Talk:Passer rating
From Wikipedia, the free encyclopedia
 |
This article is part of WikiProject National Football League, an attempt to build a comprehensive and detailed guide to the NFL on Wikipedia. If you would like to participate, you can edit the article attached to this page, or visit the project page, where you can join the project and/or contribute to the discussion. | |
| Start | This article has been rated as Start-Class on the quality scale. (If you rated the article please give a short summary on the comment page to explain the ratings and/or to identify the strengths and weaknesses.) |
|
 |
This article is part of WikiProject American football, an attempt to build a comprehensive and detailed guide to American football on Wikipedia. If you would like to participate, you can edit the article attached to this page, or visit the project page, where you can join the project and/or contribute to the discussion. |
 |
This article is within the scope of WikiProject Canadian football, an attempt to build a comprehensive and detailed guide to Canadian football and the Canadian Football League on Wikipedia. If you would like to participate, you can edit the article attached to this page, or visit the project page, where you can join the project and/or contribute to the discussion. | |
| Start | This article has been rated as Start-Class on the quality scale. (If you rated the article please give a short summary on the comment page to explain the ratings and/or to identify the strengths and weaknesses.) |
|
Contents |
[edit] Cleanup
From the article: Another criticism is the fact that rushing yards gained by a quarterback do not result in his getting a higher rating — a particularly important contemporary issue due to the recent emergence of many excellent running quarterbacks in the NFL, most notably Steve Young, Daunte Culpepper, Donovan McNabb, John Elway and Michael Vick and players like Damon Allen and Kerry Joseph in the CFL.. On the other hand, Vick holds the NCAA freshman record for passing efficiency and Young is the former record holder for passer rating in a season. This record was shattered in 2004 by Indianapolis Colts quarterback Peyton Manning.
Response: The entire "On the other hand..." statement about NCAA passing efficiency records is not only irrelevant, there's no way that Peyton Manning, in the NFL in 2004, could break an NCAA record in that year! This section is poorly worded, and seems more like a backhand attempt to promote Vick and Young.
-- MattBattison
Another Response: And was John Elway ever a mobile quarterback? He certainly wasn't toward the end of his career...
[edit] Unnecessarily complex
An obvious criticism -- I'm surprised it's not mentioned -- is that the rating is too complex and thus has no intuitive meaning for regular fans. If the TV announcer says that the starting QB has thrown 2 interceptions in each of the last 14 games, the average viewer will think that the the QB should perhaps be replaced. If the same announcer says that the QB has had a rating of 91.2 for the year, most viewers will await a further prompt ("... which is the highest in the conference!") to digest the information better. This isn't true of yards per carry, batting average, rebounds per game, etc. Even more complicated stats like slugging percentage at least measure something pretty straightforward. QB rating doesn't measure anything, it's an attempt to produce a super-stat that represents everything about a QB.Priceyeah 20:20, 14 January 2007 (UTC)
[edit] Alternate Rating
This chart [1] shows the draft results of a new rating formula I propose (min. 100 games):
TG=TDs per game YG=yards per game TD-INT=career differential
PC=%completion SBW=SuperBowl wins SBA=Superbowl appearances:
rating = [0.20 * TG + 0.20 * YG / 100 + 0.20 * (TD − INT) / 100 + 0.20 * PC + 0.15 * SBW + 0.05 * SBA] / 2.5
This rating addresses some of the criticisms of the current rating formula on the main page --Billymac00 02:55, 12 October 2006 (UTC)
[edit] Hall of Fame
as of 2006, there are 23 modern era QBs in the Hall of Fame, alphabetically:
Aikman,Blanda,Bradshaw,Dawson,Elway,Fouts,Graham,Griese,Jurgensen,Kelly,Layne
Marino,Montana,Moon,Namath,Starr,Staubach,Tarkenton,Tittle,Unitas,VanBrocklin
Waterfield and Young.
Examining the statistics, the typical Hall of Fame quarterback will exceed the following career criteria:
attempts>4500, yards >30000, TDs>150 and TD/INT ratio > 1.25, %completion >50%.
Often the HOF quarterback has a Superbowl win and appears in the all-time top 10 for at least 1 major passing category.
A quarterback appears to be a "lock" if he meets any of the following over his career:
TDs>250(Manning*,Montana,Favre*,Marino,Fouts,Moon,Jurgensen,Tarkenton,Elway,Unitas)
completions>3000 (Montana,Fouts,Moon,Elway,Tarkenton,Favre*)
TD/INT ratio > 1.50 (Young,Manning*,Montana,Brady*,Favre*,Marino,Cunningham*)
%completion>60% (Kelly,Favre*,Aikman,Brady*,Young,Montana,Manning*)
has multiple SuperBowl wins (Bradshaw,Brady*,Montana,Starr,Young,Aikman,Elway)
Considering the aforementioned, three active players, Favre, PManning, and Brady are locks to get in, and strong cases can be made for McNabb*,McNair* and Cunningham.
- indicates not yet in HOF
--Billymac00 15:13, 9 October 2006 (UTC)
[edit] College System
The article does not explain what the college formula is. Does anyone know? Makerowner 04:51, 8 December 2006 (UTC)
KISS. I propose the following formula:
rating = 100 * (((comp / att) / (quarters / 4)) + ((yards / (quarters / 4)) / 100) + ((td − int) / (quarters / 4)))
Where "yards" is equal to the QB's total yards, both QB running and QB passing... hey, the QB got the yards right?
[edit] College Rating has no upper limit???
"The NCAA passer rating system has no upper limit" I do not believe that is correct. The longest possible pass is 99 yards, so the best rating possible would occur if one threw every single pass 99 yards for a touchdown. If my calculations are correct, that would yield: (100 * 1.0) + (8.4 * 99) + (330 * 1.0) - (200 * 0.0) = 1261.6. If that is the case, then it's very inaccurate to say that the rating system "has no upper limit."Mickeyg13 06:40, 14 August 2007 (UTC)
- You're correct. I edited the section to clarify this point. I think "no upper limit" was referring to the fact that none of the NCAA passer rating components are capped (like the max of 2.375 in the NFL), but it was still rather confusing. Chaz Beckett 20:29, 2 October 2007 (UTC)
[edit] Average Rating
The following statement is pretty meaningless. "Conceptually the average rating would be 79.2..." Simply averaging the highest possible and lowest possible ratings does not get anything like a 'conceptual average'. (note: if if did the college 'conceptual average' would be in the neighborhood of 600.) The other fallacy in this statement is related to the idea that it is not a linear scale, rather "outstanding performances meet diminishing returns". I think that the historical average is a much more reasonable measure. maxsch 17:15, 1 October 2007 (UTC)
- I changed the first paragraph, removing the bit on conceptual averages. maxsch 20:17, 2 October 2007 (UTC)
[edit] Calculation Error
From the Article:Evidence for the possible overemphasis on completion percentage may be seen in the following simple example in which each quarterback has three passing plays. The first two quarterbacks completed fewer passes but got a first down for their teams while the third quarterback is rated the highest, despite only gaining five yards, due to a high completion percentage.
A counterargument to the example above is that the formula should not be applied to a sample as small as 3 passes. The minimums and maximums for each of the four passing categories attempt to modify the formula to apply to small samples. But the actual merits of the formula are best examined when the samples become large enough for the minimums and maximums to become unnecessary in calculation.
Response: Based on my calculations (C=.1111,Y=2.375,T=0,I=2.375)the first quarterback in the chart should have a passer rating of 81, not 71.5. That would make him the best of the three. However, if you still want to make your point that quarterback 3 has the higher rating, then the "Yards" for quarterback 1 needs to be 10 which will put him at a rating of 70.6. Odelphi 19:51, 15 October 2007 (UTC)
[edit] Is this the wrong wiki?
It seems this wiki should be titled "criticism and critique of the passer rating system", since 90% of it details such. No arbitrary rating system is without its faults. There is no need for thousands of words of text to merely state rushing yards are not counted, though maybe some one can fill me in on why a PPASSER rating system would care about rushing yards?
[edit] Criticism section
I've moved the criticism section from the article to this page. This section has become nearly half the article and consisted almost entirely of original research. It was also very poorly sourced and organized. In my opinion, I believe the article is fine simply describing the rating without offering opinions on its perceived shortcomings. However, I could see having a brief, well-sourced section describing its most commonly mentioned deficencies. Chaz Beckett 13:52, 21 October 2007 (UTC)
[edit] Text of criticism section
Passer rating has many critics, who have objected to it on several different grounds.
The most frequently-voiced objection is that plays on which the quarterback is sacked do not count toward compiling the rating, thereby giving a quarterback who is about to be sacked an incentive to deliberately take the sack rather than throw the ball away and have an incomplete pass lower his completion percentage, touchdown passes-per-attempt ratio and his yards-per-attempt ratio. However, many defenders have pointed out that no professional quarterback would intentionally take a sack to the detriment of his team in order to preserve a passer rating, not to mention that sacks often have more to do with the offensive line than the quarterback.
Another criticism is the fact that rushing yards gained by a quarterback do not result in his getting a higher rating — a particularly important contemporary issue due to the recent emergence of many excellent running quarterbacks in the NFL, most notably Randall Cunningham, Steve Young, Donovan McNabb, John Elway, Michael Vick, and Vince Young, and players like Damon Allen and Kerry Joseph in the CFL. On the other hand, Vick holds the NCAA freshman record for passing efficiency and Steve Young is the former NFL record holder for passer rating in a season. However, "It is important to remember that the system is used to rate passers, not quarterbacks. Statistics do not reflect leadership, play-calling, and other intangible factors that go into making a successful professional quarterback." [2]
Some observers have proposed replacing yards per attempt with yards per play as one of the rating's components, counting both sacks and rushing attempts (in addition to passes actually thrown) as plays, thereby resolving both of the above issues.
A third complaint concerns the placing of artificial limits (both upper and lower) on the points that can be earned in the various categories. Critics claim, for example, that it is unfair not to give a quarterback a higher rating for completing 90% of his passes than for completing 77.5%, or for averaging, say, 15 yards per attempt as opposed to only 12.5, although this is likely to affect only the rating for a single game and not for an entire season. As well as encountering problems when approaching the arbitrary limits for performance, the system is biased inasmuch as some of the "benchmarks" are more achievable than others: A 2.00-worthy performance in Completion % is far easier to obtain than a 2.00 in the Touchdown % category. In fact, until recent years, the league norms in each non-Completion % category were below the formula's expected "average" figures.
Yet another criticism is that the formula overemphasizes Completion Percentage. While Completion % in and of itself makes up 25% of the rating's calculation, the use of Yards per Attempt rather than Yards per Completion adds to the weight of Completion %, since incomplete passes are penalized under yds/attempt. Thus, many critics feel that the Passer Rating formula automatically is skewed in favor of quarterbacks who play in a West Coast scheme that favors many high-percentage (but low-yardage) pass plays, e.g. Steve Young, who held the single-season record until 2004.
Evidence for the possible overemphasis on completion percentage may be seen in the following simple example in which each quarterback has three passing plays. The first two quarterbacks completed fewer passes but got a first down for their teams while the third quarterback is rated the highest, despite only gaining five yards, due to a high completion percentage:
| Completions | Attempts | Yards | TD | Interceptions | Rating |
|---|---|---|---|---|---|
| 1 | 3 | 30 | 0 | 0 | 71.5 |
| 2 | 3 | 10 | 0 | 0 | 71.5 |
| 3 | 3 | 5 | 0 | 0 | 79.2 |
A counterargument to the example above is that the formula should not be applied to a sample as small as 3 passes. The minimums and maximums for each of the four passing categories attempt to modify the formula to apply to small samples. But the actual merits of the formula are best examined when the samples become large enough for the minimums and maximums to become unnecessary in calculation.
The simplest complaint may be that the weighting of each category is, on the whole, arbitrary. There is no evidence to suggest that Completion %, Yards/Attempt, Touchdown %, and Interception % are of equal value (25% each) in terms of an offense's efficiency (or if those stats, in that combination, have any correlation to scoring, or winning, at all).
The system also does not account for changing offensive conditions in the NFL over time. When the formula was devised in 1973, the standards were set based on statistical averages from 1960 through 1972.[3] For example, a score of 66.7 was supposed to represent an average quarterback (50% completions, 7 yard average gain, 5.5% interceptions, 5% touchdowns), while a rating of 100 would be considered exceptional. And in 1970, the average passer scored a 65.6, a figure almost perfectly in line with the rating-creators' vision. But in 1980, the average was 73.7; in 1992, the average quarterback had a rating of 75.3; and in 2004, a record-setting year for passers, the league's mean passer rating was 82.8. The great Johnny Unitas finished his hall of fame career with only a 78.2 rating, having played in an era when interceptions were more common than touchdown passes.
The system also counts all interceptions equally regardless of the circumstances. Not all interceptions are the fault of the quarterback as a result of a bad decision or pass. Some of the interceptions were caused by factors beyond the quarterback's control such as dropped or tipped passes by the receivers that end up being caught by a defensive player. [4]
At least part of the reason for the increase in passer ratings is that the NFL has consistently made rule changes favoring the offense, most notably the "Mel Blount" rule in 1978 that limited contact with receivers to within five yards of the line of scrimmage. Thus, scores must be interpreted with care when doing cross-era analysis.
No matter what system is used, a large part of a quarterback's success is related to the talent around him. A poorly thrown pass can be overcome with a great catch, and the perfect pass can be dropped. The offensive line also has quite a lot to do with the quarterback's success.
