Why are the Angels winning more than their "Expected Won/Loss" record, computed daily by the ghost of Pythagoras and then transmitted through K-Rod's hands whenever he exalts skyward? Why are they winning far more than their very modest "run differential" would predict? And why have the Angels been able to confound Pythagoras somewhat consistently over the past few seasons?
I don't believe it is luck and I don't believe that the predictive models are especially flawed. I believe the Angels are constructed to confound the model. My theory is that three factors, some of which have I have seen suggested on AN and some of which I haven't, are combining to help the Angels exceed the expectations that their "runs scored/runs allowed," and other similar metrics, predict:
1. Bad Long Relief Pitching This results in the Angels losing games 9-2 that they could lose 5-2. When the Angels are trailing and going on to lose, they do it in style, turning it over to "run differential killers" like Bootcheck and Moseley. This hurts the Angels' "metrics" but it doesn't hurt their won/loss record much because you can only lose a game once.
2. Excellent & Deep Short Relief Pitching While it's true that teams can't usually choose how they distribute the runs they score or give up, they can choose how to react to the flow of distribution. The Angels, compared to other teams, rarely lose games they lead going into the 9th inning. Or games they lead going into the 8th inning. Or games they lead going into the 7th inning. So if in the see-saw distribution of runs, the score should happen to tip in the Angels' favor at any point, Scioscia can pounce on the opportunity, go into "close out mode," and secure a one or two run victory. That doesn't mean the Angels are cleverly choosing which innings to score in or when to take the lead; it means whenever it happens, they can turn it, better than most teams, into a "by small margin victory".
3. "Playing For One Run" While they are not good at scoring a lot of runs overall, the Angels are adept at creating a run here, a run there, through their patented combination of running, taking the extra base, bunting/squeezing, etc. They don't get a lot of opportunities to score, but they do a pretty good job of converting the opportunities they do get to score a single run.
Why is this so helpful to the Angels? Because they have created a game in which one run is worth a lot. Let me use a little exaggeration to make a point. Let's say the average score of a major league baseball game were 6-4 (it's actually closer to 5-3) and let's say the average score of an Angels game were 3-2 (it isn't actually that low, as the Angels both score and give up more than 3 runs/game). In that scenario, in an Angels game, one run would be worth two runs, and "playing for one run" or "manufacturing one run" in an Angels game would be the same as "playing for two runs" or "manufacturing two runs".
The conversion, or inflation rate isn't actually quite 2:1 but maybe it's 1.5:1.0, meaning that the Angels have created a game where single runs are worth more than usual, and then put a team on the field that is good at securing that one precious run and can protect the precious "lead run" whenever it is procured.
That's my theory, to be refuted by every available stat, metric, and point of data. Which is fine with the Angels, because they've been winning this way for years and there doesn't appear to be a thing Pythagoras can do about it.