We hear an awful lot about a team’s “body of work” in the buildup to Selection Sunday, as though it is a uniform construct. But that body is merely an accumulation of components, each of which can vary from one another in fairly significant ways.
Freshmen are rarely the same in March as they were for a slate of games in December. Different teams progress at different rates. The quality of opponents fluctuates as the season marches on, and those foes also go through various stages in their own development. And, most significantly, teams deal with injuries that fundamentally alter both their talent level and style of play.
It’s the last of those factors that concerns us here at GK Central. It’s hard enough to evaluate a team on the basis of roughly 30 games against only a small subset of the Division I universe. It’s even more difficult when key players have been absent for major lengths of time. Can you really put your finger on a Michigan State team, for instance, that looked liked the national title favorite to start the season and then lost three key players and a bunch of games in the process? Well, we’re going to try.
To aid our effort, we turned to our team of statistical consultants -- Furman University professors Liz Bouzarth, John Harris and Kevin Hutson -- to tweak our model so that it could measure the Spartans’ performance when Keith Appling, Branden Dawson and Adreian Payne were all in the lineup as well as when any member of the group was absent. Then they tackled the same issue for Arizona with and without Brandon Ashley, as well as Oklahoma State without Marcus Smart during his three-game suspension. The results may not be definitive, but they are certainly interesting and help inform our thinking as these teams get slotted into the bracket in two weeks.
When Dawson broke his hand in late January, the Spartans were 18-1 and ranked No. 3 in the country. Since that time, they’re 5-6, with Dawson returning only for the final two games of that stretch. Needless to say, Michigan State’s performance is quite different when it is missing a key player, and the Spartans have been without three top contributors at various points, with Payne and Appling also nursing injuries. So, naturally, the Spartans have been better without their best guys.
Wait, what? As strange as it sounds, our model actually awards Michigan State a better power ranking in the 14 games when the Spartans played without one of their three injured stars than in the 15 games when they were all on the court together. That’s a product of the fact that Michigan State has played a significantly tougher schedule in the games when those players were out. The model’s “secret sauce” -- how similar a team is to previously safe or slain Giants -- hardly changes, either. As a result, MSU’s overall Giant rating (71.7) isn’t much different than what it put together with a full squad (71.3).
The one thing that’s drastically different between a healthy and injured MSU squad is its GK rating. It’s extremely unlikely that the Spartans will qualify to play a game in the tourney as a Giant Killer, but the injury-depleted group posted a massive 61.7 GK rating, compared to a 30.2 mark for the healthy bunch. Why? Because without Appling, Dawson and/or Payne, MSU played more like a Giant Killer, employing a slower pace and launching more 3-pointers.
Look, there’s no doubt that the model’s conclusions are, at the very least, really weird here. And it’s silly to believe that Michigan State isn’t a better team with all its guys. But we’re also confident that we don’t need to adjust MSU’s overall Giant rating come tourney time.
For the Wildcats, the equation is slightly different. Ashley isn’t coming back, so the real issue is determining whether their strong Giant profile was built by a squad that no longer exists.
As it turns out, Arizona still doesn’t have much to worry about. The Wildcats’ overall Giant rating of 94.1 drops only slightly to 93.0 during the eight games they’ve played without Ashley. (Stats from the eighth game, a win at Oregon State, aren’t included in that rating.) Arizona has been playing faster without its sophomore forward but otherwise features the same stifling defense, efficient offense and devastating rebounding at both ends. Simply put, there is no reason to believe that Arizona is any more vulnerable to a GK-level upset without Ashley.
Unlike the previous two teams, we’re evaluating the Cowboys as a GK, not a Giant, because Joe Lunardi projects them to be a 9-seed in his latest Bracketology. And although Smart missed only three games after being suspended following his altercation with a fan at Texas Tech, those contests are enough to depress Oklahoma State’s overall GK rating.
The Cowboys project as a solid Killer with a 24.4 rating, but when you weed out the three games Smart missed, that number climbs to 29.4. That’s not particularly surprising -- Oklahoma State is obviously a better team with Smart in the lineup -- but the degree of change is somewhat unexpected, as is the revelation of just how bad OSU was without its best player. The Cowboys team that dropped those three games falls from a power rating of 26.1 to 5.7, and the GK rating is an anemic 2.5. That’s the same as teams like Western Carolina and Colgate. And it’s strong evidence that in Oklahoma State’s case, absence makes a Killer grow weaker.