Ckutalik over at the Hill Cantons asked every GM’s three best practices. I wrote an almost answer years ago because, you know, I can see the future. But I want to give better answers this time around. Or try at least.
- Strip mine your surroundings for ideas and turn them into archetypical subverted caricatures: I really really like creating NPCs out of people by making them a subverted caricature of the original to make them into a different entity. The same can be done with anything else: places, events, items. Caricatures magnify what’s important about the entity but the subversion gives it deeper, interesting shades. For example: I need a merchant city, so I pick Dunbarton, put it in the middle of Genoa and make it ruled by a never-seen immortal queen and her secret police. When I needed a shady broker I took my dad (a salesman) and gave him a gang of thugs. My demihuman races have Italian regional accents, but have magnified idiosincrasies (my halflings are more halflingish than yours, and speak Sicilian). Filthy Phil (a really stinky ragman and garbage-sifter in my current setting) is inspired to… well, I like dumpster diving (the best stuff you can get is free). The beautiful thing is that you can do this on the fly in games: players usually are fascinated by deep NPCs and they think deep NPCs must be important hence worth their time: oftentimes they just like them, especially if you play gonzoish adventures. Which gives me enough time to actually link them with the rest of the setting in (you guess it) subverted caricatures of existing relationships.
- Evolutionary game elements rock: the reasoning is that of the 10 campaign elements (npcs, odd objects, etc) you introduced in the last settings, players will pursue maybe 2. They will pursue the most interesting, because they don’t care about the others. So if you improvise you can just come up with a good/boring element ratio of 1/4, the players will kill the bad ideas by not caring about them. The morale is: stop fretting about the quality of your improv, just care about players’ feedback. If you only come up with amazing ideas, the players will pick only the most interesting so, instead of having gold and platinum in your game, you’ll have only platinum. Of course this works only if you are not a control-freak railroader. There is only trying, diversity is king, survival of the most adaptable and so on.
- Play with your players: the game is yours, but players live in it. I suppose you don’t hate them, so your goal is to have fun with them, not at their expenses. I’m not saying to indulge in monty haulism, but if you have a choice between doing what you want and doing something that your players find interesting (and doesn’t irk you), go for the second one: players will feel engaged (read point 2 above) and they will steer the game for a while, giving you overworked GM time for thinking about other stuff. It works best when a player goes like “I wonder if Mr X is obsessed with orchids because of an ex lover or something” and you apply point 2 and yeah, it’s because his mom loved them and he has an Oedipus complex. If your bud Dan feels supergood when his PC saves a young girl and gets closely but not necessarily sexually involved with her, you have a major player driver in your arsenal that you can use every time without fault: remember to apply subversion from point 2 above to keep things different. More in general, Intuitive Continuity works extremely well, if you don’t need to have your ego fed at every step of the game.
In general, I favour player centered design: it works with software, why not for D&D?