I'm not 100% sure, I only have two inputs on the recorder I use right now and normally build my crowds with a combination of mostly mono-files, with both X/Y, ORTF and A-B added to the mix as I feel it fits for the scene, but I do have a pretty clear view of how I would probably do, considering the mikes I have:
First, a pair of Sennheiser MKH 40 cardioid mics in ORTF 90 degree-formation perfectly centered from the crowd, hanging from the ceiling, pointing at the the middle of the front half of the crowd. I chose the ORTF over X/Y as I actually do not want a center here, this sounds like a bigger event than, for example, a club, and in that case a center tends to narrow the place down.
Then a pair of Octava 012 in A-B formation, on each side of the scene, pointing pretty much like the Sennheisers, with an angle of 90 degrees between them (they should be way to far apart to risk phasing from this distance). The octavas are a lot more fuzzy than the sennheisers, but give a very good body to the sound :-)
I assume there are at least 6 channels to play with here as you wanna have distinct voices in the mix, so the final two channels I use for two Sennheiser MKH 416. As I've kept the MKH 40's and MC 012 quite distant, phasing SHOULDN'T be an issue with the 416's now. Actually, that's why I've chosen lobar-mics here, normally I use the 416's from a distance high above (like the character an clarity of 'em)!
The 416's I would position facing right ahead, about in the middle of the different stereo-pairs on each side.
I'm not totally sure about phasing with the shotguns in this configuration, but should you use it and get phasing after all, just try different position till you get one that works :-)