In the 1938639 study the evaluation technique has been extended
to inelude a defensive rating system for both the team and the individual playere
The items and their evaluation weights, as used in this study, are shown in Table Ie

Data were collected during all the home games on both the Kansas
team and the visiting teams. The technique used in the collection of these data is
the same as described in the first evaluation study.

In the 1937-38 season nine home games were played, and this
season eight home games were played, thus making a totel of 17 games on which averages
of certain activities were available. These averages are shown in Table IIIe The
17 game averages seem to be reliable as there was no great variation in the figures
computed for the two seasons. The team this year took more shots than did last
season's team, but averaged one less goal per game. The number of free throws awarded
in both seasons was practically identical, but the number made was slightly reduced
this seasone

This year the total number of positive offensive evaluation points
is lower than last year's total. This is due to tw reasonse First, there was &
change in the technique of tabulating immediate assists. In last year's study credit
was given the players for both passes aiid catches, which gave them double credit in
evaluation points. In this year's study a player receives evaluation points only
once. The second reason for the lower total is that the recovery of rebounds off the
opponent's backboard was computed with the defensive play instead of offensive play,
as was the case in last year's studye

The drop in negative offensive evaluation points indicates that
the team made fewer mistakes during this season than last season. It is possible that
the team summary posted in the team dressing ream the day following each game made
the individual players more conscious of their mistakes with the end result that fewer

were madee