Step 4: Sum each of the category points of each player. The point totals per individual player will vary based on the type of league, number of categories, and how many players are auctioned, but the best players should have a higher positive total than the worst players' negative point totals. In my 4x4 leagues the best players point totals are higher than +4.0, whereas the worst players are below -3.0. As I mentioned previously, the sum total of all category points for the entire hitter and pitcher universe must equal zero.
Also mentioned in Step 3 is that I use two techniques in determining player values. In my experiences both methods yield pretty similar results although sometimes the dollar values can differ up to $3-4 on a player. Typically I will average the results of the two techniques and then proceed to the next step.
Step 5: Determine the amount of money to be spent on hitting versus pitching. Here you can either look at your league's past data, or absent that information you can estimate 65% hitting and 35% pitching. If the league uses a $270 budget, then the 65/35 assumption in a 12-team league yields $2,106 to be spent on hitting and $1,134 on pitching. This means that the average hitter (with 0.0 total points) will be worth 2,106 / 168 = $12.54 and the average pitcher (with 0.0 total points) will be worth 1,134 / 120 = $9.45.
Step 6: Determine the "best" player who you think your league will take at the minimum bid. This is an important step as this determines the pricing of all other players. Rank all players by their points and figure out which point total determines the "best" $1 player. You might want to look at past league data to determine how many players are drafted at a buck; in my leagues there were about 20 hitters and 20 pitchers taken at a dollar.
Step 7: Linearly interpolate all of the other player's dollar values. Let's say that a hitter with total points of -3.0 is the hitter that we determined in Step 6. Since a hitter with 0.0 points is worth $12.54, then the value of one total point in hitting is worth ($12.54 - $1) / 3 = $3.85. That means that a hitter with +3.0 points will be worth about $12.54 + (3.0 x $3.85) = $24.09, and a hitter who has -2.0 points will be worth about $12.54 + (-2.0 x $3.85) = $4.84.
Also very important: it was determined the "best" $1 player would be worth -3.0 points. So if we know that approximately 20* hitters are $1 hitters, then 19 hitters with point totals lower than -3.0 points are actually worth less than $1. But since those hitters must be bid on, you need to 1) adjust your values on those hitters up to $1 and 2) you'll have to adjust all of the non-$1 players a certain percentage so that the totals of all of the hitters equal $2,106.
Step 8: Optimizing the player pool. When first determining all 288 players, you might find out players that were not first picked are actually more valuable - based on their total points - than some in your pool. This is more likely the case with pitchers more than hitters. Be sure to try to optimize who is in the player pool by taking out some players and replacing them with others. Importantly, the players chosen in auctions are usually never optimal so you should be aware of your league's tendencies to bid, for example, on a certain percentage of starters versus relievers.
Review your results. There are two checks you should make to be sure your calculations are correct. The first is that given that all player values are interpolated, make sure the best hitter or pitcher's dollar value looks "right", meaning that the dollar value isn't too high or too low. My NL 4x4 league has Ryan Braun at $46 andat $41, and these at least seem to be in line with what I would expect. Getting values too high for your top players means that you have too few $1 players, and getting values too low would mean you have too many $1 players.
The second is to review the highest point totals in each category. In my NL 4x4 league, using the "normal method" (as discussed in Step 3) the best player in each category - BA, HR, RBI & SB - has 2.74, 3.36, 2.36, and 3.68 points. This also seems to make sense, since the lowest variability around the average hitter are RBIs (highest projected RBI total of a player: 113, versus an average of 56 RBIs) and highest dispersion are for stolen bases (highest projected total: 47, versus an average of 10 SBs)
Next up: Valuing Players - The Intangibles
* This is more of a trail-and-error step. Some of the players that score a little bit better than -3.0 points will also round down to $1, so you'll have more than 20 players priced at $1. You'll have to play with the "target" point number to determine how many players should have a $1 auction value.