MARCH MADNESS LIVE: Scores, updates from Thursday's NCAA Tournament games
Everyone looks for that little edge in picking their NCAA Tournament bracket. Here's a new angle, which yields some intriguing insights.
A common thought process in basketball analytics identifies four key elements that correlate most closely with winning games. Former Sacramento Kings director of analytics Dean Oliver coined these the Four Factors of Basketball Success — shooting, turnovers, rebounding and free throw proficiency.
Each of these factors hold varying weights, and we can apply them to determine a team’s likelihood of success, in this case within the NCAA Tournament. Using a model that measures these factors — eight total stats when accounting for both a team's offense and defense — we’ve unofficially created a Four Factors Score and an adjusted ranking for each team in the Field of 68 to help pick the bracket.
You can scroll down to Key Takeaways for some bracket tips and the full rankings, or continue reading for a nerdy explanation of how this system works.
Here are the stats at play:
Offensive and Defensive Effective Field Goal Percentage (weighted 40 percent) — A shooting percentage which adjusts for a 3-pointer being worth more than a 2-pointer.
Offensive and Defensive Turnover Percentage (weighted 25 percent) — An estimate of turnovers per 100 possessions.
Offensive and Defensive Rebound Percentage (weighted 20 percent) — An estimate of the percentage of available rebounds a team grabs.
Offensive and Defensive Free Throws Made per Field Goal Attempt (weighted 15 percent) — How often a team gets to the free throw line and how often it converts (aka Free Throw Rate).
Taking every school's national ranking out of 351 in each category and reversing it (top team gets 351 points), then applying the weighted rate, gives you what we'll conveniently call the Four Factors Score, or 4FS.
When looking at each team’s national ranking in 4FS, it stacks up relatively well with other commonly used ranking systems. Here, 44 of the top 68 teams (66 percent) in Four Factor Score are in the tournament. That’s compared to 70 percent of the top 68 in KenPom rankings, 70 percent of ESPN’s Basketball Power Index and 76 percent of RPI. South Dakota and Old Dominion are the only teams to rank top-20 nationally in 4FS and not get a tournament bid.
Before utilizing this to pick the bracket, we need to install some basic logic. To counterbalance differing levels of conference competition and lend some credence to the committee’s evaluation of each team and its resume, we’ll adjust the 4FS ranking a bit based on tournament seeding. Each team gets docked four 4FS points per its number seed. This still allows for a fair share of upsets but maintains some level of sanity in the earlier rounds, and prevents situations like Murray State reaching the title game (the later rounds are where these variances will be less of a factor, the teams will be more evenly matched, and this system could pay off nicely).
A simple example of how that 4-point adjustment plays out (WARNING: math ahead): Despite No. 12 Murray State's 28-point margin over No. 4 Wichita State in 4FS score, this model would have Wichita State advancing, thanks to an eight-seed difference between them (8 times 4 equals a 32-point boost for WSU, giving them a narrow 4-point edge). However, the adjustment is not enough to counteract No. 11 Loyola (Ill.)'s 29-point 4FS margin over No. 6 Miami (5 times 4 equals a 20-point boost, which falls short of 29). The Ramblers advance.
Now we re-rank the field with this adjusted 4FS and advance teams based on who has the higher score (or higher adjusted ranking) within their matchup. First read the most relevant takeaways, then check out the bracket results using this algorithm. And if you're still with us by then, scroll all the way down for the full team rankings.
Key NCAA bracket takeaways
— Cincinnati is good. Really good. The committee was justified in granting it that two seed, as the Bearcats lead the entire field in both raw Four Factors Score and adjusted 4FS rank. They rate highly across all categories, most notably with lockdown defense and offensive rebounds. It’s sure to turn some heads, but this model has Cincy taking home the title.
— As many suspected, Kansas is a highly beatable No. 1 seed. The Jayhawks seem primed for an early exit yet again, ranking 34th in 4FS, doomed by virtually zero production from the free throw line and poor rebounding.
— The major-conference team grading out as the most underseeded is Florida State, with an adjusted 4FS that has it ranked 17 spots higher than the committee did. Thanks to a strong score in every offensive category, the Seminoles are in solid position to take down No. 1 Xavier in the second round (the Muskateers rate as an overseeded team here). This system doesn’t quite get FSU over that 9-seed hump, but you might want to provide that push.
— The top seed that appears to be most overrated by the committee is the defending champion, North Carolina. The second-seeded Tar Heels rank 62nd nationally in 4FS, dragged down by an awful free throw rate and an inability to create turnovers. Don’t be afraid to boot them in the early rounds.
— The most glaring first-round upset is No. 12 Murray State over No. 5 West Virginia. MSU ranks third nationally in 4FS and is by far the most underseeded team according to this model. On the other hand, WVU is tied as the third-most overseeded team, ranking 82nd nationally in 4FS. Unfortunately for Murray State, Round 2 could bring that aforementioned matchup with another strongly rated school in Wichita State, potentially thwarting a long Cinderella run.
— Four Factors method favors the 11 and 12 seeds in this field and suggests several upsets are brewing. Loyola (Ill.), New Mexico State and San Diego State all have a top-20 4FS ranking, while play-in game participants St. Bonaventure and
Arizona State have potential to do some damage vs. overrated opponents.
— Other strong teams that get a notable boost are Ohio State, Houston, Nevada and Montana. They run into other strong 4FS teams early on, however, to limit the damage in this bracket, but you’re on the right track if you pick any of them to make noise.
— Other teams that appear to be overseeded? See Tennessee, Kentucky, Clemson, Florida and TCU for a March letdown.
— This exercise fuels the suspicion that having a stud NBA prospect buys a bubble team some good will from the committee. Alabama and Oklahoma have the twoworst raw Four Factors Scores among major-conference teams in the tournament, and can probably give sole credit to Collin Sexton and Trae Young for their respective bids.
March Madness bracket picks
Finally, here's how all this applies to an actual bracket. Stick to it like glue, or just use it to tip the scales on some matchups that are leaving you torn.
After things play out a bit chalky in the first couple rounds, not a single No. 1 seed advances to the Final Four. Sorry to the Hoos and their fourth-ranked Four Factors Score, but the Bearcats stand in the way. The two’s get much love with Purdue and Duke joining Cincinnati and four-seeded Gonzaga in the semis (all teams ranking top-eight in raw 4FS), with Cinci taking down the Boilermakers in a ratings-nightmare title game.
Final Four picks, national champion
Creighton 77 -13
Here is the full download of info, with the teams listed in order of adjusted 4FS rank. The "seed differential" is the variation between the committee's 1-to-68 seeding and that team's adjusted 4FS rank.
|Team||4FS Rank||Seed Diff.|
|San Diego St||14||27|
|New Mexico St||20||25|
|South Dakota St||46||7|
|North Carolina St||74||-9|
|Stephen F. Austin||68||7|
|College of Charleston||81||0|
|Cal St. Fullerton||120||-1|
|North Carolina Central||187||0|