- Our projections simulate the season 10,000 times to account for injuries and player performance variations. Here are 10 of the most interesting questions this season—and how are projections account for each contingency.
Sports Illustrated's 2016–17 preview is guided by data from our College Basketball Projection System, a collaboration between economist Dan Hanner and SI's Luke Winn and Jeremy Fuchs. We project teams on a player-by-player, lineup-based level and then simulate the season 10,000 times to generate our 1–351 national rankings and conference forecasts.
The final column based on our projection system is a blend of the hypothetical and the analytical. We put 10 what-if scenarios into our model to assess how certain NBA draft decisions, injuries, NCAA eligibility rulings, transfers and coaching changes impacted (or might have impacted) the 2016–17 season. We couldn’t run every possible scenario, so the ones that made the cut either had a direct impact on the national title race, or caused the most extreme shifts in a team’s ranking.
Spellman, a 6' 9", five-star freshman who had a chance to start at center for the Wildcats, was forced by the NCAA to take an academic redshirt year. When we originally ran projections with Spellman in Villanova’s rotation, the team ranked No. 4 overall—but really, was in a virtual dead-heat for the No. 2 spot along with Kansas and Kentucky.
Removing Spellman from the lineup had less of an impact than we expected. Villanova’s offensive projection only dropped from 121.0 points per 100 possessions to 120.2, because three of the forwards inheriting Spellman’s shots—Kris Jenkins, Mikal Bridges and Daryl Reynolds—are all efficient options. The Wildcats’ defense fell from 93.7 points allowed per 100 possessions to 94.2, due to a loss of some size and rebounding. Overall, Villanova only dropped one spot to No. 5 in our rankings; it’s no longer on the same efficiency tier as the three elites (Duke, Kansas and Kentucky), but it’s still in the mix for the Final Four.
Here’s what our projection for Villanova’s top eight scorers looked like with Spellman available:
And here’s what it looks like with Spellman out:
In this week’s college basketball preview issue of SI, Arizona is ranked No. 10. That was based on a full roster, but when rumors begin percolating that star guard Allonzo Trier might be ineligible, we also ran a contingency projection—one that had the Wildcats dropping to No. 15 if Trier was indeed ruled out. While Trier is Arizona’s clear best offensive player and its projected leading scorer, the team had enough quality depth that it could afford to lose him and still be at a Sweet 16 level.
But that was before Tuesday’s first exhibition game, which had three degrees of bad news: Trier sat out, his eligibility situation still unresolved; the team announced an indefinite, academic-related suspension for forward Chance Comanche; and in one of the most heartbreaking stories of the preseason, wing Ray Smith tore his ACL for a third time, ending his basketball career. This is what Arizona’s projected stats look like if it’s missing Trier, Smith and Comanche:
That Wildcats team would drop to 29th in our rankings. That could be too pessimistic, as there are plenty of examples of teams thriving with short rotations, and Sean Miller has a history of creating solid defenses. But a lineup like this leaves little room for error: If even one of the freshmen (Rawle Alkins, Lauri Markkanen, and Kobi Simmons) plays well below his projection, there aren’t any options for substitutes, nor are there any strong frontcourt defenders.
Brooks, the Ducks’ star small forward and national player of the year candidate, sat out most of the off-season with a foot injury, is still recovering from surgery, and has yet to participate in full-contact practice. If all goes well, he could be available to play in November—but timetables on foot injuries are notoriously unpredictable, and thus we ran projections to see what Oregon looks like without him.
Our model says that the Ducks would drop from No. 4 to No. 8 without Brooks, with a significant decrease in their offensive efficiency, as his ability to play both the 3 and small-4 spots takes their offense to the next level. What would keep them in the Pac-12 title hunt is their defense, which doesn’t project to slip as much. Oregon would still have three rim-protectors in Chris Boucher, Jordan Bell and Kavell Bigby-Williams, quality defensive rebounding, and more backcourt defensive depth with the addition of Villanova transfer Dylan Ennis, who was injured all of last season.
Here’s how Oregon looks with and without Brooks:
|Oregon with Brooks||120.1||93.9||0.9447||4th|
|Oregon without Brooks||117.2||94.6||0.9215||8th|
And here’s how the top eight players in the Ducks’ rotation would project if Brooks was out for the season:
The what-if scenarios we covered for Arizona and Oregon suggested that the loss of one player—even a player as high-profile as Trier or Brooks—doesn’t crush a talented team as much as one might think. But the case of Cal and Ivan Rabb would’ve been different: Had the 6' 11" power forward entered the NBA draft after his freshman season, when he was projected as a sure-fire first-rounder, the Bears would have lacked any suitable replacement, and their offensive and defensive efficiency projections would have cratered.
The with/without Rabb splits below have Cal dropping from 17th all the way to 69th if he were gone to the NBA:
|California with Rabb||111.7||94.3||0.8760||17th|
|California without Rabb||106.3||96.4||0.7537||69th|
It’s possible that Rabb’s draft decision had an even bigger impact than what we’ve modeled. His choice to remain at Cal may have made it a more attractive destination for Columbia grad transfer Grant Mullins, one of the best point guards available this offseason, as well as for elite freshman guard Charlie Moore, who committed to Cal in May.
Marquette, like Cal, had an irreplaceable, talented freshman power forward making an NBA draft decision . . . but unlike Rabb, Henry Ellenson opted to turn pro. Without Ellenson, we projected the Golden Eagles to rank 51st, making them one of the last teams left out of the NCAA tournament field. If Ellenson had come back, Marquette’s projection would have jumped to No. 19, which would put it at third-best in the Big East, and safely in the dance:
|Marquette with Ellenson||114.3||96.6||0.8738||19th|
|Marquette current roster||113.2||98.8||0.8263||51st|
This is a scenario fit for Seth Davis’s Jigsaw Man column, and there did seem like a possibility that Peters, the nation’s top mid-major stretch-four, would consider grad-transferring away from Valpo after coach Bryce Drew bolted for Vanderbilt. If Peters hadn’t chosen to stay, one of his best fits for his senior season would’ve been Iowa State, where he could’ve stepped into the role vacated by Georges Niang and become a lethal pick-and-pop weapon for point guard Monte Morris. Peters would’ve been a major upgrade over the Cyclones’ current leading options at the 4-5 spots, transfers Merril Holden and Darrell Bowie, and made a big impact on Iowa State’s ranking.
Here are the Cyclones’ projections with and without Peters:
|Iowa St. with Peters||118.5||97.2||0.9069||9th|
|Iowa St. current roster||115.6||99.9||0.8422||39th|
Peters is clearly good enough to elevate a team into the Top 10, and his projected stat line in that situation would’ve made him Iowa State’s second most important player after Morris:
Instead, Peters will obliterate those numbers as the star of the Horizon League, where he could easily average more than 20 points per game this season. With Valparaiso returning just half its minutes from last year and breaking in a rookie coach, Peters will have less of a chance to get national recognition, but pay attention to him: His production could very well be All-America worthy.
Our projections have the Spartans ranked 28th despite many experts putting them in their preseason top 10. Those top-10 rankings are, understandably, driven by the expectation that coach Tom Izzo is amazing and regardless of what happens, he’ll make this roster click. We agree that Izzo is great, as numerous studies have shown that he’s the best NCAA tournament coach of all time. But that doesn’t mean that his teams dominate every regular season. As recently as 2014–15, Michigan State was a No. 7 seed in the NCAAs, and in ‘10–11, it was a No. 10 seed. Izzo’s teams haven’t been immune to regular-season adversity.
The Spartans’ top-10 rankings (by humans) are also based on them having one of their best recruiting classes ever. Since they return just one player who averaged more than 5.0 ppg, and no players who averaged more than 10.0 ppg, they need the freshmen to play a lot, and play well. But unlike Duke and Kentucky, Michigan State’s great recruiting class isn’t filled with one-and-dones. With the exception of hybrid forward Miles Bridges, the Spartans’ freshmen project to break out in their sophomore seasons, and even in the case of Bridges, his advanced AAU-level stats suggest he may not match his high-volume shooting with a high level of efficiency in Year 1.
State’s defense, meanwhile, is the real cause for concern. Height correlates with rim protection and two-point field goal percentage defense, and because of injuries to grad transfer Ben Carter and projected starter Gavin Schilling, Michigan State will have some lineups with a 6' 7" center and a 6' 5" power forward. Some teams overcome height issues with extreme ball-pressure, but with so many freshmen in the rotation, the Spartans’ D has the potential to struggle, particularly in the first 1–2 months. We project them as the No. 28 team if Schilling can come back and play the back end of the season, and lower if he does not return.
What would make Michigan State a top-10 team in the eyes of our projections is the addition of one frontcourt star—a guy like, say, Deyonta Davis, who made a surprise decision to leave after his freshman season. Had he come back, we would have projected the Spartans at No. 9 in the nation and in a first-place tie with Wisconsin in the Big Ten:
|Michigan St. with Davis||117.9||95.7||0.9168||9th|
|Michigan St. partial Schilling season||115.2||99.0||0.8513||28th|
Here’s how the Spartans’ rotation would project if Davis were still around:
|Tum Tum Nairn, Jr.||PG||3.6||1.6||3.4||108.6||15%||52%|
It didn’t turn out to be serious, but when freshman wing Jayson Tatum was injured at Duke’s pro day last week, we re-ran the Blue Devils’ projections to account for a few different scenarios. If only Tatum were out for the season, Duke would stay No. 1. And if only freshman power forward Harry Giles III—who had knee surgery in October and will be brought along slowly—was out for the season, Duke would stay No. 1. If Tatum and Giles were both removed from the equation, Duke would fall to No. 5.
We also wondered what would happen if the Blue Devils entire recruiting class had never enrolled. It turned out that a team exclusively comprised of Duke returnees would still be pretty good, dropping only from No. 1 to No. 12 in the national rankings. With an efficient 1-2-3-4 of Grayson Allen (averaging 23 ppg), Luke Kennard (15), Matt Jones (12) and Amile Jefferson (12), the Blue Devils’ offense would still be quite good, but their defense (especially their lack of frontcourt depth) would keep them from winning the ACC:
|Duke full roster||125.2||94.0||0.9641||1st|
|Duke without this year's recruiting class||119.6||98.7||0.9010||12th|
Kentucky, on the other hand, would not be fun to watch without this year’s freshman class. Isaiah Briscoe and Derrick Willis would do their best to carry the offense, and John Calipari’s would likely continue his track record of having strong defenses. But even if Cal got everyone to buy in defensively, Kentucky would project to drop from No. 3 to No. 49, and be the first team left out of the NCAA tournament:
|Kentucky full roster||119.2||92.0||0.9515||3rd|
|Kentucky without this year's recruiting class||108.2||94.0||0.8353||49th|
All of which means: Duke’s recruiting class may have higher draft prospects, but Kentucky’s class is more essential to the team’s success this season.
Our projections have the current version of the Orange at No. 14 overall, and they were given a serious boost by the additions of White, a volume wing scorer from Nebraska, and Gillon, a point guard from Colorado State. If that grad-transfer duo were removed from Syracuse’s roster, it drops all the way to 35th, and ninth place in the ACC:
|Syracuse current roster||113.6||93.9||0.8999||14th|
|Syracuse without White/Gillon||109.4||94.3||0.8466||35th|
It’s possible that had White and Gillon not transferred in, Syracuse might have filled their scholarships with a juco guard or a 3-star recruit in the spring, but this is how the Orange’s rotation would project without any replacements in the mix:
Until now we’ve focused exclusively on players, but our projection model also accounts for the impact of coaches. And no major-conference head-coaching change projects to have a bigger, immediate impact than former Pitt coach Jamie Dixon taking over at TCU. Not only does Dixon have a superior efficiency resume on offense and defense than Johnson, the Horned Frogs’ previous coach, but Dixon and his staff also brought in a top-100 recruit, Jaylen Fisher, who immediately improves the team’s rotation.
Our system has Dixon improving TCU’s ranking by 44 spots vs. where it would be with Johnson in ‘16-17—still on the outside of the NCAA tournament, but at least in the top 100 nationally:
|TCU with Dixon||104.9||96.5||0.7231||90th|
|TCU with Johnson||101.1||97.2||0.6111||134th|