Probably should make sure that the player played a minimum of games in that season too. While I'd love to take Rod Brookin's sophomore year where he shot 53.8% from three I don't think that 5 games is a representative sample size.
There have been at least two guys picked based on seasons that included less than half of their team's games that year. I would think that that is something that would be taken into account when thinking about which team would win a particular matchup, but who knows?