PDA

View Full Version : Range Distribution in Results Analysis



verneer
10-03-2009, 10:17 AM
When I sort data by a category (say, VPIP), I get the mean, median, and SD along with a range distribution with median winrates.

For example, on page 88 of the HEM Manual it uses a VPIP example with following:

Avg Value: 22.0
Median Value: 25.9
Standard Dev: 8.54

And then the following Range distributions with median winrate:

7.00 to 19.30: Median Winrate = -1.17
19.30 to 23.40: Median Winrate = - 1.88
etc etc ...

I would like to know how are the five ranges constructed. Why does the first one go from 7.00 to 19.30 and the second one from 19.3 to 23.4?

netsrak
10-04-2009, 05:30 AM
I'll forward this for a reply...

zoom7
06-20-2010, 10:57 AM
Bump, I would also like to know the answer to this.

zoom7
06-20-2010, 11:41 AM
Ok, I figured this one out. It is divided evenly based on number of players. So let's say there are 150 players, this would be divided to get 5 groups, each with 30 players. Then sort the players based on the selected stat (say, VPIP) and select groups of 30 players, use the minimum and maximum values from those groups as the range.