On a message board, the topic of Roethlisberger being a clutch player came up, and someone attempted to make the argument that there’s no such thing as a clutch player. Specifically this:
No such thing [as a clutch player], really. I thought [we were] above the notion that one player among the 22 starters suddenly waits until 4 minutes left to start playing really super-de-duper well and wills his team to victory - that's Around the Horn nonsense.
The first thing we have to do to be able to make an objective, quantifiable argument is to define what it means to be “clutch.” More specifically, what qualifies as a “clutch” situation, and what does not? As there isn’t much data to go on that can easily be sorted, I had to create my own. In this case, I defined “clutch” situations to be as follows:
• Games which ended with one team within one score of the other’s.
• Of those games, the final half of the fourth quarter.
Therefore, I would define a clutch player as one who plays well in those situations. I do NOT use the following anywhere in my definition of a clutch player:
• Game winner – One player can play extremely well in a clutch time, and the team still lose the game. Example: Warner played clutch in the Superbowl, but the Steelers still won.
• Plays better in clutch situations vs. other situations – If a QB plays very well in all situations, he can still be clutch. Clutch situations are situations where pressure is high on a player. Some players crumble under pressure. Others don’t. If you’re already a great player, and play just as well in clutch situations, I would argue that you are a clutch player. This point is debatable, but turns out not to matter in this argument.
• Big game player – There is no standard for what defines a “big game.” There is also no real way to tell how great a player played in a big game. Example: Roethisberger played poorly in his first Superbowl, but still played well enough to win.
In addition, there is one disclosure that must be made for the sake of making an honest argument:
• In clutch situations, defenses will often give a QB more opportunities to make plays. This relates closely to a QB playing better in clutch situations vs. other situations. Often times a QB just has to play well, and simply take advantage of what the defense is giving him. If a QB’s stats are better in clutch situations than they are everywhere else, does that mean he played better in the clutch, or that he played just as good but took advantage of the extra that the defense was giving him?
Unfortunately, we don’t have a base-line for this. To create a base-line, I would look at all NFL QBs play in clutch situations vs. their play in other situations. Given the effort I underwent here for just Roethlisberger, I’m not going to do this, simply because I think even taking the top 15 NFL QBs in ’08, such a base-line would take 25-30 hours of data mining to create. If someone else would like to undergo this process, please feel free.
Three categories are shown in the data below. First is his performance in my definition of clutch situations (titled “clutch”). Second is his performance during the first three and a half quarters of those same games (titled “games”). Third is his performance for the entire season, removing stats for the clutch situations (titled “season”). In the “season” categories, I included playoff game stats, since I included playoff games in his clutch situations.
Comparing the first with the second gives a look at how he performed in those clutch situations vs. how he was playing in that game. I think this provides a more apples-to-apples comparison, as theoretically if someone has a bad game, they should have it straight through, and vice versa for good games. If a player truly plays better in clutch situations, it will show here. Comparing the first with the third is also instructive, though because we’re looking at other games which had no bearing on clutch situations, can only tell us so much.
These are broken out year by year. In ’08, he had 9 games in which he threw a pass in the final 7:30 of the 4th quarter which had clutch situations. In ’07 he had 5, in ’06 he had 5, in ’05 he had 4, and in ’04 he had 3. For what it’s worth, the Steelers were 16-10 in those 26 games.
One caveat: In ’04, one of his three was the NYJ playoff game. The play-by-play data I looked up did not include the OT portion. Given the Steelers kicked a FG on their first possession in OT, his stats during the OT portion should be negligible…no TD, no INT, likely has minimal impact on completions and yards. My guess would be that including them would only improve his numbers, though marginally.
So then, on to the data.
Conclusion: This pretty clearly shows a deviation in how well he plays in clutch situations vs. the same game and vs. his all other times statistics. His QB ratings:
• Clutch – 99.6
• Games – 79.3
• Season – 88.1
This isn’t even accounting for the six spikes, which if you take out of the equation put his clutch completion % to 67.9%, and lifts his QB rating over 100. I don’t do that here cause I don’t remove spikes from the rest of his other stats.
Again, it’s difficult to compare this to a base-line for how NFL QBs perform in clutch situations vs. other situations. However, VERY clearly there is an improvement. He is better across every single category except sacks per attempt, and in many cases it isn’t by a small number.
So is Ben Roethlisberger what I would define as a clutch player? A player who, in tough, tight situations, performs very well and gives his team a great chance to win the game?
I think pretty clearly the answer is “yes.”