The factor analysis shows that Bartle captured 3 motivations fairly well, however, the data also shows that Bartle either included an unrelated element or excluded a related element in each case.

Bartle emphasized conversing, interacting and role-playing in his Socializing type. The factor analysis shows that while conversing with other players is part of the Socialization motivation, it should be focused on relationship formation instead. And the real appeal seems to be the intimacy and support that can be gained from such relationships. The factor analysis also revealed that role-playing should not be placed under this motivation, but in a separate factor altogether.

With regards to the Achiever type, Bartle focused on points accumulation - whether this be experience points, skill points, levels or equipment, and this part of the Achievement motivation does pan out in the factor analysis. But the data also pointed out that the underlying motivation is to gain a sense of power within the construct of the game, and that points accumulation may just be a means to this end.

When Bartle talks about Killers, he focuses on the satisfaction gained from causing distress to other players. The factor analysis reveals that causing distress may be the more extreme end of a Grief motivation that gains satisfaction from using other players for one's own gains. And that this motivation should also include the more subtle tactics of manipulation and deception.

What was very interesting was to not see the Explorer type validated. It was clear that if the Exploring motivation did exist, that the interest in mapping the world and understanding the game mechanics were unrelated. I felt that I had perhaps constructed poor statements for this motivation, and switched several statements into the questionnaire after 6700 responses had been received. I then received another 500 responses, and tried to see whether the following statements might validate an Exploring motivation that focused on game mechanics:

1. I like to think about class-balancing issues.
2. I try out a lot of things to experiment with the game mechanics.
3. I try to find bugs I can exploit.
4. What fascinates me is finding out how stuff works in the game.
5. I like numbers, charts and tables.

The smaller data set was not significantly different from the larger one in terms of age and gender composition. The factor analysis on this data set also did not validate the Explorer type. In particular, none of the above 4 statements correlated above (0.20). However, statements 1, 4 and 5 had weak loadings on the Achievement factor, but the loading was too weak to be considered significant (.21, .27, and .37 respectively).