Honestly, by comparing that list to the top pitchers of the 70s and 80s, I’m inclined to call the 90s a Golden Age of Pitching.

Whatever. None of those 90s pansies could match the grittiness and guttiness of Jack Morris.

yea, a method that reduces pedros overated dominance...heck any ##### can pitch 7 innings a game......

90s pitchers have an incredible built-in advantage over 70s pitchers in an analysis like this.

yea, a method that reduces pedros overated dominance...heck any ##### can pitch 7 innings a game......

Pedro looks worse in this comparison to the other great pitchers (Johnson, Clemens, Maddux, etc.) of his era because he pitched almost enitrely in the high offensive era of the 1990's. That said, Pedro's rep as a guy who can't pitch deep into games is somewhat undeserved. A quick look at some other pitchers who pitched mostly post-1994:

Pedro : 6.75 ip/s
Oswalt; 6.68 ip/s
Halladay 6.79 ip/s
Hudson: 6.67 ip/s
Zito: 6.38 ip/s
Pettitte: 6.39 ip/s

In 2000, Pedro averaged 7.48 IP a start. No one else (I think) in the AL averaged 7 ip/s. The one knock you can make about Pedro's greatest seasons is that he made 29-33 starts rather than 32-35.

90s pitchers have an incredible built-in advantage over 70s pitchers in an analysis like this.

How so? It's true that actual ERAs from the 1970s were much closer to the league-average ERA and replacement-level ERA, but I've re-scaled each decade onto the same scale. Tango's shown that it's just as easy to post an impressive ERA+ in a low-scoring environement as it is in a high-scoring environment.

Pedro looks worse in this comparison to the other great pitchers (Johnson, Clemens, Maddux, etc.) of his era because he pitched almost enitrely in the high offensive era of the 1990's.

Again, I don't think this is true. Offense was way up during the second half of the 90s, but Pedro's ERA is being compared against his own leagues, not the leagues of the early 90s.

If I'm missing something, please elaborate.

90s pitchers have an incredible built-in advantage over 70s pitchers in an analysis like this.

How so?

All pitchers give up more runs in the 8th and 9th innings, than in the first six.

A quick look at some other pitchers who pitched mostly post-1994:

Pedro : 6.75 ip/s
Oswalt; 6.68 ip/s
Halladay 6.79 ip/s
Hudson: 6.67 ip/s
Zito: 6.38 ip/s
Pettitte: 6.39 ip/s

Pitchers in the 90's have avoided this little issue by simply not pitching these innings very often. The age of specialization was here. We have set-up guys and closers to pitch the last two innings.

Take away the runs allowed and the IP, after the 7th inning, of the starters in the 60's and 70's and you will see markedly lower ERA+.

Your analysis is simply a less exaggerated version of comparing relievers ERA+ to starters ERA+. It's not a fair comparison at all. Pitchers who pitch fewer innings per appearmance will always be able to post lower ERA+ without actually having any more talent.

Your analysis is simply a less exaggerated version of comparing relievers ERA+ to starters ERA+. It's not a fair comparison at all. Pitchers who pitch fewer innings per appearmance will always be able to post lower ERA+ without actually having any more talent.

Yep.

Your analysis is simply a less exaggerated version of comparing relievers ERA+ to starters ERA+. It's not a fair comparison at all. Pitchers who pitch fewer innings per appearmance will always be able to post lower ERA+ without actually having any more talent.

Sounds reasonable, but has anyone actually done this type of analysis?

I think the advantage is smaller than you believe. Aren't pitch counts today actually pretty similar to the '70s? It's just that there were fewer pitches required per batter and fewer batters were faced per inning back then. Similar effort resulted in more innings. If so, today's pitchers should be just as tired in the sixth as those pitchers were in the 7th.

My guess is that any advantage today's pitchers have in ERA by not going as far into games is more than countered by the fact that they don't pitch as many innings. If Pedro pitched .5 more innings per start at a 5.00 ERA, he would still increase his value because of the additional innings. That's because I'm comparing to replacement level. Things would be different if you were comparing to league average.

Sounds reasonable, but has anyone actually done this type of analysis?

Yes.

Aren't pitch counts today actually pretty similar to the '70s? It's just that there were fewer pitches required per batter and fewer batters were faced per inning back then.

Applying Tango's Pitch Count Estimator to league-wide totals indicates a slight increase in pitches per game in the 1990s/2000s over the 1970s, but only very slight: 3-to-5 pitches per game.

If this estimation is accurate, then pitch counts today are actually pretty similar to the '70s. And if that's the case, then it cannot possibly be the case that there were fewer pitchers per batter and fewer batters faced per inning back then. The math simply doesn't work.

There are no more walks per game in the 2000s than there were in the 1970s. (There were slightly more in the late 1990s, but that was eliminated with the imposition of QuesTec.) There has been about 1/2-hit per game more in the 1990s/2000s than in the 1970s. All in all, there have not been significantly more batters per game in the modern era; the increase in scoring has come from dramatically more doubles and home runs among the hits.

Thus if there are more pitches per batter in the modern era, then Tango's estimator is not accurate for the 1970s. That may be true, but he tested it against the limited available actual pitch count data from the 1950s/60s, and it checked out as being pretty accurate.

So what's the solution? Is there an accepted adjustment? How about computing replacement level ERA by inning?

So what's the solution?

Good question.

Is there an accepted adjustment?

Nope.

How about computing replacement level ERA by inning?

Something along those lines is probably the best approach. But even that needs to reckon with the fact that a more durable pitcher adds value with his higher workload even when his per-inning rate stats suffer to some degree, simply by helping the team to deploy fewer pitchers.

I think the advantage is smaller than you believe.

Actually you're just hoping that.


It's just that there were fewer pitches required per batter and fewer batters were faced per inning back then. Similar effort resulted in more innings. If so, today's pitchers should be just as tired in the sixth as those pitchers were in the 7th.


This is simply not true. You can check it out.


My guess is that any advantage today's pitchers have in ERA by not going as far into games is more than countered by the fact that they don't pitch as many innings.

You're guesses are leading you to incorrect conclusions. There have been lots of studies done to show the inceasing ERA of starting pitchers on an inning by inning basis. This was actually the basis of the increased use of back-up pitchers, who were not as good as a starter would be in the first five or six innings, but often be better if used in short spurts in the later innings.