Originally Posted by Hoosier Red
A thought came to me last night with Lohse's performance:
For five out of seven innings, he shutout the D-backs. Of course in the other 2 innings he gave up 4 runs. But I was wondering if that's a good thing or a bad thing or nothing.
So it leads me to ask this;
1) Is there any way to track how a pitcher gives up runs, say in big spurts,(one or two multiple run innings) or 1 and 2 runs per inning over 5 innings.
2) Is one better than the other.
Can I look at Lohse's performance and predict anything for future performances?
Just a thought, I'd be thrilled if some of the better stat people could point me in the correct direction to find data for such a project.
Yes, it has been studied. The venerable Bill James did it in discussing peak seasons in his book, "What Ever Happened to the Hall of Fame?"
And the short answer is that peak seasons (and by extension, peak games and volatile performances) are worth more than consistency -- they are worth 23 or more pennants in 10,000 seasons.
The long answer is the Sutton vs. Carlton debate (they have essentially the similar career records but very different career arcs) and the discussion on pages 79-87 of the book. As the old saying goes, "a run saved is worth more than a run scored," and the old saying appears to be correct. Pitching volatility is a good thing because a shutout or 1-run allowed game is worth far more than the cost of a volatile performance (i.e., the downside is a crappy 5 or 6 runs allowed in 5 IP and your team still might have a chance to win).
Your intuition is right though, these metrics should be tracked because then we would be able to identify value in the performance volatility that others are missing by simply looking at ERAs.