We can almost always look back in a "what could have been" way and pinpoint a play or two that, if different, would have dramatically changed the landscape of a game. That's just the nature of professional football.
But it's still interesting and useful to understand just how much a touchdown here or a fumble there alters the course of a contest. We can do that with Advanced NFL Stats’ win probability graphs. Using past game data, the site graphs each team's chances of winning a game in real-time. Here's the chart from the Cowboys' Week 2 loss in Kansas City.
You can see that the 'Boys were the favorites to win for right around half of the contest – all of the second quarter and most of the third. At their peak, when they had a three-point lead with a first down at the Chiefs' 5-yard line, the Cowboys owned a 79 percent chance of winning. That is, if we were to play out the rest of the game 100 times from that starting point in the middle of the third quarter, the most likely outcome is that Dallas would win 79 times.
There really weren't many game-altering moments in this contest, but there were still a few plays that shaped the outcome. Here are three.
1: Jamaal Charles Touchdown
The first quarter Charles touchdown wasn't accompanied by a huge increase in win probability for Kansas City just because the Chiefs were already on the 2-yard line. But that entire drive, one that lasted nearly half the quarter and resulted in a 7-0 lead, gave the Chiefs a 72 percent chance to win before Dallas even touched the ball.
And in truth, the probability was likely even higher than that. Most of us had a sense that this would be a relatively low-scoring game, and in such contests, the value of each score is greater than in high-scoring affairs. We all remember what happens late in games, but the early scores are just as important.
2: Lance Dunbar Fumble
When Dunbar fumbled away the ball, the Cowboys had a second-and-4 at their own 49-yard line. Despite being down 14-13, Dallas actually owned a 53 percent chance to win at that point since they had the ball. That shows you the value of possession. As NFL offenses become more and more potent, field position will matter less and just having the ball will matter even more than it has.
In any event, the Cowboys' win probability dropped to 32 percent once the Chiefs recovered Dunbar's fumble. It was the single biggest rise or fall in the game, dropping the Cowboys' win probably by 21 percentage points.
3: Tony Romo Fumble
A few minutes into the fourth quarter, the Cowboys faced a third-and-5 at their own 38-yard line, down by four points. They had exactly a one-in-three chance of winning prior to the play. After Romo was blitzed and fumbled, Dallas had just an 18 percent chance to win the game. The Cowboys quickly rallied back to turn the game into a near coin flip, but they were never able to rebound. [embedded_ad]
Where are the big plays?
If Terrance Williams hadn't run the wrong route on Romo's Week 1 interception, the quarterback wouldn't have any picks through two games. That's a positive, obviously, but what's the cost? It seems like there's been such an emphasis placed on minimizing turnovers that it's really limited the Cowboys' ability to acquire big plays. It's like the offense is playing scared, attempting to avoid turnovers, regardless of the cost.
Look, the turnover differential is vitally important to a team's success, but you're not going to be a very effective offense if all you care about is minimizing mistakes. I could be the quarterback of the Cowboys and not have a single interception. I just wouldn't throw the ball. Problem solved, right?
The Cowboys aren't going to win games by doing everything possible to limit turnovers in the same way that they wouldn't win by just blindly running the ball because "balanced teams win games." In reality, winning teams become balanced and turnover-free offenses win games because they avoid turnovers and generate big plays.
Through two weeks, five of Romo's passes have traveled 20 yards or more. That's 5.5 percent, the second-lowest mark in the NFL, according to Pro Football Focus. The Cowboys have one of the top wide receivers in the NFL who seems to come down with (almost) every jump ball, and has a track record of success when throwing the ball deep. Yet we've seen five deep passes all year.
Romo needs to find a happy medium between being a gunslinger and playing so conservatively that the benefit of not throwing interceptions is outweighed by the cost of not moving the ball. Maybe if the Cowboys start thinking of punts as what they really are, turnovers, they'll be a bit more risk-seeking, ultimately increasing offensive efficiency.
