The Preseason Top 25 and the 16 Seed’s Burden

By Sean McAlevey

THE MINOR LEAGUE BASEBALL FIELD WHERE THE LOONEY TOON SPACESHIP FIRST LANDS, USA, Earth — With the Final Four coming to a close tonight as Kentucky squares off against Connecticut in Arlington, Texas, two quick thoughts on the NCAA tournament.

1. The Preseason Top 25 is an incredibly useful predictive tool for March Madness, despite being nearly five months removed.

Louisville, Kentucky, Duke, and Michigan State were all ranked in the preseason top five.

Louisville, Kentucky, Duke, and Michigan State were all ranked in the preseason top five.

Kentucky and Connecticut ranked #1 and #18, respectively, in October’s AP Preseason Poll. The Wildcats entered the 2013-14 season with POY candidate Julius Randle, a projected 2014 NBA lottery pick, and a slew of other five-star studs like Mr. Clutch (aka Aaron Harrison), his twin brother Andrew, and fellow freshman Marcus Lee. It would have been a travesty for Kentucky not to have been the preseason number one. A wager on Kentucky in October would have returned 7/2. Now consider that the Wildcats are one win away from winning it all, and they’re currently 2.5-point favorites to do so. Bad decision to leave the money in your pockets, right? Not nearly as bad as missing out on Connecticut.

The Huskies began the season as the boilerplate savvy veteran bunch led by a young, optimistic head coach – a lot of potential alongside a lot of question marks. Still reeling from the previous season’s NCAA sanctions, Connecticut was a far cry from a hot ticket to win it all: a wager on the Huskies in October would have returned 75-1! The best assets head coach Kevin Ollie had were his two veteran guards: star senior Shabazz Napier, a holdover from the 2011 national title team, and junior Ryan Boatright, Robin to Napier’s Batman. In hindsight, given the performance Napier and Boatright have put on in the tournament, Connecticut came into the season with arguably the best backcourt combo in the country.

These two teams were always contenders, and they are now more than ever. Who knows why they fell asleep at the wheel during conference play? On paper they’re elite programs: combined, they feature at least six future NBA players. But they were both undervalued by the selection committee as a result of their sub-par records in the latter half of the season. They both owned sterling pre-conference records but then stumbled through conference play, going 12-6 each in the two weakest “power” conferences, the SEC and inaugural AAC. Although Kentucky and Connecticut made deep runs in their conference tournaments, the selection committee viewed the Wildcats as a band of “show me the money” prima donnas and the Huskies as a group of question marks. Hence their low seeds.

But there’s good news: Kentucky and Connecticut are more or less the same teams as the ones ranked #1 and #18 in October. And if you had applied that knowledge at the beginning of the NCAA tournament, taking advantage of the vast disparities between their preseason rankings and tournament seeds, you wouldn’t be reading this because you’d be too busy picking out the water slide for your new yacht. “But what about all the valuable experience they gained along the way?” “What about their bond as brothers in arms after the season they’ve gone through?” “What about” – Spare me. Here’s the bottom line: Kentucky and Connecticut still have all those same four- and five-star players on their roster. They still have the same coaches. They might not have played up to expectations during the regular season, but they’re still immensely talented – and if there’s one forum for unrealized potential to rise to the surface, it’s the NCAA tournament.

2. I’m not sure we’ll ever see a 16-over-1 upset.

With three 15-over-2 upsets in the last three years, and seven in tournament history (7 out of 116 matchups, 6%), you would think there would have been at least one 16-over-1 upset by now, right? Well, yes, but it’s not so cut and dried. First, the history. There have been 116 1 vs. 16 matchups since the tournament expanded to 64 teams in 1985, and never has a 16 seed pulled off the upset. Fifteen of those losses were by single digits. Two were decided by one point; one was decided by two; and one went to overtime. In all fairness, we should have had one 16-over-1 upset by now. If Alonzo Mourning doesn’t unleash his inner NBA Hall of Fame ‘Zo in the closing seconds against Princeton in 1989, blocking the final two attempts by the Tigers to lift Georgetown, 50-49, we’ve got our first 16-over-1 upset. This year, in fact, Weber State (21-point underdogs) “almost” upset Arizona, falling by only nine. We’ve been that close so many times, but nay, we’ve never actually seen it happen… Since the basketball gods have smitten us for our apparently excessive hubris – how dare we expect such a holy event to occur – who knows if we’ll ever see the mythical 16-over-1 upset?

Now the “science.” After the selection committee has established the 68 teams that will make the tournament, those 68 teams are then ranked first to last. (There will be changes made to this master list in order to accommodate the NCAA’s restrictions, such as those against inter-conference matchups and regional advantages, but for the most part it is essentially a power ranking of the top 68.) The four 1 seeds are roughly the best four teams in the country, and the four surviving 16 seeds (two will be knocked out in the first round) typically rank in the range of 100th to 250th nationally (out of 351 D1 teams). On the other hand, 2 seeds have a lower and much wider range. For example, a 2 seed might be one of the five best teams in the nation or it might be 25th-best. Moreover, 15 seeds are also much more capable than 16 seeds: instead of a range of 100th-250th, a 15 seed’s range is approximately 60th-150th, with a mean somewhere around 90th-100th. The point is that 1 seeds are significantly better (“significantly” in a statistical sense) than 2 seeds, while 16 seeds are significantly worse than 15 seeds. Take a look at the graph below:

The above graph shows the average expected tournament wins per seed. 1 seeds win significantly more games than 2 seeds, and although 15 seeds don't win much more than 16 seeds, at least they win at all!

Average expected tournament wins per seed.

The graph “snakes” downwards from left to right, flattening out around the 7-12 seed range, increasing exponentially as you approach highest seeds, and dwindling down to nothing as you near the 16 seeds. This is exactly what you would expect to see if 68 teams were ranked from best to worst and then matched up in the NCAA tournament – bravo, selection committee (it’s really not that hard to do, but we can give them a little credit). The salient point to take away, however, is that 1 seeds have significantly higher expected win averages than 2 seeds (0.91 wins, almost a full game). And although 15 seeds don’t win much more than 16 seeds, at least they win some of the time.

I personally haven’t even watched a 1 vs. 16 matchup since 8th grade. I USED TO BELIEVE. Now I know better. A game where the outcome is known beforehand doesn’t interest me in the slightest; I’m just watching actors go through the motions. It’s a waste of time – which is unfortunate, because a 16 seed upsetting a 1 seed would make for great television. Instead it’s just an exercise in futility. I can watch people walk to their cars, get in them, and drive off successfully all day – it’s just not that interesting. Neither is watching Florida’s Scottie Wilbekin and Patric Young run a fast break against a vastly inferior U of Albany team. When I fill out my brackets, I write the 16 seeds off instantly, immediately penciling in all four 1 seeds in my bracket without the slightest thought, no matter how good people tell me this year’s Sam Houston State squad is. But when I pick the 2 vs. 15 games, I actually check to see if there’s any dangerous 15 seed that might be worth a look against a 2 seed that isn’t going anywhere.

I know one day I and every other sane bracket-filling being out there will be wrong when a Alabam A&M upsets Georgetown. (The funny thing is that it actually will be Georgetown that blows it first, because LOLZ GEORGETOWN SUCKS IN THE NCAA TOURNAMENT.) And I know that I, just as much as the next guy, would love to witness that glorious 16-over-1 upset – re-feel the thrill of 11-seed George Mason upsetting Connecticut in the Elite Eight in 2006 and then going toe-to-toe with Florida in the Final Four; the thrill of Ali Farokhmanesh drilling the nail-in-the-coffin three to give 8-seed Northern Iowa the upset over top-seeded Kansas in 2010; the thrill of CJ McCollum and 15-seed Lehigh upsetting 2-seed Duke in 2012. But I won’t get to. And neither will most people. Not because it won’t happen, but because we won’t be watching.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s