Who is the greatest pitcher ever? There are different answers to that question, I suppose.
When Bill James ranked pitchers a decade ago in his "New Bill James Historical Baseball Abstract," he had Walter Johnson at the top of his list, followed by Lefty Grove and Pete Alexander (Roger Clemens and Greg Maddux were still finishing up their careers).
According to Baseball-Reference.com's career WAR leaders list, it's Cy Young, followed by Johnson and Clemens.
If you go by the highest percentage of votes upon induction into the Hall of Fame, it's Tom Seaver and Nolan Ryan (who have the highest percentages of any players), followed by Steve Carlton.
If you go by Cy Young Awards, it's Clemens, followed by Randy Johnson and then Maddux and Carlton.
If you just go by career wins, it's Young, Johnson (Walter, not Randy) and then Alexander and Christy Mathewson.
That's 11 pitchers we've mentioned, all with varying claims to being the "greatest ever," and you see why there isn't an easy answer to this. And that doesn't include Warren Spahn or Bob Gibson or Pedro Martinez or Sandy Koufax ...
In the Hall of 100 vote, Clemens finished as the highest-ranked pitcher -- seventh overall -- while Walter Johnson ranked 13th and Maddux 14th. Let's focus on those three since they led the voting. It's a little easier to compare Clemens with Maddux since they were contemporaries and I think the voters got this one right. While Maddux had that incredible run of four straight Cy Young seasons from 1992 to 1995, I'm not sure how you rate him higher than Clemens, even if you emphasize peak seasons.
On a career basis, it's not that close -- Clemens towers above Maddux in career WAR. Let's look at peak value.
Best three seasons -- Clemens 31.0 WAR, Maddux 26.7 WAR.
Best five seasons -- Clemens 48.0 WAR, Maddux 41.3 WAR.
Best eight seasons -- Clemens 70.9 WAR, Maddux 59.4 WAR.
That's how awesome Clemens was -- clearly better than the wonderful, amazing Maddux. And before you start jumping on Clemens' alleged PED usage -- keep in mind that the Hall of 100 voters were told to factor in only what happened on the field -- Clemens' best seasons were 1997, 1990, 1987, 1986 and 1992. Other than that outlier 1.87 ERA season with Houston in 2005 when he was 42 (a season driven by a low .246 average allowed on balls in play), his aging pattern wasn't really all that unique. His best seasons came when he was younger and threw harder and threw more innings. Remember, this was a guy who once threw 18 complete games in a season but threw just four over the final nine years of his career. And keep in mind that many pitchers have been excellent in their 40s -- Ryan, Randy Johnson, Young. Warren Spahn won 23 games at age 42. Being a great pitcher in your 40s isn't as historically unique as being a great hitter in your 40s.
Clemens versus Walter Johnson is a more difficult comparison, considering one guy was born in 1887.
Johnson had more innings and the lower ERA, but that's a product of environment, of course. Johnson pitched most of his career in the dead-ball era, when home runs were essentially a non-factor -- in 1916, he pitched 369 innings and didn't allow a single home run. Of course, there were only 144 hit that entire season in the American League. When you adjust each pitcher's ERA to its era -- the ERA+ figure above -- it's pretty close, with Johnson having just a very slight edge.
And because Johnson pitched in an era when starters threw more innings, his best seasons are worth more then Clemens', at least in terms of WAR -- 14.3 in 1913, 12.9 in 1912, 11.5 in 1914, 10.8 in 1910 and 1915. Johnson was 32 when the "lively ball" era began in 1920, and his best after that point 6.3 WAR in 1924. If we look at how many times each pitcher led his league in WAR, we get Johnson with eight and Clemens with seven. Clemens led in ERA seven times, Johnson five. Johnson won more games and pitched more innings, but lost many more games. Clemens pitched on better teams, but the Senators of Johnson's era weren't as awful as often believed, and they later won two pennants in 1924 and 1925. Baseball-Reference estimates the winning percentage that a player would push an otherwise average team to -- for Clemens, it's .633, for Johnson it's .621.
It's close. In our voting system, I gave both players a 95. If I had to rank ... well, I give the edge to the guy from the more modern times.
You may disagree.