Actually, you know what? I ran the Stats.
Assumptions
Simple Random Sample of Guns of Icarus Players (assuming that forum attendance has no influence on spanner-mallet, mallet-spanner preference).
n(p) > 10.
n(1-p) > 10.
Population of GOIO players are greater than 460.
Hypothesis
Ho: p = .33333
Ha: p > .33333
Where p is the true proportion of GOIO players who prefer spanner-mallet.
Calculations
z= (P^ - P) / sqrt( P(1 - P) / n ) = .3779644.
p value of .0722.
Conclusion
Since p value is less than .1 (hence marginally significant), I reject the Ho. There is sufficient evidence to suggest that the true proportion of GOIO players who prefer spanner mallet is greater than .333333.
This is a little confusing (even for me who is currently learning Statistics in school), and a slightly flawed here and there, but all of it summarized to suggest that there are more players who prefer spanner-mallet than the other two choices. So, spanner-mallet!
EDIT: as more people vote, my calculation will be off. The above was calculated when there were 20 spanner-mallet, 16 mallet-spanner, and 6 others. The null hypothesis was that all three categories were preferred equally.