball issued a statistic that graded each player with a specific number, it would be picked apart within seconds and no one would pay any heed to it. But the NFL issued the passer rating and the media basically lapped it up.