A year ago I saw advanced statistics as nonsense, and no way to analyze a team sport like basketball. However; after a year of experimenting, researching, and learning, I realize advanced statistics can be extremely useful in the evaluation process.
Basketball's version of advanced stats aren't perfect, and still light years away from being as reliable, or as mainstream as baseball's. They can also be extremely misleading, and used incorrectly if not careful.
For most of the season, I've been using a combination of Basketball Value, Basketball Reference, Hoop Data, and 82games to gather information for my posts. Today I discovered another -- and in my opinion, the most accurate -- advanced stats site, Basketball Prospectus. (Technically I didn't just discover Basketball Prospectus, just their advanced stats section.) I don't know how I've gone this long without BP, considering the success of Baseball and Football Prospectus.
The cool thing about BP is the way they explain things. They don't just give you random categories and numbers, they tell you exactly what the categories mean, and how they crunched the numbers.
Here are some new stats I came across while surfing BP, along with some old ones -- explained a lot better. Italics are mine.
Offensive Rating (ORtg): Points scored per 100 possessions, with possessions defined as .96*(FGA + (.44*FTA) - OR + TO). For individuals, estimate of the Offensive Rating of a team made up of the player and four average teammates.
Defensive Rating (DRtg): Points allowed per 100 possessions, with possessions defined as .96*(FGA + (.44*FTA) - OR + TO). For individuals, estimate of the Defensive Rating of a team made up of the player and four average teammates.
Bias: Offensive/Defensive Bias. Measures whether a player is more valuable and by how much on offense (positive numbers) or defense (negative numbers). (I think this breaks down how balanced a player's game is. If the Bias is 0, or close to it, the player's impact on offense or defense is very similar. Correct me if I'm wrong.)
Win%: Player winning percentage. The per-minute component of the WARP rating system, estimating the winning percentage of a team made up of the player and four average teammates.
WARP: Wins Above Replacement Player. Based on performance and playing time, the wins a player has created as compared to a replacement-level player seeing the same minutes.
WP: Wins Produced
WP82: Wins produced per 82 games
WP3K: Wins produced per 3000 minutes
([WP, WP82, and WP3K] reflect the meat of the NBAPET player rating system. The aggregate offensive and defensive points created, saved and lost are combined to calculate the number of wins a player is responsible for. The raw total is WP, or wins produced. WP82 is wins produced prorated for the full 82-game season. However a player's availability is reflected in this total. Thus if a player has missed 50% of his team's games, his WP82 figure will reflect this--it isn't "wins produced per 82 games." WP3K, however, removes all playing time and availability differences. It's simply wins produced per 3000 minutes.)
They have a couple more that I won't mention, because they seem confusing and unperfected. But you can read about them here, if you wish.
Since these stats are brand-spanking new to me, I'll save myself the criticism and avoid analyzing them, for now. I'll simply provide you with a table of the relevant Sixers' numbers, and let you do the analyzing.
I think based on those numbers, my favorite new (to me) advanced statistics are Win% and WARP, because they seem to pass the eye-test. Based on the Win% and WARP Andre Iguodala, Sam Dalembert, and Lou Williams were our best players, and Willie Green, Jodie Meeks, and Thaddeus Young were our worst players, this season. (Green and Meeks were below league average.)
What do you guys think of the numbers? Accurate representation? Useless? Useful sometimes? Useful as long as you factor in the Lou Williams bias, because there's no way he was our third best player?