Here are some stats on how players fared using own v opponents' armies from the data I had:
Four players used own and opponents' armies three times each. Three players used own armies 4 times and opponents' two times Three players used own armies two times and opponents' four times.
There were 9 games when players won with own armies There were 21 games when players won with opponents armies Players lost with own armies 19 times and lost with opponents' armies 6 times.
There were 4 draws with own armies and 6 draws when using opponents' armies.
These are based on 11 score sheets out of 14, one missing data on one round.
This seems to go against intuition that players would do better with armies they know. I think that what be could happening is that the better players are able to pick the stronger army in their opponent's pair - it was interesting to note that the ratio of games won with opponents' armies to games won with own armies was higher amongst the players doing better in the tournament.