Number Theory: The Florida/New Mexico Paradox
While both the RPI and Friend of the Blog Seth Davis don’t want to consider scoring margin as part of the selection committee’s profile evaluation tools, it’s inarguable that decisive scoring margins are strongly correlated to the overall quality of a team. The best teams in any sport win by larger margins, on average, than lesser teams, and overall scoring margin is a great predictor of future performance. Matchups and luck can have an impact on any individual matchup, but over larger data sets, the team that pounds everyone else is going to win a lot of its forthcoming games.
With that assertion in place, we arrive at one of this season’s most interesting discussion paradoxes. You see, there’s a team this season that’s basically been destroying everyone it plays. And a lot of people don’t believe in them because they can’t seem to win a close game. That’s fine, because those folks have an alternate option, a team that plays a lot of close-ish games and wins almost every single one of those. Except no one seems to believe in that team, either. It has to be one or the other, doesn’t it? Or is this the basketball case of Goldilocks’ middle bowl of porridge, where a team has to be “just right”?
Team A in this debate is Florida, which has shredded its way through the majority of their games this season, both in nonconference play and in a relatively weak SEC. Somehow, the Gators have managed to accrue five losses along the way, felled by a combination of injuries, bad execution on a few plays, and generally bad luck. The result has been a perception that the Gators are “unclutch,” a feeling that’s been amplified by Florida’s late cave-ins in regional finals in each of the last two seasons. Kenny Boynton specializes in questionable choices in these spots, Mike Rosario isn’t all that much better in that respect, and there’s no true go-to guy who you can trust to drag the Gators to the winner’s circle when the chips are down.
Team B is New Mexico, the “unimpressive and lucky” yang to Florida’s “choking bully” yin. The Lobos are a low-curb appeal, egalitarian outfit, comfortably grinding out wins with hard-nosed defense that helps them overcome the occasional scoring drought. They have improved a lot offensively since 7-footer Alex Kirk has come on stronger during league play, and any of their five starters can hurt you on a given night, but they usually plod along, staying a few possessions ahead of the opponent before closing things out. The Lobos are 12-1 in games decided by seven or fewer points or in overtime.
Florida, currently projected as a 3-seed in SI.com’s latest bracket, has an adjusted scoring margin (ASM) of 22.1 per ESPN Bracketologist Joe Lunardi. That’s the second-highest margin in the nation, just behind Indiana, which is widely regarded as a top-level national title threat. New Mexico, currently projected as a 2-seed, has an ASM of 10.2, which is roughly the same as teams like North Carolina and two Mountain West peers, Colorado State and San Diego State. All three of those teams are in the 8-10 seed range in the latest projection. Other advanced metrics systems like Pomeroy and Sagarin have the Lobos in the mid-to-high 20s overall, equating to a 7-seed on merit.
So what should you believe? And who should you believe in? The answer to the latter question very well might be “both.”
The core of this Florida team has been to the Elite Eight in each of the last two seasons and historical data suggests that a fully healthy Florida is a very good bet to make a deep run in this year’s tournament, too. Florida currently is second in Division I in Adjusted Offensive Efficiency and fifth in Defensive Efficiency, per KenPom.com. The table below shows the teams in the last five NCAA tournaments who finished the season ranked in the top 10 in both categories.
Seven of the 10 teams made it to the Elite Eight, with six advancing to the Final Four, and three winning the national title. Only the Farokhmanesh’d Jayhawks in 2010 didn’t make the second weekend:
Florida hasn’t won a game yet this season by single digits, which is being used as evidence that they can’t execute in close games. In actuality, it’s evidence that they mash most teams that they play, a very good trait. Optimists would say they threw away the Arizona game when Boynton was forced to play too many minutes as the primary ballhandler (Scottie Wilbekin had a broken finger and was limited) and a couple other games ended in losses because of bad late decisions (notably Missouri).
The problem with the choke theory is that there appears to be no evidence that a pattern of losing close games has much impact on how a team will do in its next close game. Pomeroy posted a blog entry last week that shows almost no difference in win expectancy between teams that have won or lost their previous close games. Even if that team (ahem, Florida) has lost four or five in a row.
The undercurrent of that is that Florida’s mistakes don’t just surface in tight situations. Boynton and Rosario take questionable shots all the time. The lack of a definitive go-to post scorer isn’t isolated to final sequences, but rather when Patric Young is in the mood and when the guards elect to feed him. The point is these mistakes are being masked most of the time because they’re coming during the course of games in which Florida is crushing someone. You can survive a bad possession when you’re up 20. Isolating a couple of bad ones in key spots this season, or Boynton missing the front end of a key one-and-one late at Arizona, isn’t (in my mind) a fair representation of Florida’s future chances. Based on what they have done so far, they may only have to find a way to win one close game to go deep.
But if you don’t believe in Florida, how can you also not believe in New Mexico, which has shown the skill in closer games that the Gators supposedly lack? New Mexico may end up “overseeded” based on its numbers, but there’s another component that’s not being discussed as much within the overall suppressed ASM: The Lobos are a fairly low variance team, too.
This may be a case where a team’s style is just conducive to these types of games, and they have become very adept at managing them. New Mexico has won games by similar margins somewhat regardless of the level of competition, with the team’s overall improvement seeming to offset the uptick in opponent quality most nights in Mountain West play. It’s also worth noting that they have won their last [five] league games by 10, 9, 10, 11 and 13. The first three of those games were against top-50 opponents. The last two were against shorthanded Wyoming and a sub-150 Nevada team. It’s pretty much what they do, play fairly close for 35 minutes and then drop the closer hammer, opponent quality be damned.
In a one-loss-and-out tournament, you’re always wary of teams with this kind of overall ASM, but they’re not very often in real risk at the end of a close game. The Lobos are not winning by one or two points a night. They’re winning by five and seven and nine, still methodically but more definitively as their offensive ability has expanded. They also get to the free throw line a ton and shoot a really high rate from the stripe, which helps them bridge some in-game shooting slumps and helps them close out the close games they’re in.
Maybe we’ll be fooled. After all, single-elimination events are built for variance to snag teams in either of these molds. It only takes one loss no matter how often or infrequently you play with fire. The overall signs for both teams are good, though. In this season of upsets, you can’t really fully trust anyone, but if you need to pick a couple horses to ride, you can make much worse choices than either of these teams.