Romo has ranked among the league's top 10 rated passers seven consecutive seasons. Since the only QB with more consecutive top 10 seasons in the NFL-- Montana, Fouts, Young, Manning (whose streak continues), and Brady (whose ended last year) -- are all Hall of Famers, people think there must be something misleading about the stat, at least as it applies to Romo. Break down pass rating into its component parts, and you can see that there's nothing mysterious about it. It measures four elements on a per-attempt basis: TD, INT, completions, and yards.
Since 2010, no team has lost more close games than the Cowboys. In this analysis of why Dallas hasn't posted a winning record the last four seasons, those four elements are isolated (along with some others, including two stats involving the QB that aren't a part of pass rating -- sacks and lost fumbles) in the 4th quarter or overtime of the Cowboys' close losses.
in losses (2010 - 2013)
among NFL's top 10 rated passers with 1800+ attempts
4th qtr or OT, score + or - 8 points
(includes playoffs)
First, here they are ranked by pass rating. Again, that's just TD, INT, completions, and yards.
Player...Att/TD%/Int%/YPA/Rtg
Romo ..... 222 / 5.0 / 4.5 / 7.8 / 83.1
Rodgers...136 / 2.2 / 1.5 / 7.0 / 83.1
Roethbgr..136 / 4.4 / 2.9 / 7.0 / 77.9
Brees ......142 / 3.5 / 4.9 / 6.9 / 66.7
Stafford....201 / 3.0 / 3.0 / 5.9 / 63.2
Ryan........196 / 0.5 / 3.6 / 6.1 / 58.8
PManning.. 94 / 3.2 / 7.4 / 6.4 / 57.3
Brady........162 / 3.1 / 4.3 / 5.7 / 56.6
EManning. 116 / 5.2 / 7.8 / 6.8 / 53.2
Rivers...... 232 / 0.4 / 6.5 / 5.9 / 44.7
The first thing you may notice is that their number of pass attempts varies greatly. Some of these players were involved in more losses than others. Since these are all highly-rated QB, most of that difference is explained by the quality of the players around them, or by games missed due to injury.
Romo heads the list, hundredths of a rating point in front of Rodgers. This is traditional passer rating, so sacks and fumbles don't count. They're examples of what someone is talking about when they say "Stats don't tell the whole story." Of course, sacks and fumbles are also stats. They're recorded, they just aren't part of pass rating. So let's count them. Instead of pass attempts, I'll look at total dropbacks. Instead of INT, I'll look at all QB turnovers.
This will be our starting point. Touchdowns vs. turnovers, as a perecentage of all dropbacks. To see the effect of sacks and fumbles, compare Romo's 5.0 TD% above (pass attempts only) to his 4.8 below (pass attempts + sacks), for example. Also compare his 4.5 INT% above to his 5.2 turnover % below. Then, most importantly, look at the differential.
Touchdown Percentage
(the percentage of all dropbacks resulting in TD)
EManning 4.9
Romo 4.8
Roethbrgr 3.9
Brees 3.3
PManning 3.2
Brady 2.9
Stafford 2.8
Rodgers 2.0
Ryan 0.5
Rivers 0.4
Turnover Percentage
(the percentage of all dropbacks resulting in an INT or lost fumble)
Rodgers 1.3
Roethbrgr 3.3
Stafford 3.3
Brady 4.1
Ryan 4.4
Brees 4.7
Romo 5.2
PManning 7.3
EManning 7.3
Rivers 7.5
Differential
Roethbrgr +14.1
Romo +12.8
Rodgers +8.8
Stafford +6.4
EManning +5.1
Brees +4.3
Brady +3.9
PManning -6.8
Ryan -14.1
Rivers -27.2
If I wanted to mislead, I could mention just one of these two categories--either TD or turnovers. You absolutely must look at both of them together. Even better, see the differential.
Although we've all seen the statistic "TD/INT Ratio," TD and turnovers are not equal values, because not all turnovers either cost the offense a touchdown, or result in a touchdown for the opponent. Turnovers have been found to be worth 4 points, while TD are obviously worth 7. Accordingly, I multiply Eli's TD% by 7 and his turnover% by 4, then subtract to get his +5.1 differential.
High touchdown and turnover totals are the sign of a QB who takes more risks (see Eli) in late-and-close situations. Likewise, a QB with fewer turnovers (Rodgers) often has fewer TD as well because he is playing it safe in these situations. (As some have suggested, maybe too safe, when you look at his sack totals in the losses.) Rodgers has avoided late turnovers in close losses, but they're still losses, and the Packers' poor W-L record when trailing late is partly due to his risk aversion (and partly due to his defense, as you'll see later.)
Late in close losses, Romo's TD/TO differential is 2nd best among these top quarterbacks. What this means is that when their team lost a close game, Big Ben and Romo have played relatively well in the 4th quarter with regard to touchdowns and turnovers -- not the only parts of the game, but two big ones.
The next two categories are yards per dropback and sacks, two statistics which you don't see in a passer rating. Since sacks are very often not the fault of the QB, I did not subtract sack yardage from the QB's yardage total, I just counted the sack as a 0-yard completion.
Yards Per Dropback (sacks are counted as pass attempts)
Romo 7.6
Brees 6.5
EManning 6.4
PManning 6.3
Rodgers 6.3
Roethbrgr 6.3
Ryan 5.8
Stafford 5.5
Rivers 5.4
Brady 5.4
Sack Percentage
(the percentage of all dropbacks resulting in a sack)
PManning 2.1
Romo 3.5
Brady 4.7
Brees 5.3
EManning 5.7
Stafford 6.5
Ryan 7.1
Rivers 7.9
Rodgers 10.1
Roethbrgr 10.5
Again, you want your QB to rank in the top half of both of those lists. It's easier to keep your sacks down with short passes and throwaways, but since those count as attempts, your YPA will suffer (see Brady). Romo ranks in the top two of both lists, and is really in a class by himself in the first one. We're still talking late in close losses.
Note that two of the best QB in the TD/TO differential (Roethlisberger and Rodgers) are the bottom two in sacks. Rodgers took only half as many sacks late in close wins (5.3) as in close losses (10.1) the last four seasons. Big Ben's sack percentage was actually higher in their wins. Romo's was about the same in the wins (4.1) as losses (3.5).
If sacks and all turnovers were a part of passer rating, the net effect on Romo's would be positive. And that's late in close losses. Also, if you haven't been keeping track, Romo and Brees are the only players who rank in the top 7 of all the lists we've seen so far.
Plays of 30+ yards as a percentage of total dropbacks
EManning 5.7% (7 of 123)
Romo 4.8% (11 of 230)
Stafford 2.8% (6 of 215)
Rodgers 2.0% (3 of 151)
Ryan 2.0% (4 of 204)
Roethbrgr 1.3% (2 of 152)
Brees 1.3% (2 of 150)
Brady 1.2% (2 of 170)
PManning 1.0% (1 of 96)
Rivers 0.8% (2 of 252)
First Down Percentage
PManning 40.6% (8.36 to go)
Romo 33.5% (9.33 to go)
Brady 33.5% (9.08 to go)
Rodgers 31.8% (8.79 to go)
Roethbrgr 30.3% (9.45 to go)
Brees 30.7% (9.59 to go)
Rivers 29.8% (9.17 to go)
Ryan 27.5% (8.71 to go)
Stafford 25.1% (9.60 to go)
EManning 24.4% (9.28 to go)
Both lists are about yards, whether in large or small chunks. Yards in the 4th quarter or OT of close games that these QB's teams ultimately lost.
The first list explains itself. The second list is about possession -- it's the percentage of pass plays that result in a first down (with the average yards to go on those plays). Again we see the feast-or-famine that is Eli. A ton of big plays, not many first downs. Proving that yes you can have both, only Romo ranks among the top 3 in both lists, which is to be expected from a QB whose yards per play is on another planet.
Peyton's first down percentage dominates, but moving the chains is not his problem in close losses--it's his poor TD/turnover differential and relative lack of big plays. Unlike Romo, his teams don't lose many close games, but also unlike Romo, when they do, the loss is usually on him.
This adjusted rating that follows uses dropbacks instead of attempts, and includes all QB turnovers -- treating lost fumbles as INT.
Player...Dpbk/TD%/TO%/Sk%/YPD/Rtg
Romo ..... 230 / 4.8 / 5.2 / 3.5 / 7.6 / 76.7
Rodgers...151 / 2.0 / 1.3 /10.1/ 6.3 / 75.1
Roethbgr..152 / 3.9 / 3.3 /10.5/ 6.3 / 67.2
Brees ......150 / 3.3 / 4.7 / 5.3 / 6.5 / 63.2
Stafford....215 / 2.8 / 3.3 / 6.5 / 5.5 / 57.2
Ryan........204 / 0.5 / 4.4 / 3.9 / 5.8 / 52.5
PManning.. 96 / 3.2 / 7.3 / 2.1 / 6.3 / 56.2
Brady........170 / 2.9 / 4.1 / 4.7 / 5.4 / 54.0
EManning 123 / 4.9 / 7.3 / 5.7 / 6.4 / 50.3
Rivers...... 252 / 0.4 / 7.5 / 7.9 / 5.4 / 34.7
Even though sacks are listed and do count as attempts, again I did not subtract sack yardage, or else Rodgers would drop back closer to the pack. Anyway, you get the idea. The Dallas pass offense with Romo has clearly been the best in the league in the 4th quarter or OT of close losses.
How is it that the QB with the best numbers late in close losses has a reputation for...losing close games late? Most of it is that the team he plays on loses a lot of close games late -- games that would not be close in the first place if not for the Cowboys' passing game.
Defense - TD% Allowed in Late-and-Close Losses
Ind/Den 12% (26 drives, 3 TD)
Saints 12% (43 drives, 5 TD)
Falcons 16% (50 drives, 8 TD)
Chargers 18% (68 drives, 12 TD)
Steelers 20% (61 drives, 12 TD)
Patriots 24% (33 drives, 8 TD)
Packers 24% (51 drives, 12 TD)
Giants 27% (56 drives, 15 TD)
Lions 28% (72 drives, 20 TD)
Cowboys 30% (64 drives, 19 TD)
Having one of the league's worst defenses means two things: (1) you'll be involved in more close games than other teams (assuming your offense is good enough), and (2) if the game is close in the 4th quarter, you'll be more likely to lose it. None of these defenses has performed worse in late-and-close losses than the Cowboys' defense, and the running game hasn't been any better.
Offense - Rushing First Down Percentage in Late-and-Close Losses
Patriots 31% (55 rushes, 17 fd)
Ind/Den 29% (49 rushes, 14 fd)
Packers 28% (92 rushes, 26 fd)
Falcons 28% (58 rushes, 16 fd)
Chargers 22% (98 rushes, 22 fd)
Giants 22% (55 rushes, 12 fd)
Steelers 21% (71 rushes, 15 fd)
Saints 20% (55 rushes, 11 fd)
Lions 17% (110 rushes, 19 fd)
Cowboys 16% (67 rushes, 11 fd)
The most important run stat late in close games is first down percentage -- how well you move the chains. None of these teams' running games has performed worse in late-and-close losses than the Cowboys' running game. And even though yards per rush attempt is an almost meaningless stat, Dallas ranks 25th in that category late in close losses over the last four seasons, and 28th in lost fumbles.
You hardly ever hear this, but if it weren't for pass offense, this Cowboy team of the last four seasons (2010-13) would easily be the worst in the history of the franchise over a 4-year period. Look back, and you won't come close to finding a worse run offense, pass defense, and run defense over a 4-year stretch since 1960.
All these close-game numbers I've been giving are 4th quarter and OT only. In case you were wondering about Romo's performance in the first three quarters of these losses...
Quarters 1-3 of games lost by 8 points or less, 2010-13
Romo 93.5
Rodgers 91.9
Brady 91.8
Ryan 88.7
Rivers 84.4
Stafford 81.3
Brees 76.8
Roethbrgr 76.2
PManning 71.6
EManning 68.3
Just TD, INT, completions, and yards. This list provides some needed perspective about Brees and Roethlisberger in close losses. While they did play better than most in the 4th quarter, it wasn't enough to make up for their 76 rating in the rest of the game. More often than not, these QB (along with both Mannings) had dug the hole in which their team found itself.
And what about Brady in close losses, with a 91.8 rating through 3 quarters, and a 56.6 in the 4th despite having the best run support of all these QB? That's a collapse. Likewise, Rivers goes from a decent 84.4 through 3 quarters to a 44.7 in the 4th, despite a defense that held the opponent without a TD more often than any other. Ryan's story is similar.
The list also shows how this team has made a real habit of wasting Romo's best games. After all, he does not have the highest 1st-3rd quarter rating in all games--not even top five. Just in the close losses. And when this team loses with Romo, it's usually close. Of the 52 games in which Romo has seen significant action the last four seasons, 44 have either been wins or close losses.
This last one will blow your mind.
When QB Has Pass Rating Between 95-110
(W-L, 2010-13)
Brady 13-2 (.867)
Brees 12-2 (.857)
Roethbrgr 8-2 (.800)
Rodgers 9-3 (.750)
Ryan 10-4 (.714)
PManning 11-5 (.688)
EManning 8-5 (.615)
Stafford 9-8 (.529)
Rivers 6-7 (.462)
Romo 4-10 (.286)
Since 2010, when Romo has a 95-110 rating in a game, the Cowboys are 4-10 (.286).
When these other QB have a 95-110 rating in a game, their teams are 86-38 (.694). And that winning percentage is not what makes those quarterbacks special.
When any other QB in the league besides Romo has a 95-110 rating in a game, those teams are 243-120-2 (.669).
Pause and reflect. Response is optional.
dwmyers, perrykemp, and KJJ all (mostly unintentionally) provided ideas or info used in this post.