So here's my look at Harvard...
Big Picture
Notre Dame has made 9 straight tournaments and four straight Elite 8s (will be going for our 5th straight quarterfinals this year). In this stretch, the Irish have faced four Ivy opponents winning all four contests by an average of 5.25 goals and never by less than 3 goals. In short, when ND meats an Ivy in the NCAA tournament it's historically a comfortable victory.
ND is currently a 3.5 goal favorite in Vegas over Harvard. This is the largest margin of any first round game (Loyola is also a 3.5 goal favorite over Albany). For comparisons sake, Drexel is a toss up @ #4 seed Penn, and Cornell is a 2.5 goal underdog to a faltering #7 seed Maryland. The big takeaway here is that everyone with eyeballs thinks the Ivy League sucks. The Ivy League also went a woeful 1-7 against the ACC (the lone victory coming when Cornell [best team in Ivy] defeated Virginia [worst team in ACC]).
Harvard
Harvard is basically the most Ivy team of all the Ivies. We should fear them because ND has shown a penchant for shaky goaltending and inconsistent faceoff play which are the two most common elements in how you lose to an inferior opponent. Conversely, if you watched Harvard go 45 minutes without a goal against Penn, you have to think this is about as good of a draw as ND could've gotten.
Notre Dame finished the season 9th in adjusted offense and 11th in adjusted defense. Harvard finished 21st and 20th. All things considered, Harvard is about 5% points worse per possession than ND... so over a 60 possession game, ND is about 3 goals better on paper. ND also has a markedly better faceoff percentage, and assuming clearing/riding evens out for the game you're looking at ND having probably 5% to 25% more possessions. So do the math, and it looks good.
Moving away from the paper, here are the key traits for Harvard:
1. Generally unathletic in the defensive midfield and on close defense. Sound players, but compared to the monsters you see at Maryland, UNC, and Duke they're just not nearly as physically imposing. This is odd, because they recruit VERY well. There is one exception to this.
2. The exception is Robert Duvnjak who is a first team All-Ivy defender and was the #2 recruit coming out of HS. He's pretty darn good, and might be able to lock down Kavanagh by himself with little help.
3. They basically have one star player in each position group on offense. Devin Dwyer on attack is a feed-first attackman who is pretty darn good, and Peter Schwartz had a fantastic all-around year for a midfielder. If one guy scares me, it's Schwartz getting isolated on a guy he can beat. You better believe Harvard is going to pop in the Army tape and just try to copy their gameplan.
4. They have a decent goalie in Jake Gambitsky, but he isn't a true game changer.
Where is ND vulnerable? If Harvard does hold a possession edge, they don't play particularly fast. They have decent outside shooters and ND hasn't been able to save many outside shots this year. They don't have very many strong dodgers, but Army exposed a vulnerability in ND's defense by inverting against SSDMs. If ND slid to the dodge, then Army force fed the crease or passed to the wing for a step down shot. If ND didn't slide, the player just took a blind shot wrapping around the goal. Both were very effective.
Prediction
Notre Dame hasn't beaten a "good" opponent by more than a goal since Ohio State and Virginia in March. It's tough to get a read on why ND plays so many tight games, but I think it has to do a lot with bad goalie play and inconsistent faceoffs leading to a lack of runs. It's too hard to pull away from a team when they can get every desperation shot they take to go in.
I think ND triumphs because they're the better team, but with the added motivation for Harvard (getting told by everyone that the Ivy league sucks and that they're overrated) plus the penchant for close games I wouldn't be surprised if ND finds themselves on the wrong side of an upset either. I certainly don't think it will be the typical "destroy the overrated Ivy" walkover we've been used to.
So I'll take ND 12 Harvard 11... and I could see the game going anywhere from ND 14 Harvard 7 to Harvard 13 ND 8. I think if you played this game 10 times, ND would win 7-8.