Who's Best in the Clutch? Who Knows?
A million theories, but no clear answer

By Kenneth Broder
Los Angeles Herald Examiner
June 5, 1988

In 1987, he led the American League in RBIs, was fifth in Game-Winning RBIs, and ranked third in Go-Ahead RBIs. Now that's a clutch hitter.

He's nothing like that chump who ranked 75th in the league in driving home runs from scoring position when the pressure was on (seventh inning or later with the score tied or his team trailing by three or fewer runs -- four runs if the bases were loaded).

... Or the guy whose hitting dropped across the board (batting average, homers, slugging average, etc.) in pressure situations.

... Or the hitter who had his worst month down the stretch in the middle of a pennant race his team lost.

Nobody should really be surprised that all the above players are the same man, George Bell of the Toronto Blue Jays. Probably the most confusing, contentious arguments in baseball are whether there is any such animal as the clutch hitter and, more importantly, how do you recognize him in the wild.

Now, I'm not one of "the small group of shrill pseudo-statisticians" who, according to the authors of the Elias Baseball Analyst, have "used insufficient data and faulty methods to try to disprove the existence of the clutch hitter."

I'm one of those shrill pseudo-semi-sports journalists who is still waiting for the definitive measure of what determines a clutch hitter. Is it RBIs, GWRBIs, Go-Ahead RBIs, Batting Average in Pressure Situations, Percentage of Runners Driven in From Scoring Position or Percentage of Runners Driven in From 3rd with Less Than Two Out? Maybe it's Bill James' Very Important RBI stat, his RBI-Importance stat or WBI (Wins Batted In).

Or how about RAB (Runs Per At-Bat) or TIB (Teammates Batted In, which is RBIs minus Home Runs) or Runs Produced (Runs plus RBIs minus Homers)?

No kidding; someone actually sits around and makes up these lists. Unfortunately, the lists rarely agree unless they use the same base of comparison, like RBIs.

Last year in the American League it only took 19 players to fill out the Top 10 lists for RBIs, Game Winning RBI and Go-Ahead RBI. This unusual incestuous relationship between statistical categories was highlighted by George Bell, Mark McGwire, Gary Gaetti and Alan Trammell appearing on all three.

But not one of these fine 19 "clutch hitters" sparkles in the Analyst's benchmark test of clutch hitting. Percentage of Runners Driven in from Scoring Position in Pressure Situations. That list is topped by Ruppert Jones and includes Alfredo Griffin, Wayne Tolleson and Garth Iorg. Mark McGwire, who appeared on all three RBI charts, also appears on this Analyst list -- he was 12th worst in the league.

Would you have pinch hit Alfredo Griffin for Mark McGwire in late-inning pressure situations last year when they were teammates in Oakland?

One of the problems with using clutch statistics is that players have relatively few opportunities in a season to prove their ability; the results are skewed by a puny statistical sampling. So let's try the career leaders, among active players, in the (oh, dare I write it again) Percentage of Runners Driven in from Scoring Position in Pressure Situations?

Jose Canseco tops the list, Eddie Murray is third, Pedro Guerrero ranks sixth and Wade Boggs checks in at seventh. I've heard all these players referred to as clutch hitters through the years, except for Canseco, who hasn't been around long. But if you buy the proposition that this stat is proof these guys are great clutch hitters, maybe the best, you've got to take the whole package -- and that includes Eric Soderholm, No. 2, Jim Essian, No. 4, Jim Norris, No. 5, Pete LaCock, No. 8, Lenn Sakata, No. 9, and Eddie Milner, No. 10.

It's hard enough dealing with the illusions and biases inherent in Runs Batted In and Runs Scored without a dozen freak stats mucking up the picture. Do we really need another formula like Richard Zitrin's Teammates Batted In (RBIs minus Homers) that puts Bill Buckner on the road to Cooperstown for having the good fortune to play in Boston and hit behind Wade Boggs?

Which is not to say that there is no such thing as clutch hitting. Common sense dictates that some ball players will react to pressure situations differently than others; they're only human.

A few weeks ago I did a small study that compared the runs a team scored in 1987 with its projected runs scored based on the Runs Created formula (Hits + Walks - Caught Stealing) x (Total Bases + .55 x Stolen Bases) / AtBats + Walks). The difference in the two numbers, which by the way is usually less than 4 percent, I flippantly assigned to luck. But there could be another partial explanation.

The four National League teams with the lowest Percentage of Runs Driven In From Scoring Position were also the same four teams that failed most miserably to meet their projected runs total. The pattern also held true throughout the rest of the majors, though not nearly so strikingly. (See Team Clutch Hitting chart.)

But if clutch hitting is the explanation for this anomaly, how does one account for the fact, and it is a fact, that most ball clubs that outperform or underperform their projected runs stat one year often swing the other way the following year?

There's a lot of room for research in the area, and as more raw data becomes available to baseball stat freaks more succinct explanations of clutch hitting, albeit contradictory at times, will be published.

Until then, we'll continue to be ill-served by outfits like the Elias Sports Bureau that use one method to list George Brett among 30 players proven over a three-year period to be a poor clutch hitter and another method that assigns him the post-1974 record for Single-season Batting Average with Runners in Scoring Position, at .466.

It's the quality of the statistics that count, not the quantity. I think they choked on this one.

Team Clutch Hitting

American LeagueNational League