Thanks to 82games.com, I was able to dig into some random overlooked statistics that could help ease the pain that's been overshadowing the 2005-06 season. Areas I'd like to cover include the relationships between the combinations of players on the floor and our win/loss ratio, how each player generally performs when put on the floor with certain teammates, and other in-depth features that could help analyze what's gone wrong with this season. Please note that although I'm not an official NBA sabrmetrician, I wouldn't mind getting into that field. This is my very first attempt at anything remotely close to this, so bear with me. I'd take opinions, pointers, or other input and analyzation. What I'm doing: Taking a deeper look into the different lineup combinations that Bob Hill has put on the floor since being named coach. Ever wonder how well Luke Ridnour and Ray Allen do as a pair, compared to Luke Ridnour and Rashard Lewis; and how both combinations do against their opponents? Well here's your chance: Section 1: <font face="verdana"><font size="-1">Top Five-Man Floor Units</font></font> UNITS CONTAINING PG LUKE RIDNOUR <font size="1">Other unit sets, including SG Ray Allen, SF Rashard Lewis, PF Nick Collison, and C Robert Swift to come...</font> This is based on a total of 20 total floor units. Each pair of players below is based on the average per unit containing those two players. <font size="1">KEY: </font><font size="1">MPU: Minutes per unit; PPU: Points per unit; PAPU: Points allowed per unit </font> Ridnour-Allen: 11/20 units<font size="1">*</font> Total Minutes: 990 Total Pts: 2189 Total Def: 2167 Pts/48: 106.13 Def/48: 105.06 Wins: 41 Losses: 79 Win%: .342 <font size="1">*This means that, in 20 total lineup combinations that Bob Hill could put out on the floor, 11 of those 20 will include both Ridnour and Allen. The other three players are irrelevant.</font> Ridnour-Lewis: 15/20 units Total Minutes: 1092 Total Pts: 2406 Total Def: 2481 Pts/48: 105.76 Def/48: 109.05 Wins: 66 Losses: 97 Win%: .405 Ridnour-Collison: 7/20 units Total Minutes: 650 Total Pts: 1435 Total Def: 2481 Wins: 41 Losses: 50 Win%: .451 Ridnour-Swift: 5/20 units Total Minutes: 317 Total Pts: 694 Total Def: 733 Wins: 14 Losses: 31 Win%: .311 ----------------- If you need any more help comprehending this (or why I did it in the first place), go ahead and ask away.
I assume you are looking at the other pairs without Luke next? Just looking at one player's pairs on a losing team (particularly one forced to play a lot more minutes than desired) doesn't tell us too much. For me the value of player pair data is in looking at the whole data set over multiple years. That said, I think the effort is worthwhile...just don't stop with a single year and focus on one player. Look at the player with the same pairs from previous year's data and look at the pairs for this year with players OTHER than Luke.
You nailed the exact reason why I scratched this idea. That said, expect some more interesting stuff like this to come.
