Numbers tell grim story of what Watt loss would mean for Texans

News broke Tuesday that Houston Texans star defensive end J.J. Watt may be lost for the season with a back injury, so what does that mean for one of the league's most dominant defenses? We can quantify Watt's bottom-line effect on a game by comparing the Texans' expected points added (EPA) per play when he is on the field to their EPA/play when he's not.

Since Watt's rookie year in 2011, the Texans' defense averaged minus-0.06 EPA/play with Watt on the field, and minus-0.04 EPA/play with Watt on the sideline. (Negative EPA is good for defenses.) That difference of 0.02 EPA/play equates to about 1.2 points per game based on Houston's average of 62.4 snaps per game. Those 1.2 points per game would be roughly equivalent to 4 percent win probability each game. So at first glance, Watt's absence would be significant, but not necessarily fatal to the Texans' playoff hopes.

But Watt was not an every-down player in his first year and a half in the league, and his performance found a whole new gear starting in 2013. Since the start of the 2013 season, the difference in defensive performance with Watt on the field and off the field is a whopping 0.09 EPA/play. That equates to 5.6 points per game or roughly 20 percent win probability!

There have been only 194 snaps without Watt since '13, so consider the 20 percent win probability number as an extreme upper bound of his impact. That kind of effect on win probability is what might be more typical when a team loses an elite quarterback to injury. But if any single defensive player can have as significant an impact on his team's prospects, it might be Watt, who is both a preeminent run defender and pass-rusher.