I attended a talk by John Conway recently where he explained to us that the usual number, 26, was wrong, that there are in fact 27 sporadic simple groups. His reason was that the Tits group, which is often claimed to be a group of Lie type is not.

The reasoning seems pretty sound; it is usually lumped in with the groups ${}^2F_4(2^{2n+1})$, but these are only simple for $n \geq 1$: ${}^2F_4(2)$ is not simple.

The Tits group is an index 2 subgroup of this group, and it is simple. So it is not of this form, no matter how "almost" it is. Moreover, these groups all have a BN-pair, while the Tits group does not.

So to me it seems clear that there are 27 sporadic groups, not 26.

What I'd like to understand is this: What is the reasoning to include it in the infinite family ${}^2F_4(2^{2n+1})$? There must be some reasons that people do list it as a member of that family other than surface similarity.