Print and Go Back Baseball [Print without images]

Tuesday, June 3, 2003
Soft-tossing Rueter, Cornejo defying odds

By Nate Silver
Special to

Kirk Rueter has long frustrated opposing hitters by keeping runs off the board with a combination of slop, guts, and guile. He's long frustrated analysts for much the same reason. As you've heard people like Rob Neyer and Bill James state many times before, ERA is not the best predictor of a pitcher's performance going forward. Rather, it's his peripheral numbers -- particularly his strikeouts, walks, and home runs allowed -- that matter most.

Kirk Rueter
Starting pitcher
San Francisco Giants
12 73 6-1 22 21 3.33

This much we'll give him: Rueter is no Bob Tewksbury, but he has better-than-average command. But he gives up home runs at a rate higher than league average after adjusting for the power-dampening effects of Pac Bell Park, and his strikeout rates -- well, his strikeout rates are abominably low.

Of the 40 pitchers who threw at least 200 innings last year, Rueter had the fewest strikeouts per nine innings pitched, with 3.36. Nobody else was even close. The next name up the list, Kenny Rogers, had 4.52. As a group, the other 39 pitchers averaged 6.73 strikeouts per nine, more than double Rueter's total.

And yet, Rueter succeeds. Consistently. He was a key part of a pitching staff that made its way to the World Series last year, leading Giants qualifiers in ERA. He's off to a good start this year, too, tied for the team lead in wins (six), and was maintaining an ERA barely above 3.00 until he learned about the Coors Field Effect first hand on Memorial Day. What's the secret to his success? Do his big ears bring him good luck?

Well, maybe. But a lot of Rueter's accomplishments also have to do with good timing.

Here's what I mean. Let's look at Rueter's numbers since 2000 in three different situations: bases empty, runners on, and runners in scoring position. (We'll look at Rueter's numbers based on percentages of batters faced, rather than innings pitched, which is a helpful way to evaluate situational data).

Rueter ('00-'03) H HR BB SO
Bases Empty 27.4% 3.0% 5.9% 8.6%
Runners On 23.4% 2.2% 9.8% 10.0%
Runners in Scoring Position 23.0% 1.8% 12.6% 10.0%

With the bases empty, Rueter is a control artist, conserving his pitches, avoiding walks, and challenging hitters. He pays a price, allowing a lot of hits and a whole lot of home runs. But since those long balls are solo shots, they don't harm him all that much.

With runners on base, and especially runners in scoring position, Rueter's approach changes markedly. He becomes as annoying as Gilbert Gottfried, nibbling incessantly at the corners, triying his best to induce double play balls, running up his pitch count. The result is a few more strikeouts and a lot more walks. But he's rewarded with a dramatic reduction in his hit rate, and his homer rate.

You'll hear baseball analysts say all the time that There's No Such Thing as a Clutch Hitter -- it has practically become a slogan for our profession. Countless studies have been performed on the subject; none has turned up any evidence that there are certain hitters who consistently perform better when more is on the line.

But what about clutch pitchers? Is that what Rueter is, with his demonstrable ability to avoid big hits when it matters most?

Well, it depends on what the meaning of 'clutch' is. But more to the point, there's a fundamental difference between a pitcher and a hitter -- the hitter can only react to what's thrown at him, whereas the pitcher can be proactive. Within certain boundaries, he can manipulate the odds that various outcomes occur against him by changing the way that he pitches. If he's smart, he can adapt those outcomes to optimal levels based on the presence of baserunners, the number of outs, and the score of the game.

Rueter is smart. He challenges hitters when the bases are empty, giving up hits and home runs in the situations where they hurt him least. And he dodges them when the bags are full, giving up walks in the situations where extra base hits would be far more damaging.

Rueter, nicknamed Woody, is no blockhead.

In fact, he may have learned a thing or two from Tom Glavine, another pitcher who has consistently undermined the analyst reliance on strikeout rates. In the interest of brevity, we won't run the numbers here, but you'll see the same pattern reflected to an even greater degree when looking over Glavine's situational splits: hittable control pitcher with the bases empty, nibbling groundballer with men on. And since Glavine has better stuff than Rueter, he's been able to achieve even more success.

There's a Woody-wannabe this year, too: the Tigers' Nate Cornejo. Like Rueter, Cornejo comes complete with a movie tie-in, looking like a cross between the title character from Powder and the Stay-Puff Marshmallow Man. Also like Rueter, Cornejo doesn't fool anyone: through Monday, Cornejo has struck out just 11 batters in 60 innings (1.65 K/9), a rate that makes Woody look like the second coming of Nolan Ryan.

For purposes of comparison, no pitcher has maintained a strikeout rate of fewer than two batters per nine in a season in which he started at least 25 games since 1949, which is so long ago that Strom Thurmound was a Democrat. And yet, Cornejo keeps on keeping on with a 3.00 ERA and eight quality starts in 10 tries, putting him on pace to be the Tigers' obligatory All-Star representative. Is Cornejo a bigger and better version of Woody? Let's run his situational splits and find out.

Cornejo (2003) H HR BB SO
Bases Empty 28.7% 4.2% 7.0% 5.6%
Runners On 18.2% 2.0% 6.1% 3.0%
Runners in Scoring Position 17.6% 2.0% 5.9% 3.9%

Cornejo's been better at preventing hits with runners on base -- a lot better, and that's played a key role in producing his low ERA. But there's no evidence that this is the result of any deliberate strategy on Cornejo's part. Rather, his walk and strikeouts rate are actually somewhat lower with runners on. He's challenging hitters -- and he's getting away with it.

He won't for very much longer. With very few exceptions, pitchers have only limited ability to prevent balls in play from becoming hits -- that's mostly the job of the defense behind them. As a team, the Tigers have surrendered hits on 27.1 percent of balls in play against them. With the bases empty, Cornejo has yielded hits at a rate of 27.1 percent, exactly matching the team average. But with runners on, Cornejo's hit rate allowed has dropped to an unfathomably low 17.0 percent. There's no modern pitcher who has sustained a rate anywhere near that low, and there's no ability that can account for that degree of difference. Cornejo has simply been lucky.

Nate Cornejo
Starting pitcher
Detroit Tigers
10 60 3-3 16 11 3.00

Whether attributable to skill, luck, or some combination thereof, the performances of Glavine, Rueter and Cornejo stand out because they're unusual in today's game. We're living though an era in which strikeouts are as common as reality dating shows and Ron Artest temper tantrums.

Last year, the National League averaged 6.77 strikeouts per nine innings. Fifty years ago, in 1952, only one pitcher in the entire league who qualified for the ERA title managed that many (Wilmer "Vinegar Bend" Mizell, whose fastball must have been as disagreeable as his nickname).

Now, there are a couple of obvious explanations you can offer up for this:

1) today's pitchers are better;
2) today's hitters are worse.

It goes without saying that neither one holds up to even a cursory analysis. Run scoring, even if it's off a little bit, is still high by historical standards, and most of the advances in training, equipment and medical science that have taken place over the past 30 years have favored hitters. In fact, I'd like to make exactly the opposite case: today's hitters are better, and that's why strikeout rates are so high.

Take a look at these numbers. These are the correlations between strikeout rate (K/9) and ERA for each decade since the end of the deadball era, including all pitchers with at least 100 IP.

1920s -.16
1930s -.23
1940s -.15
1950s -.15
1960s -.29
1970s -.27
1980s -.26
1990s -.30
2000s -.38

Correlation is simply a tool to measure the strength of a relationship between two variables; all the numbers here are negative because, as we'd expect, there's an inverse relationship between strikeout rate and ERA. But there's never been such a strong inverse relationship as there is right now. It's no accident that we don't see too many Tewksburys and Quisenberrys anymore; strikeouts have always been important, but they're fast becoming a prerequisite for pitching success.

The next question, of course, is why. Except for player salaries and ticket prices (cough, cough), the only other thing that have increased as consistently as strikeouts throughout the course of baseball history are home runs. Batting average waxes and wanes, the stolen base comes into fashion as often as bellbottoms, and walk rates flutter up and down. But home run rates have increased steadily over time.

Here are another set of correlations. This time, we're looking at the relationship between strikeout rate (K/9) and home run rate (HR/9) for the same groupings of pitchers.

1920s -.06
1930s -.11
1940s -.07
1950s +.05
1960s -.15
1970s -.10
1980s -.08
1990s -.16
2000s -.22

Historically, there has been a very weak relationship between strikeout rate and home run rate. When we think of the classic power pitcher of the past, the home run was always his Achilles' heel. Tom Seaver placed in the top 10 in his league in home runs allowed seven times. Bert Blyleven, five times. Bob Feller, six times. Steve Carlton, nine times.

With just a couple of exceptions, that pattern doesn't hold any longer. Between them, Pedro Martinez, Randy Johnson and Roger Clemens have placed in the top 10 in their leagues in home runs just twice. The correlation figure is still relatively low because home run rates for individual pitchers can be fluky, but it's growing markedly stronger. Home runs used to be the price that a strikeout pitcher paid for challenging a hitter with his heater. Nowadays, hitters are so strong, ballparks are so small, the ball is so juiced -- whatever -- that the surest way to avoid a home run is to try and strike a guy out. The risk of challenging a hitter is simply too high, even for a power pitcher. (Jeopardy! Question: What is the reason why pitch counts are so high today?).

The operative concept here is competitive adaptation. Kirk Rueter adapts to his limited natural ability by making sure that his choice of pitch and location are as optimal as they can be given the game situation. Roger Clemens has tremendous natural ability, but adapts to today's stronger hitters by avoiding throwing hittable pitches when he's ahead in the count, even if it means putting more strain on his arm, and running up his pitch counts. And teams adapt to all this adapting by building deeper bullpens, using five-man rotations, and training their hitters to lay off anything out of the strike zone. Power begets power, and so the cycle continues.

You can check out more work from the team of writers of the Baseball Prospectus at