From the FanPosts -Joel
Over the last day or two, a fan shot by JayhawksNChiefs and a fan post by juvatbull have sought to prove statistically that certain ideas about Alex Smith's play, particularly the comparisons to Matt Cassel, are demonstrably false. I think they proved rather definitively that Smith is a far more efficient quarterback than Cassel, but at the end of his post, juvatbull made the following statement:
Alex Smith's future looks bright--at least as bright as Joe Flacco's.
Now, this type of statement makes my brow furrow, my eyes squint, and my brain say, "I wonder." After all, I have not been a fan of the trade for Smith. I preferred a simpler, more Maddenesque solution to the problem at quarterback. Andy Reid and John Dorsey did not consult me on this, however, and now we have Alex Smith. But, is he as good as Flacco? That's the question.
My problem with Alex Smith, is primarily that his numbers are inflated by very selective application of his skills in practical situations. In other words, his stats are only as good as they are because Harbaugh used him only in situations that he had a very good chance to be successful. That, if asked to pass more, his efficiency would be significantly worse than they have been over the last two seasons.
So, here was the basis of my meta-analysis of the stats of these two quarterbacks. First, I am only looking at the 24 games encompassing the regular season of 2011 and the first 8 games of 2012. This might make my analysis seem like cherry picking; selective to Smith's best years. As the data will show, though, this doesn't really matter, a point I will come back to later. Regardless, there is enough data from these games to form a basis for comparison.
I then broke down the quarterback ratings by either quarterbacks average number of attempts per game. The idea being, if we compare either players' games by the number of passes they attempted, it will tell us how efficient Alex Smith would be if he had a Joe Flacco work load and vice-versa. So my demarcations are 27 attempts per game and 34 attempts per game, Smith and Flacco's averages respectively.
Here is the result of my analysis (note that ratings for >27 pass attempts also include data from games with >34 pass attempts):
Rating in games with <27 attempts: 109.4 (13 games)
Rating in games with >27 attempts: 81.8 (11 games)
Rating in games with >34 attempts: 75.0 (3 games)
Rating in games with <27 attempts: 88.9 (8 games)
Rating in games with >27 attempts: 76.9 (16 games)
Rating in games with >34 attempts: 73.3 (10 games)
Rating in games with <27 attempts: 103.2 (6 games)
Rating in games with >27 attempts: 92.6 (18 games)
Rating in games with >34 attempts: 94.5 (7 games)
From these breakdowns, it's clear that both quarterbacks are more efficient when they have fewer attempts in a game. When they pass more than 34 times in a game, both quarterbacks are significantly less efficient and, I would say, about equal in their relative efficiency at that level. When demarcating games by Smith's average of 27 passes per game, we likewise see greater efficiency as the number of attempts go down, and less efficiency as the numbers go up. Smith as compared to Flacco, is consistently more efficient regardless of the number of attempts, but has a far more precipitous decline when he exceeds his average of 27 passes per game. Flacco, though not as efficient overall, is more steady in his efficiency as his attempts increase.
With these data, I believe my initial belief, that Smith's extraordinary efficiency is a product of a highly-selective coach is true. In another circumstance, particularly one where he is expected to pass more, we would expect him to be significantly less-efficient. However, even in so doing, he would still be at a reasonably average-to-above-average level of efficiency. As such, I think juvatbull's claim that Smith's future is at least as bright as Flacco's is not only reasonable, it may even be understated a bit.
More to the point, though, there lies a significant question that has been asked multiple times since Alex Smith was traded for: Can Alex Smith lead the Chiefs to a Super Bowl? If Flacco can quarterback a Superbowl winning team, why shouldn't Smith also be able?
There were two other arguments against Alex Smith that I will bring up: his 3rd-down conversion percentage, and his performance before Harbaugh. Taking the second first, the reason why I didn't include previous data is I believe a game by game analysis will substantiate the above findings, that Smith's efficiency was hindered by the number of throws he was forced to make. While the assertion of JayhawksNChiefs and juvatbull, that Smith's stats trend up over the course of his career, is absolutely correct, we don't have to defend Smith on the basis of him improving over that time. He wasn't as bad as those stats suggest, even from the beginning, and has improved steadily in his performance over that time. His older numbers were as deflated as his recent numbers are inflated.
As far as the 3rd down efficiency percentage, an article from Inside the 49ers by PressDemocrat.com does a good job of comparing Smith to Flacco. While ostensibly it seems to suggest that Flacco is the much better quarterback on 3rd down, what's important to note about the article, is that it was written a year ago... a year before Joe Flacco both won the Super Bowl and had the lowest 3rd-down efficiency percentage of his career, ranking 21st in the NFL.
Before I started doing this analysis, I was a doubter. Even now I wouldn't necessarily identify myself as a Smith Fan; not until he starts playing here up to his potential. But, what I have proven to my own satisfaction is that Smith is probably about middle-third NFL quarterback. How he performs this year will be tied to how much he is expected to throw. My prediction is his quarterback rating will be about 100 if he averages 25 attempts per game or less, and will drop by about 15 for every 5 more passes he averages above that. And ultimately, if he remains healthy, he can lead our team to a Super Bowl and win it.
I have added some numbers for Matt Cassel for comparison. What is important to note about these stats for Cassel: They are 15 games from Cassel's best season (2010) and 9 games from his second best season (2008). Also important to note is that in these games Cassel was playing in two different systems. The first group he was playing under Todd Haley/Charlie Weis, the second group he was playing under Belichick. All but one of the games he threw over 34 attempts were from the latter group.
Edit #2: just realized I messed the subtitle up.