Art McGee is author of the book, "How to Value Players for Rotisserie Baseball." A futures and options trader in Chicago, McGee's fantasy teams have finished in the money in 10 of 11 years. We have adopted his valuation model for the roto values you'll find here at RotoHQ.com. Order his book here.

Q. Is dollar valuation really so important? Isn't it more important to have good forecasts of player performance?

McGee: Good forecasts of player performance are essential, but when most of the other owners come to the draft with equally good forecasts, then having the most accurate dollar values to go with those forecasts can make the difference.

Q: To begin with, how can you take production in these diverse categories -- homers, batting average, wins, and so on -- and combine it all into a single dollar value?

M: The key is, how much will a player's production in each category help you in the final standings? For instance, it may take 8.5 more home runs, on average, to move up one notch in that category in the typical league. In wins, it may take only about 3.5 to advance one spot. If it takes 3.5 wins, then a pitcher whom we might expect to win 14 games will move us up four points in the wins category, all other things being equal. Using a term coined by Alex Patton, we can say that this pitcher contributes four Standings Gain Points in the wins category. We can apply the Standings Gain Point -- or SGP -- concept to each category, and then convert each player's production into a single number, his SGPs.

Q: It is easy to see how SGPs would work in a category like wins, but what about ERA or batting average?

M: It's the same idea: how much will a player's performance change the team total and how much will that move the team in the standings? The calculation gets more complicated, though.

Q: Let's say we have the SGPs we expect each player to produce. Then I guess we look at the total dollars available for the draft, which is $3,120 in a standard 12-team league, and we allocate those dollars to each player according to his SGPs, right?

M: Not quite. Actually, that's a mistake that a lot of people make. You see, if an owner has $260 to spend in his draft, then he really doesn't get to allocate all 260 of those dollars. He knows that he has to spend at least $1 on each roster spot. Since he has 23 roster spots to fill, he really only gets to allocate the remaining $237. So each of the 276 players good enough to be taken in a standard 12-team draft is worth at least $1. The next question is, how do we allocate the remaining $2,844 ($237 x 12)? I call those dollars the marginal dollars.

Q: So then we allocate the 2,844 marginal dollars to each player according to his SGPs?

M: Well, actually no. We don't want to give players credit for all of their SGPs, because we know that, no matter what, we can get a certain number of SGPs if we just spend the minimum $1 per player. If we let every other team fill up its roster, and then we spend $23 to get the best 23 guys left, we would still get some production on our roster. Those are the baseline SGPs that we can expect to get without spending any of our marginal dollars. The question then becomes, how do we allocate our marginal dollars to get the most additional, or what I call marginal, SGPs?

Q: So we need to allocate the 2,844 marginal dollars to each player according to his marginal SGPs?

M: Exactly. That's why I call my method "Marginal SGP Pricing."

Q: But how do you come up with the marginal SGPs for each player?

M: You have to look at how many SGPs a player will contribute above the last available alternative player. Let's take pitchers, for instance. In a standard 12-team league, 108 pitchers are drafted. If I wait until the end of the draft, when 107 pitchers have already been taken, then I can get the 108th- best pitcher for $1. Let's say that I think the 108th-best pitcher in the NL is John Burkett, and I expect Burkett to produce 1.1 SGPs. Then that's my baseline. For all other pitchers, I need to subtract 1.1 from their total SGPs to get their marginal SGPs. If I expect Scott Elarton to produce 6.1 total SGPs, then he's worth 5.0 marginal SGPs.

Q: So, you add up all of these marginal SGPs for the players who will be selected in the draft. And let's say, to keep the arithmetic simple, the total comes to 711. With 2,844 marginal dollars to spend, then each marginal SGP is worth $4, is that right?

M: Exactly. And each player would be worth $1 plus an additional $4 for each of his marginal SGPs. Usually, there are somewhat more than 700 marginal SGPs available, and each one will be worth somewhere between $3 and $4. But that's really just a starting point. From there, we can make adjustments for things like position scarcity and potential value as a keeper for future seasons.