If the Celtics keep their current winning percentage they will finish with 32.8 wins.
Rondo's average win shares over his career are 6.27 a year. If Rondo plays half the year you can add 3 wins to that total, if he plays 3/4 of the year he would add around 4 wins to the total.
Using this rough estimate, the Celtics will win around 36 games this year (assuming they play at a similar level than in the first quarter). If 36 wins means winning the division I will be shocked, then again the Atlantic looks to be a historically bad division.
The numbers and the fact they have barely played the Western conference yet suggests they won't win 32 games this year.
Rondo's win shares take away Crawfords win shares, who's currently at 1.8 (just under 2).
It's basically impossible to predict how many wins Rondo will add because of his recovery schedule as to when he'll be 100%, then how much the team must adjust to his game, then how much it will effect us with Crawfords scoring out of the game...
The reality is with our stats, over an 80 game season, eventually our bad team is going to have a bad results orientated season, compared to the rest of the NBA- because we are consistently bad- compared to the rest of the NBA- in almost all areas except defense.