This past weekend was big: 16 teams threw hands and eight managed to avoid considering a future as a professional wrestler. This weekend is even bigger: Those that remain will chart a course through either Sports Authority Field at Mile High or Navy-Marine Corps Memorial Stadium to earn an invitation to Championship Weekend at Lincoln Financial Field. This is one of the few times in history that people have actively attempted to spend and extended period of time in Philadelphia.
I turned on the genius lacrosse machine to set the scene for the NCAA Tournament quarterfinals. Here's how things shake out in various ways.
The Computing Machine Favors . . .
There is a twofold problem with publishing win probabilities: (1) Recency bias and confirmation bias are almost impossible to avoid when it comes to internet feelings ("The defense submits 'Towson-Notre Dame' into evidence"); and (2) The stronger team doesn't always win in a single elimination tournament (or, alternatively, win by as much as expected). The former is associated with insane people being insane; the latter is a fact of reality that can be difficult for non-insane people to systematically understand. What follows are probabilities, not concrete assertions of absolute truth. Don't send me a severed head through the United States Postal Service because the probabilities did not align with your particular interpretation of an eventual outcome.
I assembled three sets of probabilities here: Brian Coughlin ran probabilities for Inside Lacrosse based on an efficiency model (the strongest of the three models in my opinion); the Massey Ratings probabilities based on the site's prediction model; and a probability determined by running the Massey Ratings' power ratings through a Bradley-Terry model. There aren't drastic differences in the three models, although the Massey Ratings' matchup probabilities maintain odd outliers:
|GAME||COUGHLIN||MASSEY (POWER)||MASSEY (MATCHUP)|
|Albany // (1) Notre Dame||46.80% // 53.20%||48.13% // 51.87%||45% // 55% (13-14)|
|Ohio State // (4) Denver||31.50% // 68.50%||40.35% // 59.65%||22% // 78% (9-12)|
|(6) Maryland // (3) North Carolina||41.90% // 58.10%||44.79% // 55.21%||34% // 66% (9-11)|
|Johns Hopkins // (2) Syracuse||37.30% // 62.70%||44.89% // 55.11%||33% // 67% (12-14)|
Some brief notes:
- Albany-Notre Dame is pretty much a tossup game across the board. The opening game of the quarterfinal round has earned the anticipation surrounding it, and a replication of the two teams' incredible quarterfinal meeting in 2014 is not out of the question. Maryland-North Carolina and Johns Hopkins-Syracuse slot in behind Danes-Irish as the tightest games of the weekend while Ohio State-Denver is projected as a fairly imbalanced tilt.
- When the Massey Ratings' matchup model likes a favorite it really likes a favorite. Proceed with caution.
- I considered running probabilities using KRACH ratings, but the ratings were yielding hyper sensitive results. Pertinently, no favorite was given lower than an 81 percent chance of victory against its opponent, and considering that KRACH is based on the Bradley-Terry model, that felt impossibly off-base. (I also ran probabilities based on the site's adjusted winning percentages and the results were similarly eccentric.) So, that's a no go (but it is another way to consider the probabilities for the weekend's games).
The Computing Machine Likes . . .
The weekend is tricky: It's not miserable outside and you should get some sun. This may lead to circumstances in which you can't devote eight hours of your precious time away from your office desk to watching lacrosse, thus requiring you to pick only a few games to actually witness. To maximize your enjoyment of lacrosse games, I came up with something called the "Fun Factor": This measures a game based on overall competitiveness, the quality of the teams, and if the teams play a style conducive to making your eyes bulge out your head like a Muppet. Unsurprisingly, Albany-Notre Dame and Johns Hopkins-Syracuse -- the tensest games on the board and the ones with the highest potential for pace and offense -- lead the way:
|RANK||GAME||FUN FACTOR||RANK||GAME||FUN FACTOR|
|1.||Albany v. (1) Notre Dame||7.78||3.||(6) Maryland v. (3) North Carolina||4.82|
|2.||Johns Hopkins v. (2) Syracuse||5.83||4.||Ohio State v. (4) Denver||4.58|
Some brief notes:
- Albany-Notre Dame not only has the potential to be the game of the tournament, but it could also be the game of the year. Anything at or above a 5.00 is rock solid and when the value approaches 7.00, things start to look like a fireworks show at a nuclear power plant. The hype for Danes-Irish is legitimate, and even if the game fails to meet expectations, the evidence is there that Albany-Notre Dame could impact an audience that is relatively new to the game (or only occasional puts lacrosse on the tube). This is a defiantly must-watch situation.
- Maryland-Notre Dame is the only seeded-seeded matchup this weekend, but it doesn't have the kind of juice that Albany-Notre Dame and Johns Hopkins-Syracuse are running with. This has little to do with the quality of the two teams and more to do with Maryland's profile: Both the Terps and Tar Heels are among the best half-dozen or so teams in the nation, but Maryland plays at a notably reduced pace and the team has struggled to find consistent offense. Maryland-North Carolina hasn't earned a bad score -- it's stronger than three games from last weekend -- but it's also not bringing the kind of heat associated with a seeded-seeded game.
- If there was an "Unbearable Fan" aspect to the "Fun Factor," Johns Hopkins-Syracuse would destroy the metric. Fortunately, fans of the Jays and Orange at least have a leg to stand on this weekend as they echo their familiar refrain: "LOOK AT US, WE'RE IMPORTANT!" Hopkins-'Cuse isn't the game this weekend, but it’s clearly among the best.