One writer, Jason Alt, promotes building your deck to about 75% power level. I really like optimizing decks, however, I realize I could easily steamroll over other players doing that if I am not careful. As such, rather than building to 75% I like to pick a "75% strategy", as in a not commonly used general, or taking a more commonly used general in a different direction. This leads to my decks being at roughly this 75% power level, without sacrificing the optimization I really like.
Beyond this I've always adhered to a few general theories for multiplayer:
- synergy over raw power: it's very rare for me to play cards just because they are good, but they will get in if they synergize with the rest of the deck and you'll have better draws in the late game
- if you're playing with a "
gentlemanly" group: skimp a little on the land count as they're bound to not take full advantage of a mana screw
- try to hide as much functionality as possible in lands: mass LD is frowned upon in such "gentlemanly" groups, I really like to have man lands in my decks for example
- aim for the long game: I may not have a strong plan early game, but again counting on "gentlemanly" opponents I may very well get away with that, however you can pretty much count on me having a plan to take over the game if it runs long
- play politics/diplomacy even beyond the current game: also known as influencing the meta game, I will carefully craft a reputation such as that I will come for you if you come for me if I can (discouraging players from just hitting me when they can, supporting "gentlemanly" behavior), I've even gone so far as to let Magic games influence other games played with the same people and vise versa
Though these theories may not look like it, I always try to encourage fun and be a pleasant player to have at the table.