How Badly Will Big Ben's Suspension Hurt The Steelers

Brian Burke at the excellent blog Advanced NFL Stats tries to tackle this question. In the write up below, Burke notes that the replacement of Roethlisburger with an average quarterback would probably result in one extra loss over four games. With Big Ben we might go 3-1, but without him, we would go 2-2. But, the Steelers back ups have played much better than average in the last four years. I hope that trend continues for at least four to six more games.

Over his career, Roethlisberger has averaged 6.7 net yards per attempt (NYPA) and a 3.4% interception rate. In 2009, the league average was 6.2 NYPA and 3.1% int rate. According to my game probability model, all things being equal, Roethlisberger's passing stats would turn an average team into one with a 57% chance of beating an average opponent.

If we replace him with a substitute for a few games, it's hard to predict how the Steelers' passing game would do. But just as an exercise, say their passing game drops by one standard deviation. Their NYPA would be 5.7 and their int rate would be 4.5%. All else equal, this would drop them from a 57% team to a 40% team.

If the suspension, which is expected to be announced later today, were for four games, you could think of it as a 4 * 0.17 = 0.7 win penalty for Pittsburgh.

But there's something unusual about Roethlisberger's replacements over the past several years. Since 2006, Steeler backup QBs have averaged 7.4 NYPA and a 2.1% int rate, a substantial improvement over Roethlisberger's numbers! Charlie Batch, Byron Leftwhich, and Dennis Dixon combined for 1930 yards on 243 attempts. They gave up 5 ints and 54 sack yards. The backup QBs differed from Roethlisberger primarily in int rate and sack rate.

Roethlisberger was sacked on 9.1% of his drop backs, while his replacements were sacked at a 4.0% clip. Assuming a backup, whoever it is, continues to perform so well, he would make a 50% team into a 67% team.

The backups were sometimes playing in lower leverage situations, so it's not fair to expect they would continue to perform as well as they did. It's also likely they would regress to the mean, but it's hard to say by how much.

Still, these numbers might make it easier for the Steelers to consider putting Roethlisberger on the trading block.

For reference, you can see how the Steeler backups compare to Roethlisberger in terms of Win Probability Added (WPA) and Expected Points Added (EPA) over thepast several seasons here.

What do you think? What will the Steelers' record be after four games?