Earlier today, I came upon this 2008 piece from The Hardball Times about risk in baseball.
It is a well thought out, well-written piece by Victor Wang, but in Rotisserie we do have to make decisions regarding "risk" before we compile our bids.
I agree with Wang's general point that there isn't a good or even mediocre metric to measure risk. But we have to look at the following components when we decide what to pay for a player:
Health: Is this a player who stays healthy every year? Does he miss a little time now and again due to injuries? Or is he the kind of player who misses 40-50 games a year?
There most likely isn't a corollary between all of Ian Kinsler's various maladies, but the fact that he always misses a good chunk of time every year should have been factored into your bid in March even before he got hurt. Yes, you might be letting another owner get a bargain if Kinsler played 150 games, but the odds of that are poor.
Performance/Consistency: CC Sabathia is probably on his way to his fourth consecutive $30+ season (4x4). While he might not earn $30 for a fifth year in a row in 2011, you have to give some consideration to his earning power from 2007-2010 when putting a bid together. Likewise, it probably would have behooved you to knock Zack Greinke's bid down a few bucks from his Cy Young caliber 2009. He hadn't reached that level of performance in the past and while he certainly could have done so again, the odds weren't great that he would.
Playing Time: A player who is going to play every day should get a higher bid limit than one who isn't. This is a general proposition. Don't overpay for a scrub catcher or middle infielder because he's going to get 500 AB. But don't overpay for back-ups. I see this mistake time and time again. Jeremy Hermida seemed to be this winter's darling in terms of paying for performance and not playing time. I know that this advice goes against the mentality of buying ability and not stats, but if you can get a team that is going to get 6,500 to 7,000 AB you're going to win every time.
While I agree with Wang that we can't measure risk well, we do have to attempt to account for it. Not doing so is a Rotisserie recipe for failure.