Here's a review of our 2006 projection for closer Mike Gonzalez.
| Community | 72.2 IP, 84 K, 31 BB, 2.68 ERA |
| ZiPS | 44 IP, 52 K, 21 BB, 3.27 ERA |
| Actual | 54 IP, 64 K, 31 BB, 2.17 ERA |
Who was closest depends on what metric you use. Momen 21 got pretty close to Gonzalez's K's, BB's and IP but was off on his ERA; Azibuck and SteveZ were close on the ERA but off in the other categories.
We got closer than ZiPS here, but this may not prove a lot more than that guessing what ERA a reliever will have in a single season is a pretty strange and somewhat meaningless exercise. And ZiPS at least guessed correctly that Gonzalez would miss time with an injury, whereas we didn't think so.
Anyway, Gonzalez was dominant when he pitched. He suffered some control problems in the first half that didn't end up hurting his ERA or preventing him from keeping runs off the board - despite the relatively high number of walks, he was only a little worse than average at keeping inherited runners from scoring, and only slightly fewer of Gonzalez's runners scored after he left than would be expected.
Before Gonzalez was shut down on August 29 with elbow tendinitis, he was having a ridiculous second half - 18.1 innings after the break, he struck out 29 batters and walked 10, allowing only three runs.
Gonzalez is a great reliever, but history tells us that great relievers tend not to stay great for long. For example, only four of the top ten leaders in relief strikeouts in 2004 were good just two years later, in 2006. (If you're wondering, I counted Scot Shields, Justin Duchscherer, Tom Gordon and our own Salomon Torres.) Reliever performance is highly variable. As I've said before, Gonzalez, who is injury-prone and who will be 29 next season) is not likely to be good next time the Pirates could possibly be competitive. Now is the time to trade him.