Tuesday, March 04, 2008

Down on the Upside

At the end of a piece on Noah Lowry, Rob Neyer implies that while the Giants may look like they'll go 70-92 "on paper," they possess extreme downside, and that makes them the best bet of any team to lose 100 games. Do the Giants actually have more downside than other teams? And if so, why?

I think most people have an inaccurate opinion of what downside means. Firstly, downside refers to deviation from the projected outcome, not a bad projection. The Orioles do not automatically have the most downside of any team in the league, simply because they are the worst team. Put another way, I'd say a world-class poker player in a $100/$200 game has greater downside than a punter in a $.01/$.02 game, even though he's expected to do better on average. The world-class player can easily lose several thousand bucks in a short session, which is impossible for the punter.

Further, teams with "more" downside will generally also have more upside, because additional downside implies greater volatility in their potential outcomes. In other words, their projections have a variance that is greater than normal.

For example, let's say Barry Bonds signs with some team tomorrow. Bonds is going to increase that team's downside right off the bat: he's injury-prone, his playing time is uncertain, and his numbers could fall off a cliff due to age. However, Bonds is also a threat to come to the plate 500 times and post an OPS of 1.050. That's big, BIG upside.

Even this effect is overstated. Perhaps a team with "low downside" will have a standard deviation of 8.5 wins from their projected total, and a "high downside" team's S.D. will be 9.5 games. This is a significant difference, but not a huge one. Besides, if your team isn't projected to make it to the playoffs, you'd rather have the higher downside, because with it comes higher upside, and you'll need that upside to advance to October.

To maximize a team's downside, we want as much of that team's value as possible to be tied up in one or two players that are injury-prone and far, far better than their backups. In my mind, that team is not the Giants, but the Cardinals. Even though they likely won't make the playoffs this year, St. Louis still has baseball's most valuable player, and reports keep coming out that he'll need major surgery at some point this year. If Pujols goes down, who takes his at-bats? Ryan Ludwick? That's a 10-win dropoff.

If we're talking pitchers, I think Mark Mulder has about as much downside as any starter in baseball. I don't think anyone would be too surprised if any of Joel Pineiro, Braden Looper, or Matt Clement put up an ERA over 6.00 this year.

Perhaps we're not considering the Cardinals for this "honor" because they're projected to finish near .500. But that's not the way we should be looking at it.

No comments: