If it weren't for those four bad starts, eh?

Over on USSM today, Derek pointed out an article on the Hardball Times by Craig Burley called Phranklin, which is an interesting read, regardless of whether you're a Mariners or Phillies fan, or just like to read people discussing whether a particular pitcher is bad or seems bad.

The upshot of the THT article is that "Pat Gillick claims if you just remove Ryan Franklin's four awful starts, he really wasn't that bad a pitcher" -- and Burley sets out to prove that no, really, he still was that bad a pitcher. Which makes you wonder whether it's a valid claim at all, and contrariwise, how much worse would a good pitcher be if you just took out a few of their good starts?

Well, because I just finished tallying up the 1940 game log for Hughie Mulcahy tonight (though, sadly, my numbers, the NYT season numbers, and baseball-reference's numbers don't agree -- I have him at 91 walks and 81 strikeouts, NYT had him at 90 and 81, and b-ref has him at 92 and 81; I'm going to check the Baseball Encyclopedia tomorrow), let's see if I can do this four-worst and four-best thing on him. Since, after all, my original thought was that he wasn't such a bad pitcher either, but was just cursed by run support and by being on the 1930's Phillies. This is a guy who got named to the 1940 All-Star team despite a 7-10 record at the break, who would work his way to a 12-10 record before dropping 12 straight decisions to go 13-22 for the season.

Here's Hugh Mulcahy in 1940, with all his games, and then adjusted to remove his four worst starts - I'm using baseball-reference's numbers for now; I added in FIP and WHIP to Craig Burley's numbers:

          GS  W-L    IP      H   R   ER  HR  BB  SO  ERA   FIP   WHIP
Real      36  13-22  280    283 141  112 12  91  82  3.60  4.15  1.336
Adjusted  32  13-18  263.7  253 111  83  9   80  77  2.83  3.97  1.263

For the day, Mulcahy really wasn't so bad a pitcher, if you think about it. In 1940, you could lead the NL in K/9 with less than 5 (as opposed to today's near-10s), could lead in K/BB in the mid-2's... of course, there weren't really relief pitchers back then, either, at least not as we know them today. (The guys leading the leagues in saves had 7.) The AL had slightly higher numbers to lead their pitchers that year - well, higher K/9 but lower K/BB and such.

Yes, a guy with a 2.83 ERA still would have lost 18 games. Why? Because the Phillies that year suuuuuuuuuucked. This is a team that scored 494 runs, allowed 750 (87 of which were unearned -- Bobby Bragan alone made 49 errors that year in 132 games) and averaged 3.23 runs scored per game. As I calculated it, the team scored 115 runs of support while Mulcahy was on the mound, and he allowed 141 runs total, 112 of which were actually earned.

If you calculate the Pythagorean expected W/L on his games, real and adjusted for removing the 4 worst:

             RA    RS     PW-PL          ER      RS      PW-PL
Real        141   115     14-21          112    115      18-18
Adjusted    111   109     16-16           83    109      20-12

Ahh, and there he is, a 20-game winner, if not for all those damn errors and those four bad outings! The funniest part is, there was an article in the July 31, 1940 New York Times actually saying "Will Mulcahy be the first 20-game winner the Phillies have had in the last 24 years?" Naturally, that was right before his 12-game losing streak.

Now, I'm going to do this the other way, taking out his four best starts (I'm going on earned runs to pick them -- he had three shutouts, and I'm including what I thought was his best start outside them -- a 13-inning complete game on May 23 where he gave up 15 hits, walked none, struck out 10, and of course, the game-losing run came in on Chuck Klein's error):

          GS  W-L    IP    H   R   ER  HR  BB  SO  ERA   FIP   WHIP
Real      36  13-22  280  283 141 112  12  91  82  3.60  4.15  1.336
Adjusted  32  10-21  240  250 137 109  11  85  67  4.08  4.30  1.396

             RA    RS     PW-PL          ER      RS      PW-PL
Real        141   115     14-21          112    115      18-18
Adjusted    137    99     11-21          109     99      15-17

Hmm. Yeah. Still bad, of course, but really not as atrociously bad as one would expect a guy to be to win a moniker like "Losing Pitcher". And if he'd been playing on a team like the Pirates where they were averaging 5.19 runs scored per game, he might have even had a .500 record.

You can't compare Losing Pitcher Franklin and Losing Pitcher Mulcahy easily, to be honest, but I'm working on it. They both definitely had one thing in common: they went out there and threw a metric crapload of innings for a pathetically bad team (albeit one with a much better defense, that's for sure) and came out looking much the worse for the wear.