I can't say I found these results all that surprising; in my experience of quizzing the most common team size limit is six, and I doubt that's a coincidence. Still, it's mildly interesting to note that even numbers seem to be preferred, while it's good to see the often under-represented non-integer team size getting a mention as well.
|Not pictured: pretending it isn't a problem|
The latter case is probably more easily dealt with: one ad hoc fix is to let big groups play but explain they can't win the top prize (or, as I've seen at some venues, have a separate 'big team prize'). Another quite common choice is to ask the big team to split in two, although this needs a quizmaster who's either alert to collusion or good at persuading them to be competitive.
Handling a situation where large teams are an established part of the quiz requires more creativity. A careful balance has to be struck between the smaller teams feeling hard done by without the larger teams (and their all-important larger spending power) feeling unfairly targeted. The one successful solution I've seen (and discussed in the comments here) are multi-answer questions where the number of required answers is the same as your team size. ("For five points, name as many landlocked African countries as your team size", for example.) Choosing the correct questions for this is difficult, but it ties the difficulty directly to your team size making it feel perfectly fair.
If you have any other suggestions to balancing team sizes do let me know, either in the comments or on Twitter, and keep an eye out for another Pub Quiz Poll coming soon!