The Yankees might have just traded their best pitcher

Well, their best pitcher not named Philip Hughes.

The Yankees have agreed to trade Randy Johnson to the Arizona Diamondbacks. Many writers and Yankee fans are praising this move; the Yankees are getting younger, continuing to stockpile prospects, and saving money. Most importantly, however, the Yankees are getting a pitcher who had an ERA of 5.00 last season out of their starting rotation.

Foxsports.com’s Dayn Perry writes that the Yankees got the far better end of this deal. He cites seven “grim indicators” about Johnson’s future with one conclusion: Randy Johnson is visibly declining.

No one will argue with the idea that Randy Johnson is declining. However, the key is that Randy Johnson used to be so good that even a Randy-Johnson-in-decline is still pretty darn good.

Johnson did have a 5.00 ERA last season; however, this ERA is misleading. I would argue that his ERA was rather fluky---if we look at the underlying statistics (which are far better indicators of future performance than ERA)---we see a pitcher who is in a clear decline, but also pitched far better in 2006 than people might realize.

Johnson’s strikeout rate was low for him, but still an above-average 7.7 strikeouts per nine innings. His walk rate rose, but only to 2.7 walks per nine innings. In other words, Johnson obviously was not as good as he had been in the past, but was still quite good, posting an overall K/BB ratio of 172/60. He also gave up a lot of home runs for him, allowing 28 in 205 innings (of course, Curt Schilling allowed 28 in 204 innings).

Johnson’s main problem was pitching with runners on base. He stranded only 61.8% of the runners who reached base---meaning that almost 40% of all runners who reached base ended up scoring (for comparison’s sake, his previous two years’ strand rates were 74.1% and 69.5%). In fact, Johnson’s stand rate was the worst for any starting pitcher in all of baseball, at least 2% worse than any other starter’s.

With nobody on base, opposing hitters hit .206 with a .595 OPS against Johnson; with runners on base, the batting average rose to .321 and the OPS rose to 927. With runners in scoring position, Johnson allowed a batting average of .348 and an OPS of 998. However, Johnson’s K/BB ratio was very similar with the bases empty and runners on base---with the bases empty Johnson struck out 2.91 batters per walk; with runners on base, he struck out 2.78 batters per walk.

Only 38% of the plate appearances against Johnson came with runners on base, yet 15 of the 28 homers Johnson allowed were with runners on base. Hitters were 9-for-15 against Johnson with a runner on third; 7-for-14 with runners on first and third; 6-for-15 with runners on second and third; and 3-for-6 with the bases loaded. With runners in scoring position and two outs, batters hit .346 against Johnson.

Perhaps this is an indication that Johnson is not as good of a pitcher when there are runners on base. If this were true, we’d expect to see similar splits from 2005, when he was also old, also with the Yankees, and also hampered by back problems (although not to the same extent).

Year	2005	2006
BA with bases empty	0.228	0.206
BA with runners on base	0.251	0.321
OPS with bases empty	0.653	0.595
OPS with runners on base	0.766	0.927
OPS with runners in scoring position	0.717	0.998
Runner on 3rd	5-for-24	9-for-15
Runners on 2nd and 3rd	1-for-10	6-for-15
Runners on 1st and 3rd	9-for-43	7-for-14
Bases loaded	2-for-8	3-for-6
% of PAs with runners on base	42%	38%
% of homers allowed with runners on base	50%	54%
strikeout/walk ratio with bases empty	4.06	2.92
strikeout/walk ratio with runners on base	4.65	2.78

If Johnson was not as good of a pitcher with runners on base, we’d likely see a significant change in Johnson’s K/BB ratio with runners on base, as compared to his ratio with the bases empty. In 2005, Johnson was slightly better with runners on base; in 2006 he was slightly worse. However, neither difference is extreme---it’s fair to say that Randy Johnson was essentially the same pitcher whether there were any runners on base or not. (Additionally, I must note that successful veteran pitchers---especially those without power stuff---sometimes post worse ratios with runners on base because there is less incentive to challenge hitters---thus leading to more walks---with runners on base).

Nate Silver at Baseball Prospectus did a similar evaluation of Johnson and also noted that Johnson’s BABIP with the bases empty was .239, while with runners on base it was .369. “This is almost certainly a matter of luck,” Silver writes.

Certainly Johnson was not as a good of a pitcher in 2006 as he was in 2005. However, the change in Johnson’s abilities was not nearly enough to warrant a rise in ERA from 3.79 to 5.00. Yes, the Yankees are getting younger and cheaper, but they’re also losing a pitcher who will likely perform well in 2007 from an already-questionable rotation.

Perry writes that “Arizona must now pin their hopes on Johnson's ability to defy expectations.” He is right that Arizona is hoping that Johnson defies Perry’s expectations; however, Perry’s expectations are misguided.