About a week ago we started this series on the NFL's most underrated quarterbacks by looking at San Francisco's Alex Smith; and next up is someone who took quite a different path to being under appreciated: the Dallas Cowboys' Tony Romo.

Unlike Smith, Romo was not the first overall draft pick.  In fact, he was never picked at all in the 2003 Draft, and ended up signed by the Dallas Cowboys as an undrafted rookie free agent, where he was almost cut before the season even began.

Unlike Smith, who spent his first couple seasons getting pummeled behind a poor offensive line, throwing way too many interceptions, and generally losing game after game; Romo spent all of '03, '04, and '05 on the bench behind Vinny Testaverde and Drew Bledsoe, respectively.

Finally, in '06, Romo got his first start, winning the quarterback battle between him and Bledsoe and replacing him as the starter for the final ten games of the season.  The Cowboys would make the playoffs that year and be a 19-yard field goal away from taking the lead against the Seattle Seahawks with roughly a minute to go.  

Romo, however, botched the hold (you can't make this stuff up) and, and as he scrambled left and appeared a sure-thing himself for the endzone, Jason Babineaux wrapped him up one yard short of the goal line (seriously, cannot make this up).

This was the all-too-pertinent beginning of Tony Romo's legend in the eyes of America.  Most starting quarterbacks never even have to handle a hold, and such a fate would therefore never befall them - but Romo started the season as the back-up to Bledsoe, and so it had been his duty up to that point, anyway.  The pure comic oddity of it all...

As the years went on, the Dallas Cowboys' inability to win the "big game" reflected poorly on Romo.  In the world of football, quarterbacks are judged most by the close games they win or lose and the Super Bowl rings they acquire or don't.  And when you play for a franchise as storied as the Cowboys, that criteria becomes magnified.  On top of that, the 17 weeks prior to the playoffs are quickly forgotten by a fan base, and the media, when the season ends on a sour note.  Similarly, losing late in the regular season, with playoff hopes on the line, is something people tend to remember (i.e., selection bias).  It is in these games the case against Romo is made.

Stats below courtesy of pro-football-reference:

Romo is 1 - 2 in career playoff games with a rating of 80.8, and 11 - 17  in regular season games played in December/January (ratings of 86 and 106.5, respectively).  Given that the NFL prefers to hold divisional games during the last few weeks of the season (i.e., during December/early-January), many of those 17 losses came against Romo's NFC East foes in games that helped dictate who would and would not make the playoffs.  In games against the NFC East in general (another thing people tend to remember: games against your rivals), Romo is 18 - 19 with a 90.5 rating, boasting all his least favorable stats when compared to his play against the rest of the NFC.

Still, if you average all those out, you come up with an above-average passer rating of 89.1 (math below for those wondering):

(Average rating in each category times games played in that category) added together and divided by total games played = average rating in "high-visibility" games
((80.8 times 3) + (90.5 times 37) + (86.0 times 25) + (106.5 times 3)) / 68 = 89.1 rating (record: 30 - 38)

This is eerily close to Romo's numbers in another category of games that people tend to remember: close ones.  In contests decided by seven or less points, he has amassed a rating of 89.4 and a record of 22 - 25.

If you take either of the ratings above (89.1 of 89.4), or both, it is clear that in matches where Romo's value is most measured and criticized - and most visible - he achieves a rating markedly less than his career average, but still high enough to put him among the top ten passers during most NFL seasons.  Even if that was all Romo was good for, it should be good enough to have a better record than 30 - 38 in those games.  But Romo does better than that.  For his career, believe it or not, he is the second highest rated passer of all time.  Yet, somehow, that only nets him a very pedestrian career record of 59 - 40.

Passer rating is an oft-criticized metric, and my consistent use of it here may seem curious...

SIDE NOTE:  If it bothers you that much, you may take comfort in playing around with the table here and noticing Romo is among the best all time during a quarterback's "prime" years in both AY/A and ANY/A.  And accounting for "prime" years actually hurts Romo's standing.  The reason for this is that Romo spent his first few years - a time when quarterbacks make many mistakes - on the bench.  

Also, Romo is just now exiting his "prime" years, at the age of 32, where one would expect such numbers to decline.  In other words, Romo has only ever had the opportunity to play during his peak years.  This makes his career averages appear better, so to account for that the table above only looks at other quarterbacks during similar years (notice Rodgers benefits from the same career path).  Romo's standings go up even further if you take out such age considerations.

However, as Cold Hard Football Facts has noted, passer rating is the Mother of All Stats in the NFL when you compare the two ratings of the winning and losing quarterbacks - and it is this point here we want to illuminate.  If Romo is performing among the best in his position, year-in, year-out, on a very consistent basis, even in losses and close games, then what exactly has his defense been doing against opposing quarterbacks?  The answer: not much.  Remember, if 25 of Romo's 40 losses have been within one touchdown, then while any one mistake by him is costly, so too is any one mistake by the defense (or special teams).

Let's take a look at just this past year to illustrate some points (ARL = Average Rating in Losses):

Quarterback Passer Rating Wins Losses Win% ARL
Aaron Rodgers 122.5 15 1 94% 80.1
Drew Brees 110.6 13 3 81% 85.4
Tom Brady 105.6 13 3 81% 87.7
Tony Romo 102.5 (104.2)* 8 7 53% 97.8
Matthew Stafford 97.2 10 6 62% 79.3
Matt Schaub 96.8 7 (11)+ 3 (5)+ 70% 86.4
Eli Manning 92.9 9 7 56% 81.3
Matt Ryan 92.2 10 6 63% 72.2
Alex Smith 90.7 13 3 81% 74.1
Ben Roethlisberger 90.1 12 4 75% 62.1

* Romo went 0 for 2 for no yards in Week 16 against the Eagles.  His rating was 39.6.  This actually affects his official rating since the attempts were still made, and the NFL does not account for what is an obvious outlier, giving the two attempts here the same weight as, say, his 47 attempts against Detroit.  Accounting for this ourselves bumps his rating up a couple points, for what that may be worth.  You will notice I also did not give Romo "credit" for that loss against the Eagles, making his record 8 - 7.  Similarly, that game did not affect his ARL.

+ Matt Schaub's win/loss record prorated for the rest of the season, assuming he stayed at a similar win percentage.  Since the Texans finished 10 - 6 without Schaub for six of those games, it seems fair to assume that with Schaub they would have won at least one more.  This was done for the sake of consistency in the table.

Out of the above quarterbacks, Romo posted the best passer rating in defeats by a significant ten point margin, and he also accrued the worst win percentage despite having a top four passer rating overall.  For every other quarterback the data is clear that losses came when their own play was poor - or, at best, average.  But Romo performed admirably all season long and still lost eight games.

Romo was the only quarterback last year to lose at least three games in which his rating was over 95.  In fact, he lost four.  Two of those came in the closing weeks of the season in contests against the eventual Super Bowl champion New York Giants.  If the Cowboys had managed to win either of those games, they would have made the playoffs, the Giants would not, and history would look immeasurably different.

Romo posted ratings of 141.3 and 106 in those games.  It really would not be fair to have asked him to play much better.  In particular, that 141.3 rating came at home with four touchdowns, no interceptions, and 34 points on the board - but the Cowboys' defense let the Giants put up 37.  With 0:46 seconds left, from his own 20, Romo drove the Cowboys down the field, within field goal range for a 47-yard try from Dan Bailey and a chance to force overtime.  But Jason Pierre-Paul blocked the kick and the Giants took the game.

In the Cowboys' eight losses last year, opposing quarterbacks averaged a rating of 102.2 - basically meaning while Romo's 97.8 ARL should be enough to get by more often than not, the Dallas defense had other ideas.  Winning the passer rating battle is a huge indicator of game success, and even though Romo does his part to get Dallas on the good side of that battle on a quite consistent basis, his defense does not.

It might be time to appreciate where the blame for Dallas' lack of success as a team truly lies: with the team.

Thus, when we ask the question of who is overrated in the NFL, it is important to know who exactly is doing the rating, as well as what criteria they are using.  When you are a quarterback in this league, a lot of the focus rests on you; and when you play for "America's Team" and are projected to make the playoffs every year by "experts", and expected to win the Super Bowl by fans, any less than that makes you a terrible quarterback, regardless of how well you play or how bad the rest of your team plays.

Such is Tony Romo's fate.  He is underrated because most the media and most the fanbase expect way too much from him.  Only one quarterback wins the Super Bowl every year, and it is usually not the best one.  Until people step back and take a look at Romo for his outstanding career averages, and not his alleged lack of "clutchness", he will continue to attract a lot of criticism.

Unfortunately, such calm, even-handed analysis from modern NFL pundits is few-and-far between - even in retrospect - let alone in the heat of the moment when Romo is still a starting quarterback and the Cowboys are still sitting at home on Super Bowl weekend.