MHB Correlation to Winning question

[vballer]

I have a bunch of data for volleyball and I am trying to figure out how correlated certain stats are to winning a game. Here is a small example of the data set.

[TABLE="width: 500"]
[TR]
[TD]Pts
[/TD]
[TD]Err
[/TD]
[TD]Diff
[/TD]
[TD]EFF%[/TD]
[TD]Win
[/TD]
[/TR]
[TR]
[TD]21
[/TD]
[TD]10
[/TD]
[TD]11
[/TD]
[TD].360
[/TD]
[TD]1
[/TD]
[/TR]
[TR]
[TD]21
[/TD]
[TD]10
[/TD]
[TD]11
[/TD]
[TD].350
[/TD]
[TD]1
[/TD]
[/TR]
[TR]
[TD]18
[/TD]
[TD]11
[/TD]
[TD]7
[/TD]
[TD].250
[/TD]
[TD]1
[/TD]
[/TR]
[TR]
[TD]14
[/TD]
[TD]6
[/TD]
[TD]8
[/TD]
[TD].280
[/TD]
[TD]1
[/TD]
[/TR]
[TR]
[TD]19
[/TD]
[TD]10
[/TD]
[TD]9
[/TD]
[TD].380
[/TD]
[TD]1
[/TD]
[/TR]
[TR]
[TD]17
[/TD]
[TD]6
[/TD]
[TD]11
[/TD]
[TD].300
[/TD]
[TD]1
[/TD]
[/TR]
[TR]
[TD]12
[/TD]
[TD]9
[/TD]
[TD]3
[/TD]
[TD].200
[/TD]
[TD]0
[/TD]
[/TR]
[TR]
[TD]14
[/TD]
[TD]10
[/TD]
[TD]4[/TD]
[TD].100
[/TD]
[TD]0
[/TD]
[/TR]
[TR]
[TD]11
[/TD]
[TD]8
[/TD]
[TD]3
[/TD]
[TD].050
[/TD]
[TD]0
[/TD]
[/TR]
[/TABLE]

I am trying to determine which of these items are most correlated to winning (the last column). As you can see if it is a win the value is 1 and if not then it is a 0. In addition, to determining the most useful stats for determining a win, I would like to know what levels each of the first 4 columns should be in order to generate a win at a certain confidence level.

Any help with this is greatly appreciated.

Thanks
Jamie

 

[CaptainBlack]

I have a bunch of data for volleyball and I am trying to figure out how correlated certain stats are to winning a game. Here is a small example of the data set.

[TABLE="width: 500"]
[TR]
[TD]Pts[/TD]
[TD]Err[/TD]
[TD]Diff[/TD]
[TD]EFF%[/TD]
[TD]Win[/TD]
[/TR]
[TR]
[TD]21[/TD]
[TD]10[/TD]
[TD]11[/TD]
[TD].360[/TD]
[TD]1[/TD]
[/TR]
[TR]
[TD]21[/TD]
[TD]10[/TD]
[TD]11[/TD]
[TD].350[/TD]
[TD]1[/TD]
[/TR]
[TR]
[TD]18[/TD]
[TD]11[/TD]
[TD]7[/TD]
[TD].250[/TD]
[TD]1[/TD]
[/TR]
[TR]
[TD]14[/TD]
[TD]6[/TD]
[TD]8[/TD]
[TD].280[/TD]
[TD]1[/TD]
[/TR]
[TR]
[TD]19[/TD]
[TD]10[/TD]
[TD]9[/TD]
[TD].380[/TD]
[TD]1[/TD]
[/TR]
[TR]
[TD]17[/TD]
[TD]6[/TD]
[TD]11[/TD]
[TD].300[/TD]
[TD]1[/TD]
[/TR]
[TR]
[TD]12[/TD]
[TD]9[/TD]
[TD]3[/TD]
[TD].200[/TD]
[TD]0[/TD]
[/TR]
[TR]
[TD]14[/TD]
[TD]10[/TD]
[TD]4[/TD]
[TD].100[/TD]
[TD]0[/TD]
[/TR]
[TR]
[TD]11[/TD]
[TD]8[/TD]
[TD]3[/TD]
[TD].050[/TD]
[TD]0[/TD]
[/TR]
[/TABLE]

I am trying to determine which of these items are most correlated to winning (the last column). As you can see if it is a win the value is 1 and if not then it is a 0. In addition, to determining the most useful stats for determining a win, I would like to know what levels each of the first 4 columns should be in order to generate a win at a certain confidence level.

Any help with this is greatly appreciated.

Thanks
Jamie

I'm not sure that you strictly mean correlated, the best predictor for winning for this data is EFF% where EFF%>=0.25 predicts a win and EFF%<=0 predicts a loss and between 0.2 and 0.25 is a no-man's land.

Diff will also provide a perfect predictor for this data but the no-man's land is relatively wider.

Without knowing what the real question is there is little more that it is worth saying.

CB

 

[vballer]

Let me try this again with the actual data set attached.

Most importantly I am not looking for the actual answer but more how to derive the answer. I guess when you say predictor how did you determine this?

To recap, I have 7 columns of stats and the 8th column is 1 if the game was won and 0 if it was lost. So how do I determine what the best predictors of a win are?

Thanks

View attachment 148