And years from now, when we’re debating who deserves the Roy Oswalt Award every winter, we’ll look back fondly on this post and know that we were there when gonzo-metrics got its start.
With a ball-peen hammer.
Yookeroo–Did I write “Maddux should want to complete games”? If so, the presumption that pitching 25 games and 200 innings is exactly as good as pitching 35 games and 300 innings is just wacky, if the pitching is at the same level of quality. How few innings would Maddux have to have pitched for you to concede that “Yeah, pitching more innings and starting more games does your team some good”? It’s sorta the definition of “effective,” isn’t it? This looks like hand-waving to me.
Working the umpires definitely is a skill. It’s just a far less impressive skill than blowing people all away is. I’m not interested in coming out to the ballpark to watch the umpire thinking well of some people, and less well of other people. I want to see Willie Mays, Hank Aaron, Roberto Clemente et al swinging and missing at pitches they can’t see rather than seeing batters taking called strikes six inches off the plate and cursing the umpires on their way back to the dugout. Maybe that’s just me.
This has got to be one of the most misleading stats ever, BTW, so much so that I’d almost call it a deliberate lie. Maddux twice had full season ERAs under 1.65? That’s what we call setting self-serving retrospective cut-off points, where you find someone’s level and set the cut offs for that level exactly where he falls on the list. IOW, this stat diminishes Guidry, Koufax and every other killer ERA-pitcher because they were a hair above 1.65 and Maddux was a hair below. It’s misusing stats to argue for victory, instead of looking at what the stats tell you.
But the almost-lie come in the modifier “full-season”–Maddux’ two best seasons, ERA-wise anyway, came in strike-shortened years. I haven’t raised the point explicitly yet, but no one else is coming forward to point out that maybe it’s not quite as impressive to pitch for part of a season as it is to pitch a complete 162-game season (or in Koufax’s 1962, a full 165-game season). To include that in "full-season"s may be technically true but 1994 and 1995 weren’t full seasons in any reasonable sense of that term. They were strike-shortened full seasons, and while the strike wasn’t Maddux’s fault, he shouldn’t get the credits his fans demand for it, either, when they compare it to non-strike-shortened seasons. It’s tainted.
I haven’t seen anyone’s response to my earlier point about Win Shares and the purpose of them, to be able to compare performances across all sorts of lines including time lines. If you 'd like argue that they’re invalid, can you show me how? I hope you’re not thinking that I’ve cherry picked hundreds of ways to measure performance and selected this one way because it favors Koufax. In fact, Win Shares hurts Koufax in one material and obvious way: it doesn’t account for post-season work, which would puff up Koufax Win Shares in his peak seasons. But WIn SHares excludes the post-season.
Post-season is important, though, in one way I haven’t touched on yet. Koufax didn’t pitch superbly in the World Series after carefully resting his arm, and spending time on the DL, and coming out of games all season long. No, he put the maximum strain on his arm in the regular year, taking his turn up to over 40 times per season, finishing up to 27 games, and then he pitched a murderous number of innings in the World Series on ridiculous short rest and threw shutout ball for the most part against some of the best lineups in baseball history (well, not the 1959 White Sox, granted). This may have been dumb, short-sighted, even self-destructive, but I think he deserves credit for this near-heroic choice he made, not to be equated with a year where Maddux prudently took care of his arm, had less impressive stats, had far less impressive results (as far as World’s Championships go) and was able to extend his career by declining to take on the challenges that Koufax assumed. That was pitching, and if you missed it, you missed seeing something amazing.
Missed the edit window. This goes at the end of this paragraph:
Koufax gets zero credit for his post-season pitching in WIn Shares, yet that’s the standard I chose. If you want to see something that will positively blow you away, look at his World Series record sometime. I’d start with the 1965 Series against the Twins–it will just amaze you. You’ll think you’re looking at a typographical error–no modern pitcher would even consider pitching on that schedule, for that long, with such excellence. Then consider that he did that after such a grueling regular season’s workload.
Here’s a link to the 1965 World Series, probably the most impressive athletic performance I’ve ever seen:
Forgetting for just a second about the two final shutouts on short rest (Game Seven coming on two days’ rest), his first start in this series was a loss in which he gave up one earned run in six innings, which considered alone would be a Series-MVP nominator. Will you ever possibly see a similar performance? No current manager would even consider giving a player this sort of chance to succeed, much less the unlikeliness that a player could come through like Koufax did.
If you want to argue that we shouldn’t give Koufax any credit at all for how he MIGHT have pitched beyond 1966 if only he’d been healthy, which I agree witih, then how can you credit Maddux for pitching well over the parts of the 1994 and ‘95 seasons that never got played? I would think you’d have to see those as diminished seasons, though no fault of Maddux’, exactly as you’d have to view Koufax’ career as diminished, through no fault of his own. You don’t take away points for how someone MIGHT have done if given the chance, but you certainly don’t give him credit for earning those points, either.
Koufax won 25, 26 and 27. There are a lot of 20 game winners. Few get to 25 and 27 is in rare territory. It is the difference between a 5 handicapper and a scratch shooter in golf. You have to get there to see the difference . Koufax was a better pitcher. Maddux was a very solid pitcher for a long time.
I already discussed how the cross-era comparisons can be problematic, in that differing offensive levels can affect how many innings/pitches/batters a pitcher could safely throw before he’s tired and/or at risk for an injury. There’s no effin’ way Koufax could throw 335 innings in a league which is scoring 5+ runs a game-no way to do it without blowing his arm out that is. For one thing innings in 1994 likely required more pitches than they did in 1966, and Maddux actually could be said to have an advantage in that his pitches/batter were unusually low. Nobody’s tinkered with using a four-man rotation since the late seventies, so we haven’t had the privilege of seeing if said rotation could cope with very high offensive levels or not. That’s probably the main reason most teams switched tho.
As I said it is more a matter of taste than anything else, but your way isn’t the only valid way of looking at the issue. In the context of his time Maddux was a very durable pitcher at his peak, as durable as Koufax was in his time. In addition to that, his ERA’s, again as compared to his peers, were more impressive. What would you rather have, a 1.74 ERA in a league (and in a tough ballpark to score runs in) which averages 4.01 runs per game, or a 1.56 ERA in a league with 4.63 runs per game (more or less neutral park)? And he maintained that level of performance over a two year stretch, 400 innings worth, which makes it less of a fluke. I have no problem giving him credit for games missed due to the strike, in which case even his Win Shares would end up looking pretty good, considering his two best seasons were 94-95. James was dangerously inconsistent on this point, giving the likes of Ted Williams credit for missing entire seasons during World War II, but giving players in strike-shortened seasons no extra credit at all (something about spilled milk, as irksome a comment as I’ve ever seen from the man, and I’m otherwise a big fan of his work).
The main question I ask when looking at how good a pitcher is, is as follows: how much did he stand out above his peers, both in terms of opportunities (innings/batters faced) as well as performance (ERA mainly*, tho not limited to that)? Maddux beats Koufax pretty handily when looked at in this way (as does Clemens, and perhaps another killer lefty in the Big Unit). Koufax has a better argument against Pedro Martinez, despite the latter’s insanely low ERAs for his time and ballpark, because Pedro often had trouble staying in the rotation over an entire season, missing a few starts here and there.
*[As I said in that other thread, ERA I don’t believe scales down or up linearly when you are trying to match performance with talent across eras.]
All cutoff points are arbitrary. But yeah, let’s look at what that ERA stat tells us-Maddux’s ERA for '94-95, (53 Starts) was 37.6% of the league average. Gooden’s phenomenal '85 netted him an ERA of 43.8 % the league average. Koufax’s Best showing was in '66 with 52.7% of the league average, his second best was '64 (when he started only 28 games) at 53.5%, and his third was '63 when at 62.9%.
**Maddux’s ERA for the 1990’s was about 61.5% of the league average, better than all but one of Koufax’s full seasons.
**
Dude, trust me, I’m with you on the declining role of starting pitchers. I want them to be real men and take the mound everyday, and I’ve got the torn rotator cuff to prove it. But you cannot hold the transformation of a whole game against one of the many men who has benefited from it.
Yes, Maddux pitched less than Koufax et. al. But did he pitch less than his peers? I seriously doubt. The best you can do is lament that Koufax wasn’t afforded the same opportunity, and argue the phantom numbers he might have put up.
When someone puts up a solid decade of numbers that compare favorably to the best 3 seasons of one of the brightest shooting stars to ever play any sport, I’d call that pitching to.
Well, Maddux pitched 5008 innings. In the entire history of baseball only 12 other men have pitched more, and five of them pitched at least half their career in the dead ball era.
Dude, you’re going awfully deep into the looking glass if the only criticism of a guy you can muster is a half-assed rumor that he didn’t pitch enough innings when he was in fact one of the most prolific pitchers in the history of the sport. Might I suggest that this Maddux argument here has gone hopelessly off the rails; if people are seriously questioning his greatness they need their heads checked.
I’ll need a cite here. James gives WIlliams a lot of verbal credit, but not a single Win Share does he award Williams for missing seasons. I think you’re being remarkably inconsistent in granting Maddux phantom Win Shares for the games he would have pitched in 1994 and 1995 but giving Koufax no extra Win Shares for the games he would have pitched if he hadn’t gotten injured. I think all three cases (Williams, Maddux, Koufax) should be decided on the basis of the games they actually played, and you can award moral or virtual credit all around–all three guys deserve credit for coping with various issues–but I think crediting anyone for games they might have played but didn’t play is just an avenue to subjective arguing.
There’s actually a pretty legitimate argument to be made here, which is that Maddux’s Win Shares in 1994 and 1995 represent a more significant contribution to his team than they appear, simply because the season was shorter.
Take, to use another player as an example, Mike Schmidt in 1981. Schmidt, according to Win Shares, accumulated 30 WS in that season. That sounds pretty good on its own, but he did it in a season only two thirds as long as normal. The season wasn’t just two thirds the normal length for Mike Schmidt, it was two thirds the entire length for THE ENTIRE LEAGUE. In the context of how much it helped the Phillies that season, Schmidt’s 30-share season had just as much of an impact as a 45-share season would in a normal schedule. He was ridiculously far ahead of all other players by any metric. 30 win shares is much more impressive when there’s only two thirds as many of them up for grabs. Schmidt’s 30-share season propelled his team to a winning season more than any of the (many) years he had more than 30 shares.
By comparison, when Sandy Koufax did not get any win shares in 1967, the Dodgers still played 162 games. Koufax getting 0 win shares is not an illusion caused by the league shutting down, or the season being shorter than normal, like it was in 1981 or 1995; it was the result of Sandy Koufax being unable to pitch.
We can overdo the Maddux bit in 1995 because he didn’t miss THAT much time - the league only lost one-ninth of its schedule. But it certainly has a big impact in 1994. Of course, there’s the philosophical question of whether a season with no championship matters, but Maddux had lots of other great years.
Like you know for a fact that Maddux wouldn’t have gotten bombed out of every start he made during the games he never got to play in 1994 and -5. You can’t do that. That is why they actually play the games, to get actual results. You cannot credit someone with a performance based on games he woulda-shoulda-coulda played in.
People want to do it all the time, but as I say, it’s pure speculation and opens the door to all sorts of subjective nonsense based on one’s own wishes and fantasies.
The reason is you’re counting only the positives, but negatives happen all the time. Schmidt might have suffered a career-ending injury in one of those unplayed games, and you’d be saying “Mike who?” now. But that doesn’t figure into your fantasy, only the good stuff.
To take a less drastic example, say Schmidt would have played those games but simply gotten injured for the year–in the remaining 1/3 of a season, George Foster could have played at his actual WS rate and wound up with 36 WS, or Dawson with even more. As Walking Underwear famously noted, baseball can be summed in one word: YouNeverKnow.
This is truly bizarre.
You’re right that Schmidt might have had a career-ending injury in one of those games. But your own argument here actually supports RickJay’s position.
Because we cannot know what might have happened if the rest of the season had been played, all we can do is evaluate the players’ performance in the games that were actually played. And, as RickJay notes, Schmidt’s 30 WS in 1981 are even better than they initially appear precisely because he accumulated them not in a full season (which would have been incredibly good anyway), but in two-thirds of a season.
The Phillies won 59 games in 1981, meaning that there are 177 WS to be distributed between all Phillies players for that season. So, if RickJay is correct and Schmidt’s WS for 1981 is 30, that means he accounts, by himself, for better than one-sixth of all his team’s WS.
And, as RickJay suggests, there’s an important difference between missing a games that is played, and missing a game that is not played. If the league shuts down, a player has no opportunity to play, even if he’s healthy; if a player is injured for a whole season, but the team still plays, that’s quite a different thing. The player is missing games not due to something external (i.e., no games being played), but due to something directly related to him as an individual (his injury).
I’m not claiming that Schmidt had anything less than a spectacular 2/3rds of a year. But there’s no reason to credit him with having anything more than that. 30 WS (which is the correct figure) IS a spectacular performance for 108 games, but it’s nothing particularly distinguished for a full season–it’s James’ minimum MVP-candidate level for a full season, of which there are usually at least two or three, sometimes many more.
Schmidt’s 30 WS wouldn’t have been incredibly good by Schmidt’s own standards btw: the year before and the year after, he got 37 WS.
No matter why a game isn’t played by a player, all you can do is note it–off to war, arrested for exposing his weenie, colon cancer, dudn’t matter. He cannot get any credit, other than moral credit, for what he did in a game he did not play in. I don’t see what you want me to do other than credit Schmidt with a league-leading 30 WS for 1981, which I’m glad to do.
Other problem here is: where does it end?
if you’re going to make a shortened season’s excellent performance (as in Maddux’ or Schmidt’s strike-shortened years), then how do you deal with a season shortened off the charts? Say the season ends, for whatever reason in mid-May 40 games in, and someone has hit 20 HRs already, and accrued 20 Win Shares–is this a great season, or just a hot start? If you’re not prepared to say this is the bestest season ever, then where are you willing to draw that line? Seems to me you’d have to think of this as an 80 Win Shares year, a new record, or just credit him with the 20 and say “Too bad. Bummer.” I’d easily choose to do the latter. How about you?
How about if the season lasted just a week, and someone went nuts in that week? Do you give him a pass on the 155 games that never got played?
I honestly don’t understand why you’re reacting so angrily, and your counterargument doesn’t address anything I said. Please, I’m honestly trying to make a point here, not hurt your feelings or win some argument.
Now, please understand: **I’m not trying to imagine what Schmidt would have done in 162 games instead of 107. ** I have no interest in guessing how he would have played. What I’m pointing out is that the 1981 season was only 109 games long, but it was still a complete season that counts as much as any other. It matters not a whit what Schmidt would have done in the other 53 games. There were no other 53 games.
What matters is what Schmidt DID in an entire season. The 1981 season was a full blown major league season that anointed a World Champion and various other lesser titles and all that jazz. Schmidt’s 30-Win-Share season was not just the best year any player had that year, it was the best year by a wide margin. Win Shares are a direct function of the length of the league schedule. The league had only 2/3rds as many Win Shares to dole out, so a season of enormous value (like Schmidt’s) will only account for 2/3rds as many Win Shares. Schmidt’s 30 Win Shares had as much impact on a 109-game season as about 45 Win Shares would have on a 162-game season, because there are fewer Win Shares, and each one means more.
The issue here is not what Schmidt MIGHT have done. I don’t care. What matters is what he did do and what it meant to his team. 30 Win Shares means, basically, ten wins. Ten wins obviously means more in a season of 109 games than it does in a season of 162 games, doesn’t it? It’s a much larger portion of the schedule. Win Shares (or any other measure) don’t have any meaning apart from how they tell you how much the players helped thier teams. Mike Schmidt’s 30 Win Shares in 1981 helped his team more than 30 Win Shares would have in a 162-game season.
The same logic applies to Maddux’s 1994 and 1995 seasons. Now, you could make the argument that 1994 is just as good as nothing because the season was cancelled and effectively meant nothing at all, like exhibition games. But that’s another argument.
Not angry at all, RickJay. Sorry that I’m coming off that way.
What I’m trying to stress here is we have to go with what IS.
It’s a shame, for Greg Maddux, that baseball adopted the five-man rotation and began using relief pitchers so much that fewer CGs and 30O+ IP seasons were no longer possible. Instead of pitchers being able to put up 30 WS numbers like Koufax and Gibson and Marichal all did at their peaks, Maddux in an excellent season topped out in the high 20s, as the other Win Shares went increasingly to relievers. But that’s what happened, and it probably allowed pitchers of his generation to have longer, healthier careers, which was the goal of reduced pitching loads.
But Maddux can’t claim to have contributed as much to his team’s success as the pitchers who “benefitted” from the heavier workloads, and I think that’s what I’m hearing. If we went to a ten man rotation, with leagues routinely being led with 16 or 17 starts, would you say that the guy who led that league in starts year after was as much a workhorse as Koufax was, because he led his own contemporaries more frequently?
I wouldn’t. I’d say the guy didn’t get as many chances to excel, and his performance is diminished for that.
Likewise, if Maddux has a season where he starts 25 games, there;s got to be a limit on how much credit he gets for that performance, and it’s got to be less credit than someone gets for the same performance over 50% more games. I don;t see a way around that, and neither do Win Shares.