by benji on Thu Oct 17, 2013 2:09 pm
It's because of the bell curve in modeling. What happens in reality is teams "give up" so the superior teams will run up the score while most teams that are down at 15-25 wins really could win more games if they tried but there's no point. Injuries have the same effect. Simulations assume neither of these happen. (Actually, that they don't happen unexpectedly.)
You could probably add four wins to every team at 50+ except the Spurs*, two wins to every team 43-50 and then the do the corresponding opposite to every team under 37 wins and every team under 30 wins. Then it would look more like you would expect (for the unreleased teams I just reversed the top teams records to fake a poor record and they're in random order**):
1. Heat: 58-24
2. Bulls: 54-28
3. Nets: 54-28
4. Pacers: 51-31
5. Pistons: 51-31
6. Hawks: 47-35
7. Knicks: 37-45
8. Raptors: 37-45
9. Bucks: 35-47
10. Wizards: 33-49
11. Cavaliers: 31-51
12. Magic: 31-51
13. Bobcats: 28-54
14. Celtics: 24-58
15. Sixers: 18-64
1. Spurs: 60-22
2. Clippers: 58-24
3. Rockets: 58-24
4. Thunder: 55-27
5. Timberwolves: 55-27
6. Grizzlies: 51-31
7. Nuggets: 46-36
8. Mavericks: 46-36
9. Warriors: 36-46
10. Blazers: 36-46
11. Lakers: 31-51
12. Pelicans: 27-55
13. Suns: 24-58
14. Kings: 24-58
15. Jazz: 20-62
*Because they're already almost on the upper bound of what you'd want to ever project at. (Only seven teams in NBA history, EVER, have had scoring margins over 10 points a game. The number over nine is not much larger, and most seasons nine falls at about 61-63 wins.)
**You can actually use the player profiles posted earlier to "guess" at the projections within a game or two. At least they've matched up with the 16 released so far pretty closely.
(As for Live and 2Ks modeling, look what they use for inputs...)