You missed the best part:

Overall, he was 7-1 in the regular season and 2-0 in the 1990 American League Championship Series. Clemens' record against Stewart was 1-8, his only win coming in their first showdown, in 1984 when Stewart pitched for Texas.

Nice reference to the Costanza method of creating yourself an alternate truth.

Stewart must read a limited amount of baseball sites if he thinks nobody brought Roger up before Mitchell. Though he's right that Clemens did not get the attention of Bonds.

This looks spot on to me:

“Now we see Roger actually buying what he’s saying. He has dealt with an unrealistic situation so long and has a hard time thinking he really did anything wrong. He could probably pass a lie-detector test.”

Stewart apparently doesn't realize that Adam Everett is the reason for Clemens success, not drugs.

Besides the BALCO paper trail, Stewart also neglects the fact that Roger wasn't counting down to breaking any all-time records, the way Barry was. Sadly, most people still see PED's as a route only to more power - hence the spotlight on the hitters.
None of which is to say that there aren't some black/white issues at play also.

Overall, he was 7-1 in the regular season and 2-0 in the 1990 American League Championship Series. Clemens' record against Stewart was 1-8, his only win coming in their first showdown, in 1984 when Stewart pitched for Texas.

Yeah, and Stewart's Oakland Athletics did it drug-free!

Stewart resented Clemens getting more respect in the late 80s? Sure, cause you media types kept printing silly things that helped him believe, as he said then, that he (not the Rocket) was the best pitcher in the game, when it was obvious to anyone with half a brain that the Oakland 600-ft-to-fences-and-out-of-play-areas stadium, a good defense, and a great offense helped Stewart win a ton of games. Dave Stewart was a workhorse and good pitcher for a few years, tremendously helped by cirucmstances, who got way too much credit for it (check out the Cy Young voting in 1990 and tell me how he got even that many votes), and then bragged about it. And now apparently it still han't gone away. As for using 9 games to measure two pitchers; gee, Barry Bonds only slugged .262 against Tim Belcher in 42 ABs; obviously Belcher was better.

check out the Cy Young voting in 1990 and tell me how he got even that many votes

By being the best pitcher in the league that year?

He is so correct. This is why McGwire sailed into the HOF on his first ballot.

And now apparently it still han't gone away. As for using 9 games to measure two pitchers; gee, Barry Bonds only slugged .262 against Tim Belcher in 42 ABs; obviously Belcher was better.

It's not really the same. Stewart and Clemens probably faced 300-400 batters each in their head to head matchups. Obviously Clemens was a far better pitcher than Stewart against everyone else, but Stewart's personal domination of Clemens and his team is still impressive.

Roger was Dave Stewart's beeyatch.

It's almost like the way Bill Russell dominated Wilt Chamberlain in playoff game 7's, despite Wilt being the statistically dominant one.

It's almost like the way Bill Russell dominated Wilt Chamberlain in playoff game 7's, despite Wilt being the statistically dominant one.

It was uncanny. Clemens was so obviously better, but Stew got so pumped for their meetings with Clemens that he elevated the intensity of the entire team.

I'd rather Stew not dogpile on Clemens now, though. There's no gain in it.

You picked a fine time to leave me, Lucille.

Stewart never liked Clemens, and is probably enjoying piling on now.

Yeah, they never liked each other, I think. All the more reason to refrain from the dogpile. Nothing stings an enemy more than when you take the part of the better man.

It's not really the same. Stewart and Clemens probably faced 300-400 batters each in their head to head matchups.

But Roger was facing the A's lineup, and Stewart was facing the Red Sox lineup.

Look at 1990, for example

C: Steinbech > Pena
1b: McGwire >>>> Quintana
2b: Randolph < Jody Reed
ss: Weiss > L. Rivera (who?)
3b: Lansford << Boggs
OF: McGee/Jose < Brunansky
OF: Rickey >>> Burks
OF: Henderson = Greenwell
DH: Canseco >>>> Dwight Evans

The way I see it, the Red Sox had a big advantage at 3b, and the A's had huge advantages at 1b, 1 OF spot and DH. And I mean huge advantages.

Rickey Henderson had a 188 OPS+ in 1990.

It was uncanny. Clemens was so obviously better, but Stew got so pumped for their meetings with Clemens that he elevated the intensity of the entire team.

The way he was able to rise to the occasion and pump pitch after pitch right by Luis Rivera and Carlos Quintana was remarkable.

I was living in the Bay area aroudn the time of the majority of those Stewart-Clemens matchups, and it was uncanny. Stewart had this big-game-tough thing going on, and it was one of those things that causes people to discount statistical evidence in favor of what they see with their eyes.

As for the rest of this article, well, you know, blind squirrels and what-not. Dave Stewart sees racism in a lot of places I don't, but I think in this case he's got a leg to stand on.

The way he was able to rise to the occasion and pump pitch after pitch right by Luis Rivera and Carlos Quintana was remarkable.

Yeah, those late 80's to 1990 Red Sox teams were just awful. Those were truly dark days in Boston.

Seriously though -- why wasn't Clemens suspended for throwing that bat? As I recall, he wasn't even removed from the game. I found myself completely nonplussed at the time and remain so when I reconsider those events. Steroids or not, how can throwing a bat directly at a player not result in any kind of punishment? I'll say this much: if it had been Jose Offerman, things would have played out very differently.

RDF, sj.

I've been putting off asking this, but it's finally getting to me.

What does RDF stand for?

What does IIRC stand for?

Do have my internet taken away now?

It was uncanny. Clemens was so obviously better, but Stew got so pumped for their meetings with Clemens that he elevated the intensity of the entire team.

And Clemens was the opposite, throwing a fit in the ALCS so he wouldn't have to continue taking a beating.

DAS RED SOX NICHT ES VERGUT

Between 1988 and 1990, the Red Sox scored more runs than the A's. The A's were primarily a pitching/defense team, even if the Bash Brothers got all the attention.

Parks didn't have anything to do with that? Position by position, the As were a ton better on offense.

Stewart pitched great in some great games, but baseball games do not happen in a vacuum. It was not Stewart vs. Clemens, it was Stewart vs. the Sox and Clemens vs. the A's.

And Clemens was the opposite, throwing a fit in the ALCS so he wouldn't have to continue taking a beating.

One of the dumbest post-season meltdowns I've seen until the bat throwing, which luckily for him, didn't cost the Yanks. Don't get me wrong, though, Clemens is great and I've enjoyed watching him all these years. The fact he was so great made those matchups with Stew so fun and I appreciate him for that alone. I don't want to make it seem like I have all this Clemens hate. Even now, he provides me more entertainment than I deserve.

DAS RED SOX NICHT ES VERGUT

With a name like Belizean Thug I feel you should be pushing the nanny limits in espanol.

Something along these lines would do nicely...

Hijo de Puta! Chi Chis! Maricon! Culo!

Parks didn't have anything to do with that? Position by position, the As were a ton better on offense.

Position by position comparisons are fun, but mostly worthless.

OPS+

check out the Cy Young voting in 1990 and tell me how he got even that many votes

Dave Stewart got a lot of Cy Young votes he shouldn't have, but it's strange that you pick 1990 as the year to illustrate this. He was a lot better than the actual winner that year.

Dave Stewart got a lot of Cy Young votes he shouldn't have, but it's strange that you pick 1990 as the year to illustrate this. He was a lot better than the actual winner that year.

That was an interesting Cy Young vote. Clemens was the best pitcher, but Stewart and Chuck Finley had great years, too. I can see somebody voting for Stewart because of the extra innings he threw. Welch had a good year--not great-- but those 27 wins were an overwhelming number. 27-3 is hard to rationally overcome. It just pops out at you and I don't blame the writers too much for giving in to it. I wonder if we'll ever see a 27 game winner again?

Welch obviously didn't deserve it, but he was pretty damn good in 1990. He was a better choice than Colon in 2005.

How good was that 1990 A's bullpen? They had a 2.35 ERA in over 400 IP. Eck led the way, but everyone was very good.

According to this piece by Rany Jazayerli from back in the day, the 1988-90 A's had the best bullpen ever, and so good that after "tak[ing] Eckersley out of the equation completely, the A's still have a 2.49 ERA, better than any other team on the list."

Clemens was actually the runaway favorite most of the year but he got hurt and missed the end of the season, taking him out of consideration for most voters.

Eckersley should have won it, if you take Clemens out.

Eckersley's pitching won the award for Bob Welch.

Position by position comparisons are fun, but mostly worthless.

OPS+
Year OAK BOS
1988 109 113
1989 104 107
1990 107 103

Would you rather have 9 guys with a 113 OPS+ or 3 superstars, a couple guys around 110, and some crappy players.

SG ran a few sims studying team competition, I am not sure if he ever published it. It might be on the RYLW blog.

I always loved Stewart, one of my favorites, i think i said that on here before too. And he is spot on with a number of his statements concerning the media's coverage here. I have always felt that way.

I am kind of surprised Dave Stewart is not in anybodies front office anymore though. That was short lived, huh? ...

one and done? No 'old boys network' there ... :)

I've defended Clemens on the bat throwing issue because I didn't feel it was his intent to hit Piazza (he wouldn't have been looking along the ground and skimmed it along the ground if that was his intent). But regardless of what you think his intentions were, the act was at the very least ridiculously reckless and deserved an immediate ejection. I wouldn't have disagreed with the idea of a suspension either.

Danny's numbers are interesting. I was always under the impression that Clemens was facing much tougher lineups than Stewart was in those matchups, but the numbers don't bear that out apparently. Was it possible that by the postseason the lineups were more tilted toward the A's?

“Now we see Roger actually buying what he’s saying. He has dealt with an unrealistic situation so long and has a hard time thinking he really did anything wrong. He could probably pass a lie-detector test.”

This whole thing is funny in a couple ways. First, the headline could read "Dave Stewart sees [pick your own word] in black and white" and it would be just as accurate. And the quote above is amusing because if any player every believed his own hype, it was Dave Stewart. The guy has spent decades saying how he outdueled Clemens and it actually appears that he believes the big-game hype enough to convince himself that he's a better pitcher than Clemens.

And the lie-detector thing is beautiful. It's the perfect next step in the discussion of Clemens's possible steroid use: it takes something that people have been calling for Clemens to do and disqualifies it as a way to clear his name. This neatly follows the pattern that was established with "Clemens needs to deny this publicly!" and "If he was really innocent, he'd sue him!" and "I want to see Clemens get up there and testify under oath that he didn't do it!" Once Clemens does these things, and even before, he doesn't get any credit from those who wanted him to do it.

Hey, Dave Stewart scuffed the ball and used steroids starting in 1987. It's obvious from his stat line--no 30-year-old suddenly becomes a top starter after being a mediocre reliever for so long. He was on the A's for crying out loud! Of course, he'll probably say he didn't, but he's so far gone that he probably actually believes that.

Lastly, does anyone remember Dave Stewart saying such nasty stuff about Bonds? I assume he must have because he claims to be able to look beyond race in such issues and would surely have been just as harsh on Bonds as he is being on Clemens.

Ah, the polygraph, everybody's favorite pseudoscience.

I'm a phrenology man myself.

Give me 5 minutes with Clemens and McNamee's skulls and I could have this whole thing wrapped up by lunchtime.

Give me 5 minutes with Clemens and McNamee's skulls and I could have this whole thing wrapped up by lunchtime.

That's the same thing a guy with a claw hammer would say.

did someone actually claim above that Stewart was the best pitcher in the AL in 1990? With Clemens and Eck (sub 1 ERA) and Thigpen (MLB saves record), never mind Welch, in the league? You're kidding, right??

did someone actually claim above that Stewart was the best pitcher in the AL in 1990? With Clemens and Eck (sub 1 ERA) and Thigpen (MLB saves record), never mind Welch, in the league? You're kidding, right??

Stewart: 267.0 IP, 2.83 RA, 71 VORP
Clemens: 228.3 IP, 2.33 RA, 77 VORP

Clemens was great, but, as Baudib noted in #41, he missed most of September. I can certainly see a case for Eckersley, who I probably would have voted for, but not really for Thigpen or Welch.

Stewart appears to have benefitted from defense quite a bit. He had 166 Ks, 83 BBs, and 16 HRs allowed. Clemens was 209/54/7(!). I don't feel like doing the math, but Clemens's DIPS has got to be a ton better than Stewart's.


Parks also come into play, which Clemens's 213 ERA+ to Stewart's 145 ERA+ illustrates. And WARP3 has Clemens at 12.7 and Stewart at 8.9. Not sure why that's so different from VORP (does it account for park?).

I don't see a good case that Stewart was even in close proximity to Clemens.

The A's allowed a ridiculously low .258 BABIP in 1990. Stewart was at .260, while Clemens was at .292. I think Oakland pitchers get double "penalized" for low BABIP, though. Oakland is a pitchers park solely because it suppresses BABIP. So once one has adjusted for park, it doesn't make sense to further ding the pitcher for having allowed a low BABIP.

Anyway, the VORP comparison is closer than IP/ERA+ would indicate because 17% of the runs Clemens allowed were unearned, compared to 9.5% of Stewarts' runs. I have no idea why WARP differs so much from VORP for pitchers (aside from including fielding and batting, which doesn't seem to matter in this comparison), so I've always just used VORP.

I think Clemens was better than Stewart, but missing those starts in September made it much closer than it otherwise would have been. I can see how voters would ding Clemens for missing a few September starts in the middle of a close Division race. Stewart was closer to Clemens in VORP than he was to the third place finisher.

The VORP-WARP disconnect seems to stem from Davenport's view of defensive replacement level. It appears to "infect" the pitching side of things because he assumes replacement-level defensive support, and the elevated rate of hits on balls in play amplifies the importance of strikeouts. Strikeout pitchers get overrated relative to non-strikeout pitchers.

Isn't DIPS better as a predictive tool than an evaluative tool?

Better predictive tools are better evaluative tools. If a skill is not repeatable, how do we know it's a skill and not just good fortune?

Anyway, the VORP comparison is closer than IP/ERA+ would indicate because 17% of the runs Clemens allowed were unearned, compared to 9.5% of Stewarts' runs. I have no idea why WARP differs so much from VORP for pitchers (aside from including fielding and batting, which doesn't seem to matter in this comparison), so I've always just used VORP.

I wasn't too clear, but that was sort of my point. Earned runs are not perfect, but their better than raw runs. I'd give more credence to ERA-based stats. It's okay to use VORP generally, but in this case, it's the only stat that has Stewart within shouting distance of Clemens. Everything else pointed to Clemens being waaay ahead.

I think Clemens was better than Stewart, but missing those starts in September made it much closer than it otherwise would have been. I can see how voters would ding Clemens for missing a few September starts in the middle of a close Division race.

I could see them doing that too but a) even after a dinging he's still way ahead of Welch, the guy who they chose instead (who still only threw 10 more IP) and b) I don't think they dinged him so much as said "Holy crap! 27 winzzz!"

Better predictive tools are better evaluative tools. If a skill is not repeatable, how do we know it's a skill and not just good fortune?

Because sometimes we are interested in what actually happened, not what was likely to have happened.

DIPS basically does two things:

1) Normalizes BABIP
2) Normalizes context.

If you're using it as a value stat, think (1) is a crude way to adjust for quality of defense. But I think a pitcher should get credit if he steps up with runners on, which (2) takes away from him.

I wasn't too clear, but that was sort of my point. Earned runs are not perfect, but their better than raw runs. I'd give more credence to ERA-based stats. It's okay to use VORP generally, but in this case, it's the only stat that has Stewart within shouting distance of Clemens. Everything else pointed to Clemens being waaay ahead.

I generally prefer RA to ERA. There is skill in reducing errors for pitchers (FB pitchers allow fewer unearned runs because FBs are less likely to result in errors than GBs).

And, again, I agree that Clemens was clearly better than Stewart. But I'd say the extra IP, especially in September, put it within shouting distance.

Because sometimes we are interested in what actually happened, not what was likely to have happened.

What actually happened is obvious, though. Clemens (or Stewart or Welch or whoever) combined with his defense/park to allow a certain number of hits/runs/balls/Ks/etc in a certain number of IP. It's how you dole out the credit for what actually happened that we're discussing here.

We don't credit the concessions sales person with preventing runs and winning games because the way they perform at their job does not have a predictive effect on the outcomes. So even though the A's were 41-10 in games Dave the hotdog guy worked the rightfield bleachers, we don't give him the MVP because we can easily see that it's a coincidence (or is it....).

There's nothing theoretical about DIPS--every stat that it involves are things that have actually happened. They're also things that have been shown to be the most consistently repeatable skills for pitchers. Thus, they are the things that we should be most certain to credit to the pitchers. (Not that it's perfect by any means.)

But I think a pitcher should get credit if he steps up with runners on, which (2) takes away from him.

This is another case where I would ask if you're confident that this is a repeatable skill. It would certainly make sense that some guys would be better at working out of the stretch than others are, but I haven't seen the data on it--has anyone done a study of it?

This is another case where I would ask if you're confident that this is a repeatable skill. It would certainly make sense that some guys would be better at working out of the stretch than others are, but I haven't seen the data on it--has anyone done a study of it?

It does appear to be a skill some pitchers have.

But, more importantly, we give batters credit all the time for unrepeatable events. Whether it's a flyball being lost in the sun, a strange bounce off an OF wall, or just a lucky amount of seeing eye singles. I don't see why it should be different for pitchers.