"You are educated when you have the ability to listen to almost anything
without losing your temper or self-confidence." - Robert Frost

Wednesday, June 27, 2012

New Quarterback Rating - 2011


Last summer I introduced a new quarterback rating formula; something that could be achieved using available statistics and not arbitrary or based on film study (like ESPN’s QBR)

You can see the article here

I went through all the quarterbacks in the league with 10 or more starts and ran their statistics through my formulas and thought I’d compare and discuss the differences of the two ratings.

The Elite

                                New Rating                       Old Rating               ESPN QBR
A. Rodgers                 139.7 (1st)                    122.5 (1st)               85.2 (1st)
D. Brees                    134.7 (2nd)                   110.6 (2nd)              84.0 (2nd)
T. Brady                    126.0 (3rd)                   105.6 (3rd)              74.2 (3rd)

Comments:
The top three performers remain the same in both calculations; however in my system they separate themselves from the pack much more so than the old system.  This is primarily due to my use of team offensive proficiency (points per possession) as a key statistic in my analysis.  The Packers (3.05), Saints (2.98) and Patriots (2.79), all greatly exceeded the NFL average (around 1.85) and even lapped the 4th highest (San Diego @ 2.32).

We are watching historic offensive production these days and my quarterback ranking system reflects that.  I have always felt the efficiency of an offense is very closely tied to the quarterback and I want a rating system to reflect that cooperative relationship.

I don’t really think anyone can argue rewarding Rodgers, Brees and Brady for their accomplishments in the 2011 season.  All three produced historically great seasons.

Tier 2

                                  New Rating                   Old Rating                ESPN QBR
M. Schaub                 111.2 (4th)                     96.8 (6th)                66.7 (6th)
T. Romo                    105.4 (5th)                    102.5 (4th)               70.1 (4th)
C. Newton                 104.8 (6th)                    84.5 (15th)              56.6 (15th)
M. Stafford               103.7 (7th)                     97.2 (5th)                65.2 (7th)
E. Manning               102.9 (8th)                     92.9 (7th)               61.0 (12th)
P. Rivers                    101.8 (9th)                    88.7 (11th)               64.3 (8th)
M. Ryan                    99.1 (10th)                     92.2 (8th)                67.5 (5th)
M. Vick                      98.1 (11th)                    84.9 (14th)              63.2 (10th)

Comments:
You start to see some differences now with the how categories are weighted and the fact I use running statistics and team offensive statistics rather than just passing yards and passing touchdowns.  Both Newton and Vick make big jumps from the traditional passer rating system and rightly so.  Both produce a lot of yards out of the passing game with their legs and both led productive offenses.  Cam Newton and Carolina’s 5th rated points per possession (2.26) is an impressive accomplishment for a rookie as was Newton’s +36 comparison between plays of 20+ yards and sacks (5th in the league).  If Newton can continue to generate that type of big play offense while improving his efficiency even slightly (bump up completion percentage and reduce turnovers), he is going to be a very special quarterback in this league.

Matt Schaub’s jump is due to his all-around impressive season (in his limited 10 starts).  He was 13th in completion percentage, 5th in yards per pass play, 11th in points per possession, 5th in turnovers per pass play and 2nd in big play vs. sack ratio.  Schaub has a great all around season and it is reflective in extremely important statistics that win football game (7-3 in his starts).

I think most of the teams above are very confident in their quarterback situation with the exception of those with health or durability concerns (Eagles/Texans).  While there is only one Super Bowl winner (Manning), most are confident if any new quarterback breaks into that club it will come from the above list.

Median

                                       New Rating                      Old Rating                ESPN QBR
B. Roethlisberger      90.0 (12th)                    90.1 (10th)               63.3 (9th)
M. Hasselback          87.1 (13th)                    82.4 (16th)              61.6 (11th)
A. Smith                    86.8 (14th)                     90.7 (9th)               46.4 (19th)

Comments:
Steelers fans aren’t going to like seeing this, but Ben Roethlisberger had a very mediocre season in 2011.  The offense he led was average (1.86 ppp) and while he was good in the overrated completion percentage and underrated yards per pass play, he was bad in important statistics like turnovers and big play vs. sack ratio.  All-in-all, a very pedestrian season for Roethlisberger and the Steelers offense (and the likely reason for Bruce Arians’ departure).  How and if Roethlisberger improves with Todd Haley can/will make or break the 2012 season.

Matt Hasselback was surprisingly good at avoiding negative plays (only 19 sacks and 15 turnovers).  He was also average at completion percentage.  What hurts Hasselback the most was Tennessee’s terrible offensive production (1.65 ppp), which was 21st in the league.

Alex Smith was very similar to Hasselback in completion percentage and committed even fewer turnovers (7 all season, 2nd in the league).  Smith and the 49ers also ranked a surprising 12th in points per possession (helped by the best starting field position in football).  What hurts Smith the most is his below average yards per pass play (6.28) and very poor Big Play vs. Sack ratio (Smith took 44 sacks this season compared to 41 plays of 20+ yards).

Alex Smith’s season reminds me eerily of David Gerrard’s 2007 season and Matt Cassell’s 2010 season where an abnormally low turnover rate propelled mediocre teams into the playoffs.  Neither of those teams had the defense the 49ers put on the field in 2011, but I fully expect an adjustment year in Smith’s turnovers and some struggles on the offensive side of the ball.

Are we Sure?

                                  New Rating                   Old Rating               ESPN QBR
J. Cutler                     82.7 (15th)                    85.7 (13th)              59.5 (13th)
A. Dalton                   82.3 (16th)                    80.4 (18th)              47.3 (18th)
R. Fitzpatrick             80.0 (17th)                    79.1 (20th)              51.2 (17th)
J. Flacco                    78.7 (18th)                    80.9 (17th)              57.9 (14th)

Comments:
This is a scary tier to be in as a team.  All these players seem to lack “something” and have a concern that might never be fixed.  Cutler continues to take too many sacks, can be turnover prone and the offense isn’t top-15.  Dalton flashed, but has accuracy issues and Cincinnati’s offense was worse than some realize (1.66 ppp, 20th).  Flacco struggled with big plays, fumbles (6 lost) and completion percentage (less than 60%).  Fitzpatrick doesn’t have any one thing that really stands out.

Dalton, as a rookie, gets the most leeway.  Cutler (7-3) and Flacco (12-4) obviously led their teams to very good records under their watch, but both give you the feeling of playing with fire while watching them compete each week.  Fitzpatrick might never be more than a middle of the road, system QB.

There are a lot of expectations for the teams of this group (Chicago, Cincinnati, Baltimore and Buffalo).  Most pundits expect bigger and better things for many of them.  But that is only going to happen if the four names above improve significantly into the 90+ new QB rating.

Need to Worry

                                  New Rating                   Old Rating              ESPN QBR
M. Moore                 70.7 (19th)                    87.1 (12th)              54.0 (16th)
T. Jackson                  67.3 (20th)                    79.2 (19th)              37.4 (23rd)
R. Grossman              67.3 (21st)                    72.4 (25th)              42.2 (21st)
J. Freeman                64.2 (22nd)                   74.6 (22nd)             43.3 (20th)
M. Sanchez               61.5 (23rd)                    78.2 (21st)              33.6 (25th)
C. McCoy                  57.9 (24th)                    74.6 (23rd)             39.8 (22nd)
C. Ponder                  57.7 (25th)                    70.1 (27th)              35.9 (24th)
T. Tebow                   53.9 (26th)                    72.9 (24th)              27.2 (27th)
S. Bradford                42.2 (27th)                    70.5 (26th)              27.3 (26th)
B. Gabbert                37.4 (28th)                    65.4 (28th)              20.5 (28th)

Comments:
The bottom of the starting quarterback barrel in 2011, these are the teams with the most scrutiny and change associated with the position.  Matt Moore falls hard when you add his 6 lost fumbles in 12 starts to his ledger.  My bell curve also goes much lower than the traditional ratings (you will often see a range of 20-175 for individual games as opposed to 50-140 you see now).  Bradford and Gabbert lead the two move inept offenses in the league with Jacksonville at 1.15 points per possession and the Rams at an abysmal .92 points per possession.  To think the Packers score three times as many points per drive as the Rams shows how far apart the haves and have-nots are in the league.  In fact, the ratio between Bradford’s QB rating (42.2) and Rodgers (139.7) is almost identical to the ratio of points scored by the Rams (.92 ppp) and Packers (3.05 ppp).


Conclusions

The one statistic having the most impact on raising or lowering a QB’s rating between my system and the old system is fumbles lost.  Since I consider a fumble lost equivalent to an interception, that greatly impacts some ratings (turnovers per pass play is roughly 25% of your grade in both systems).  The second major factor is my use of team offense vs. passing touchdowns.  Some teams just score better with the run (see Carolina and San Francisco), while some the touchdown pass statistics are not indicative of the team’s offense (see Jets, Buffalo and Lions).  I do not reward a quarterback for one over the other.

The player that most surprised me (in a positive way) was Cam Newton.  I had not realized watching last season that the Carolina offense was that potent.  In 2010, Carolina average .95 points per possession (last) and improved that to 2.26 (5th).  Over a typical 175 possession season, Newton “created” almost 230 points (two touchdowns/game)!  Obviously there were other factors (health, new coaching staff), but that’s a historic improvement.  Even Dan Marino and Ben Roethlisberger’s rookie years improved offensive output by only around 5 points per game from their former levels.

Lastly, remember this is just a statistic.  This is not a ranking of quarterbacks.  The new QB rating doesn’t take in to account health, clutch play, surrounding talent, or strength of schedule.  Just like the old QB rating system it is just based on raw numbers and only from the 2011 season.

My purpose was to just fix some of the flaws with the old rating system.

Remember the old system is based on the following stats:

Completion Percentage (25%)
Yards per Attempt (25%)
Touchdown Percentage (25%)
Interception Percentage (25%)

Each of those categories is equally weighted towards your final grade.  My new system uses the following stats with slightly different importance to your overall grade:

Completion Percentage (17%)
Yards per Dropback (25%)
Points per Possession (25%)
Turnovers per Dropback (20%)
Big Plays (20+ Yards) vs. Sacks Ratio (13%)

Overall, I think I am accomplishing what I hoped by giving a “bump” to the running quarterbacks and aggressive downfield throwers that produce points and penalizing quarterbacks that are fumble prone or too conservative.

I hope you enjoyed the analysis.  Thanks for reading.







No comments:

Post a Comment