I plan on posting the 2006 run distributions in three phases: against the 2005 team; against the AL average; and against the rest of the AL East.
Run Scored Distribution - % of Games
I posted these charts at the end of August but wanted to post the final numbers. Obviously, the 2006 did a much better job scoring runs. The 2005 squad was able to score fives and six runs many more times than the 2006 team. While that was often good enough for the 2005 team, the offensive output of the 2006 was still superior.
Run Scored Distribution - Win %
The winning percentages of the 2005 team in low scoring games was phenomenal and not repeated in 2006. As you can see in the chart above, the Sox just needed to score 3 runs in 2005 to achieve at least a .600 winning percentage and they didn't do too shabby when they only scored two runs. The reason they had so much succes in low scoring games in 2005 was because the pitching was much better.
If we look at how many games each team would be predicted to win with league average pitching (run distribution X league winning percentage at that run level)* the 2006 Sox come out ahead 88.6 to 80.4, a full 8 games better.
Run Allowed Distribution - % of Games
You can see why the Sox won so many of those low scoring games last year. The pitching staff yielded 3 runs or less in 49.38% of their games last year compared to 35.42% in 2006. As for the high scoring games, the Sox gave up 7 or more runs in 32 games last year. In 2006 that number was to 46.
Run Allowed Distribution - Win %
This chart shows how the 2006 Sox were able to win some games even when the pitching gave up runs. In 2005, the Sox only won 7 out of 45 games (15.6%) in which they gave up more than 6 runs. In contrast the 2006 Sox won 14 out of 61 (23%).
Overall, the expected wins by the 2005 pitching staff with a league average offense would have been 92.11 games compared to 82.04 for the 2006 team, a full 10 wins better.
What I found interesting however, is the fact the 2005 team actually had a better winning percentage than the 2006 Sox in games where they allowed either 4 or 5 runs. This should not be the case as the 2006 team had a much better offense. I'm not sure if the discrepancy can be fully explained. I would chalk it up to the small sample size (42 games in 2005, 37 in 2006), lucky bounces (or clutchness if you believe in such things), and a lockdown bullpen. In 2005, it always seemed that the Sox played with the lead and did enough to hold on for nine innings. In 2006, it always seemed, the Sox were digging out of holes created by their starters.
Like I said, I don't think it can be fully explained. Yet, at the same time it doesn't really surprise me that the 2005 squad had a better record when they allowed between 4 or 5 runs. I always expected the 2005 team to pull out the close ones and felt the exact opposite watching last year's squad.
* for example, AL teams won 59.1% of their games when they scored 5 runs. The Sox played 14 of these games and would be expected to win 8.27 games (14 x 59.1%). Adding up all their games would give their offense 88.6 wins.