Wednesday, August 20, 2008

Evaluating Pitchers

Anonymous asks:
I'm trying to assess some free-agent pitchers in my NL-only league. Generally -- with hitters -- I look at the last seven to 14 days of production so I can see who has the hot bat. How should a judge pitching performance, since SPs might have only two starts in that 14-day period? How would I judge middle-relievers? Thanks.
I find an analysis of the last seven, 14, and 21 days useful when it comes to batters who are either rookies or bench players who are on the cusp of working themselves into extra playing time. If Paul Konerko isn't hitting in an A.L.-only league, I'm stuck with Paul Konerko and I'm not going to drop him for Randy Ruiz. For a player like Ruiz, though, it's worth looking at his short-term line: not as an indicator of future performance, but more for an idea of whether or not Ron Gardenhire will keep finding AB for him down the stretch.

For this reason alone, three or four start trends aren't all that useful when it comes to pitchers, particularly since some organizations might have a quick hook after a couple of starts, while others might let a pitcher twist in the wind for 10-15 starts before pulling the plug.

On the other hand, there many more opportunities to grab a starting pitcher who is taking a regular turn in the rotation than there are to grab a position player with a regular job. The flip side of this equation is that most of the starting pitchers who are sitting in the free agent pool week after week are dreck, and there's a good reason that team near the bottom of the pile in wins won't touch any of those guys; his ERA and WHIP stand to crater if he does.

I'd agree that one or two week samples are very small. However, I also can't help but look at them, in the hopes that there will be something promising in the tea leaves, since the pitchers available in deep leagues are either marginal filler guys or rookies who will be aggressively FAABed right out of the gate.

What I'm looking for are pitchers who fit the LIMA profile: good K/IP, K/BB, BB/IP, and HR/IP rates. Walks, strikeouts, and home runs - in that order - are the three things I like to look at in terms of potential value going forward. Even if a pitcher isn't a strikeout machine and is prone to the occasional long ball, throwing strikes is an obvious key to success and keeping runners off base via the walk is the best way to fall prey to the big inning. I like strikeouts next, since a pitcher who can keep the bats off the ball is more likely to throw up a dominant outing. Failing both of these, a pitcher who keeps the ball in the park can at least allow his defense to pick up the slack, or at least get lucky and stay out of the big inning by allowing a single and two doubles as opposed to two singles followed by a long ball or two.

Beyond that, I like to try and get pitchers on good teams if those pitchers are available. These pitchers have a greater chance at picking up wins (or saves as relievers), and also of having better defenses behind them. They're also less likely to be left in a game to wither and die on the vine.

It'd certainly be interesting to take a look at LIMA pitchers and see if guys who didn't do well in ERA/WHIP but had good K/IP, BB/IP and HR/IP numbers were successful the following year. However, my instincts have generally been good in this regard, and I've done pretty well sticking with this basic strategy for scouting and picking up pitchers.

No comments: