How We Value Players, Part One

Sammy Sosa, $44.

Sometime in February or March, most fantasy participants head down to their corner national bookstore chain and buy a baseball magazine or two (or twelve) with a similar recommended bid. One or two publications will provide projected statistics, but most won't have them and some will only have rankings without any dollar value at all.

You might be tempted to go into your draft armed only with one of these publications, but you do so at your peril. Sosa's price from above might as well read Box o' Kleenex, $44, or Sammy Sosa £3.25 for all the good it will do you. Without knowing the league context, that dollar value has no meaning at all.

Every league is different. Many league constitutions differ dramatically from the original Rotisserie constitution written over twenty years ago. Some online leagues don't even have constitutions; they instead rely on a set of rules designed to please the majority of that site's users.

Each fantasy owner knows his or her league specifications, and these specifications give you the ability to compute accurate player values.

Traditional fantasy baseball employs a 23 player roster using four offensive categories (home runs, runs batted in, stolen bases, and batting average) and four pitching categories (wins, saves, earned run average, and the ratio of walks and hits allowed per inning pitched). Many leagues have added runs and strikeouts for a total of ten categories. You might prefer a league that uses different categories altogether, or a points-based format like the popular challenge games, or even computer-simulated games like Scoresheet baseball.

Regardless of your preference, you likely know that Sosa's $44 price only has relevance for a league with a draft salary cap and relatively low player values (normally under $50 with occasional exceptions like an Alex Rodriguez and Pedro Martinez). His value is representative of the quality of the statistics that you expect him to contribute to your team.

In order to calculate a reasonable bid for Sosa, you must know and understand your league's rules. Twelve team, 23-man rosters are somewhat standard, as is a $260 team draft cap and a minimum bid and incremental increase of $1. In 4x4 leagues, Sosa will likely contribute significantly in home runs and RBI, somewhat in average, and very little in steals.

Two primary ways exist to determine the exact value of this contribution.

Standings Gain Points, employed by such notable rotisserie stalwarts as Alex Patton, Art McGee, and Sandlot Shrink's John Coleman, attempt to measure the contribution of a player using the number of standings' points you can expect him to earn your team in each category. SGP lets you incorporate marginal values, position scarcity, and keeper options into your actual draft prices, but unfortunately the system has two significant weaknesses.

First, it has no way of determining category inflation. If 8 closers are kept in your league, the value of the remaining bullpen studs dramatically increases since saves are now scarcer at the draft table. Without the ability to calculate exactly how scarce, you have no way to adjust your bids to insure your team competes in that category.

Second, and far more damning, SGP relies on two very volatile variables. All value projections lack precision and even the most acceptable and accurate projections will be up to 20% different FOR EACH PLAYER IN EVERY CATEGORY. The discrepancy between projected value and actual value is a problem that every fantasy manager and every general manager has to overcome, but SGP adds a second uncontrollable variable to the calculations.

You can determine SGP denominators for the counting categories by dividing each player's projected stat in each category by the stats expected to move you from last place to next-to-last place. For example, if 5 steals are needed to move up one place in the standings and Ichiro has 56 steals, he earns 11.2 SGP in steals. The calculations are more difficult in the ratio categories, but the results are the same. Each player's total SGP are then divided by the net SGP for the top 276 players in a standard league, and then multiplied by the $3120 total available budget to get his dollar value. A player earning 15 SGP in a league with a total of 1200 SGP would be worth $39 (15 SGP / 1200 SGP * $3120 = $39).

However, the SGP denominators themselves are based on last year's standings and other historical figures. Some established leagues can see some consistency in their SGP, but the yearly numbers that someone like Mr. McGee publishes for everyone to use are as much a projection as another 73 HR would be for Barry Bonds. While they are generated using thousands of leagues and incorporate significant calculations of league trends, adding this second variable adds an unnecessary level of uncertainty to your calculations.

Many people use SGP since it is the only way they know to convert statistics into dollar values. This fallacy is based on the assumption that a standings' point in one category is worth as much as a point in any other category. While true in and of itself, this statement has no relevance in determining draft prices, as the important fact is that all dollar values are equal. A dollar earned in any category is just as valuable as a dollar in any other category, and this statement allows for the calculation of value independent of projected standings. Adding projections of SGP denominators dramatically increases the inherent risk in the final calculated dollar values.

As long as one knows how to draft a well-rounded team and can display patience in the draft room, Category Dollars can be used instead to calculate player values just as accurately as SGP. They also allow you to significantly decrease projection risk since a single projection type remains more internally consistent than multiple projections.

Tomorrow, we will discuss how to calculate Category Dollars and why we believe they are preferable to SGP.

Today's Fantasy Rx: Do you believe that the benefits of SGP outweigh the drawbacks? Have you found a way to modify SGP to correct the flaws? We'll be happy to discuss SGP and any other pricing system further after reviewing Category Dollars, so please write in to tell us how you price your players.