The following is present at ESPN Insider. If anyone pays for that, we'd appreciate some info on what's in that article, as its teaser gives an indication that some quarterbacks with ... wait a minute, Jeff Garcia already has great passer ratings, so that's even more of a case for him -if- a change makes his passer ratings go up (assuming the change is reasonable).
As for David Carr and Quincy, they could certainly benefit from a change that would improve their low ratings thus far in their careers ...
"In search of a more complete QB rating -- The NFL's passer rating formula gets a makeover, and Jeff Garcia, David Carr and Quincy Carter benefit. ..."
x-- 100 Proof News - http://www.100ProofNews.com x-- 3,500+ Binary NewsGroups, and over 90,000 other groups x-- Access to over 1 Terabyte per Day - $8.95/Month x-- UNLIMITED DOWNLOAD
By Ryan Early NFL Insider Tuesday, February 24 Updated: February 25 11:25 AM ET
Quarterback is easily the most difficult position in football, perhaps in all of sports. No other position requires a combination of intelligence, strength, durability and quick reactions.
In 1973, the NFL devised a formula to calculate how well its quarterbacks were playing. The result, called the NFL passer rating, was so complex that figuring out how to calculate it might be harder than actually playing quarterback. The formula balances four different statistics to produce a passer rating:
a.. Completion percentage
b.. Passing yards per attempt
c.. Touchdown percentage
d.. Interception percentage But the formula was designed using league statistics from 1960 through 1972, and needless to say, the game has changed since then. Teams pass more often these days, and they use more high-percentage, short passes than high-risk, long passes. Rules also have been changed to make it easier to pass, thus opening up the game and increasing offense.
But the formula is weighted heavily in favor of the offensive trends of the day, with teams more inclined to run the ball, but tending to throw deep when they did throw. Thus the number of passing yards per attempt and the percentage of passes that produce touchdowns were actually higher in the early '70s. But it was harder to complete those long passes and easier to throw interceptions.
When it was introduced, the passer rating formula produced an "average" passer rating of 66.7. Because of the current emphasis on shorter and more controlled passing, passer ratings now average out almost 14 points higher.
But modern offenses really cry out for more modern evaluation tools. What follows is an attempt to produce just that - a more complete passer rating.
Sacks The passer rating formula was created before the sack became an official NFL statistic, thus it wasn't included, and a sack is not considered a pass attempt on the QB's stat sheet. If a quarterback throws the ball away to avoid a sack, he protects field position and down-and-distance, but his completion percentage, yards per attempt and touchdown percentage all take a hit. Moreover, his passer rating suffers. But if he takes the sack, hurting his team's chances of making a first down and maintaining possession, he can protect his passer rating.
With the passer rating stat often used to mark incentive thresholds in contracts, shouldn't the sack, which has become an integral piece of the passing game, be included in the formula? A quarterback shouldn't have incentive to take a sack. The situation can be easily rectified simply by classifying a sack as a pass attempt and subtracting the sack yards from the gross passing yards.
Of all quarterbacks with at least 224 pass attempts, Detroit's Joey Harrington was the least-sacked QB for the second season in a row. He went down just 9 times in 2003, or once every 62.6 times he dropped back to pass (the league average is once every 16.1). Harrington also was flagged twice for intentional grounding while trying to avoid a sack. Adding both of these stats into the passer rating formula would lower Harrington's QB rating just 1.8 points.
At the other end of the spectrum is Buffalo's Drew Bledsoe, who was sacked once every 10.6 pass attempts. His 49 sacks for 371 yards lost would lower his QB rating a whopping 9.7 points.
Factoring in Sacks Player Team Sacks + Off. Penalties Yds Lost Att/ Sack NFL Rating Adj. Rating Diff. Top 5 Joey Harrington DET 11 75 51.4 63.9 62.1 -1.8 Brad Johnson TB 20 111 29.5 81.5 78.1 -3.5 Peyton Manning IND 20 123 29.3 99.0 94.8 -4.2 Trent Green KC 20 130 27.2 92.6 88.2 -4.3 Brett Favre GB 20 142 24.6 90.4 85.6 -4.8 Bottom 5 Patrick Ramsey WAS 30 206 12.2 75.8 67.5 -8.4 Daunte Culpepper MIN 41 226 12.1 96.4 86.7 -9.7 Donovan McNabb PHI 46 280 11.4 79.6 70.6 -9.0 Kordell Stewart CHI 26 156 10.7 56.8 49.3 -7.5 Drew Bledsoe BUF 49 371 10.6 73.0 63.4 -9.7
Certainly sacks aren't entirely the fault of the quarterback, but a statistic designed to measure a QB's efficiency should reward those who avoid sacks and penalize those who don't. Some quarterbacks, like Bledsoe, have shown throughout their career the propensity to hold onto the ball too long. Bledsoe's attempts per sack have been near the bottom of the league every season.
Running Quarterbacks In addition to not penalizing a QB for taking sacks, the passer rating formula also ignores positive yards a quarterback gains with his feet. Designed to measure how well a quarterback passes the ball, the formula doesn't account for mobile quarterbacks who can escape the pocket and pick up yards on their own. If we change the NFL's formula to include sacks, why not also include QB running plays?
To avoid rewriting the formula entirely, we should try to keep it as close to the original formula as possible: A rush attempt could be included as a pass completion. Rushing touchdowns could be added to passing touchdowns, and fumbles lost could be counted along with interceptions.
Running Quarterbacks Player Team Rush Att Rush Yds Fumbles Lost NFL Rating Adj. Rating Diff. Top 5 Kordell Stewart CHI 59 290 3 56.8 65.4 +8.5 Jeff Garcia SF 56 319 3 80.1 86.2 +6.0 David Carr HST 27 151 0 69.5 74.9 +5.4 Josh McCown ARZ 28 158 2 70.3 73.3 +3.0 Donovan McNabb PHI 71 355 3 79.6 82.6 +3.0 Bottom 5 Jake Delhomme CAR 42 39 6 80.6 76.9 -3.7 Patrick Ramsey WAS 15 62 5 75.8 72.0 -3.8 Kyle Boller BLT 30 62 5 62.4 57.9 -4.5 Aaron Brooks NO 54 175 11 88.8 82.9 -5.9 Drew Bledsoe BUF 24 29 10 73.0 66.8 -6.2
Daunte Culpepper led the league in rushing yards by a quarterback with 422, yet he also lost six fumbles, canceling out his hard-won statistical gains.
There were several quarterbacks who saw their rating drop after we added rushing stats. These were the traditional pocket passers who need a great offensive line to protect them. Aaron Brooks, who gained 175 rushing yards but led the league with 11 lost fumbles, is the exception -- a mobile QB who actually lost rating points when his rushing stats were included.
Penalties Regaining Lost Yards Player Tm Penalties Yards Donovan McNabb PHI 27 288 Matt Hasselbeck SEA 16 151 Kerry Collins NYG 15 149 Tommy Maddox PIT 15 134 Marc Bulger STL 15 113 Jeff Blake ARZ 14 139 Aaron Brooks NO 13 218 Jon Kitna CIN 13 139 Byron Leftwich JAX 12 189 Peyton Manning IND 12 126 Another play that can make a huge impact on a game but doesn't show up in individual stats is the pass-interference penalty. In fact, penalties aren't considered "plays" at all. While most penalties come in 5-, 10-, and 15-yard increments, pass interference advances the ball to the spot of the infraction on the assumption that if there hadn't been interference, the receiver would have caught the ball. So the quarterback makes a good throw and the receiver beats his defender, but neither the QB nor the receiver get credit for the play in their individual statistics.
As an example, consider Kyle Boller's debut performance in Week 1 against the Steelers. He completed 22 of 43 passes for 152 yards, with one TD and one INT, for a less-than-stellar 57.5 passer rating. Not included in those stats, though, are three pass interference penalties and one defensive holding penalty totaling 100 additional yards gained. If those four non-plays were added to Boller's stat sheet, his yards-per-attempt would have increased 1.6 yards and his passer rating by 11.3 points.
Defensive pass interference, defensive holding and other such penalties only occur a couple of times per game, so including them in our formula only adjusts the overall ratings by a couple points. But still, a couple of points here and there can make a big difference in evaluating QBs.
The Complete Quarterback Rating The end result of our tinkering with the passer rating formula takes into account everything a QB does on the field to help (or hinder) his team. Here's what we have done to the formula:
e.. Pass attempts: This becomes "total plays," which consists of pass attempts, sacks, runs and penalties.
f.. Completions: This becomes "positive plays," which consists of pass completions, rushes and penalties.
g.. Passing yards: This becomes "total yards," which consists of passing yards, rushing yards, net penalty yards and sack yards lost.
h.. TD pct.: This becomes "total touchdown pct.," which adds passing TDs and rushing TDs, then divides by total plays.
i.. INT pct.: This becomes "turnover pct.," which adds interceptions and fumbles lost, then divides by total plays. Stats Comparison: Passer Rating vs. Adjusted Rating Official Passer Rating Stats (Average for top 36 qualifiers) Pass Att. Pass Com. Comp.% Pass Yards Yds / Att Pass TD INT QB Rating 393.4 233.6 58.7% 2648 6.66 15.9 12.7 78.0 Adjusted QB Rating Stats (Average for top 36 qualifiers) Total Plays Gains Gain % Total Yds Yds/Play Total TDs Total TOs Adj. Rating 461.3 275.1 59.1% 2700 5.76 17.4 16.1 72.7
The adjusted stats increase significantly the total number of plays upon which a QB's performance is measured. There are modest increases in yards gained, percentage of plays with gains, and total touchdowns over the corresponding stats from the traditional passer rating. And there are large decreases in yards per play and total touchdowns when runs, sacks and fumbles are incorporated. Because of that, the average adjusted rating is 72.7, more than 4 points lower than the average NFL passer rating but still higher than the 66.7 rating the
...
Thanks for posting that. It takes a complicated passer rating formula, rethinks additional factors, and comes up with an even more complicated passer formula, mixing non-passing data with passing data for a different rating, not really a passer rating, but instead, a passer-runner-sack-penalty-fumble rating.
Even though the article is written in a very "rah-rah, isn't this much better" way, when you think about what is proposed, it doesn't really come across as strong as the writer is trying to make it seem.
It does mention some aspects of passer rating touched on in this newsgroup for the past few months, such as the fact the average passer rating has gone up over the years.
It overly penalizes quarterbacks for sacks, counting those against a quarterback as if it's a pass attempt. It disregards the fact that potent passing attacks have outlet receivers -and- the more accurate passers can avoid sacks and make completions (not always, but much of the time) with accurate reading of blitzes, passing accuracy, and well-designed of- fensive schemes. Throwaways are quite often results of deficiencies in the offensive scheme, the offensive adjust- ments, the quarterback's inability to accurately throw the ball, and the quarterback's shortcomings in reading blitzes.
Counting a quarterback being sacked by a yard or more as a pass attempt and counting quarterbacks running it in as passing TDs is very contentious, heavily favoring running QBs over pocket passers. A quarterback running it gets credit, in the running stats, not the passing stats, and that is as it should be. A running quarterback should benefit in the passing stats simply by posing a running threat. Some do, some don't. That's a matter more of passing accuracy than anything else, and if a running quarterback is still a lousy passer, he shouldn't gain passing credit for his run- ning. It's simply confusing the issue. It would be like adding passing yardage to running back stats for the few times a runner throws the ball, or adding rushing yardage to a wide receiver's receiving yardage stats for the few times a wide receiver runs the ball.
It's mixing apples and oranges.
Counting pass interference and defensive holding as pass completions, also very contentious, especially since a signifi- cant number of pass interference calls are disputed (it's one of the most disputed calls there is). In addition, many feel it shouldn't be a 'spot' foul, but instead, should be a 10 or 15 yard penalty. Defensive holding counting as a completion is highly suspicious being that there's no way one can know if a pass would've been made and completed.
Fact of the business is, the writer is adding much more com- plexity, and doing little but mixing run & pass & penalty & fum- bling stats (for QBs) and penalizing pocket passers (counter to the way in which pocket passers are the most successful quarterbacks in the league). I suspect the writer's bias is to favor running QBs, and to attempt to achieve that, he's trying to throw running type information into the passer rating. I sup- pose, next he'll want to start adjusting runner yards based on defensive penalties that, had they not occurred, the runner would've made much more yardage.
Oh, and what about nit-picky penalties that are major game changers, like a movement penalty on a wide receiver apart from the play, yet changing a 44 yard TD to a 5-yard loss, -or- an incidental penalty like a defensive lineman lining up an inch over the line marker, being called for offsides, and costing his team an interception, and so on and so forth.
Perhaps the writer would want to add interceptions-that-should've been (had the interceptor not dropped the ball, or had a penalty not been called) -or- receptions-that-should've-been (had the receiver not dropped the ball). I then, there's accuracy, perhaps the writer would like to judge how well the quarterbacks hit he receivers in stride. How about missed open receivers, the de- gree to which quarterbacks miss open receivers, perhaps that factor should be involved, if accuracy is the goal, regardless of complexity, here.
- - -
Plain and simple, if you want to see quarterback running stats, simply go to a good stats page, list the quarterbacks, and voila, there you have it. Sacks? Same deal, they're listed on good stats pages, though what's not listed are the whys and where- fores.
One clear factor in passer ratings not touched on by the writer, quality of competition. That, more than any other factor I've run across, can have a significant impact on passer ratings, and in a wise upgrade to passer ratings, that factor would have to be the primary one added, so that a passer with a rating of 70 against tough competition was bumped up, and a passer with a rating of 70 against weak competition was bumped down. However, assigning level of competition would be very subjec- tive and contentious, unless a simple formula (like end of year strength of schedule -or- end of year passer defense ratings) was arrived at.
And, speaking of complexity, what about the running game? Many say that directly impacts a passer's performance, yet that isn't included in the writer's list of "adjustments".
All-in-all, as mentioned previously in a detailed discussion of this topic, very disputatious, any effort to change the passer rating, and nowhere near as straight-forward as the writer tried (in vain, in my opinion) to make it appear to be.
x-- 100 Proof News - http://www.100ProofNews.com x-- 3,500+ Binary NewsGroups, and over 90,000 other groups x-- Access to over 1 Terabyte per Day - $8.95/Month x-- UNLIMITED DOWNLOAD
> Thanks for posting that. It takes a complicated passer rating > formula, rethinks additional factors, and comes up with an > even more complicated passer formula, mixing non-passing > data with passing data for a different rating, not really a passer > rating, but instead, a passer-runner-sack-penalty-fumble rating.
> Even though the article is written in a very "rah-rah, isn't this > much better" way, when you think about what is proposed, it > doesn't really come across as strong as the writer is trying > to make it seem.
Are you crazy? The new rating ranked the QBs a LOT closer to the reality of how they played this last year than the "normal" QBR. This rating should be the one they use in the future.
observer <obser...@observer.net> wrote: > Fact of the business is, the writer is adding much more com- > plexity, and doing little but mixing run & pass & penalty & fum- > bling stats (for QBs) and penalizing pocket passers (counter > to the way in which pocket passers are the most successful > quarterbacks in the league). I suspect the writer's bias is to > favor running QBs, and to attempt to achieve that, he's trying > to throw running type information into the passer rating. I sup- > pose, next he'll want to start adjusting runner yards based on > defensive penalties that, had they not occurred, the runner > would've made much more yardage.
C'mon, Observer. He's just quoting hard statistics and giving a reasonable interpretation to them. On what basis do you suspect a bias on his part? What makes his use of statistics any more biased than the current QB rating system?
So much for the idea that using statistics are intrinsically 'objective...'
On Sun, 29 Feb 2004 15:33:29 -0600, "observer" <obser...@observer.net> wrote:
>Thanks for posting that. It takes a complicated passer rating >formula, rethinks additional factors, and comes up with an >even more complicated passer formula, mixing non-passing >data with passing data for a different rating, not really a passer >rating, but instead, a passer-runner-sack-penalty-fumble rating.
It's a Quarterback rating, rather than a passer rating. Much more encompasing, being as how a QB does much more than just pass.
>Even though the article is written in a very "rah-rah, isn't this >much better" way, when you think about what is proposed, it >doesn't really come across as strong as the writer is trying >to make it seem.
>It does mention some aspects of passer rating touched on >in this newsgroup for the past few months, such as the fact >the average passer rating has gone up over the years.
>It overly penalizes quarterbacks for sacks, counting those >against a quarterback as if it's a pass attempt. It disregards >the fact that potent passing attacks have outlet receivers >-and- the more accurate passers can avoid sacks and make >completions (not always, but much of the time) with accurate >reading of blitzes, passing accuracy, and well-designed of- >fensive schemes. Throwaways are quite often results of >deficiencies in the offensive scheme, the offensive adjust- >ments, the quarterback's inability to accurately throw the ball, >and the quarterback's shortcomings in reading blitzes.
>Counting a quarterback being sacked by a yard or more as >a pass attempt and counting quarterbacks running it in as >passing TDs is very contentious,
It wouldn't be called passing attempt or passing TD. It would be total plays and total TD's.
>heavily favoring running >QBs over pocket passers.
Nonsense. One of the least sacked QB's in history was ALSO one of the least mobile...or haven't you heard of Dan Marino? The point was, and is, that taking a sack is better for the current QBR than throwing the ball away...even though the latter is MUCH better than the former for the team.
> A quarterback running it gets >credit, in the running stats, not the passing stats, and that >is as it should be.
Not if you are going to insist on using QBR as a measure of a QB's worth to his team it isn't.
> A running quarterback should benefit in >the passing stats simply by posing a running threat. Some >do, some don't. That's a matter more of passing accuracy >than anything else, and if a running quarterback is still a >lousy passer, he shouldn't gain passing credit for his run- >ning.
You are taking this way too far. He SHOULD get credit for whatever positive (or negative) yards HE is responsible for in a true QBR. What is needed for a semi-fair comparison is a comprehensive QUARTERBACK rating system. Not just a passer rating. Currently the QBR DOES only take into account passing, therefore it is NOT good for rating a quarterbacks overall performance.
>It's simply confusing the issue. It would be like adding >passing yardage to running back stats for the few times a >runner throws the ball, or adding rushing yardage to a wide >receiver's receiving yardage stats for the few times a wide >receiver runs the ball.
The author didn't do a good job of explaining their position. They are looking for total yards, not literally calling rushing yards passing yards and not literally calling a rushing TD a passing TD.
>It's mixing apples and oranges.
No it isn't. It's giving the QB credit for EVERYTHING they do. Not JUST when they put the ball in the air.
>Counting pass interference and defensive holding as pass >completions, also very contentious, especially since a signifi- >cant number of pass interference calls are disputed (it's one >of the most disputed calls there is).
Huh? What are you talking about? You CAN'T dispute a PI call as it's a judgement call.
>In addition, many feel >it shouldn't be a 'spot' foul, but instead, should be a 10 or >15 yard penalty.
What moron thinks that? If THAT was the case, every time a player was downfield with a chance to catch a ball he'd get mauled. Only an idiot would advocate changing the PI rule to a 10 or 15yd. penalty.
>Defensive holding counting as a completion >is highly suspicious being that there's no way one can know >if a pass would've been made and completed.
I'll agree with that.
>Fact of the business is, the writer is adding much more com- >plexity,
Quite neccessary in such a complex position which is currently judged in a very simplistic manner now.
>and doing little but mixing run & pass & penalty & fum- >bling stats (for QBs) and penalizing pocket passers (counter >to the way in which pocket passers are the most successful >quarterbacks in the league).
How do you figure? The top 3 passer rating guys in the league would stay in exactly the same position in either system. The current system rewards guys who hold the ball too long and take a sack over the guys who get rid of the ball and actually help their team.
>I suspect the writer's bias is to >favor running QBs, and to attempt to achieve that, he's trying >to throw running type information into the passer rating.
Great. Except a QB is much more than just a "passer". He's supposed to be the field general, the leader, the coach on the field, etc. Passer rating ONLY addresses what he does when he throws the ball. Nothing else.
> I sup- >pose, next he'll want to start adjusting runner yards based on >defensive penalties that, had they not occurred, the runner >would've made much more yardage.
Uh, there is no runner rating system...
>Oh, and what about nit-picky penalties that are major game >changers, like a movement penalty on a wide receiver apart >from the play, yet changing a 44 yard TD to a 5-yard loss, -or- >an incidental penalty like a defensive lineman lining up an inch >over the line marker, being called for offsides, and costing his >team an interception, and so on and so forth.
What about 'em? Dead plays. The ONLY penalty that makes sense is the PI gain, but I'd even agree to throw that one out, too.
>Perhaps the writer would want to add interceptions-that-should've >been
>(had the interceptor not dropped the ball, or had a penalty >not been called) -or- receptions-that-should've-been (had the >receiver not dropped the ball). I then, there's accuracy, perhaps >the writer would like to judge how well the quarterbacks hit he >receivers in stride. How about missed open receivers, the de- >gree to which quarterbacks miss open receivers, perhaps that >factor should be involved, if accuracy is the goal, regardless of >complexity, here.
>- - -
>Plain and simple, if you want to see quarterback running stats, >simply go to a good stats page, list the quarterbacks, and voila, >there you have it. Sacks? Same deal, they're listed on good >stats pages, though what's not listed are the whys and where- >fores.
>One clear factor in passer ratings not touched on by the writer, >quality of competition. That, more than any other factor I've >run across, can have a significant impact on passer ratings, >and in a wise upgrade to passer ratings, that factor would have >to be the primary one added, so that a passer with a rating of >70 against tough competition was bumped up, and a passer >with a rating of 70 against weak competition was bumped down. >However, assigning level of competition would be very subjec- >tive and contentious, unless a simple formula (like end of year >strength of schedule -or- end of year passer defense ratings) >was arrived at.
>And, speaking of complexity, what about the running game? >Many say that directly impacts a passer's performance, yet >that isn't included in the writer's list of "adjustments".
>All-in-all, as mentioned previously in a detailed discussion of >this topic, very disputatious, any effort to change the passer >rating, and nowhere near as straight-forward as the writer >tried (in vain, in my opinion) to make it appear to be.
IOW anything that addresses anything other than pure passing is a waste of time, eh?
> > Fact of the business is, the writer is adding much more com- > > plexity, and doing little but mixing run & pass & penalty & fum- > > bling stats (for QBs) and penalizing pocket passers (counter > > to the way in which pocket passers are the most successful > > quarterbacks in the league). I suspect the writer's bias is to > > favor running QBs, and to attempt to achieve that, he's trying > > to throw running type information into the passer rating. I sup- > > pose, next he'll want to start adjusting runner yards based on > > defensive penalties that, had they not occurred, the runner > > would've made much more yardage.
> C'mon, Observer. He's just quoting hard statistics and giving a > reasonable interpretation to them. On what basis do you suspect a bias > on his part? What makes his use of statistics any more biased than the > current QB rating system?
The passer rating system measures passing, not sacks, not fumbles, not running, not penalties. He's simply trying to present a case that adding non-passing data will yield an accurate quarterback rating which, point-in-fact, it won't. It would yield a confused rating adding arbitrary complex- ity and non-passing information into the passing stats. It would bias the ratings towards running QBs. It would change the rating system from being about purely pas- sing to being about passing, fumbles (not a passing activity), running (only performed by some QBs), pen- alty impacts (with the most controversial penalty of all, pass interference, along with defensive holding).
As stated previously (in a response to Astaroth on January 3, 2004, regarding passer ratings, and ideas to change it, ideas that did not include penatly-adjust- ments or sack-adjustments) ...
The thing is, when it comes to quarterbacks, 'QB rating' *is* 'passer rating', presently, and changing it to 'passer/ runner/receiver/touchdown scorer' rating would beget, I suspect, a split in QB categorizations: 'pocket QBs' rated separately from 'scrambler/runner QBs'. At the present time, with most 'scrambler/runner QBs' being black, and most 'pocket passers' being white, that would bring up divisiveness based on race, and I assure you the NFL will *never* create split categorizations that would promote a race-divisive approach to QB ratings.
Most in the NFL, even those with 'scrambler/runner QBs', want their quarterbacks to be perceived as passers, with something extra. Perhaps it's best to keep it simple, rating all QBs based on their passing, only, on a level playing field, and leaving the something extra as something extra, separate from 'QB ratings'.
If you combine the two, giving 'scrambler/runner QBs' a one-up in the rating system, that's an endorsement of 'scrambler/runner QBs' over 'pocket passers'. That's an evaluation of a QB as a passer -and- as a scrambler/runner, something many (most?) feel is not a valid measurement of QB skills, for that exists on an uneven playing field, as most QBs are *not* 'scramblers/runners'.
As stated, in another post, -if- a 'scrambling/running' QB benefits, in passing, from his 'scrambling/running' ability, therein resides his gain from that skill, in the QB rating. If he does not, if he in fact is hindered or injured by 'scrambling/running', it would be erroneous to give him bonus 'points', in QB passing evaluation, for having an extra ability that detracts from what QBs are paid 'big bucks' to do, and that's complete passes.
I really don't think the NFL has an interest in *encour- aging* running QBs. Instead, I think the NFL is interested in *encouraging* efficient passing, and therein resides the vested interest in keeping the QB rating system the way it is, a passing measurement, not an overall measure- ment beyond passing.
- - -
=======================
> [...]
x-- 100 Proof News - http://www.100ProofNews.com x-- 3,500+ Binary NewsGroups, and over 90,000 other groups x-- Access to over 1 Terabyte per Day - $8.95/Month x-- UNLIMITED DOWNLOAD
"JustForFun" <justforfu...@bellsouth.net> wrote ...
> "observer" wrote ...
> >Thanks for posting that. It takes a complicated passer rating > >formula, rethinks additional factors, and comes up with an > >even more complicated passer formula, mixing non-passing > >data with passing data for a different rating, not really a passer > >rating, but instead, a passer-runner-sack-penalty-fumble rating.
> It's a Quarterback rating, rather than a passer rating. Much more > encompasing, being as how a QB does much more than just pass. > [...]
1) Adding penalties to passer ratings is dubious 2) Sacks? Already tracked, and creditable to innumerable factors -- already part of passer ratings to the extent than pressures/sacks impacts the passer's efficiency 3) Running? Already tracked, rushing stats 4) Intangibles? Not tracked, by the current or proposed systems, simply due to its highly subjective nature
Quarterback's secondary tasks? Handing off, audibling (by some), running (by some). Comebacks? Primary when a team is trailing, and not part of either rating system. Wins-losses? Something often mentioned when discussing passer ratings, but not part of the proposed or previous rating system -- if it is added to a "new" rating system, some subjective assess- ment would have to be implemented whereby a quarterback's percentage credit for a win or blame for a loss could be arrived at.
Once you start down the trail of "perfecting" the passer rating, or, if you prefer, quarterback rating, much more attention should be paid to factors not mentioned in the writer's proposal, and much less use of the arbitrary factors he chose should be imple- mented, in my opinion, especially when dealing with the attri- butes unique to 'running quarterbacks', as those attributes are disputed as to their advantage/disadvantage in quarterback effectiveness.
- - -
=========================
x-- 100 Proof News - http://www.100ProofNews.com x-- 3,500+ Binary NewsGroups, and over 90,000 other groups x-- Access to over 1 Terabyte per Day - $8.95/Month x-- UNLIMITED DOWNLOAD
> > > Fact of the business is, the writer is adding much more com- > > > plexity, and doing little but mixing run & pass & penalty & fum- > > > bling stats (for QBs) and penalizing pocket passers (counter > > > to the way in which pocket passers are the most successful > > > quarterbacks in the league). I suspect the writer's bias is to > > > favor running QBs, and to attempt to achieve that, he's trying > > > to throw running type information into the passer rating. I sup- > > > pose, next he'll want to start adjusting runner yards based on > > > defensive penalties that, had they not occurred, the runner > > > would've made much more yardage.
> > C'mon, Observer. He's just quoting hard statistics and giving a > > reasonable interpretation to them. On what basis do you suspect a bias > > on his part? What makes his use of statistics any more biased than the > > current QB rating system?
> The passer rating system measures passing, not sacks, > not fumbles, not running, not penalties. He's simply trying > to present a case that adding non-passing data will yield > an accurate quarterback rating which, point-in-fact, it won't. > It would yield a confused rating adding arbitrary complex- > ity and non-passing information into the passing stats. It > would bias the ratings towards running QBs. It would > change the rating system from being about purely pas- > sing to being about passing, fumbles (not a passing > activity), running (only performed by some QBs), pen- > alty impacts (with the most controversial penalty of all, > pass interference, along with defensive holding).
It's no more confusing or arbitrary than the current system. You can argue that it's not a more accurate measure of QB effectiveness, or that there are good reasons to focus on the passing aspect of a QB's performance, but the fact that such judgments can be debated only proves that there is no single, objective measure of a QB's performance.
> As stated previously (in a response to Astaroth on > January 3, 2004, regarding passer ratings, and ideas > to change it, ideas that did not include penatly-adjust- > ments or sack-adjustments) ...
> The thing is, when it comes to quarterbacks, 'QB rating' > *is* 'passer rating', presently, and changing it to 'passer/ > runner/receiver/touchdown scorer' rating would beget, > I suspect, a split in QB categorizations: 'pocket QBs' > rated separately from 'scrambler/runner QBs'. At the > present time, with most 'scrambler/runner QBs' being > black, and most 'pocket passers' being white, that would > bring up divisiveness based on race, and I assure you > the NFL will *never* create split categorizations that > would promote a race-divisive approach to QB ratings.
Like some people (in this very NG, no less) don't already divide up QBs based on exactly these categorizations. Don't kid yourself -- such folk don't need the QBR system to be overhauled to be 'race-divisive.' As for what the NFL will do,
> Most in the NFL, even those with 'scrambler/runner QBs', > want their quarterbacks to be perceived as passers, with > something extra. Perhaps it's best to keep it simple, rating > all QBs based on their passing, only, on a level playing > field, and leaving the something extra as something extra, > separate from 'QB ratings'.
Sez you. Perhaps it's better, as the writer of the article asserts, to include other aspects of a QB's overall contribution when assessing a QB's effectiveness.
> If you combine the two, giving 'scrambler/runner QBs' a > one-up in the rating system, that's an endorsement of > 'scrambler/runner QBs' over 'pocket passers'. That's an > evaluation of a QB as a passer -and- as a scrambler/runner, > something many (most?) feel is not a valid measurement > of QB skills, for that exists on an uneven playing field, as > most QBs are *not* 'scramblers/runners'.
So what? If you don't combine the two, you're not counting the legitimate contribution of scrambling/runner QBs to their teams. The ability to make plays with your feet is a skill that many (including JJ himself) value in a QB. As for some QBs not being able to run as well, you could make that statement about ANY skill a QB possesses. Some guys can't throw long; maybe we should only count short passes that they can all make.
I noticed that running QB Kordell Stewart didn't fare too well in this system, while the much less mobile Brad Johnson seems to come out fine. Is Trent Green as mobile as Anthony Wright? I doubt it, and Green is rated MUCH more highly. I think passing QBs would get a fair shake. The top rated guy is Peyton Manning, and he's not really coveted for his amazing running ability. For the record, neither is QC, who seems to be able to move around to avoid a sack and make an occasional play with his feet, but seems like nothing special as a runner (he certainly isn't in Michael Vick's category as a runner, who BTW, is rated only slightly higher).
> As stated, in another post, -if- a 'scrambling/running' QB > benefits, in passing, from his 'scrambling/running' ability, > therein resides his gain from that skill, in the QB rating. > If he does not, if he in fact is hindered or injured by > 'scrambling/running', it would be erroneous to give him > bonus 'points', in QB passing evaluation, for having an > extra ability that detracts from what QBs are paid 'big > bucks' to do, and that's complete passes.
> I really don't think the NFL has an interest in *encour- > aging* running QBs. Instead, I think the NFL is interested > in *encouraging* efficient passing, and therein resides > the vested interest in keeping the QB rating system the > way it is, a passing measurement, not an overall measure- > ment beyond passing.
So you acknowledge that the current QBR system isn't a bias-less evaluation of QB effectiveness, but actually can be used to 'encourage' efficient passing?
The QBR as it's presently constituted may already be a good enough measure of how a QB is playing, and you've made that argument well enough. But it isn't objective.
On Sun, 29 Feb 2004 21:31:23 -0600, "observer" <obser...@observer.net> wrote:
>"JustForFun" <justforfu...@bellsouth.net> wrote ... >> "observer" wrote ...
>> >Thanks for posting that. It takes a complicated passer rating >> >formula, rethinks additional factors, and comes up with an >> >even more complicated passer formula, mixing non-passing >> >data with passing data for a different rating, not really a passer >> >rating, but instead, a passer-runner-sack-penalty-fumble rating.
>> It's a Quarterback rating, rather than a passer rating. Much more >> encompasing, being as how a QB does much more than just pass. >> [...]
>1) Adding penalties to passer ratings is dubious
As I addressed and you snipped.
>2) Sacks? Already tracked, and creditable to innumerable > factors -- already part of passer ratings to the extent > than pressures/sacks impacts the passer's efficiency
No it doesn't. Not in the least. QB1 is under a lot of pressure all game long and takes 6 sacks. QB2 is under a lot of pressure all game long and takes 0 sacks, instead throwing the ball away 6 additional times. QB1's actions has negative impact on his team but ZERO impact on his QBR. QB2's actions have a positive impact on his team a severely HURT his QBR.
>3) Running? Already tracked, rushing stats
Not tracked in any meaningful way where the QBR is concerned.
>4) Intangibles? Not tracked, by the current or proposed > systems, simply due to its highly subjective nature
>Quarterback's secondary tasks? Handing off, audibling (by >some), running (by some). Comebacks? Primary when a team >is trailing, and not part of either rating system. Wins-losses? >Something often mentioned when discussing passer ratings, >but not part of the proposed or previous rating system -- if it >is added to a "new" rating system, some subjective assess- >ment would have to be implemented whereby a quarterback's >percentage credit for a win or blame for a loss could be arrived >at.
Also not addressed is things like how the WCO if FAR more QBR friendly than other systems.
>Once you start down the trail of "perfecting" the passer rating, >or, if you prefer, quarterback rating, much more attention should >be paid to factors not mentioned in the writer's proposal, and >much less use of the arbitrary factors he chose should be imple- >mented, in my opinion, especially when dealing with the attri- >butes unique to 'running quarterbacks', as those attributes are >disputed as to their advantage/disadvantage in quarterback >effectiveness.
Fine. Then the QBR is a meaningless stat that can NEVER be truly defined. I'll buy that.
>> > > Fact of the business is, the writer is adding much more com- >> > > plexity, and doing little but mixing run & pass & penalty & fum- >> > > bling stats (for QBs) and penalizing pocket passers (counter >> > > to the way in which pocket passers are the most successful >> > > quarterbacks in the league). I suspect the writer's bias is to >> > > favor running QBs, and to attempt to achieve that, he's trying >> > > to throw running type information into the passer rating. I sup- >> > > pose, next he'll want to start adjusting runner yards based on >> > > defensive penalties that, had they not occurred, the runner >> > > would've made much more yardage.
>> > C'mon, Observer. He's just quoting hard statistics and giving a >> > reasonable interpretation to them. On what basis do you suspect a bias >> > on his part? What makes his use of statistics any more biased than the >> > current QB rating system?
>> The passer rating system measures passing, not sacks, >> not fumbles, not running, not penalties. He's simply trying >> to present a case that adding non-passing data will yield >> an accurate quarterback rating which, point-in-fact, it won't. >> It would yield a confused rating adding arbitrary complex- >> ity and non-passing information into the passing stats. It >> would bias the ratings towards running QBs. It would >> change the rating system from being about purely pas- >> sing to being about passing, fumbles (not a passing >> activity), running (only performed by some QBs), pen- >> alty impacts (with the most controversial penalty of all, >> pass interference, along with defensive holding).
>It's no more confusing or arbitrary than the current system. You can >argue that it's not a more accurate measure of QB effectiveness, or that >there are good reasons to focus on the passing aspect of a QB's >performance, but the fact that such judgments can be debated only proves >that there is no single, objective measure of a QB's performance.
BINGO! That is EXACTLY the point I've been trying to drive home with Observer forever.
>> As stated previously (in a response to Astaroth on >> January 3, 2004, regarding passer ratings, and ideas >> to change it, ideas that did not include penatly-adjust- >> ments or sack-adjustments) ...
>> The thing is, when it comes to quarterbacks, 'QB rating' >> *is* 'passer rating', presently, and changing it to 'passer/ >> runner/receiver/touchdown scorer' rating would beget, >> I suspect, a split in QB categorizations: 'pocket QBs' >> rated separately from 'scrambler/runner QBs'. At the >> present time, with most 'scrambler/runner QBs' being >> black, and most 'pocket passers' being white, that would >> bring up divisiveness based on race, and I assure you >> the NFL will *never* create split categorizations that >> would promote a race-divisive approach to QB ratings.
>Like some people (in this very NG, no less) don't already divide up QBs >based on exactly these categorizations. Don't kid yourself -- such folk >don't need the QBR system to be overhauled to be 'race-divisive.' As for >what the NFL will do,
>> Most in the NFL, even those with 'scrambler/runner QBs', >> want their quarterbacks to be perceived as passers, with >> something extra. Perhaps it's best to keep it simple, rating >> all QBs based on their passing, only, on a level playing >> field, and leaving the something extra as something extra, >> separate from 'QB ratings'.
>Sez you. Perhaps it's better, as the writer of the article asserts, to >include other aspects of a QB's overall contribution when assessing a >QB's effectiveness.
Exactly. How can ANY ONE think it is the same for Hutch to stand back there, hold the ball too long and fumble when he gets hit as it is for QC to take off and run for 3 yards in the same situation? Both situations affect the QBR exactly the same way...they don't.
Hell, I'd submit that throwing the ball is the EASY part of being an NFL QB. They can ALL throw the ball quite well.
>> If you combine the two, giving 'scrambler/runner QBs' a >> one-up in the rating system, that's an endorsement of >> 'scrambler/runner QBs' over 'pocket passers'. That's an >> evaluation of a QB as a passer -and- as a scrambler/runner, >> something many (most?) feel is not a valid measurement >> of QB skills, for that exists on an uneven playing field, as >> most QBs are *not* 'scramblers/runners'.
>So what? If you don't combine the two, you're not counting the >legitimate contribution of scrambling/runner QBs to their teams. The >ability to make plays with your feet is a skill that many (including JJ >himself) value in a QB. As for some QBs not being able to run as well, >you could make that statement about ANY skill a QB possesses. Some guys >can't throw long; maybe we should only count short passes that they can >all make.
>I noticed that running QB Kordell Stewart didn't fare too well in this >system, while the much less mobile Brad Johnson seems to come out fine. >Is Trent Green as mobile as Anthony Wright? I doubt it, and Green is >rated MUCH more highly. I think passing QBs would get a fair shake. The >top rated guy is Peyton Manning, and he's not really coveted for his >amazing running ability. For the record, neither is QC, who seems to be >able to move around to avoid a sack and make an occasional play with his >feet, but seems like nothing special as a runner (he certainly isn't in >Michael Vick's category as a runner, who BTW, is rated only slightly >higher).
>> As stated, in another post, -if- a 'scrambling/running' QB >> benefits, in passing, from his 'scrambling/running' ability, >> therein resides his gain from that skill, in the QB rating. >> If he does not, if he in fact is hindered or injured by >> 'scrambling/running', it would be erroneous to give him >> bonus 'points', in QB passing evaluation, for having an >> extra ability that detracts from what QBs are paid 'big >> bucks' to do, and that's complete passes.
>> I really don't think the NFL has an interest in *encour- >> aging* running QBs. Instead, I think the NFL is interested >> in *encouraging* efficient passing, and therein resides >> the vested interest in keeping the QB rating system the >> way it is, a passing measurement, not an overall measure- >> ment beyond passing.
>So you acknowledge that the current QBR system isn't a bias-less >evaluation of QB effectiveness, but actually can be used to 'encourage' >efficient passing?
>The QBR as it's presently constituted may already be a good enough >measure of how a QB is playing, and you've made that argument well >enough.
I'd disagree. It benefits guys who do things that are detrimental to their team and penalizes guys who do things that are beneficial.
> >The QBR as it's presently constituted may already be a good enough > >measure of how a QB is playing, and you've made that argument well > >enough.
> I'd disagree. It benefits guys who do things that are detrimental to > their team and penalizes guys who do things that are beneficial.
BS. The passer rating is well-founded, and the best determinant of passer performance, extant qualitative judgements on competition. Add those, and it'd be as near-perfect as any non-intangible non-subjective passer rating could possibly be.
- - -
==================
x-- 100 Proof News - http://www.100ProofNews.com x-- 3,500+ Binary NewsGroups, and over 90,000 other groups x-- Access to over 1 Terabyte per Day - $8.95/Month x-- UNLIMITED DOWNLOAD
>> >The QBR as it's presently constituted may already be a good enough >> >measure of how a QB is playing, and you've made that argument well >> >enough.
>> I'd disagree. It benefits guys who do things that are detrimental to >> their team and penalizes guys who do things that are beneficial.
>BS. The passer rating is well-founded, and the best >determinant of passer performance, extant qualitative >judgements on competition. Add those, and it'd be >as near-perfect as any non-intangible non-subjective >passer rating could possibly be.
Uh...so?
My point is pretty simple...a QB is much MORE than a "passer".
>Are you crazy? The new rating ranked the QBs a LOT closer to the reality of >how they played this last year than the "normal" QBR. This rating should be >the one they use in the future
YES...ANY rating that shows Queercy coming out better MUST be used!
We should ONLY look at Queercy's POSITIVE plays, not his negative ones.
>QB2 is under a lot of pressure all game >long and takes 0 sacks, instead throwing the ball away 6 additional >times. QB1's actions has negative impact on his team but ZERO impact >on his QBR. QB2's actions have a positive impact on his team a >severely HURT his QBR.
Uh...they subtract off yards for sacks, you dipshit. And, six incomplete passes are of NO significant effect on QBR.
On 01 Mar 2004 22:43:48 GMT, travis...@aol.com (TravIsGod) wrote:
>>QB2 is under a lot of pressure all game >>long and takes 0 sacks, instead throwing the ball away 6 additional >>times. QB1's actions has negative impact on his team but ZERO impact >>on his QBR. QB2's actions have a positive impact on his team a >>severely HURT his QBR.
>Uh...they subtract off yards for sacks, you dipshit. And, six incomplete >passes are of NO significant effect on QBR.
>You are SO incredibly stupid.
ROFLMAO...lets look at an example and find out, shall we?
Let's use 30 attempts, ten yards per completion, O int's and 0 TD's. The first two numbers are pretty average, the last two don't skew it one way or the other.
On 02 Mar 2004 18:16:25 GMT, travis...@aol.com (TravIsGod) wrote:
>>>You are SO incredibly stupid.
>>ROFLMAO...lets look at an example and find out, shall we?
>>Let's use 30 attempts, ten yards per completion, O int's and 0 TD's.
>Yes. Let us cherrypick an example that is completely unrealistic.
30 attempts and ten YPC are pretty average, dude. I HAD to leave int's and TD's out the equation entirely because your claim didn't involve them.
What do you want? 40 attempts and 30 yards per completion? 10 attempts and 3 yards per completion. Matters not. There is a HUGE difference in 6 completions per game.
>That's what you do EVERY time, isn't it?
>Get your head out of your ass and Quincy's dick out of your mouth.
Quit making claims without checking to see if they are close to reality. QBR isn't that hard to check out.
