NewsBot
New Member
- Messages
- 111,281
- Reaction score
- 2,947
Shouldn’t we evaluate Dak Prescott based on what he’s done, and take the subjectivity out of it?
All offseason, there have been so many quarterback and player rankings done that it’s impossible to keep track.
For example, MIke Sando of ESPN, and a group of 50 members of an “expert panel” ranked all the quarterbacks, putting them in Tiers 1-5. It’s an NFL Insider piece, so you have to pay to read it, but we learned that Dak Prescott is at the top of Tier 3, behind Eli Manning and Kirk Cousins, among others. Then Field Gulls, our sister SBN site for the Seahawks, had this to say about Mr. Sando’s ranking of Russell Wilson, which calls into question the whole enterprise.
On the other hand, the NFL players voted who the top-100 players were, regardless of position. They voted Dak Prescott #14, or the fifth quarterback on the list, behind Tom Brady (1), Aaron Rodgers (6), Matt Ryan (10), and Derek Carr (11). Once again, let’s cite our sister Field Gulls site for questioning whether Dak deserved to be behind Mr. Carr.
I’m sure there were many more outside rankings that we needn’t go into.
The only way I think quarterbacks should be ranked is by statistics, and more specifically, the statistics that best correlate with winning.
Here’s a five-year old article that directly addresses this point. (Forgive me for not having the most up-to-date correlations, but I don’t think things have changed.) It ranks the stats in the following order: ANY/A, passer rating, NY/A, touchdown/attempt, yards/attempt, completion percentage. The people at ESPN also like their proprietary stat - QBR - and it also appears to correlate with winning. This Bleeding Green Nation post makes the following observation:
The important metrics here for the most part measure how well the ball is thrown, not how often it's thrown. In fact, the passing play percentage metric has a negative correlation with winning, meaning that losing teams threw the ball more often.
In other words, when you evaluate quarterbacks, you can dismiss all the raw numbers like yards, completions, etc. You account for these by inserting them into formulas like ANY/A (adjusted net yards per attempt) and passer rating because they discount for interceptions and sacks, while also calculating yards gained per attempt.
As we wrote this offseason, real quarterback rating, and it’s corollary, real quarterback rating differential, as developed by Cold Hard Football Facts, showed numbers superior to passer rating in evaluating quarterback play.
We introduced Real Quarterback Rating before the 2011 season as a way to quantify all aspect of QB play. It’s been a TOTAL score behind your wildest dreams. Teams better in Real QB Rating in 2011 and 2012 went an incredible 441-70 (.863), proving that winning in the NFL is almost always about more efficient play at the QB position.
The difference between this rating, and the standard quarterback rating, is that it takes into account the value of quarterbacks as runners, bringing rushing yards, rushing touchdowns, and fumbles into the equation. Unfortunately, Cold Hard Football Facts’ website with this data is no longer active.
We also looked this offseason at the nine quarterback stats that Pro Football Reference indexes. It showed that Dak Prescott was one of only four quarterbacks in the NFL above average in all nine categories, and was third overall when the rankings were averaged. We made the following observation:
One thing you notice from these numbers is the ANY/A column is almost an exact mirror of the overall average ranking. There is some variation, but less than any other column, which suggests it’s a good proxy for evaluating a quarterback’s overall value.
Last year, in our weekly Dak and Zeke report, we provided the following stats for Dak Prescott: completions, attempts, completion percentage, yards, ANY/A, TDs, interceptions, and passer rating. The ones that correlate the best with winning are passer rating and ANY/A. The others stats are there because they feed into calculating passer rating and ANY/A. We didn’t use QBR because we don’t know precisely how ESPN calculates it, and it wasn’t out in time for our weekly reports.
So, if we had to boil down the key quarterback stats that make a difference with winning, these three primarily come to mind — passer rating, ANY/A, and QBR. If they were still calculating it, Cold Hard Football Facts’ real quarterback rating would be better than passer rating, but we don’t have access to those numbers.
How did the NFL’s top quarterbacks stack up in these categories last year?
QB
Rate
QB
ANY/A
QB
QBR
Ryan
Ryan
117.1
Ryan
9.03
Ryan
83
Brady
Brady
112.2
Brady
8.81
Brady
83
Prescott
Prescott
104.9
Prescott
7.86
Prescott
82
Rodgers
Rodgers
104.2
Cousins
7.45
Rodgers
77
Brees
Brees
101.7
Brees
7.27
Cousins
72
Bradford
Bradford
99.3
Rodgers
7.24
Brees
72
Cousins
Cousins
97.2
Carr
7.2
Luck
71
Carr
Carr
96.7
Mariota
7.14
Stafford
71
Based on this, it’s quite clear why Matt Ryan won the MVP award, and why Ryan and Tom Brady squared off in the Super Bowl. Solidly in third in all three categories was Dak Prescott.
Accordingly, by the stats that matter, Dak Prescott is a top-five NFL quarterback.
But then, how do you account for the fact he’s done it only one year? Why not discount him and require him to do it again before you pronounce him a top-five quarterback? After all, RGIII and Nick Foles had fantastic rookie seasons, and then dropped off the map.
You could look at it that way. But why?
Isn’t it more fair to Dak Prescott to rank him where his play on the field has established he deserves to be ranked, instead of substituting one’s own speculation about where you think he ought to be ranked? The latter just introduces a lot of subjective nonsense, as was obviously done by the ESPN “experts” Mike Sando relied on above.
If Dak Prescott’s performance falls off in 2017, then we’ll evaluate his ranking based on that new data, coupled with what he showed in 2016. But until then? Dak Prescott is a top five NFL quarterback.
Continue reading...