The AL's DL factor and it's application to totals is hard to measure for a couple of reasons. First, let me give you a few facts.
NL average ERA -- 4.45 (all 16 teams)
AL average ERA -- 4.53 (all 14 teams)
Hence, the OVERALL ERAs from league-to-league are negligible (less than .1 run per game)
However, the figures are skewed because of a couple of factors. First -- ballparks in Colorado (and to a lesser extent Houston) have skewed the statistics downward. Scores of 12-9 are common in Colorado and Houston, as well. For those who may think Texas' and Tampa's pitching numbers would offset this skewing -- keep in mind that Pittsburgh and Cincinnati (bad pitching staffs) offset this somewhat.
Let's look at STARTING PITCHERS and not just ERAs but actual runs allowed per staff/per 9 IP:
NL average RA: 5.02 runs per game
AL average RA: 5.25 runs allowed
This indicates AL games produce about .23 more runs per game -- on average (I would have to do the same breakdown for bullpens and extrapolate the figures in correlation with IP to get a precise figure). This does not necessarily mean the DL factor ALONE accounts for this difference. NL starters could simply be marginally stronger overall (more Maddux's and Johnson's in relation to the rest of the league). But the DL certainly plays a role. Since the average teams sends ~40 batters to the plate in every game, it follows that with pitchers hitting 3-4 times in those spots, that the NL loses 2-3 percent in terms of hitting effectiveness compared to AL counterparts. Assuming the average MLB game produces 9 runs, this 2-3 percent would be in line with about a .23 differential between the two leagues.
Finally -- in terms of the realtive value of the DL rule in totals one should probably raise the figure a bit because of the Colorado (and Houston) factor.
My conclusion is that, based on a mix of stats and subjective judgment, AL totals should be (on average) about a half-run higher than NL totals. Hence approximately .5 is my answer.
Other points of view welcome.
Nolan Dalla
