They are now tied for 1st place with the Yankees yet they have a run differential of -19. Their record is 76-59 and 24-7 in 1-run games. Here is one thing that might be the difference. Their batters have an OPS of .722 while their pitchers have allowed an OPS of .724. That should give them just about a .500 record. But in close and late situations their batters have an OPS of .697 (probably a typical dropoff) but their pitchers allow just a .598 OPS. Maybe that is why they wins so many close games.
In the AL this year the overall batting OPS is .733. In CL situations it is .683. So the Orioles only drop off .035, not the norm of .050. For AL pitchers, the overall OPS allowed is .729. In CL cases it is .674. So the Oriole drop off is .126 while it is just .055 for the league.
In non-CL situations the Orioles have about a .727 OPS and their pitchers allow a .747 OPS. Here is my formula for predicting winning pct. using non-CL OPS and CL OPS (both for batters and what is allowed)
PCT = 0.501 + 0.918*NONCLOPS + 0.345*CLOPS - 0.845*OPPNONCLOPS -
That would give them a .526 Pct. Then they should have 71 wins but they really have 76. So there are still 5 wins not explained.
Baseball Reference gives the Orioles hitters a clutch rating of 2.9, meaning that they tend to hit better the more runners on base, the closer the score and the later the game. So they have won 2.9 more games than you would expect from their hitting than you would normally expect. Pitchers have 6.8. That adds up to 9.7. If they were to win only half their games, it would be 67.5. But if you add 9.8 to that you get 77.3. Very close to their actual win total of 76. They just perform better when it matters more.