Friday, October 29, 2010

Squinting at a Small Sample

It's hard to analyze big matchups early in the season statistically, because the sample sizes are so small. The teams and their opponents just haven't played enough games against enough varied types of teams to establish definitive trends on why some teams have big numbers and why other teams don't (incidentally, that's one of the reasons why the Pairwise Rankings are useless until mid-January at the earliest).

There's only one thing that's clear - both of these teams have had relatively easy schedules thus far, with most of the easy teams coming at home, where both teams are unbeaten. The four teams Union has beaten have a combined record in other games of 1-8-5 (.250), while RPI's three wins have come over teams with a combined record in other games of 0-6-4 (.200).

Both teams have some gaudy numbers - Union on pretty much everything, RPI on defense and the penalty kill. So is that it? Does Union's offense give them the edge? It might. Let's look at the numbers and see if there are signs of overperformance against the weak schedules both teams have had.

Overperformance?
Offense
Union is tops in the nation in offense. 37 goals in 7 games translates to 5.29 per game. The flip side, of course, is that the teams they've played are among the worst in defense.

* Sacred Heart (played twice, 9 and 7 goals) - 52nd (last) in the nation on defense, with a team GAA of 5.75 in 4 games (half of their games against Union)
* RIT (7 goals) - 48th in the nation on defense, team GAA 4.60 in 5 games
* Alaska-Anchorage (4 goals) and Niagara (6 goals) - tied for 40th in the nation on defense, team GAAs at 4.00, UAA with 6 games, Niagara with 4.

Those are the wins. In the blemishes, they faced Alaska (tied for 3rd on defense with Union at 2.00) and UConn (39th, 3.75). In those games, they scored 1 and 3 goals respectively, their worst offensive outputs of the season.

Defense
It's the same thing for the Engineers. RPI enters tomorrow's game with the 2nd best defense in the nation - 1.67 team GAA in 6 games. But... look at the opponents.

* Bentley - 52nd (last) in the nation on offense, with a team goals per game rate of 1.50 in 4 games (all against ECAC opponents, RPI shut them out)
* Northeastern (2 goals) - 50th in the nation on offense, team GPG 1.80 in 5 games
* RIT (1 goal) - 40th in the nation on offense, team GPG 2.20 in 5 games
* Colorado College (played twice, 2 goals both times) - 37th in the nation on offense, team GPG 2.33 in 6 games
* Niagara (3 goals) - tied for 31st in the nation on offense, team GPG 2.50 in 4 games (half against RPI/Union)

That's the entire season for RPI so far. None of their opponents is scoring 3 goals per game, which is usually a benchmark for a solid scoring club.

Union's just behind RPI in 3rd on defense, but theirs is the same story here as well. We've already touched on RIT (2 goals) and Niagara (1 goal), but look also at Sacred Heart (51st, 1.75), Alaska (43rd, 2.00), UConn (T-31st, 2.50), and Alaska-Anchorage (T-27th with RPI, 2.67).

The bottom line? In a combined 13 games, neither RPI nor Union has faced a team in the top half of the nation in scoring. That changes for RPI tomorrow.

But the question is - are Union and RPI's defenses putting up big numbers because they're playing weak teams, or are their opponents' offensive numbers terrible because they're playing solid defenses? The sample size really isn't big enough at this point to be able to distinguish between the two. The reality is more likely in the middle - both teams have good defenses that are being made to look better by the caliber of the opponents.

Underperformance?
So it appears that Union's offense and defense are probably overperforming, at least to some extent, as is RPI's defense. That's not to say that those elements are not that good, they just may not be as good as they appear on paper, since none of them have been challenged to any major degree this year (or in the case of Union's offense, when challenged by what appears to be solid defense, they shrunk significantly).

But what about the RPI offense? It's still struggling to reach the golden mark of 3.00 GPG. But check out two of the defenses the Engineers have been up against this year, which account for half of the games:

* Northeastern - 2.00 GAA (tied for 3rd with Union)
* Colorado College - 2.17 GAA (9th)

That might explain the 5 goals they scored in those three games to some extent. The offense naturally did better against the 33rd, 40th, and 48th rated defenses at home. So it might be safe to say that the RPI offense is underperforming. Union's strong defense probably won't help that the way the last few opponents did.

Special teams?
Honestly, it's too early to get a feel for whether special teams are over- or underperforming. That's because special teams are a win/lose proposition. One team is going to be successful, while the other team will fail. It's pretty black and white. Thus, the sample size really isn't big enough to get a good look at whether Union's top rated power play (41.2%) is a product of lousy penalty killing (SHU, UAA, and RIT all in the 10 worst penalty kill ratings in the nation) or vice versa. The same goes for RPI's 9th ranked penalty kill (88.1%). The Engineers have faced some lousy power plays (all but RIT rated in the lower half nationally).

Suffice it to say that Union probably has a pretty good power play and RPI probably has a pretty good penalty kill. Especially given the Engineers' proclivity for taking penalties (tops in the ECAC, what else is new), this is going to be a key matchup, with the winner having a huge edge overall. Union faced UConn's 4th rated PK last night, going 1-for-3. They seem to know what they're doing. Is that Ben Barr's doing? We're just going to go ahead and assume so.

Analysis
Neither of these teams are realistically among the top 20 in the nation despite both being ranked this week. They've been skating by on what has been largely inferior teams, and each are facing probably their second most difficult team of the year (RPI having seen CC, Union having played Alaska). Both teams play their games on NHL sized rinks and have a pair of games this season on the Olympic sheet, so the bigger ice is probably a wash. If there's one advantage to be had on the larger surface, it's RPI's speed and up-tempo, WCHA style play. The larger ice surface will make it difficult for Union's defense to gum up the neutral zone - which is one reason why games out west (where there are a lot more Olympic sheets) are more run-and-gun. That could mitigate the offensive advantage that the Dutchmen appear to have.

Choosing a winner in this game is crazy. It's going to be Keith Kincaid vs. Allen York, the split preseason favorites for the Dryden Award, and the Union power play against the RPI penalty kill. Whoever wins those battles, wins the game. Goals are likely to be at a premium throughout, so scoring first is key, and if a team reaches the three-goal mark, they're going to be in pretty good shape.

No comments:

Post a Comment