MU KenPom Luck rating 364 of 365

[mileskishnish72]

Luck is the deviation in winning percentage between a team's actual record and their expected record using the correlated gaussian method.

Well, thank God for that! The uncorrelated Gaussian method wouldn't have meant squat.

[brewcity77]

Basically, being 364th in Luck means that our actual wins (12) are way less than our expected wins (much more than 12).  Which either means we just shat the bed in clutch time all season (we did) and lost games we shouldn't have or we had some really good wins that make our computer numbers look better (also true).

This felt like a team that made their own luck.

[muwarrior69]

I may be misunderstanding what you're saying, but we were not 0-4 in games decided by 3 points or less. We were 3-4 (Wins vs Valpo/Xavier/Providence, Losses vs Oklahoma/Nova/@Nova, vs Xavier). 3-5 if you count the OT loss to Dayton.

I think extending it out shows the bigger picture:

Record In games decided by 1 possession or OT:
3-5

Record in games decided by 2 possessions:
1-4

Record in games decided by 7-11 points:
0-5

Record in games decided by 14 points or more:
8-6

When we won, we tended to blow teams out of the water in no doubt fashion. When we lost, the game was usually at least somewhat competitive. I may be misremembering, but I think all of those 7-11 point losses were in question within the last few minutes.

This data could indicate that we were better a team than our record showed (and some of the underlying analytics do support that). It could also indicate that our coaches, players, or both have trouble executing in high pressure situations. I think that's a fair  concern moving forward.

Which underlying analytics supports that we were a better team than our record and is there a single analytic that measures trouble executing? Not a metrics guy, just an eye test guy and curious about your statement.

Appears the team had trouble executing just about everything: can't shoot, rebound, make layups and play defense.

[panda]

This felt like a team that made their own luck.

Poor roster construction, bizarre in game management, not enough depth etc. Luck favors the prepared and this team was nowhere near prepared.

[TAMU, Knower of Ball]

Which underlying analytics supports that we were a better team than our record and is there a single analytic that measures trouble executing? Not a metrics guy, just an eye test guy and curious about your statement.

Appears the team had trouble executing just about everything: can't shoot, rebound, make layups and play defense.

The major services, KenPom, Bart Torvik, Evan Miya, all ranked us between #83 (KenPom) and #73 (Evan Miya). If you look at the other high major teams ranked in the same range as us, you will see much better W/L records despite us playing similar strengths of schedule:
Colorado 17-15
Creighton 15-17
Cal 22-12
Minnesota 15-17
Georgetown 16-18
USC 18-14
Oklahoma State 20-15
Wake Forest 18-17
Butler 16-16
Syracuse 15-17

Oversimplifying it a bit, but essentially a team who put up the metrics we did against the strength of schedule we played you would expect to be around a .500 record. Actually, according to all three, we were the highest rated team with 14 or less wins.

I'm not aware of any single metric that measures a team's performance in high pressure end of game situations and compares it against other teams. I think it's fair to assume that if you have poor record in close games, you likely have some issues performing high pressure end of game situations.

[1SE]

Economists call it "Total Factor Productivity"