if there are 11 clicks .. wouldn't you want a 12 point scale ??
(analogy .. a 2 point scale would have 1 click between 1 and 2)
(a 10 point scale would seem to fit with a 9 click scale)
Additionally, it seems like there is some problem in statistically separating the manager from the team "context" ... to really test it out well, you'd need to come up with your best numbers for tendencies, and then play test it using historical rosters to see that the stats came out roughly correct... A lot of work to get it really good. Though certainly estimates made from observation of stats would be MUCH better than random assignment of tendencies.
it seems like the most useful data might be the data at the two transitions of a managing assignment... how did his tendencies compare with the manager immediately prior.. and how did they compare with the manager immediately after.
Good luck
Steven "HoustonReplay" Copley
scopley409@aol.com