I still can't believe Jack Morris tossed a 10-inning game in the '91 Series. What a bizarre time to revive a dead art.

I won't lie when I say I miss the complete game, but as King pointed out in his column and his gentle response to an earlier comment of mine, these pitchers are thinking their way through the nastiest hitters the Western Hemisphere has to offer. No wonder they're tired; they don't have pockmarked baseballs dying at infielders 90 percent of the time. If Shoeless Joe were cranking out 35 jacks a year, maybe Chesbro would've ceded to a reliever.

Dadgummit! While we're complaining about lost arts, whatever happened to boxing matches? You hardly ever see fights going over 50 rounds any more! I remember when I was a lad and John L. Sullivan knocked out Jake Kilrain in 75 rounds. We almost felt cheated when that pantywaist Kilrain didn't finish all 80 rounds as originally planned.

This reminds me of something I've wondered about for years: why bother with starting pitchers at all, especially in the NL? If your whole pitching staff is middle- and short-relievers, you can mix and match all you want and use the exact same pitchers tomorrow, or at least the day after.

Comparing today's game to the game hundred years ago is setting up a comparison to a strawman. A more relevant comparison is to the game played in the 70s and 80s. As far as I can tell it's basically the same game with the same quality of players as it was then. Why less complete games now? I think it's just a change in strategy, nothing else. Use of the entire pitching staff has changed drastically in the last 30 years, with more innings being given to the worst pitchers. starters are pitching fewer innings and the best relievers have been relegated to the nebulous ninth-inning closer role rather than the far more useful "fireman" role they used to occupy. All changes for the worse in my opinion

I think that KK used 2004 as an example b/c Phil Pepe used them in his article as follows:

The Yankees, as an example, range from a high of 123 complete games in 1904, when they were still the Highlanders, to a low of one complete game exactly 100 years later. It’s interesting to note that the number of saves recorded by the Highlanders in 1904 is the same as the number of complete games recorded by the 2004 Yankees, 1.

Also interesting is that even with their staggering total of 123, the Highlanders were eighth in an eight-team American League in complete games. That season, of the 1,252 games played, the AL recorded 1,097 complete games, or 88 percent of games started, and only 12 saves.

Therefore, it seems unfair to accuse KK of using 1904 to create a strawman argument.

Good point by Mark but that still doesn't explain why managers take out one of the best pitchers on their staff (a starter) in the sixth inning with men on base so that they can bring in Jorge Julio, or Scott Proctor

Our friend Mr. Kaufman did not do himself a great service by his snide and pointlessly dismissive article on the collapse in pitchers completing their games. If he had a coherent argument, I would like to have seen it. You don't have to go back to 1904 to remember complete games--I saw Seaver, Carlton, and Hunter manage it just fine, and they faced black and hispanic ballplayers of exceptional merit, along with a few good white players, too. I agree with him about the overall pool of talent rising faster than the number of teams, but I also know that the NBA and NFL today skim off a lot more talent than they did thrity or fourty years ago, so the net result is probably a wash. A little sarcasm is one thing; showing contempt for your elders is just bad manners.

When exactly did it become de rigeur to track pitch counts in the course of a game? The managers we heard back when I cared -- up until, oh, '91 or so -- would almost always explain removing a pitcher in terms of how many throws they'd made, if there was any reason for them to be asked. (When the guy's getting shelled, it's obvious.)

The idea was to pull starters before their effectiveness tailed off, and in the long run to prevent injuries.

People are saying the 60s and 70s would be the better comparison, 1904 having been a very different game. Does anyone know if the teams themselves even kept track of pitch counts by the 1970s? Or were managers just looking for telltale beads of sweat on their guys' furrowed brows?

drlimerick: This reminds me of something I've wondered about for years: why bother with starting pitchers at all, especially in the NL? If your whole pitching staff is middle- and short-relievers, you can mix and match all you want and use the exact same pitchers tomorrow, or at least the day after.

It's an interesting question. It would be a tough sell to the fan base, for one thing. Also, generally speaking, while there are enough middle relievers around to cover two or three innings most games -- say, one or two competent, no need to hide your eyes when he's pitching type per team -- I don't know that you can stockpile enough good ones to cover 50 or so innings every week for one team. Would you rather have 200 innings from a solid starter or try to find three or four guys to do the same thing, and keep them all healthy, etc.? Times four. It's politically difficult, if not impossible, to convince starting pitchers to submit to such a rotation. They get paid for wins. They don't want to pitch 2 or 3 and come out. They'll sign elsewhere.

One of the stathead books I wrote about last week, "Baseball Between the Numbers," suggests that the gain from not letting a lousy pitcher bat, pinch-hitting every time the spot comes up, would be almost a half a run a game -- akin to replacing a nothing hitter with Vlad or A-Rod. It suggests using the 4th and 5th starter, plus a middle reliever, as a three-man team to cover starts number 3 and 5 in the rotation, pinch-hitting for the pitcher's spot every time around. They each throw 1 to 3 innings both of those days, and and turn the ball over to the bullpen around the 5th or 6th, which back of the rotation starters do anyway.

David Novak: King, you can do better. Comparing today's game to the game hundred years ago is setting up a comparison to a strawman.

As marktgarten points out, I didn't do that. Phil Pepe did. I made fun of him for doing that.

A more relevant comparison is to the game played in the 70s and 80s. As far as I can tell it's basically the same game with the same quality of players as it was then.

I don't think that's true. The hitters are much better today, and they're playing in MUCH more hitter-friendly parks. It's a lot harder to pitch today. There's power everywhere in most lineups. Where are the Mario Mendozas of today, the Johnnie LeMasters and Mark Belangers, the guys who have 10-year careers or longer without even hinting at hitting? You can't coast through the bottom of a lineup anymore. Look at HR per game:

1975: 1.40

1985: 1.71

2005: 2.06

And it takes more pitches to get through a lineup. There are no good pitch-count figures from the '70s and '80s, but look at the strikeout figures per game:

1975: 9.97

1985: 10.68

2005: 12.61

I think it's just a change in strategy, nothing else.

It's certainly a change in strategy, and as you point out, a lot of the usage of bullpens is less than optimal. But it's not "nothing else."

More David: Good point by Mark but that still doesn't explain why managers take out one of the best pitchers on their staff (a starter) in the sixth inning with men on base so that they can bring in Jorge Julio, or Scott Proctor.

There's also the well-documented effect of high pitch counts on future performance. ( Ianscot, no one talked about pitch counts in the '70s, and if teams kept track of them, they kept it quiet.) After a certain point, leaving a pitcher in, even if he does well in this game, increases the likelihood of poor performance in the future, as well as injury.