I agree completely with the "halftime adjustments" crowd, but I would amplify it a bit.
Not just this year but ever since UConn has returned to the top of the charts (i.e., circa 2009), a common pattern is for good teams to play a close game for at least the first quarter and often for the entire first half. Duke often did that. Last year in Florida, Wisconsin Green Bay did that. Even Texas in the Sweet 16 last year did that for the first 12-15 minutes before eventually losing by 50. Dayton was up by 1 point at the half before enduring a second half that was more typical of UConn.
Opposing coaches are often good enough to come up with a game plan that will surprise UConn a bit and will work for a period of time. By contrast, I don't have the sense that Geno tries to fool or ambush opponents with big surprises at the start of the game, although he did seem to do that to Ohio State in this year's opener. (Those of us who attended the open practice before that game could have offered an accurate prophecy if Kevin McGuff had made it worth our while.) This often causes the game to be relatively close for the first quarter and sometimes for part of the second quarter. In some cases, it lasts until halftime. But by then, the opponent's strategy has been completely diagnosed, and the countermoves have been pulled off the shelf to stop it. Opponents don't seem to be able to predict or practice for the adjustments, so they usually lose big in the second half -- more specifically the 3rd quarter before the game descends into garbage time.
If the game remains close into the fourth quarter, that is a game that UConn could easily lose -- for example, last year's Stanford game, the numerous close Notre Dame games a few years ago, and the St. John's loss at Gampel.
An interesting question would be what UConn's record is in games that are within 5 points after 30 minutes. I will bet that it might be under .500.