March 30, 2005
After my colleague Jeff Erickson worked out all the projections for our 2005 Fantasy Baseball Guide, he asked me for help turning those numbers into dollar values. After a bit of research and work, I came up with the following system: (1) Figure out what stats constitute replacement value in a given league; (2) Subtract the replacement value stats from a given player's stats to find out how much he exceeds them in each category; (3) Figure out the standard deviation among draftable players for each category; (4) Divide the amount by which a player exceeds replacement value in a category by the standard deviation in that category; (5) Add up the quotients to come up with a player's total unadjusted value; (6) Multiply unadjusted value by a constant to generate dollar value.
We generated the constant by adding the unadjusted values for all drafted hitters in a standard league of x teams, and dividing by (180 * x). (180 is the number of dollars to be spent on hitting. For pitchers, we divided the unadjusted values by (80 * x). (This assumes a standard $260 league in which a typical team spends $180 on hitting, $80 on pitching).
There are two components to the value of a set of player stats: How much better they are than what is freely available on the waiver wire; and what being 20 home runs or 20 steals or 20 strikeouts better than replacement value is worth. We understand intuitively that 20 steals above replacement value is worth more than 20 strikeouts above it, but what about plus-46 points of batting average in 546 at-bats versus plus-17 in saves? That's why the value above replacement has to be divided by the standard deviation for each category: to compare value accurately across categories.
What results did this method of valuation yield? For the hitters, there were few big surprises: In a 12-team 5x5 mixed league, we projected Carlos Beltran at $43, Albert Pujols at $41, Alex Rodriguez at $37 and Vladimir Guerrero at $36. Results for the pitchers were a little surprising, though. Forget that Johan Santana came out to $38 (remember that's not necessarily what you should pay for him, but merely what his projected stats were worth relative to the rest of the league's projected stats) or that Jason Schmidt and Ben Sheets clocked in at $35. What was more surprising is that Eric Gagne was worth just $24, and Francisco Rodriguez just $21. Our fifth-best closer, Mariano Rivera, was worth just $14 (keep in mind that he's projected for a Rivera-like four wins, 44 saves and an ERA of just 2.15).
So why the low values for these stud closers and even lower ones for guys like Troy Percival ($9)? Because in a 12-team mixed 5x5, where each team drafts nine pitchers, there is a lot of starter depth among the 108 pitchers drafted. So much so that closers, while helping in saves, cost you a lot in wins and strikeouts, and barely help you in WHIP and ERA, given their low innings-pitched totals. In fact, at the rate closers go in mixed leagues, you'd be better off buying nine starters, dominating wins and strikeouts and tanking saves. In a theoretically perfect league with 12 informed onwers trying their hardest from start to finish, that's what you ought to do, assuming people want to pay $15-plus for anyone who saves more than 25 games, and $25-plus for the top guys. Because in a league like that, the standings tend to be very close. Even the top teams have weaknesses, and if yours is saves, so be it; you've spent your money on hitting, and your nine starters ensure that you're near the top in wins and strikeouts even though you're last in saves.
But in most leagues this won't work. Why?
Because most (almost all) leagues are less than ideal; people at the bottom give up early on, and many of the owners don't really know or care what they're doing. As a result, there are often two or three smart teams with almost no significant weaknesses. And you won't be able to beat those teams while tanking a category. Which means that you, like everyone else, will have to overpay for saves.
Incidentally, in 12-team AL-only 4x4 leagues, Rivera is worth $33, both because strikeouts aren't an issue, and in a single-league universe, there isn't the depth of quality starters. As a result, there are fewer guys who get wins without giving up a ton of value in ERA and WHIP. Closers are therefore far less of a liability in wins and can actually make a significant contribution in the averaging categories despite their low IP totals.