Alright so let's assume that the SOS is similar (within 50 spots).
The following breakdown has occurred between the two teams:
Team A:
33 games:
7 games (20+ point victories)
12 games (between winning by 20, and losing by 6+)
14 games (between +5 and -5)
This means that they played 14 games which they should on average be around 7-7, which would have placed their record on average at 24-9.
Team B:
32 games:
16 games (20+ point victories)
7 games (between winning by 20, and losing by 6+)
9 games (between -5 and +5)
So Team B has blown out exactly half their opponents, had 7 games which were in a reasonable margin, and 9 games which were close. They went 5-4 in these games (an average record) and if they played a similar schedule to Team A had 9 more games where they blew the opponent out.
Obviously matchups would need to be taken into account, but Team B by most metrics would be a far superior team based on a season long body of work rather than your record in 13 or 14 close games providing a much smaller sample size.