Busting Brackets
Fansided

NCAA Basketball: Why Michigan, BYU and NC State are underrated in 2019-20

ANN ARBOR, MICHIGAN - DECEMBER 14: Franz Wagner #21 of the Michigan Wolverines reacts to a second half three point basket while playing the Oregon Ducks at Crisler Arena on December 14, 2019 in Ann Arbor, Michigan. Oregon won the game 71-70 in overtime. (Photo by Gregory Shamus/Getty Images)
ANN ARBOR, MICHIGAN - DECEMBER 14: Franz Wagner #21 of the Michigan Wolverines reacts to a second half three point basket while playing the Oregon Ducks at Crisler Arena on December 14, 2019 in Ann Arbor, Michigan. Oregon won the game 71-70 in overtime. (Photo by Gregory Shamus/Getty Images) /
facebooktwitterreddit
Prev
1 of 5
Next
NCAA Basketball
PITTSBURGH, PA – MARCH 21: A detailed view of a Wilson college basketball (Photo by Justin K. Aller/Getty Images) /

Evaluating NCAA Basketball teams is challenging enough. When injuries get thrown into the mix, it’s even more likely mistakes will be made.

Whether it be the AP Poll, the NET, or analytics sites like KenPom, Barttorvik or Haslametrics, there are countless ways to rank college basketball teams. In terms of strictly evaluating a team’s resume, tools like the AP Poll and NET are decent (albeit flawed) methods.

(Credit to KenPom, sports-reference and hoop-math for statistics, Big 12 Conference for GIFS and ESPN for chart)

For the purposes of projecting a squad’s performance moving forward, however (all any amateur gambler cares about), there’s a general consensus that the analytics sites are the way to go.

Of course, none of KenPom, Barttorvik, nor Haslametrics is perfect. There are challenging questions, like how much to value recent performance, that these rankings try to answer. There are other factors, however, like injuries for example, that these sites can’t/don’t really account for.

Imagine a team that plays really well to start the season and climbs up an analytics ranking to number 10 overall. Unfortunately, this team suffers a two-month injury to its best player. The squad struggles over this stretch, and the ranking drop them to 40th.

When the player returns from injury, the team gets back to playing well again. With the strong close to the regular season, it jumps up to 25.

This is a realistic scenario, and things not too dissimilar to this play out every season. What do we make of it? Analytics sites are generally well-respected, but in the above example, something seems to be clearly off. The team shouldn’t be punished so much for missing its best player, right? Isn’t it probably better than 25th in the nation?

College basketball fans CAN reasonably disagree with a team’s analytics ranking; perhaps they’ll even be proven correct. When it comes to evaluating teams with significant injuries, however, even analytics diehards NEED to be skeptical of a team’s placement. How much to adjust a ranking is subjective, but the need for some sort of adjustment seems unquestionable.

There are loads of college basketball teams with injuries this season, but I’ll touch on three for the purposes of this article. How should fans adjust their perceptions of these squads?