Thursday, June 5, 2008
Count the Winzzzz
Mike Mussina is 9-4 with a 4.01 ERA after last night's 5-1 victory over Toronto. Moose was great, going six innings and allowing just one run, with a season-high six strikeouts.The problem is, wins are not really a good indicator of how he's pitching. Wins are at least partially dependent on the run support his team has given him. Here are some run support numbers for the Yankees' starters.
Mussina: 5.30
Rasner: 4.81
Pettitte: 4.79
Wang: 4.58
Kennedy: 3.91
Hughes: 3.74
The other thing is, Moose's ERA of 4.01 ignores nine unearned runs he's allowed, which means his RA is 5.21, which is below average. We shouldn't penalize Mussina for the errors that were made behind him, although he's responsible for the events that happen post-error. We can use linear weights for the batting events behind Mussina to estimate how many runs he should have allowed this year.
Here's a simple table that shows the offensive events that have occurred against Mussina this year, and their linear weights values as far as runs scoring.
| Event | Count | Weight |
| H | 75 | 0.47 |
| 2B | 11 | 0.38 |
| 3B | 1 | 0.55 |
| HR | 9 | 0.93 |
| BB | 11 | 0.33 |
| HBP | 4 | 0.33 |
| ROE | 6 | 0.47 |
| SB | 6 | 0.22 |
| CS | 1 | -0.38 |
| Outs | 192 | -0.1 |
| BR | 37.86 |
What these numbers mean is that the average hit is worth .47 runs, with each extra base worth a certain bit more, walks are on average worth 1/3 of a run, etc., Reaching on an error is weighed the same as a single. So when you include the 6 ROE against Moose, he's allowed the linear weights equivalent of 38 runs this year, which would translate to an RA of 5.08. If we remove those 6 ROE, then his RA drops to 4.69.
A quick check of Moose's FIP (fielding independent pitching) and xFIP (expected fielding independent pitching) indicate that he's pitching better than that. FIP is calculated using the formula 13 x HRs + 3 x (BB + HBP) - 2 x SO, divided by innings pitched and then added to 3.2. This gives you an ERA-like number that removes the impact of the non-HR hits a pitcher allows (regresses BABIP to exactly average). xFIP is the same formula, but replaces HRs with the number 0.11 times fly balls, which is the major league average for HRs per fly ball. Moose's FIP is 4.35 and his xFIP is 3.84. With a below average defense behind him we'd expect him to have a worse RA/ERA than FIP/xFIP, so this isn't really surprising.
The point of this entry isn't to be a wet blanket. I hope Moose wins 20 games and puts up an ERA under 4. But it's just a caution that his numbers may be a little superficially good. Here's hoping he can keep it going.
And Mariano Rivera is awesome.
Page 1 of 1 pages: