Making Projections into Values: Rotisserie Scoring
Valuation Issues in Rotisserie Scoring
Assessment of true player value is somewhat difficult in Rotisserie scoring. It is important to note that player value is partially determined by the makeup of the rest of the team. For example, if a fantasy basketball team already has a lot of rebounds from other players, but has only a moderate number of steals, a rebound specialist might be virtually worthless, but a steal specialist could be of enormous value. Because a team scores points based on its ranking in various categories, and some sort of team-wide balance is required, you will need to pay some attention to your team-wide statistical strengths and deficiencies. As a result, although you should assess a single number value to each player, a player will actually have a number value for each category to which they contribute. You should keep these numbers handy while drafting – in this way you can see in which categories you are deficient and try to rectify these issues during the draft.
There is a separate issue in fantasy sports leagues where players contribute in different categories, depending on their classification – namely fantasy hockey and fantasy baseball. In a fantasy hockey league, skaters and goalies contribute in entirely separate statistical categories, while in fantasy baseball pitchers and hitters do the same. This can lead to a big value difference. The clearest example of this is in fantasy hockey, where typically two goaltender positions contribute to four statistical categories, while typically ten skaters contribute in six categories. It is critical to take into account the fact that a good goalie has dramatically more impact than a good skater, because of the fact that each goalie determines about half of four of his team’s categories, while each skater determines about one tenth of six of his team’s categories. When evaluating the relative values of players in different classifications (like skaters and goalies – not like left wings versus defensemen, who contribute to the same set of statistics), I like to modify player value by this multiple. I multiply each player’s value by the number of categories to which that player contributes, divided by the number of players who contribute to that set of categories.
Basic Rotisserie Valuation
The first step in assessing player value in a rotisserie scoring system is to discover the projected statistics for an average player without reference to position (perhaps two waiver wire players for certain fantasy sports, one for each major player group, e.g. a pitcher and a hitter in fantasy baseball, or a skater and a goaltender in fantasy hockey). In doing this, I usually start by separating out all players who are projected to play a certain number of games, minutes or innings – just enough so that you are taking all players who are meaningful for your league. Take an average of each statistic being used in the league across all players in your list. You now have the statistics for your “average” imaginary player.
Next, determine a standard deviation from the average for each category. Basically, for a given category you should measure
the average amount of difference between each player in your list and your average player. Once you have these two pieces of
information for each category – an average and a standard deviation – you are ready to measure the level of contribution of
each player in each category. A given player’s benefit in a given category is the player’s projected statistic in that
category, minus the projection of the average player, divided by the standard deviation. In the case of a category which
is an average, you should multiply a player's simple value by a ratio of that players attmpts per game divided by the average
player's attempts per game. For example, if a goalie's simple value in save percentage is 1.2, and averages 30 shots per game while
the average goalie averages 25 shots a game, then the goalie's modified value in the save category should be 1.2 * 30/25, or 1.44.
A player’s total value is the sum of their values in each category to which they contribute.
An Example of Player Valuation in a Rotisserie League
An example is probably warranted here. Let’s say your list of relevant players to your fantasy basketball league is four players long (obviously this number is too small to include all meaningful players, but this is just an example). Let’s also say that your league uses only two measured categories – points and rebounds. Your players have the following projected points and rebounds over the course of a season: Player A should get 2200 points and 250 rebounds, Player B should get 1900 points and 600 rebounds, Player C should get 1100 points and 400 rebounds and Player D should get 1000 points and 750 rebounds. Taking an average of all players, your average player should get (2200 + 1900 + 1100 + 1000)/4 = 1550 points, and he should get (250 + 600 + 400 + 750)/4 = 500 rebounds. Now that we have our average player (1550 points and 500 rebounds), we find the standard deviation for each category. This is an average of the absolute value of the difference in a given category between each player and the average player. Here, the standard deviation in points is equal to (650 + 350 + 450 + 550)/4 = 500, and the standard deviation in rebounds is equal to (250 + 100 + 100 + 250)/4 = 175. Given that the formula for value of each player for each statistic is (projection – average)/(standard deviation), the players should have the following values: Player A should have a value of 1.3 in points and -1.4 in rebounds for a total of -0.1, Player B should have a value of 0.7 in points and 0.6 in rebounds for a total of 1.3, Player C should have a value of -0.9 in points and -0.6 in rebounds for a total of -1.5, and Player D should have a value of -1.1 in points and 1.4 in rebounds for a total of 0.3. That total value number is the one you should use to rank your players, and you should try to have access to all value numbers during the draft.
Once you have a gross value number for each player, you proceed in the same way that the fantasy point leaguer does. You need to determine differential value – value over waiver wire. First, you separate all players by position, and rank players within their position according to player value. For each player position, you guess at the total number players at that position who will be on some team in your league after the draft. The waiver wire player at each position is your guess of the best player at that position who would be available after the draft has been completed. Each player’s differential value is that player’s gross value minus the value of the waiver wire player at their position. Once you have differential value, you are ready to go.
Next Section - Making Projections into Values: Thoughts on Risk Management