Men’s basketball | What might have been: Simulation picked Ohio State to win NCAA tournament
Note: A simulation conducted by the Washington Post of the NCAA men’s basketball tournament picked Ohio State to win it all. Here’s how the data-aided simulation played out.
With the cancellation of the 2020 men's and women's NCAA basketball tournaments due to the covid-19 pandemic, this marks the first year the NCAA tournament has not been played since its inception in 1939. It's a tough break for the players and coaches who worked hard all season to get a chance to hear "One Shining Moment" as they cut down the nets in Atlanta, the scheduled site of this year's Final Four.
Get the news delivered to your inbox: Sign up for our BuckeyeXtra newsletter
In the wake of the cancellations, some governing bodies have taken it upon themselves to award the title of "national champions," as the Florida state senate did in declaring the Seminoles champs. We have a slightly more objective solution.
We have a pretty good sense of which teams were likely to make the field thanks to the Bracket Matrix, which compiles data from 81 different bracketologists, and roughly where each one should be seeded according to Patrick Stevens' last bracket projection before cancellation. All that's left is to simulate the What-Might-Have-Been 2020 Bracket.
To make it simple, we're going to start with an estimated number of possessions for each team using the definition outlined by Nylon Calculus, including any additional possessions to be had via offensive rebounds and turnovers. Then, estimated scores are derived by multiplying those respective possessions by the team's adjusted offensive rating found at Ken Pomeroy's site. Once we know the projected scoring margin, we can infer an implied win percentage. For instance, teams that are favored by two points would have an expected win probability of 57%. That rises to 77% if the predicted scoring margin is seven points. Finally, these win rates are then used to decide which team would win in a head-to-head matchup with luck factoring into the equation just like it does during March Madness. To simulate luck in this tournament, we are going to compare each team's expected win rate in a head-to-head matchup to the expected win rate if we based its chances on seed matchup alone. Any team that had a win rate higher than expected for a seed pairing was given a chance at the upset victory. That individual game was then simulated 5,000 times, with the result from the 2,020th game used for this bracket.
A word of caution. Don't expect many upsets this year. The NCAA's NET rating, used to rank teams during the regular season and seed them in the tournament, was one of the best predictive measures of 2019-20. The higher ranked team in NET went on to win over 67% of the time. The best ranking system for predictive performance this season, created by Mark Moog, predicted a win 68% of the time. That's a difference of less than one game per 68-team field. The average ranking was right 61% of the time.
So, what could have this year's NCAA men's basketball tournament looked like? Here's a region-by-region breakdown.
The First Four teams are the four lowest-seeded teams earning an automatic bid to the tournament (North Carolina Central, Prairie View A&M, Robert Morris and Siena) plus the four last at-large qualifiers (North Carolina State, Richmond, UCLA and Wichita State). The former will be No. 16 seeds and the latter will be No. 11 seeds.
No. 11 Richmond beat No. 11 North Carolina State
The Spiders were one of the best shooting teams in the college ranks (effective field goal rate of 53%, 30th) with good discipline and an ability to generate steals (12%, 19th in country) for extra possessions.
No. 11 Wichita State beat No. 11 UCLA
A top-10 defense (90.4 points allowed per 100 possessions after adjusting for opponent) centered around rim protection helped Wichita State move on to the field of 64.
No. 16 Prairie View A&M beat No. 16 North Carolina Central
Prairie View created turnovers (24%, 10th best) and was aggressive getting to the free throw line (24th in nation), two key ingredients for success in the tournament.
No. 16 Siena beat No. 16 Robert Morris
Point guard Jalen Pickett and forward Manny Camper were each above-average players for Siena this year. Pickett led the team in points (15.1) and assists (6.0) per game while Camper cleaned the glass with over 10 rebounds per contest. Robert Morris was overmatched.
This region was stacked in the aggregated bracket and went largely as expected. The Dayton Flyers, who earned the top seed in the East after winning the Atlantic 10 Conference behind an offense that is the second-best in the nation after adjusting for opponent, beat No. 16 Winthrop. No. 2 San Diego State, the last undefeated team in the country, moved past No. 15 Eastern Washington. No. 3 Duke was the 11th most efficient offense and ninth most efficient defense during the regular season and had no trouble moving past No. 14 Belmont.
The first top seed that ran into trouble was No. 4 Kentucky. The Wildcats lost to No. 13 North Texas in the simulation, and you'd have to imagine senior DJ Draper, one of the best three-point shooters in the country, would have played a key role in the upset. The backup shooting guard hit 42 of 77 three-point attempts this season (55%) and that included 13 for 26 on guarded catch-and-shoot attempts from behind the arc. No. 7 Houston was also sent home early. The Cougars were bested by a No. 10 Texas Tech squad that boasted the ninth-best defense in the county per Pomeroy's ratings.
No. 6 West Virginia took care of business in the first round against Richmond and then stunned Duke to reach the Sweet 16. That wasn't a complete surprise, the Mountaineers were the third-best defense in the nation. North Texas also defeated No. 5 Auburn, one of the luckiest teams during the regular season, to continue a Cinderella run. North Texas couldn't, however, make it past Dayton in the Sweet 16. West Virginia claimed the other side of the bracket, casting aside San Diego State for an Elite Eight spot.
That left Dayton and West Virginia battling it out for a Final Four berth. Dayton won 71 to 67 in our simulation, giving the school its second Final Four appearance in school history and first since 1967.
No. 1 Kansas, the top overall seed in the tournament, easily cruised through the first round. As did No. 2 Creighton, No. 3 Michigan State, No. 4 Oregon, No. 5 Ohio State, No. 6 Iowa and No. 7 Arizona. Although not technically an upset, No. 9 Rutgers won over No. 8 LSU 74-72.
The Buckeyes, however, were a sleeping giant. They were the eighth-most efficient team in the country after accounting for strength of schedule, with skilled three-point shooters (37%, 23rd) that could score in bunches. They beat Oregon in the round of 32 and knocked off Kansas 72-65 in the Elite Eight to make this year's Final Four. The Jayhawks' defense was stout (second-best in the country) but they allowed way too many shots from behind the arc (41% of all field goals, 288th out of 353 teams). Once they ran into a team like Ohio State that could make them pay, it caught up to them.
Then Ohio State beat Michigan State to make the Final Four. Pace played a key role in this simulation. The Buckeyes' slow play (280th in the nation) added emphasis to their three-point shooting. Their rebounding prowess (25.2%) also stopped Michigan State from fully utilizing its own strength on the offensive glass, propelling Ohio State to a 64-52 win.
Three transfer students helped No. 1 Baylor get its highest seed in the tournament since the Bears' first appearance in 1988. Point guard Davion Mitchell, center Freddie Gillespie and forward MaCio Teague combined to score 96 points per 100 possessions during the regular season, 11 points per 100 possessions more than the rest of the roster. Gillespie was also stellar on the offensive glass, grabbing nearly 16% of the team's misses, the 10th highest rate in the country. Baylor moved on easily against No. 16 Boston University and No. 9 Florida (who beat No. 8 Providence).
No. 5 Butler didn't succumb to No. 12 Stephen F. Austin as many thought (5-seeds are popular picks to fall in the first round) nor did the Bulldogs have trouble with No. 13 New Mexico State, who had upset No. 4 Wisconsin in the round of 64 using extra possessions from offensive rebounds and turnovers, two fingerprints found in many first-round upsets. Had Butler met Wisconsin the outcome would have been the same. The Bulldogs were the 21st best team in the country per the consensus of dozens of rating systems, including the NCAA's NET ranking, with an ability to score around the rim and defend others from doing the same.
On the other side of the South bracket, No. 6 Michigan made its way past No. 3 Seton Hall to gain entry to the Sweet 16 and then No. 2 Florida State for a spot in the Elite Eight. No. 10 Oklahoma, the 39th best team per the consensus rankings, upset No. 7 Illinois, the 28th best, due to not giving away easy points via free throws (Illinois shot an above-average 73% from the line this season). The Sooners were the most-disciplined team of 2019-20 in terms of how often they sent opponents to the free throw line.
Baylor beat Butler 65-59 and then beat Michigan 70-65 to earn the region's Final Four spot.
No. 1 Gonzaga boasted a solid resume (5-2 against Quadrant 1 teams and 30 against Quadrant 2) and the most-efficient offense in the country after adjusting for opponent (120.6 points per 100 possessions). However, on paper, Gonzaga was the most-likely 1-seed to lose in the round of 64, nearly joining Virginia as the only No. 1 seeds to ever lose to a No. 16 seed in the opening round. Luckily for them, Gonzaga narrowly beat Siena 86-83.
No. 10 Arizona State got revenge against No. 7 Virginia for a low-scoring regular season loss (48-45) in November. That would have been the first win for Arizona State in the tournament since the field expanded to 68 teams in 2011.
The biggest surprise was No. 3 Maryland, perhaps the best team Mark Turgeon has coached in College Park. Almost three-fourths of the 62 ratings systems audited by Kenneth Massey had the Terrapins ranked as at least the eighth-best team in the country. Pomeroy's college basketball ratings estimated the Terps are outscoring opponents by 20.5 net points per 100 possessions after adjusting for strength of schedule, the 13th-best rate in the country. Before this season Turgeon had never seen Maryland rank higher than 22nd, and that was in 2016. You would have to go back to Gary Williams's coaching days to find a Maryland squad ranked better by Pomeroy's algorithm than the one Turgeon coached in 2019-20.
Turgeon saw his squad beat No. 14 UC Irvine, No. 6 Penn State and then No. 2 Villanova to face off against Gonzaga in the Elite Eight. Maryland didn't have the offense Gonzaga did but the Terps were more balanced, especially on defense, leading to a 91-84 victory to give the school what would be its third Final Four appearance ever in the men's basketball tournament.
No. 5 Ohio State (Midwest) beat No. 1 Dayton (East) 80-64
The Buckeyes allowed opponents to shoot just 44% on two-point shots during the regular season, the 17th best rate in the country. That let Ohio State do the one thing better than most against Dayton and Obi Toppin, the best dunker in the nation: defend the rim.
No. 1 Baylor (South) defeated No. 3 Maryland (West) 72-58
Baylor's ability to own the offensive glass and force turnovers likely would have proven Maryland's undoing. The Bears forced a high number of turnovers (23% of defensive possessions, 22nd) and grabbed close to 36% of their misses for easy putbacks off offensive rebounds during the regular season. Both served them well in this matchup.
National title game
No. 5 Ohio State wins 74-69 over No. 1 Baylor
The Buckeyes were the better shooting team all year (effective field goal rate of 52% vs. 49% for Baylor) and close enough defensively to allow Ohio State to win its second national championship and first since 1960.