Is Woody Williams done?

added 3/18/2008 by Scott Barzilla

Some people out in Astros Nation would say that 2007 was really the beginning of the end for Williams. However, he wasn't that bad outside of a brutal April in May. In fact, management and some fans were cautiously optimistic that he would pitch like he did in the second half this season. For most, those hopes have been dashed.

Most of you reading this are aware of the power half hour we have with David Nuno and 1560 AM on Sundays. I was asked about Woody Williams and I gave a flippant answer. Woody deserves more than that after his distinguished career. Any pitcher with 132 career wins and 102 ERA+ deserves more than a flippant answer. I was going to do an entire column on the competition for the last two spots in the rotation, but I decided Woody Williams deserved more.

Anyone that has been an Astros fan for awhile remembers the dark days following Nolan Ryan's departure. John McMullen refused to sign him to a similar contract as the one he had. He thought the all-time leading strikeout leader (and defending strikeout champion) deserved a pay cut. The team tabbed Jim Clancy instead. Of course, we all remember the 9-22 record and exploding ERA. However, the treatment of Clancy was unfair. He was expected to replace a legend and there was no way he could have lived up to that.

In a similar way, the Astros brought in Woody Williams and Jason Jennings when Andy Pettitte and Roger Clemens were let go. You talk about impossible expectations. It didn't help that Williams was coming off of a good season where his ERA was artificially low and his win total was artifically high. Throw in the fact that he was finally coming home and you could have predicted how everything was going to turn out.

In order to see whether these Spring stats mean anything we need to take a look at where Woody Williams has been. Most baseball fans are now familiar with the DIPS theory that Voros McCracken put forth. When you look at defensive independent baseball statistics you can see the upward or downward progress of a pitcher. You can also see if a pitcher was really good or really lucky.

        INN     ERA     SO/9     BB/9     HR/9     BABIP     FIP  
2005    159     4.85    6.00     2.89     1.36      .299     4.60
2006    145     3.65    4.77     2.17     1.30      .275     4.91
2007    188     5.27    4.84     2.54     1.68      .292     5.57

In plain English, you can see that 2006 really does not belong for a number of reasons. First, even if you went by ERA alone you would have to admit that it falls outside of the normal range. Secondly, his walk totals were lower than normal. More importantly, his batting average on balls in play was abnormally low. He was begging for a drop off in 2007 and he had one.

The fielding independent pitching metric is an interesting one. In accordance with McCracken's beliefs, it shows that you cannot use past ERA to predict future ERA. FIP does a much better job. The following shows what Williams produced and the predicted FIP for him that season.

          ERA     FIP     PERA
2005      4.85    3.95    4.18
2006      3.65    4.60    4.85
2007      5.27    4.91    3.65
Diff      ----    0.74    1.16

So, FIP (while not perfect) is a better predictor of future ERA. This is important because The Hardball Times has a FIP of 5.57 for Williams in 2007. If that is true, he is predicted to be even worse than he was in 2007. This cannot be a good sign for the Astros. Yet, I cannot just take The Hardball Time's opinion and leave at that. One of my biggest pet peeves is when analysts use one or two sources and call it comprehensive analysis. There are four major publications and two websites that do projections. We will see projections from Bill James, Ron Shandler, Baseball Prospectus, The Hardball Times, CHONE, and Marcel. They will be abbreviated to save space.

            INN     ERA     SO/9     BB/9     HR/9
BJ          140     4.37    5.59     2.51     1.29
RS          131     5.10    4.40     ----     1.40
BP          106     4.84    4.58     2.63     1.36
THT         169     5.18    4.42     2.61     1.60
CHONE       178     4.80    4.85     2.68     1.52
Marcel      167     4.72    5.07     2.75     1.46
Average     149     4.84    4.82     2.64     1.44

You get such a clearer picture when you look at multiple projections. Essentially, most of them see Williams as the same kind of pitcher. I think Bill James may have been out to lunch when he did his projection, but otherwise they are all fairly close. The differences are minor and depend on minor differences in strikeout rates and home run rates (the walk rates are extremely similar).

This brings us to the spring numbers. People say not to put too much stock in them and some fans wonder why. Isn't a 13+ ERA indescribably bad? Yes it is, but there are some interesting factors at work that people don't necessarily see. I will include BABIP in these numbers to illustrate my point. Keep in mind the BABIP numbers of the last three years for Williams.

           INN     H     HR/9     SO/9      BB/9       BABIP
Spring     11.2    26    3.86     5.40      3.09       .457

Obviously, anyone that gives up more than two hits per inning is someone to worry about. Yet, no one in history has given up a BABIP over a full season like that. To put this in perspective, Williams would have given up 13 hits that weren't home runs in lieu of the 21 that he has this Spring if his BABIP were at career norms. Now, as for the home run rate, well, who here believes that's going to continue? I sure don't.

There is no sugarcoating the fact that Woody Williams isn't going to be a good pitcher in 2008. As we will see with Doug Brocail, the Astros made the same mistake they usually make when they read too much into the previous year's performance. Williams wasn't going to repeat 2006 anymore than Jason Jennings will repeat what he did last year for the Rangers. If they feel they have better options (we will see next time) then they should go that route, but they shouldn't flip out over these numbers.