GVC
Well-Known Member
Fair. As assumptions/corrections go, players progressing gradually/having a fairly consistent impact from year to year seems reasonable. The large standard error in adjusted +/- renders comparisons across players/positions/teams fairly meaningless; this correction/assumption helps (provided you think it's reasonable).The very notion that these statistics are expected to be consistent year-to-year is subjective. If a statistic is wildly variable in consecutive years, that you can reduce that somewhat by considering multiple years, but adjusting an individual year based on previous years is a subjective influence.
What I like about +/- statistics in general is that, unlike box score stats and the "advanced" stats derived from them, +/- attempts to measure what should be measured: How a player's presence on the floor affects his team's chances of outscoring their opponent/winning the game. Basketball is not baseball, where one player faces another, virtually independent of what the other players on the opposing teams are doing. Basketball is played with 10 players on the court who all affect the outcome of most possessions in some way. With the STATS, inc. cameras now in every arena, we'll soon have much better metrics to measure the absolute and relative value of the play of players without the ball, who are mostly ignored by box score statistics despite representing 90% of the play on the court.
Last edited: