Pitching match-ups

I read what I thought was an interesting post at the The Bard's Room - a White Sox blog on the MVN network of blogs - the post is an analysis of the value provided by 'slot' starters - that is, the relative performance of guys in the #1, 2, 3, 4 and 5 slots:

Here are the MLB average ERAs for each slot:

#1: 3.60
#2: 4.14
#3: 4.58
#4: 5.10
#5: 6.24

The post then provides the same ERA information for the AL:

#1: 3.70
#2: 4.24
#3: 4.58
#4: 5.09
#5: 6.22

The lack of difference between the two leagues actually surprised me quite a bit. However, the difference in quality between the slots really got me thinking - a 68% decline in ERA between the #1 and #5 slots is over 2 runs a start... a crucial difference in winning a game.

How did the Red Sox match up against this?

#1: 3.97
#2: 5.01
#3: 4.63
#4: 5.73
#5: 6.04

I used Schilling as our #1, but put Beckett in the #2 slot, and Wakefield in the #3 - I know they didn't start in that order, but I think it reflects the reality. For the #4 and #5 slots, the ERA is an amalgam of the numbers put up by Clement and Lester (for the #4), and Wells, Tavarez, Johnson, DiNardo, Gabbard, Jarvis, Pauley and Hansack (for the #5), but unadjusted for any bullpen numbers that they put up.

Clearly not pretty - better than average in the #5 slot, but everywhere else... below average, and quite significantly below in the #1 and #2 slots.

What interested me in all this, was how relevant is the slotting - we all in our minds have an idea of who is our #1 starter, #2 and so on - but does that matter - how often do they match up against the other team's #1, #2 etc...? (and this really is under the mildly interesting, but not particularly useful category of analysis)

The following table shows this data for the Red Sox match-ups in 2006:

	Opponent
	1	2	3	4	5
1	7	10	3	1	10
2	13	5	6	3	6
3	4	7	5	2	5
4	6	1	3	7	10
5	8	4	11	10	15

(With the Red Sox rotation rank on the vertical axis)

I allocated a rotation 'rank' by looking at the start of season rotation order, with some adjustments made for guys who would have started in a specific slot, but for injury - AJ Burnett being a good example - he would presumably have been the Blue Jays #2 behind Halladay had he been fit. For any team (including the Red Sox) where a starter was used who was not in the initial rotation, and would not have been in it had he been fit (pick any one of the gang of 8 above for the Red Sox for example), I ranked them as a #5.

As I said further up, this is more interesting than useful, but does perhaps show that we shouldn't get too caught up in any Red Sox spring training debate on the rotation order - all of our starters will see their fair share of 'ace' match-ups. The run differential between starter slots is really only crucial when they are matched up against a starter slot where the differential exists... but as the table above shows, those significant mis-matches are really quite common.