I don't think you understand simulations.  A large group of simulations show how two teams are likely to fare against one another.  During the regular season there are a limited number of matchups, so the results may not be even close to what they likely would be in a larger sample.

Matt DID beat Peavy three straight times last season -- by one-sided scores of 8-2, 8-4 and 4-0.  But that certainly doesn't mean the reverse might not occur in their next three meetings.  Prior to that 8-2 loss Peavy had gone 5-1 with a 1.82 ERA against the Giants since 2004.

A year ago one might very well have said that Peavy's domination over the Giants would make it ridiculous that Jake would lose three straight decisions to a Giants rookie.

We must remember that even though a 162-game seasons evens a lot of things out, if that same season were played ten different times, there would very likely be ten different results.  The results would likely be similar in some case, but very different in others.

If a simulation is based on reasonable assumptions and is based on many different occurences, it is likely to be at least as accurate a predictor of what would happen if the "experiment" of a season were repeated many times and averaged as the actual season results themselves are.

When we talk about one team being better than another, that would be true in most seasons if our analysis is accurate.  But if we ran enough 162-game seasons, the other team would finish higher in some of them.

A player gives a pretty accurate picture of what he's all about over a 10- or especially 20-year career.  But if we looked at any one of the 10 or 20 seasons, we might get a very wrong impression of the player's abilities.

For instance, is Randy Winn the highly-impressive player of 2005 or the unimpressive player of 2006?  In actuality he's probably neither, but somewhere in between -- closer to his career averages.

Your comment that "All of those numbers are pretty ridiculous" was itself pretty ridiculous.  Not because you don't know your stuff.  But because you don't truly seem to understand simulations and possibly don't understand that even a 162-game season doesn't necessarily provide an accurate picture of a team's abilities.

As an example, Detroit was a much better team in 2005 than its record showed.  And look at how they improved in 2006.  Without even looking at the moves the Indians made, I know that they were great underachievers last season and could become the improved Detroit-type team of 2007.

Part of the reason the Giants didn't improve more last season over 2005 is that the Giants played a tiny bit over their heads (believe it or not!) in 2005.  Yes, without Barry, they were a few games WORSE than their record indicted.  Last season they played pretty close to the sum of their individual performances.  In 2005 they played BETTER than expected from the individual performances.

I happen to think that the Padres and Dodgers will be better than the Giants in 2007.  The Diamondbacks could well be better too. Maybe even the Rockies, although I wouldn't go so far as to predict that (at least not at this time).

Now the Giants COULD be the best team in the NL West.  But they would likely need to have many players have career or near-career years in order for that to happen.  And it isn't likely that a high percentage of them will do so.

The Giants weren't very good last season.  They weren't horrible either.  But just how did they improve themselves by 10 or 15 games?  They will likely need a good bit of luck to make that happen.

I hope they receive that luck and go on to win the World Series.  But the odds on their doing so will be high -- and that is appropriate.