1. Players are based on TrueWARP, which is an attempt to measure the current actual value of each player's performance at this point of their career. The "WARP" part of the moniker comes from wins above replacement, while the "True" was added because you have to call it something. Henry Abbott was not consulted during the process of naming this metric, but hopefully he won't mind.
2. TrueWARP is a blend of a player's preseason baseline forecast and his 2013-14 results. The degree to which his season winning percentage has regressed against his forecast depends on his experience. Projections for younger players, especially rookies, have a much higher error bar than those for veteran players. Thus, the younger a player is, the more weight his 2013-14 results carry.
3. WARP is a calculation based on winning percentage, or per-possession efficiency, and playing time. For the playing time component, I've used actual minutes per game for this season to capture the size of role each player is holding down. Each player's TrueWARP is calculated on a per-82-game basis to filter out durability issues. We're strictly looking at a player's quality of on-court performance, not how often he's able to go out and apply it. Obviously in the real world, the durability issues regarding players like Derrick Rose, Kobe Bryant and Brook Lopez have to be carefully considered.
4. Grouping players by position is always an inexact science, but I focused on how players have been deployed this season. The stats website 82games.com uses play-by-play data to track the percentage of time a player is used at each position, and I've used that as the basis for my grouping. Position classification is nebulous, but think of it like this: Players occupy a certain space on the floor, and the space they are assigned is usually determined by the players they share the court with. So Carmelo Anthony may be a natural 3, but the majority of the time the Knicks put him in lineups in which he is a 4.