Game thread: #15 Lady Vols host Ole Miss, Sun., Feb 16, Noon ET. SECN

[Orange.]

 

[Rickster]

No,,,I do it by going to "live stats",,,setting the tab on "play by play",,,filter teams to TN ...filter plays to "substitutions",,,then go to each period and cut and paste the data in to a spreadsheet, where I can manipulate the data
View attachment 722741

then I narrow it down to the data I actually want:
View attachment 722742

My occupation was inventory control and data analyst
I did this for three games BC I wanted to know the rate of mass subs per game and their duration (time on the court per subbings)

Thanks for the extra effort in putting this together.

 

[GUNTERSVOL]

I noticed that aberration as well and it did concern me.
We need more shots, more rebounds and fewer TOs than our opponents because we won’t always shoot lights out. I wish Cora would ask CKC about those stats from Ole Miss game.

Coach addressed that in the post game presser that is still available at the official site for all to see and hear. Just listened to it again moments ago tagged just below the recap.

 

[Lucy]

Coach addressed that in the post game presser that is still available at the official site for all to see and hear. Just listened to it again moments ago tagged just below the recap.

Thanks!!

 

[Lucy]

#389

Thanks!!

 

[RespectAPA]

Ironically, +_ is a not the best measure of a single player. It actually measures how a player does in combination with the other four players on the court.

Which is to say, it 'actually measures' the players performance in the only context that they appear: in a group with 4 teammates.

Aside from shooting free throws, 5-on-5 basketball is fundamentally a team effort, and your value as a player (in terms of generating wins for your team) is what your value is in the context of sharing the court with 4 other teammates and 5 opponents. What's being presented here as a flaw is really the main strength of plusminus, which does come with the downside of having higher variance, and therefore needs a more data to lower the standard deviation to a reasonably tight spread. But the claim isn't that plusminus converges on the limit value quickly, and there's many obvious reasons why it doesn't [because its primary strength is its primary weakness; not trying to contextualize at all means that there's nothing in the player's style that it can miss, at the expense of converging on player value quickly] rather the claim is that what plusminus is converging on is the player's value, since it's the only thing that matters: whether our team beats the other team while that girl is on the court, and by how much.

If you think a girl is good, but the Vols are consistently trailing in her time on the court, then there's something about her that you're missing. If you think a girl is bad, but the Vols are consistently moving ahead in her time on the court, there's something positiv about her that you're not noticing.

But people tend to not think that way about players they really like, or players they really dislike. If metrics which evaluate the player's value opposite of what they think, they throw out the metric instead of reconsider if they're not judging the player too harshly (Puckett, who has been net positiv in quite a few SEC games this season) or too leniently (Spear, before the team adjusted how they utilized her in the fullcourt; before, she was the furthest back player in the press, and a very easy 1-on-1 for a lot of our opponents' forwards, meaning they could go in instead of settling for a 5-on-5 set, resulting in her being net negativ in her first 5 power conference games, some of which were 20+ scoring performances!!) which isn't rational, but there's no rationality requirement to be a sports fanatic.