Your tipped pass scenario is exactly the type of thing that may allocate credit improperly, but there's not too much I can do.
If I can recall the play when I'm going thru the play-by-play, I would definitely split the point for the steal between Nicholson and Berry.
Much of the time I don't have that in my head - sometimes I do. I do know that UCLA keeps an official stat for "deflections" (and has to be using it).
There are always a few team rebounds every game and an occasional team turnover. Most of the time your could split credit or blame between 2 players if you went and watched a replay. As it is, I'm at the mercy of the recorded play by play.
The negative credit depends on how you think about it. Its all tied to expecting one point per possession. If the opponent comes down and Berry steals the ball, the opponent was supposed to score 1 point, but they got zero. The fact that we now have the ball means we are expecting to take a 1-0 lead (on average) because we have possession. Now we take a shot. Lets say Martinelli scores a 2. The team is ahead 2-0. All 5 guys are "raw" +2. Martinelli gets adjusted +1 for scoring. Berry still gets adjusted +1 for the steal. All 4 guys who aren't Berry get adjusted -0.25 for Berry's steal and all 4 players who aren't Martinelli also get adjusted -0.25 for his solo scoring play. The "Net +/-" is thus Berry +2.75, Martinelli +2.75, Nicholson +1.5, Barnhizer +1.5, Leach +1.5 (that adds up to 10 because all 5 guys are "raw" +2) Martinelli and Berry are receiving extra credit for the 2-0 lead, whereas Barnhizer, Nicholson and Leach are also getting credit, just not as much.
Of course, if Martinelli misses that shot and we lose possession - Berry still keeps his +1, but Martinelli get tagged with a -0.8.
Its a scoreless game and Berry is +1.2, Martinelli is -1.05 and the other 3 are -0.05.
I hope that helps a little - but thats the way the whole thing is built when it comes to "player adjustments." It may not be perfect, but it has some merit.
