"Win-or-go-home" from a team point of view

[Dodger]

I posted it here a couple of months ago, after finding it myself fooling around with pfr's player-game finder.

http://www.pro-football-reference.com/play-index/pgl_finder.cgi

If it's been published or broadcast anywhere, I'd be interested in knowing where.

It was probably your post where I saw it.

 

[Hoofbite]

Not according to the few stats that matter. You're overestimating the significance of the 163 passing yards, because passing yardage doesn't correlate to winning. Here, for example, those yardage numbers you mentioned are low because they basically reflect what the Giants did through three quarters. Eli didn't throw a pass in the last 14 minutes of that game because he didn't have to. If he had, figure the effect of every added attempt with a 132.4 rating. The yardage numbers will change, but the rating won't.

It's not the quantity of stats that you can pile up, it's the quality. "Yards + TD" isn't a quality stat. Quarterbacks with the kind of yardage and TD numbers Eli had that day win playoff games almost as often as they lose. Since 1960, teams whose QB passed for no more than 163 yards and 2 TD in a playoff game are 131-148.

Over the last ten seasons in the NFL, teams whose QB passed for no more than 163 yards and 2 TD in a game are 558-817-1. But teams whose QB had exactly those same numbers with a rating over 125 are 15-0. Say what you want about that game, but Eli's 132.4 rating is the elephant in the room.

No but it generally scores points which was my point. Regardless of how high that number is, if that was the total output for the opposing QB most teams would take it because that isn't much and generally isn't enough to win unless your own team gives up scores on special teams or their defense scores. Having a great rating doesn't really mean your team has scored a lot of points. 70 QBR separates Romo and Eli but only 4 points on the scoreboard.

Eli not having to pass is somehow a point in favor of QBR? Had he thrown only 1 more pass and completed it for 7 yards his QBR drops even though he improved his production in terms of completion percentage and yardage. QBR rewards efficiency (in terms of YPA), not production, but production is what generally wins game. Having a high QBR could mean that you have a stud QB who carried you or it could mean you have a dud QB who was given limited opportunities but hit on a big play.

They'd both have have QBRs but I'd bet one situation is more preferable than the other for just about anyone who was asked.

In this case the Giants won the game not because they were awesome on offense but because Dallas made more mistakes and went flat. QBR was a product of the way the game played out, not the other way around.

 

[BigMac6]

Wow, 0-67 is a hell of a stat.

I will say that I don't agree that Romo would have had to have a 125+ to win the Giants game. If you look at his numbers, 18/36, 201, 1 TD, 1 INT....you only need to shift that INT to the TD column to win that game. Dallas was close to the Giants 20 and that INT came on the second to last play of the game.

If that is a TD instead of an INT, his numbers look like 19-36, 224, 2 TDs, 0 INTs. That's only a 90.5 QBR. Not only does that swing his QBR considerably just from a single play but Dallas wins.

The Giants scored 21 points and Manning was 12/18, 163, 2 TDs, 0 INTs and 52 yards and 1 TD came in the first 3 minutes of the game. If that big play to Toomer isn't completed and Manning finishes with 11/17, 111, 1 TD, 0 INTs.....his QBR is 102 instead of 132.

If you were to tell someone that their playoff opponent would throw for 163 yards and 2 TDs while gaining only 253 yards (230 after sacks), most people would chaulk that one up as a win provided there weren't defensive scores. Not really a big deal but I think saying that Romo needed a monster QBR for the team to win kind of marginalizes his actual role in the loss. The offense moved the ball less than 60 total yards on their last 4 possessions and only 115 yards in the entire second half.

By just about every stat in the book, Dallas should have won that game. Far more yardage, 13 minutes more of ball control and no turnovers until the 2nd to last play of the game. Running game was very good with 150 yards and 4.7 per carry. The only way a team loses with those sort of advantages is if you have each unit fail at some point throughout the day. Defense gave up the big play early and New York converted on a couple of clutch 3rd downs, one of which was right before the half on 3rd and 10 and the Giants scored right after going into the half. Offense went cold and finished the game pretty badly by having to possession start at midfield in the last couple minutes of the game and they came away with nothing. And Dallas had 11 penalties.

I think the stat is still impressive but how many of these 67 games hinge on a single play? A tipped or dropped pass, a wrong route, a QB getting hit as he throws or just a bad bounce? I think the Dallas/Giants game is a statement to the idea that making plays when they count the most and when your team needs them the most is kind of what determines the outcomes of games. I also think another thing that is lost in statistics are these sort of individual anomalies that sort of help boost the numbers. Games where a certain team has no right in winning based purely on being outperformed to such a great extent yet they are able to pull it out because their opponent made too many mistakes or couldn't capitalize on it's advantages.

Great discussion and great points overall.
Talking about Dal vs NYG game, keeping in mind that the Cowboys didn't play well as a team and Hochuli's crew (if a I'm correct) had a really bad day, how about Patrick Crayton who dropped a wide open pass and then stopped a route in the endzone in the last but one play of the game? Or Reeves, that couldn't cover my grandmother in a wheelchair all day?
It's incredible, at times ridicolous, how the media and people who has played the game at professional level forget how football is always a team effort.

 

[percyhoward]

Having a great rating doesn't really mean your team has scored a lot of points.

In a general sense, maybe. Specifically, though, having a 125+ rating in the playoffs really means your team won the game.

Eli not having to pass is somehow a point in favor of QBR? Had he thrown only 1 more pass and completed it for 7 yards his QBR drops even though he improved his production in terms of completion percentage and yardage. QBR rewards efficiency (in terms of YPA), not production, but production is what generally wins games.

What you're doing there is called research by proclamation.

Production (total yardage) is not what wins games. By that I mean it's been proven over and over that it doesn't. You can see this proof in correlation studies and random samples. Completion percentage is a measure of efficiency, and it does have a slight correlation with wins, although nowhere near that of passer rating or yards per attempt. Passing yardage is a measure of production and it has no correlation at all.

I put some specific examples in the last two paragraphs of my previous reply to you. The correlation info you can find anywhere.

 

[IrishAnto]

A common faulty assumption that I've seen in the discussions of this team is that, based on their record, this was an average team the last two years. To call the 2011-12 Cowboys an average team is only technically true. It can't be overstated: Take away the passing game in 2011 and 2012, and the teams that the Cowboys put on the field (especially after key injuries on defense) were bad football teams.

ESPN and others have declared that "Romo's" record in win-or-go-home games is 1-6. To accept that this record belongs to the QB as much as (or more than?) it does to the team requires that you assume that all teams are equal except at that one position. Of course, if that were true, then the only players who would ever be drafted or traded would be quarterbacks.

Let's instead call the 1-6 "the Cowboys'" record, and look into what made it -- but from a team perspective, so we can get an idea of how other QB have fared with teams that have performed similar to Romo's. I think something is not quite right about the way four of the games are perceived.

First, the two playoff losses (Giants in 07 and Vikings in 09), and the role of Dallas' defense in those two games.

Since 1960, teams whose defenses allow more than a 125 rating in a playoff game are 0-67. In other words, no matter who your QB was or what he did in the game, when the opposing QB goes over 125.0, you lose. In those games, our defense allowed well over a 125 rating (132.4 and 134.4). The predictable result was two losses.

Romo's numbers in those games (ratings of 64.7 and 66.1), are the kind that Brady and Roethlisberger have won with in the playoffs (they're 6-2 when their own ratings are in the 60's or lower). It's still 6-2 for them and 0-2 for Romo, even though they won because their teams ran the ball well and/or their defense made big plays.

That 0-67 tells us, that, unless he had put up a couple of 125+ ratings himself in those two games, Romo's performance would not have even mattered. Every QB since 1960 who has found himself in that situation in two playoff games is 0-2 -- not just Romo. It happened to Marino three times (0-3). It happened to Elway five times (0-5). It's happened to Romo and Matt Ryan in every year that they've made postseason appearances since their first (0-5).

When all teams are considered equal except for the QB, W-L records like these give the QB the reputation of someone who can't win big games. These were team losses, attributable mostly to the performance of that QB's own pass defense. And Dallas' pass defense has only gotten worse since then.

Which brings us to the week 17 losses the last two seasons. Again, Romo gets the blame for something that no other QB has been able to do. In 2011 and 2012, the Cowboys finished 29th and 25th in defensive passer rating. Those same two years, Dallas also ranked 30th and 27th in rushing TD. Defensive passer rating and touchdowns both have very high correlations to wins, so teams that ranked 25th or worse in consecutive seasons in even one of those categories have historically been bad teams. Teams that ranked 25th or worse in BOTH categories two years in a row have always had losing records. Always, as in, every single time.

Since 2000, there have been 58 teams that ranked 25th or worse in defensive passer rating. Their combined record is 270-607 (.308). There have been 38 teams that ranked 25th or worse in rushing TD, and they combined to go 186-398-1 (.318). Bad teams. And Dallas was 25th or worse in BOTH categories BOTH seasons.

Since the league expanded from 28 teams to 30 in 1995, these are all of the teams that ranked 20th or worse in consecutive seasons in those two categories. These are the dregs of the NFL over the past two decades, because to do it in consecutive seasons in both categories means it wasn't a fluke of scheduling. There is no doubt. Your team was among the league's worst at pass defense and rushing TD.

* = overlapping
2011-12 Dal (8-8, 8-8)
2010-11 Was (6-10, 5-11)
2008-09 Rams (1-15, 2-14)
2008-09 Sea (4-12, 5-11)
2008-09 Chiefs (2-14, 4-12)
2008-09 Lions (0-16, 2-14)
2007-08 Rams (3-13, 2-14)
2007-08 Bengals (7-9, 4-11-1)
2006-07 Falcons (7-9, 4-12)
2003-04 Lions (6-10, 5-11)*
2002-03 Lions (2-14, 6-10)*
2002-03 Cards (5-11, 4-12)*
2001-02 Lions (3-13, 2-14)*
2001-02 Cards (7-9, 5-11)*
2000-01 Cards (3-13, 7-9)*
2000-01 Falcons (4-12, 7-9)
1999-2000 Chi (5-11, 6-10)
1999-2000 Falcons (5-11, 4-12)*
1998-99 Saints (6-10, 3-13)
1998-99 Bears (4-12, 6-10)
1997-98 Colts (3-13, 3-13)
1996-97 Cards (7-9, 4-12)*
1995-96 Cards (4-12, 7-9)*

What you won't see there is a winning record. That's 37 different team-seasons represented. The only team to finish as high as .500 was the Cowboys.

And the Cowboys did it twice.

There have been 35 losing seasons, and two .500 seasons, both by the same team.

Remember, these are teams that finished 20th or worse. Dallas' highest ranking was 25th. So about half these teams were as good as or better than the Cowboys in those two categories, and they still couldn't manage to go .500.

As should be obvious by the W-L records, none of those teams was ever in a win-and-in game the last week of the season. And yet the Cowboys somehow managed to get into a win-and-in game each of the last two years.

Somehow.

If you're wondering why the team awarded a new contract to a QB who is 1-6 in win-or-go-home games, these are the reasons why. The team knows all teams aren't equal minus their quarterback. The team knows it's not really "Romo's" record.

Sit back and watch what Romo and Dez do the next few years, and hope we can improve enough (and stay healthy enough) in the other areas so that what they do actually means something.

Great Post Percy and I can’t disagree with anything you’ve said.

I will add one thing though.

Turnover differential.

I bet if you add that stat to your list of teams, you’d find they all were on the wrong side of that particular fence.

 

[dallasfan4lizife]

This argument on Romo being a choke or clutch is old.

If his career ended today, he would be remembered as a choke in most peoples' eyes, other than those on this forum.

I don't care what his 4th quarter QB rating is. I don't care about his furious comebacks that usually fall just short. I care about stacking post-season wins which he has yet to do.

It's sad to see people settle for peace of mind regarding why they aren't winning. Trying to make up excuses for why this team has failed and failed and failed when it counts the most.

 

[IrishAnto]

It's sad to see people settle for peace of mind regarding why they aren't winning. Trying to make up excuses for why this team has failed and failed and failed when it counts the most.

You said it right there "team" of which Romo is only a part (even if it's the most important part).

 

[ABQcowboyJR]

I see many of your point of views in regards to defining moments in games. However there is statistical support for regression to the mean. In the end the stats will prevail despite whatever anomalies occur.

 

[dallasfan4lizife]

You said it right there "team" of which Romo is only a part (even if it's the most important part).

He's contributed to the failures, but there are 21+ other guys that generally contribute as well.

 

[xwalker]

It's hard to argue that if your defense is bad beyond a certain point, then it's basically impossible for the QB to win.

It's even harder to argue that if your defense is really bad and your running game is really bad, that losing is probably not the QB's fault.

The only possible counter argument that I can think of would be in regards to how interceptions by your QB affect your defense's performance. If your QB throws 5 INTs then your defense's passer rating is probably going to be worse than if your QB threw zero INTs.

 

[percyhoward]

Great Post Percy and I can’t disagree with anything you’ve said.

I will add one thing though.

Turnover differential.

I bet if you add that stat to your list of teams, you’d find they all were on the wrong side of that particular fence.

Thanks.

Dallas ranked 10th in turnover differential in 2011, then ranked 27th in 2012. That kind of flip-flop in the rankings from one season to the next happens about 40% of the time, so it doesn't work very well as a two-year stat.

 

[percyhoward]

The only possible counter argument that I can think of would be in regards to how interceptions by your QB affect your defense's performance. If your QB throws 5 INTs then your defense's passer rating is probably going to be worse than if your QB threw zero INTs.

Yes, the shorter fields for your opponent would make TD more likely and boost their passer rating. The way around that effect would be to use a different stat that, like passer rating, correlates highly to wins, but that, unlike passer rating, does not include touchdowns. Yards per attempt, for example. Turnovers by your own QB don't affect the opposing QB's yards per attempt in any way.

Since 1960, teams whose QB averaged at least 10.5 YPA in a playoff game are 61-5.

As for how the scenario you described relates to the Cowboys, neither the Giants nor Vikings converted a Romo turnover into a passing TD to help their quarterback's rating. And season-wise, the Dallas pass defense was 24th and 26th in YPA allowed the past two seasons.

 

[Picksix]

This argument on Romo being a choke or clutch is old.

If his career ended today, he would be remembered as a choke in most peoples' eyes, other than those on this forum.

I don't care what his 4th quarter QB rating is. I don't care about his furious comebacks that usually fall just short. I care about stacking post-season wins which he has yet to do.

It's sad to see people settle for peace of mind regarding why they aren't winning. Trying to make up excuses for why this team has failed and failed and failed when it counts the most.

What is truly old, is constantly putting everything on Romo. It's not "he", it's "the team". And yes, I do want to have a greater understanding of why we're not winning. Repeatedly being fed false correlation, and then have people (Cowboy fans or not) regurgitate that same misinformation, is very frustrating to those of us who want to have intelligent conversation about our team, and know what we're talking about.

 

[5Stars]

This is such an interesting post, but Sadistics was not my favorite class in college. I even had a hard time with Algebra 101.

But, to see how you guys put the info together is cool to read. Plus, it totally makes sense.

 

[dallasfan4lizife]

What is truly old, is constantly putting everything on Romo. It's not "he", it's "the team". And yes, I do want to have a greater understanding of why we're not winning. Repeatedly being fed false correlation, and then have people (Cowboy fans or not) regurgitate that same misinformation, is very frustrating to those of us who want to have intelligent conversation about our team, and know what we're talking about.

That's what I said. But QB's generally take all the blame when the team chokes, and all the glory when the team wins.

 

[Red Dragon]

Whenever I think Cowboys fans are being absurdly negative about Romo, I go to other NFL team's message boards, read what their fans are saying about their team's starting QB, and then suddenly Cowboys fans almost seem reasonable by comparison.

 

[jobberone]

This argument on Romo being a choke or clutch is old.

If his career ended today, he would be remembered as a choke in most peoples' eyes, other than those on this forum.

I don't care what his 4th quarter QB rating is. I don't care about his furious comebacks that usually fall just short. I care about stacking post-season wins which he has yet to do.

It's sad to see people settle for peace of mind regarding why they aren't winning. Trying to make up excuses for why this team has failed and failed and failed when it counts the most.

Well that says a lot.

 

[percyhoward]

Well that says a lot.

"My mind's made up, don't bother me with the facts."

 

[IrishAnto]

Thanks.

Dallas ranked 10th in turnover differential in 2011, then ranked 27th in 2012. That kind of flip-flop in the rankings from one season to the next happens about 40% of the time, so it doesn't work very well as a two-year stat.

Maybe on a Year by Year basis, but I wonder what you would find if you went on a game by game basis.

2011 was an anomaly for the Cowboys. In the first 8 games of the season the defense only generated 10 turnovers (pretty much on par with most of their last 6+ years) and the offense turned the ball over a good deal more than the previous year (where Romo was very good at avoiding turnovers). This was a big reason we were 1-7.

However when Phillips got fired and Pasqualoni took over the running of the defense they generated 20 turnovers in the last eight games and Kitna and the offense in general reverted back to 2010 form and we went 5-3.

So even though we were 10th overall in turnover differential on the year, I would bet if you looked at the 10 games we lost we were on the wrong side (as far as winning goes) of that differential in each game.

In a nutshell you could go three games in a row generating 3 turnovers in each and conceding none in each of those games winning all three, then go the next three games generating no turnovers but conceding two in each game loosing all three. The overall differential would be +3 which would look good until you looked at each individual game.

Just my two cents.

 

[percyhoward]

2011 was an anomaly for the Cowboys. In the first 8 games of the season the defense only generated 10 turnovers (pretty much on par with most of their last 6+ years) and the offense turned the ball over a good deal more than the previous year (where Romo was very good at avoiding turnovers). This was a big reason we were 1-7.

You're mistaking 2010 for 2011.