The Hawks won 47 games last year by beating up on worse teams, more specifically, the Hawks won 47 games by beating up on worse teams in the Eastern Conference. They won 2 of 12 games against Boston, Cleveland, and Orlando. They won 8 of 18 against the nine Western Conference teams with 46 or more wins. That 8-10 record against the good Western Conference teams isn't bad in and of itself, but it only brings the Hawks' record against equal or better opposition to 10-20. That's essentially a league-average record against those 12 opponents. The rest of the league (cumulatively) would have managed 9.3 wins in the 30 games the Hawks played against those opponents.
The Hawks won 9 of 12 against the dregs of the Western Conference (Golden State, Minnesota, Memphis, Oklahoma City, the Clippers, and Sacramento) but that .750 winning percentage only marginally bettered the rest of the league's .723 winning percentage against those six teams. Collectively, against the Eastern Conference's three elite teams and the entirety of the Western Conference, the Hawks were one game better than average.
The other five extra wins were earned by virtue of an extraordinary 28-12 (.700) record against teams five through fifteen in the Eastern Conference. The rest of the league won just 56.5% of games* against Miami, Chicago, Philadelphia, Detroit, Indiana, Charlotte, Toronto, New York, New Jersey, Milwaukee, and Washington.
*the league's record against those teams when converted to 40 games: 22.6-17.4
So that's where the Hawks got wins 42 through 47 last season. The question remains: Does this breakdown by opponent have any bearing on predicting the number of wins they'll amass in 2010? Frankly, I'm not sure. It seems unlikely to me that the Hawks will go 2-10 again* against Boston, Cleveland, and Orlando and should better last year's pace** in those 12 games regardless of those teams making themselves better or worse in the off-season. It also seems unlikely to me that an essentially unchanged*** Hawks team can again be seven games better than average**** against the other 26 teams.
*For the optimists out there, the Hawks' pythagorean record in those 12 games last season rounds up to 4-8.
**Bear in mind that should those three teams cumulatively average 60 or more wins again next year, that third win out of twelve against them would just get the Hawks to the league average. They'd have to win at least four games to get ahead of the rest of the league on that sliver of the schedule.
***Even if you count the swap of Flip Murray for Jamal Crawford as a change, the Hawks return players accounting for 83.6% of last year's minutes played.
****For the pessimists out there, the Hawks' pythagorean record in the other 70 games (actual record: 45-25) was 42-28. They may be due a similar downward correction against the rest of the league as they are an improvement against Boston, Cleveland, and Orlando.
In the on-going pursuit of solid reasoning on which to base an accurate prediction of the Hawks' 2010 record, I still see little reason to make a guess that varies more than a couple of wins in either direction from last season's total.