Winning with a 1-2 pitching punch

By Kenneth Broder
Los Angeles Herald Examiner
July 31, 1988

When the Minnesota Twins stumbled to a pennant -- and eventual World Series championship -- last year, the critics had a field day.

The club's .525 winning percentage was the worst ever by a Series winner, and the second-lowest for any team even appearing in the Fall Classic. They were eighth in the American League in scoring runs and were lucky to get that many. (The Twins scored 29 more runs that the Runs Created formula predicted.) And their starting pitchers were basically a two-man rotation Viola and Blyleven.

Whitey Herzog, manager of the Cardinals team that lost to Minnesota in the Series, was quoted in the Kansas City Times as saying, "They've got two pitchers, that's it." Whitey was mostly bitter about how far you could get in post-season play without much depth on your team, and in that respect he was right, though he didn't think much of the Twins in general.

But his crack about two-pitcher teams echoes the conventional wisdom that successful teams usually have deep pitching staffs. He was wrong.

The Twins bionic Violevencombo had a winning percentage of .593, 102 points better than the rest of the team, and accounted for 38 percent of their wins. But, as Bill James points out, the historical average gap on championship teams is 106 points with the two top pitchers averaging about 40 percent of the wins.

The Twins weren't an anomaly and neither is the Dodgers' fearsome twosome of Orel Hershiser and Tim Leary.

As of Thursday, the Dodger duo had a won-loss record of 24-12, (.666), 119 points better than the rest of the squad (35-29, 547), and was responsible for 40.6 percent of the team's wins. That's a profile of true championship balance.

As a team, the Dodgers have watched their ERA shrink from 3.72 last year to 3.01 in '88. Some of that is attributable to the fact that the average team is scoring about half a run less this year than last. But, surprisingly, that drop has not resulted in a higher percentage of Quality Starts by the Dodger staff. (A pitcher gets credit for a Quality Start by pitching six or more innings and giving up less than four earned runs.)

Last year, the Dodgers ranked second behind Pittsburgh in the National League with a QS in 60.5 percent of their games. That's compared to the '87 NL average of 49 percent. Their percentage this year is 59.4. The Dodger ERA in those games is 1.80, which isn't all that phenomenal. In '86, (the only year I have stats for) the National League average ERA in games with Quality Starts was 1.94.

The big difference for the Dodgers so far this year is how they've fared in games their starters stunk up the joint. In the 27 games where the starter didn't have a Quality Start but got the decision, that decision was a loss 23 times. However, the Dodgers are an excellent 12-3 in non-QS games where the starter did not get a decision, for an overall mark of 16-26 and a .381 percentage.

That might not seem like much to you, but last year the Dodgers won a total of 14 non-QS games for a .197 percentage; both marks were last in the major leagues. Only three teams in each league in '87 had winning percentages above .381 in non-QS games above.

A lot of people don't like the Quality Start as a stat. They point out that a pitcher isn't going to stay in the "Show" for long with an ERA of 4.50 (that's what the minimum QS works out to) and they're right. But most pitchers who wrack up a QS pitch far better than the minimum, as the numbers above illustrate. The stat is a good indicator of quality, and that's enough.

In that regard it's similar to the Game Performance stats, known in some quarters as the Nolan Ryan Rating because it evaluates a pitcher's game performance on a scale of roughly 1-100 irrespective of whether he won or lost. Bill James conjured it up -- and I've added a few wrinkles of my own -- and this is how it works.

Start with a nine-inning perfect score of 100 points, subtract 47 points (27 outs, 20 strikeouts);add one point back for each out and strikeout; and subtract four points for each earned run, two for each unearned run, two for each hit and one for each walk.

Come to think of it, maybe this stat ought to be renamed for the Dodger who owns it, Tim Leary. Despite Hershiser's Cy Young performance, Leary owns seven of the Dodgers' top Game Performances and only one of the 14 worst.

In fact, if his Run Support by the hitters wasn't worst on the team, Leary's won-loss record might be much better. I say might because even with the crummy support, the Dodgers have outscored opponents in his games by a margin of 73 to 48. Using an obscure formula I won't detail here, that kind of scoring ratio would result in a 12-5 record, not 10-7.

The Dodgers, who score 4.25 runs per game, have only scored 3.48 in games Leary started. Next lowest is Hillegas (3.57), followed up the ladder (in a not-so-clever attempt at a segue) by Sutton (4.15), Hershiser (4.48), Belcher (4.86) and, oh no, Fernando (4.90).

You can blame Valenzuela's troubles on a lot of things, but hard luck isn't one of them. Although 12 of Fernando's 20 outings have been Quality Starts he's had five of the worst eight Dodger Game Performances. For the most part, when he's good he's OK, but when he's bad he's wretched.

A couple of months ago I noted John Benson's observation that pitchers who have a seasonal ORA -- (Hits + Walks) / IP -- as high as 1.50 "seldom return to form." Valenzuela, who's lifetime mark before last year was 1.18, ballooned to 1.51 in '87 and is currently at 1.58 -- and climbing.

DODGERS GAME PERFORMANCE
15 Best Starts
NameDateOppIPHitsRunsERBBSOTotal

Leary4/18SD9.030021183 W
Leary5/25Phi9.01002682 W
Hershiser6/29Hou9.02001479 W
Hershiser4/05SF9.03002678 W
Leary5/19Mon9.070001076 W

Leary6/21Atl9.04111774 W
Leary7/18StL9.05001473 W
Leary7/27SF9.052131071 L
Leary7/23Pit9.05221970 W
Hillegas6/27Hou6.02001369 W

Belcher7/22Pit6.01112568 W
Hershiser5/18Mon7.05112866 L
Valenzuela4/24SF7.65003366 W
Belcher6/18SD6.33003366 W
Three tied at 65 points

15 Worst Starts
NameDateOppIPHitsRunsERBBSOTotal

Howell6/16Atl6.015981713 L
Valenzuela7/20StL5.011772420 L
Leary5/01StL0.64663121 L
Hershiser5/12Pit7.012773423 L
Valenzuela5/22NY1.66552024 L

Valenzuela5/17Mon7.010666428 L
Valenzuela6/25Cin0.65441028 -
Valenzuela7/01Cin7.08666230 L
Sutton5/20NY4.38443233 L
Valenzuela6/03Cin2.35442234 -

Belcher5/03Pit5.07553435 W
Belcher6/07Hou2.66522335 L
Sutton4/12SD4.67441036 L
Belcher6/12SD4.06442237 L
Leary6/17SD5.06553538 L

CHART FACTS
* Opp is Opponent
* Total is Game Performance, on a scale of 1-100. Game Performance is derived by starting with 53 points (100-27 outs - 20 SO). Add one point for each out recorded and one for each strikeout. Subtract four points for each earned run, two for each unearned run, two for each hit and one for each walk.