Maddux in the postseason
Colin William
HoF, I decided to do an analysis of Maddux's postseason numbers with the
Braves. I've wondered for a while if there were any patterns in either
his or the team's performance that were behind his relatively weak W/L
record in the postseason; it's a relatively small sample, but ultimately
it's the one blemish on his career. I'm only looking here at his
postseasons with the Braves.
So, first, Maddux's career stats with ATL:
--194 wins, 88 losses, 2.63 ERA, 3.01 RA, 2273 hits, 383 BB, 152 HR and
1828 K in 2526 IP; 8.10 H/9IP, 1.36 BB/9IP, 6.51 K/9IP, 0.54 HR/9IP,
And in the postseason:
--11 wins, 13 losses, 2.82 ERA, 3.91 RA, 171 hits, 42 BB, 12 HR and 116
K in 181 2/3 IP; 8.47 H/9IP, 2.08 BB/9IP, 5.75 K/9IP, 0.59 HR/9IP
Most of the differences are about what one might expect given the better
teams a guy faces in the postseason; frankly, I'd expected a higher
postseason ERA than that. But one stat really leaps out, that RA. In the
regular season he allowed 0.38 unearned runs per 9IP, but in the
postseason that skyrockets to 1.09. It works out to 22 unearned runs in
27 postseason starts, 0.81 a start, as opposed to 0.29 a start in the
regular season.
Comparing team wins to team losses in Maddux's starts:
Losses:
--4.04 ERA, 5.85 RA, 10.1 H/9, 3.08 BB/9, 6.06 K/9, 0.64 HR/9
--Team averaged 2.28 runs in losses against pitchers with an average ERA
of 3.67, and allowed 5.0 runs/loss
Wins:
--1.80 ERA, 2.27 RA, 7.18 H/9, 1.23 BB/9, 5.58 K/9, 0.57 HR/9
--Team averaged 5.85 runs a game in his wins against starters with an
average ERA of 3.76, gave up 2.23 runs
FWIW, in his career Maddux averaged a 1.82 ERA/2.05 RA in his wins, vs a
5.62 ERA/6.45 RA in his losses. So, we see a few things here - Maddux's
postseason ERA in wins was as good as in his regular season career,
though he allowed more unearned runs. In his postseason losses he was
also better than in the rest of his career, despite the better
competition (though with a caveat - his Braves postseason career came in
his prime). But here's where we see the impact of those unearned runs -
he averages as many unearned runs in his losses as he does earned runs
in his wins, 1.81 overall. That's huge.
Looking closer, we find that he gave up unearned runs in 6 of the team's
14 losses, giving up 5, 5, 2, 2, 2 and 1. In three of those games the
number of unearned runs was equal to or greater than the team's margin
of defeat.
I wondered if Maddux suffered from frequently facing the opponent staff
ace, but as I looked over the postseason matchups, the matchups were
pretty mixed. Maddux wasn't often the game 1 starter for postseason
series. I averaged the regular season ERAs of all the opposing pitchers
Maddux faced. The average season ERA of pitchers he faced was 3.71.
Maddux's ERA for his Braves career was 2.62. So, theoretically he had a
full run advantage over his postseason opponents.
Looking at score breakdowns, when the Braves scored 3 or fewer runs,
Maddux was 4-10 and the team was 4-11; when the team scored 4 or more
runs Maddux was 7-3 and the team was 9-3. The Braves won every game in
which Maddux held the opposition to 0 or 1 run, and in games when maddux
allowed 4 or more runs the team and Maddux went 1-7. But here's a very
striking result - in games in which Maddux allowed 2 runs, the Braves
went 5-6 (Maddux went 5-5). In those games he averaged 7 IP/start with
an ERA of 2.08 and an RA of 2.54, and the team was below .500 in those
games. There's not a big sample to work from here, but we see that the
offense only once bailed out Maddux when he failed to have a quality
start, and the offense squandered a number of games in which he pitched
very well.
I've often felt as if teams did better in the postseason against Maddux
in the regular season because they got to face him twice in close
proximity; essentially, their memory for his pitching caught up with his
memory for their hitting. If that's the case, Maddux should fare more
poorly in his second start against a team within a series (he never
started three games in any 7 game series).
In first starts of postseason series:
--9 wins, 8 losses, 2.22 ERA, 3.54 RA
In second starts:
--2 wins, 5 losses, 4.33 ERA, 4.85 RA
So, that pattern is borne out here; in terms of rate stats, he allowed
only 7.98 H/9IP in first starts vs. 9.69 in second starts. His walk rate
was lower and K rate higher in second starts, but his HR rate was higher
in second starts (0.87 vs. 0.49). Notice that huge RA difference in
first starts, though - 1.31 runs/9IP. 19 of Maddux's 22 unearned runs
allowed in the postseason came in his first start of the series, so even
with his stellar 2.22 ERA he barely broke .500 in those games (the team
went 11-8 overall). But, as noted above, those unearned runs were
concentrated in a handful of starts.
There is, of course, the catcher issue. Unfortunately, there's not a
whole lot to work with here in terms of comparative data, because he had
so few postseason starts with Lopez behind the plate. Lopez only started
behind the plate four times out of Maddux's 27 postseason starts; on two
other occasions Lopez was brought in to catch after Maddux had had a
particularly bad first inning. Here's how the numbers look overall:
With Lopez behind the plate (only 36 IP:
--2 wins, 2 losses, 1.69 ERA, 2.17 RA, 7.96 H/9, 1.45 BB/9, 6.05 K/9,
0.24 HR/9
With other catchers behind the plate (144 IP):
--9 wins, 11 losses, 3.12 ERA, 4.36 RA, 8.60 H/9, 2.24 BB/9, 4.58 K/9
0.69 HR/9
Again, small sample caveats apply above. But what about hitting, how
much did the backup catchers weigh down the offense behind Maddux?
Lopez in Maddux games:
--4 hits and 1 BB in 15 PA (286/333/357), 1 R, 0 RBI. 1 double, no HR.
All 4 hits came in the games he started, he went 0-8 in the two games
where he was brought in after rough first innings. The team did not come
back to win either of those games.
Other catchers in Maddux games:
--14 hits and 4 BB in 66 PA (.226/.272), 4 R, 11 RBI. Don't know SLG,
but they did manage 4 HR, so I'm guesstimating that SLG at somewhere in
the mid-high 400 range.
Lopez catching other pitchers in the postseason
--277/324/514 - 10 HR and 28 RBI in 205 AB
So, it looks like Javy never hit much when he was catching Maddux, but
he hit to his career numbers with other pitchers. Those other catchers
mostly sucked, but 4 HR in only 62 AB is good rate, it prorates to a 30
HR season. Overall the offense averaged only 2.5 runs in Javy's starts
vs. 4.36 runs in starts with the other catchers. Again, small sample,
but interesting. Given that Lopez hit very well for the other pitchers,
one might expect that over a larger sample with maddux he'd have done
better; but, though the backup catchers were weak on OBP, they did okay
on slugging and the overall gap isn't as huge as one might think I'd
guess less than 100 points of OPS separates their performance from how
Javy did with everyone else. All that said, there's no real compelling
evidence that the backup catchers helped Maddux pitch better than he
would have with Lopez behind the plate.
That's all I have so far. Not quite sure what to make of it all; I am,
after all, trying to account for a relatively small difference between
expected results and actual results. But it seems like Maddux had bad
luck on offense, pitching well often on days when the other guy pitched
stepped way up or the offense just sucked; and, he got victimized by his
defense, and did not do a good job of damage control. The end result is
that on rate stats postseason Maddux was much better than his record.
Colin
Tom
five unearned runs? The first was I believe the opening game of the
NLCS against the Marlins, with all the unearned runs coming in the
first inning on a weak ground ball that got by McGriff, and then
another play I can't remember. This certainly would have gone
differently in that series if he had had defensive support, and I
don't think this was a situation were damage control was an issue. The
other, a dropped fly ball by Andrew against the Cardinals? A meltdown
against the Diamondbacks? I don't know . . .
It's easy to see things could look very differently if a couple of
these things hadn't happened like that.