Maryland has slowly but surely separated themselves from the ACC pack, jockeying for position with just Duke and Virginia Tech in the top tier. But no one plays for second or third place - you play, to paraphrase Herman Edwards, to win the league. But what exactly is the probability that they pull the unlikely and win the ACC? I got intrigued and starting punching numbers.
The result? Not great. For reference, here's the ACC standings to this point, including Duke's recent win over Virginia Tech:
Atlantic Coast Conference Standings
(updated 2.21.2010 at 2:26 AM EST)
|Duke Blue Devils||11||2||22||4|
|Virginia Tech Hokies||8||4||21||4|
|Wake Forest Demon Deacons||8||5||18||7|
|Florida St. Seminoles||7||5||19||7|
|Georgia Tech Yellow Jackets||6||7||18||9|
|Boston College Eagles||4||8||13||13|
|North Carolina Tar Heels||3||9||14||13|
|N.C. State Wolfpack||3||10||15||13|
As you can see, Maryland's in a bit of a tight spot, but they're the only legitimate challenge for Duke's ACC reign. For the record, the remaining ACC schedules for the team are below
|2-24 - Clemson|
|2-27 - @ Virginia Tech||2-28 - @ Virginia|
|3-3 - Duke||3-3 - @ Maryland|
|3-6 - @ Virginia||3-6 - UNC|
Duke only has three more games, although they do have an out-of-conference matchup with Tulsa. Maryland has four games, two of which will be played in Virginia.
Maryland has two possibilities to finish #1 in the ACC: having a better record than Duke straight up or tying Duke and winning the tiebreaker (duh). For option one, let's just assume that they need to beat Duke to make that happen (techincally they don't, but then it gets really confusing for probabilty and would be pretty negligible statistically). Assuming that happens, Maryland would also need to win out, and the Blue Devils would have to fall in at least one other ACC game.
Using the KenPom predictor for percentages - far from entirely accurate, but hey, it's what we got - the probability that Duke would lose at least one of their remaining two non-Maryland games is about 18%. Ok, not likely but not impossible. The probability that Maryland would win out, including the Duke game, is a less likely 9.6%.
Combined, that means there's - statistically speaking - a 1.73% of this actually occuring. So...you're saying there's a chance? Well, there's a better shot than option #2, believe it or not.
The only way that Maryland would tie Duke and win the tiebreaker would be to have a better record than Duke for the #2 ACC team. If those are the same, which they would likely be, it would go on until there was a difference in record, meaning Georgia Tech (whom Duke lost to) would need to finish above Clemson and Wake Forest in the standings. I'll be nice and only look at finishing above of Wake Forest, since they're higher up right now anyway.
For this one, Wake Forest would have to lose out, Georgia Tech would have to win out, and Maryland would still need to win out. The probability of all of those things happening, according to KenPom, is a measly - prepare for this - .137%. Not 1.37, not 13.7 - .137%. And no, I didn't forget to multiply.
So that means that the probability that either one of those would happen, thus giving Maryland the regular season crown, is 1.89%. And that's optimistic, because we're just giving Clemson a poor finish.
And yet, it still seems oddly possible. Sure, Duke would beat UNC 95% of the time if it was played in a bubble, but stats don't take a rivalry into account. The second one probably can't happen, but it only takes one game for Duke to lose. And that's a good thing, considering they only have two more.
I guess, then, Maryland should be happy the game isn't played on paper.