Which NCAA teams are outperforming offensive expectations?

With the vast majority of college hockey teams at least a third of the way through their regular-season schedules, some statistical trends are beginning to emerge.

As is often the case at all competitive levels of the sport, one of the biggest trends is teams either under- or outperforming preseason projections. We can look at teams, even those we might have expected to be "pretty good," and say their record has far surpassed what even the biggest homer might have predicted.

Take, for example, Dartmouth. Many had them in the top three in the ECAC, but potentially a bit back of favourites Cornell and Quinnipiac. The latter two teams haven't been quite as good as expected, but Dartmouth is the last remaining undefeated team in the country (5-0-1) with a plus-11 goal differential in just six games. UMass Lowell, too, is off to a strong start at 9-3-0 and has already surpassed its win total from the entirety of last season. Meanwhile, other teams are dramatically underperforming expectations, perhaps none more so than Wisconsin, which was No. 10 in the preseason national poll but is currently just 4-8-0. UNH, coming off a 20-win season, is just .500, including an abysmal 1-3-2 run in conference play. Northeastern has only two wins in 11 games (albeit against the second-toughest schedule in the country) after finishing at or above .500 every season since 2013-14.

Another thing that's extremely common in hockey is that when teams are either above or below expectations, one of the biggest culprits you can easily point to is "percentages."

Now that College Hockey News is publishing game-by-game expected-goal numbers based on shot-location data, it's easy to see which teams are punching above their weight beyond just looking at shooting percentage—and then put together visuals that look like this:

Here's where I'll do the acknowledgement that these numbers aren't perfect science or anything, so if you think these stats unfairly malign your favorite team or something, I guess that's your right. But personally, I think they're better than the "no stats based on both shot quantity and location" we had before.

A lot of the teams I mentioned above are in one of the two interesting quadrants: the upper left and lower right. The less-interesting quadrants are not interesting for the simple reason that if you generate a lot of expected goals and then score a lot of goals, that's fine. Same thing if you don't generate a lot of good looks or shot volume and consequently don't score, that's just what should be happening. (I do, however, want to shout out Niagara, which is the closest team to dead-even in terms of scoring as many goals per game as expected goals per game: 3.2 actual goals versus 3.207 expected.)

So let's first take a closer look at the teams in the upper left, who are notably underperforming their xG totals.

Harvard (fourth), Wisconsin (seventh), and UNH (ninth) are all top-10 in the country in terms of generating expected goals. 

There's more to hockey than just getting a lot of looks, of course, but one thing we may be able to point to is the finishing talent for these teams. It would be unfair to say they don't have guys who can put the puck in the net, but as a group, they aren't among the most skilled teams in their respective conferences right now, and certainly that explains some of the underperformance despite how well-coached they all are.

That said, they're all shooting a little north of 7 percent, some of the lowest in the country, and they aren't unskilled, especially not to the extent that you'd expect them to undershoot their xG totals by like 1-plus per game. Air Force, for example, has the biggest difference in the country between xG and actual goals (minus-18.3) and that feels like a trend that will hold up a little longer than what's happening for the Wildcats, Badgers, or Crimson.

Another team you can see in the mix here is Northeastern, which has generated 32.6 expected goals but only scored 25, the latter of which is heavily improved by putting up four in a win over BC this past Saturday. Their schedule has been tough, but their coach, Jerry Keefe, talked on Friday about how much more he needed from his offense. They delivered the next night against arguably the best team in the country, and their first three goals were at 5-on-5, which is certainly a step in the right direction.

There are a few other teams currently in the top 15 in the Pairwise that are actually under-performing their xG that are worth highlighting: Maine (minus-6.8 but 8-2-2 with their only losses coming at Boston College), St. Cloud (minus-6.1 but 9-4-0), Colorado College (minus-4.6 but 9-2-1), and BC (minus-3.7 but 9-3-0, and also No. 1 in the PWR). It's scary to think those teams have a little more to give offensively, especially BC, but there might be something there. Leaving aside the quality of their team goaltending among that group — which for the most part I would call "reliably strong" or even "elite" — it's something to keep an eye on.

Let's move on to the teams that are getting more done offensively than the stats say they should:

The big standout here, obviously, is Dartmouth. Only two teams (Denver and Minnesota) are scoring more goals per game than the Big Green’s 4.0, and they’re tied with Michigan. Here, we must point out that sample size is the issue, as Dartmouth has only played six games so far, tied for the fewest in the nation. Their strength of schedule is middling, but the puck is flying into the net for them—they’re shooting a national-high 17 per cent, and no one else is north of 14. Only Denver (15 goals above expected) and Sacred Heart (12.9) are doing more to outrun their xG total, and they’ve each had eight more games in which to do it.

Dartmouth is nearly doubling its xG total: 24 actual goals against 13.6 expected. We can reasonably expect this number to come down, but again, everyone thought the Big Green would be pretty good this year. Eight different players are shooting at least 22 per cent, and that’s not going to last. What likely will last, to some extent, is their strong goaltending (currently .920 as a team). I think .910 is a reasonable number for them, and with that kind of reliability, they’ll remain competitive. However, they’ve also been outshot by their opponents over the course of the season (minus-21), and while you can call it score effects, they’ve also been out-xG’d by a decent margin (13.6-16.5). Based on per-game expected-goals totals, they should be about .500 instead of undefeated. They need to play better, for sure, but they have the talent and the coaching to do it.

Elsewhere, UMass Lowell doesn’t really have the same problems as Dartmouth; they hold a plus-55 shot advantage in their 12 games and a 31.8-27.6 edge in expected goals. Only 11 teams are scoring more per game than the River Hawks, but given how bad the offence was last season, it’s probably safe to say they shouldn’t be plus-9.2 goals scored above expected. That’s tied for the fourth-highest mark in the nation and probably won’t last. The other issue, too, is that they’re only getting .904 goaltending, which is solid but unspectacular. This leaves a little wiggle room for troubles to creep in, but at the same time, that number is at least sustainably acceptable. They’ve actually allowed more goals than expected. So, while it’s hard to see Lowell keeping up this torrid pace on a number of fronts, all you can really ask of any team that only won eight games last year is to hold a positive edge in expected goals every game, and they’re doing it.

The broader context here is that expected goals becoming—or not becoming—actual goals only matters for the games played so far. Wisconsin and UNH could break out, regress to normal performance, and have strong runs in the next few months, at least positioning themselves well for the postseason. However, their records to date are going to hinder a big push for an NCAA tournament appearance, barring some kind of miracle or, of course, winning their conference tournaments. Likewise, Dartmouth and UMass Lowell have staked themselves to really strong starts that have them well above 50 per cent odds to make the tournament.

Regardless, the future is in their hands.