The other day on the MLB Network Brian Kenny and his panel again excoriated Met manager Terry Collins for letting Met ace Matt Harvey pitch the ninth inning with a two run lead at home in finals game five against Kansas City. Harvey blew the game and the series. The intro was Collins sounding like a jackass with recent statements about not wanting to use data.
Kenny showed video clips of the things that supported him but as usual omitted the fact that Jeurys Familia, the Met relief ace of the moment, who had really good numbers in the regular season, had blown the lead in finals game one just a few days before by allowing a home run to the number eight batter Alex Gordon in the ninth. Collins made the mistake of having Familia pitch the 9th two days before in a 9-3 Met win, then also used Familia the previous day in a Met loss. Familia was not fresh for game five.
I don't like the Terry Collins old school nonsense. It's not old school v. new school. It's no school v. school.
But I would have stuck with Harvey every time for reasons mentioned in previous posts in October 2015. This post takes an anecdotal look at a few previous playoff games that I happened to be looking at.
The resurrection of the Harvey matter made me wonder how Sandy Koufax had done it. Harvey supposedly melts down as soon as he throws pitch number 100. I don't think there is pitch number data that detailed from 50 years ago but we do have performance by innings.
I think there are two reasons pitchers threw more innings years ago:
- fewer walks and strike outs
- pitchers didn't throw that hard.
Koufax threw tons of innings, walked few but struck out tons and finished many games, including those in the old World Series. Like many current pitchers Koufax threw very fast pitches; he also threw a big curve ball. All of this made Koufax the first I wanted to examine to determine if there were any parallels to the Harvey situation, which represents the current dogma.
Generally, Koufax pitched quite well in innings 7,8,9. However did he do it? In one sense he didn't. Koufax retired at age 30 so that his arm wouldn't fall off. But that only accounts for the toll taken overall. It does not address his overwhelming success in big games against really good hitters. In those games Koufax almost always went all the way. It was unthinkable to remove Koufax even for the Dodgers ace relief pitcher and fellow lefty Ron Perranoski. Perranoski started his career with the Dodgers and pitched for them ages 25-31. He started one game in his career and it was in his rookie season 1961. In 1963 Perranoski led the National League in winning "percentage" with .842 (16-3) and games with 69; he threw 129 innings but with only 75 SO; 43 BB. Koufax won his first Cy Young award in 1963 and pitched games one and four against the two time defending champion Yankees in the World Series. Yankee lefty ace Whitey Ford opposed Koufax in both those games.
Koufax at age 27 (one year older than Harvey) in 1963:
25-5 (Yeah, I know, kill the win.)
40 starts, 20 complete games, 11 shutouts, 311 innings
306 SO, 58 BB (7 IBB)
led NL: Wins, ERA, SHO, SO, FIP, WHIP, H9, SO/W
Koufax was dominant in the World Series but not unhittable. With today's thinking, Koufax could have been removed in both games.
The Dodgers swept the Yankees and used only four pitchers: three starters and Perranoski who pitched only in game two in relief of lefty Johnny Podres: RBI single, the the final two outs in the 4-1 Dodger win. Don Drysdale started game three and beat Jim Bouton 1-0.
Wednesday, October 2, 1963, Yankee Stadium I game one
Attendance: 69,000, Time of Game: 2:09
Dodgers 5, Yankees 2
Koufax threw about 148 pitches. That was determined using the Tom Tango pitch count estimator. See my own research using this:
Pitch Counts and the pitch count estimator for seasons past. Monday, May 11, 2015
The Dodgers scored 4 in the second and another in the third. Koufax struck out the first five Yankees and 15 for the game, breaking the World Series record of 14 set by Brooklyn Dodger Carl Erskine exactly ten years before also against the Yankees but in Ebbets Field. Erskine struck out Joe Collins and Mickey Mantle four times each; Dodgers won 3-2. Erskine went all the way and threw an estimated 146 pitches.
Koufax was cruising but less dominant as the game wore on. He struck out the side in the fourth giving him 9 SO. He struck out the first and last batters in the fifth but allowed three consecutive singles in between. In the sixth inning with one out Koufax dangerously walked Richardson and Tresh but got Mickey Mantle and Roger Maris to pop out. Seventh: three up, three down; one SO. Eighth: SO, single, SO, HR by Tresh. The Dodger lead is cut to three with Mantle and Maris coming up. Do you pull Koufax? That would surely be the consensus today. Koufax must have been well past 100 pitches by then on his way to about 148. We don't know what pitch number may have been a point of no return for Koufax in 1963, if there was one.
In 1963 Koufax completed half of his 40 starts, so obviously he pitched fewer innings later in games. 38 innings in each of the first four, then 37, 33, 30, 27, 23. He became less effective as the game wore on. OPS for times facing an opponent:
1 - .431
2 - .507
3 - .532
4 - .568
Still great, just less so. For groups of three innings:
1-3 - .470
4-6 - .525
7-9 - .506
OK, you now have more information, probably way more than Dodger manager Walter Alston had back in 1963. What do you do? Do you pull Sandy Koufax? Come on, make a decision. Mickey Mantle's waiting to bat.
Well, we know what happened. Mantle walked. What now? Isn't this eerily similar to Matt Harvey walking the leadoff batter in the 9th of his ill fated game? Harvey wound up throwing 111 pitches. Alston stuck with Koufax and Maris grounded out.
But do you send Koufax out for the bottom of the ninth? Elston Howard led off with a line out to second. Pull him now? Joe Pepitone grounds a single to right. Now? Clete Boyer flies out and Koufax strikes out pitch hitter Harry Bright for the record. Koufax threw about 148 pitches.
Koufax innings 1960-1963: 175, 255, 184, 311. Koufax was in uncharted waters.
Maris was injured in game two and did not play again. Ford and Koufax faced each other again in game four with the Dodgers poised for a sweep.
Sunday, October 6, 1963, Dodger Stadium game four
Attendance: 55,912, Time of Game: 1:50
Dodgers 2, Yankees 1
Both pitchers were sharp. Koufax struck out two Yankees in the first and two more in the third. In the fourth Richardson got a bloop double for the first Yankee hit. In the 5th Howard led off with a single but the next three went down, one by SO. Frank Howard homered off Ford in the fifth: Dodgers 1, Yankees 0.
The Yankees were retired in order in the sixth but with one out in the seventh Mantle homered to tie the game. Pull Koufax? Then Howard singled. Pull Koufax? Hector Lopez, the replacement for Maris, and Pepitone made out.
The Dodgers scored an unearned run in the seventh to lead 2-1. Pull Koufax to protect the lead? Eighth: Koufax struck out Boyer but Phil Linz pinch hit for Ford and singled. Pull Koufax? Tony Kubek grounded into a double play.
Do you let Koufax take the mound in the ninth to protect the 2-1 lead? That's half the lead that Harvey had in 2015. Richardson led off with a single. Pull Koufax? Tresh and Mantle were called out on strikes. But then the second baseman makes an error on Howard's grounder putting runners on first and second. Koufax is now pitching in inning number 328 of his season. Lopez grounded out to short. Game over. Koufax completes the game and the Dodgers sweep. Koufax threw about 147 pitches, about the same as in the first game he pitched.
Without going through detailed melodrama here are three other games with comparable decisions to be made.
Thursday, October 15, 1964, Busch Stadium I - game seven
Attendance: 30,346, Time of Game: 2:40
Cardinals 7, Yankees 5
Cardinal starter Bob Gibson staggered to a complete game victory, giving up home runs to Mantle, Boyer and Linz, before retiring Richardson for the final out with Maris and Mantle waiting to hit. Like Familia of the 2015 Mets, the main Cardinal relief pitcher Barney Schultz had allowed a ninth inning home run earlier in the series: walk off to Mantle in game three. The Cardinal manager was quoted as saying his heart belonged to Gibson who threw about 148 pitches.
Saturday, October 11, 1986, 5:25, Anaheim Stadium game four
Attendance: 64,223, Time of Game: 3:50
Angels 4, Red Sox 3 11 innings
Red Sox starter Roger Clemens had a 3-0 lead going into the 9th but was relieved by Calvin Schiraldi. Doug DeCinces had led off with a homer, then out, single, single. Enter Schiraldi. Double, intentional walk, HBP. Tie game. Schiraldi lost the game in the 11th on a single to Bobby Grich. Clemens threw about 139 pitches. Series tied 2-2.
Sunday, October 12, 1986, 12:05, Anaheim Stadium game five
Attendance: 64,223, Time of Game: 3:54
Red Sox 7, Angels 6 11 innings
Angels starter Mike Witt may not stand out now but in 1986 he was 18-6, 2.84 ERA, 144 ERA+, 269 innings, completed 14 of 34 starts. He was the ace of the Angels. Top 9, Angels leading 5-2: Witt: single, SO, HR, popup. Angels leading 5-4. Witt was initially relieved by Gary Lucas who hit Rich Gedman in the head. Then Donnie Moore relieved. Moore blew the lead and then later lost the game. Moore gave up a two run homer to the first batter Dave Henderson: Red Sox took the lead 6-5.
The Angels tied the game in bottom of 9th but Moore allowed the winning run in the 11th on a sac fly by Henderson. Witt threw about 117 pitches. Witt had pitched a complete game 8-1 victory over Clemens in game one on Oct. 7 in Boston, so Witt was pitching on four days rest. The Red Sox also won the final two games in Boston to win the series 4-3, overcoming a 3-1 deficit.
Angels manager Gene Mauch had famously blown the lead for first place in 1964 while managing the Phillies by starting his two leading starting pitchers as much as possible in the last week. Now, 22 years later, Mauch was making opposite but comparably ineffective decisions.
The current conventional wisdom is that Terry Collins blew that finals game five by not believing the data that indicated that his ace Matt Harvey would turn into a pumpkin once he had throw pitch number 100. Personally, I would have stuck with that bull Harvey and I think that Collins would have lost the confidence of his players had he pulled Harvey after eight innings, especially if Familia had blown a lead for the second time in five games.
It's like the Pedro Martinez incident that got his Boston manager fired. At the time I thought that when Martinez is inducted into the Hall of Fame no one will remember the name of the relief pitcher they thought should have replaced him against the Yankees in the 2003 ALCS. Same with Harvey. It's a variation on one of my ideas: never trade a somebody for a bunch of nobodies. Here: never relieve a somebody with a nobody.