Manipulating Mike Glennon's statistics to tell a story

Kim Klement-USA TODAY Sports

Using statistics to tell a story can be very useful, but not if you cherry-pick the statistic you like.

Mike Glennon had good numbers as a rookie, say the Tampa Bay Buccaneers. They highlighted Glennon's passer rating as a rookie, which ranked third behind Robert Griffin III and Russell Wilson among rookies since 2010. That sounds good, but it's simply a result of looking at only a small part of his statistics. Annoyingly, Pat Yasinskas of ESPN decided to run with that stat anyway, claiming that Glennon's stats compare to Russell Wilson's and Robert Griffin III's.

The fact that the Buccaneers chose this stat is no surprise: teams always and exclusively focus on the positives. The fact that Yasinskas felt the need to reinforce a team blowing smoke up its own player's butt is more surprising, but Yasinskas' strengths aren't in the area of statistics. Mistakes happen -- but that doesn't mean we need to let them stand.

So, why does Mike Glennon come out looking so good in passer rating? Because the statistic ignores rushing and sacks, and heavily emphasizes the categories Glennon found success in, while ignoring or marginalizing those he did not do well in. It is once again an incomplete viewing of the statistical record, taken to be a complete representation.

The choice of Robert Griffin III and Russell Wilson as the relevant player is especially unfortunate, given the duo's combined 1,304 rushing yards and 11 rushing touchdowns, compared to Glennon's 37 rushing yards and no touchdowns. That alone should tell you that this is not exactly a relevant statistic.

There's more, though. Passer rating is an over-engineered mess of a statistic, based on numbers that were average in the 1960s. Because of that, today's statistics make it go wonky in weird places. It heavily rewards high completion percentages, because the average completion percentage back in 1960 came in around 50%. It hates interceptions and loves touchdowns. It doesn't take into account sacks, fumbles or rushing attempts and actual yardage is only a small part of the formula.

Certainly, Yasinskas isn't the first one to bring up these statistics, or similar ones, he's just the most recent one. Other people may point to his eight games with multiple touchdowns -- a rookie record, sure, but another example of focusing on a very, very small portion of his statistical record.

To illustrate this point, here's a table of all rookie quarterbacks with at least 200 pass attempts from 2010 through 2013, courtesy of Pro Football Reference. Every column is sortable. Go ahead and play around with them, to see where Mike Glennon lands in each category.

Rk Player Year G GS Cmp Att Cmp% Yds TD Int Rate Sk Y/A AY/A ANY/A
1 Russell Wilson 2012 16 16 252 393 64.12% 3118 26 10 100 33 7.93 8.11 7.01
2 Andrew Luck 2012 16 16 339 627 54.07% 4374 23 18 76.5 41 6.98 6.42 5.66
3 Cam Newton 2011 16 16 310 517 59.96% 4051 21 17 84.5 35 7.84 7.17 6.24
4 Andy Dalton 2011 16 16 300 516 58.14% 3398 20 13 80.4 24 6.59 6.23 5.65
5 Robert Griffin 2012 15 15 258 393 65.65% 3200 20 5 102.4 30 8.14 8.59 7.47
6 Mike Glennon 2013 13 13 247 416 59.38% 2608 19 9 83.9 40 6.27 6.21 4.98
7 Sam Bradford 2010 16 16 354 590 60.00% 3512 18 15 76.5 34 5.95 5.42 4.73
8 Brandon Weeden 2012 15 15 297 517 57.45% 3385 14 17 72.6 28 6.55 5.61 4.98
9 Christian Ponder 2011 11 10 158 291 54.30% 1853 13 13 70.1 30 6.37 5.25 4.25
10 Blaine Gabbert 2011 15 14 210 413 50.85% 2214 12 11 65.4 40 5.36 4.74 3.68
11 Geno Smith 2013 16 16 247 443 55.76% 3046 12 21 66.5 43 6.88 5.28 4.17
12 Ryan Tannehill 2012 16 16 282 484 58.26% 3294 12 13 76.1 35 6.81 6.09 5.23
13 EJ Manuel 2013 10 10 180 306 58.82% 1972 11 9 77.7 28 6.44 5.84 4.87
14 Matt McGloin 2013 7 6 118 211 55.92% 1547 8 8 76.1 6 7.33 6.38 5.96
15 Nick Foles 2012 7 6 161 265 60.75% 1699 6 5 79.1 20 6.41 6.02 5.13
16 Colt McCoy 2010 8 8 135 222 60.81% 1576 6 9 74.5 23 7.1 5.82 4.73
17 Jimmy Clausen 2010 13 10 157 299 52.51% 1558 3 9 58.4 33 5.21 4.06 2.98

Glennon's performance across statistical categories varies heavily. Yes, he was one of the best rookies at throwing touchdowns and avoiding interceptions, and his passer rating was really good, too (as a consequence). But he lands near the bottom in yards per attempt and sacks and his completion percentage is average. If we look at advanced statistics like DVOA and Total QBR, he was one of the worst starters in the NFL. And we haven't even touched added value on the ground.

None of this is to say that Glennon will fail to improve, or that he's definitely worse than those players, or that he had a bad season given the circumstances. It's just that this statistical manipulation has to stop.

The need to try to make his statistics seem better than they are by looking at incomplete records is a little weird, because none of this signifies anything. None of these statistics say anything about his circumstances, or his ability to improve going forward. And that's what matters, not some attempt to seek positives no matter the relevance.

X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

Join Bucs Nation

You must be a member of Bucs Nation to participate.

We have our own Community Guidelines at Bucs Nation. You should read them.

Join Bucs Nation

You must be a member of Bucs Nation to participate.

We have our own Community Guidelines at Bucs Nation. You should read them.

Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9341_tracker