A couple of weeks ago I did a post about solo playtesting, while hinting that it was part of a series. So, finally, here is part two, about taught playtesting.
Once you've got the prototype into a workable state (i.e. it's not totally broken, or dull!). The next stage is to play it with others. In the first instance, it's best to just play the game with you teaching it to the other players.
Why bring others on board? Two main reasons: other impressions and other strategies. Playing the game by yourself is not a huge heap of fun, especially if the game has any hidden information (you have to pretend you don't know what's in the other players' hands and what hidden actions they've taken). With real players it's easier to see how the game works when the information is truly hidden. You know what you think about the game, what about everyone else? Publishing is all about guessing whether other people will like a game enough to buy it. The more accurate your guess the fewer duds you'll publish and the better hits you'll get. Playtesting lets you get a sample of the gamer public and see what they think. If they all love it, then hopefully it will be a success, if they're not bothered or hate it then maybe you're best off dumping that game, or significantly re-working it.
I should note at this point that the playtesters' feedback is a sample of the public as a whole, but they are not a random sample and like any other sample can be affected by bias. This is especially true when you do the majority of your playtesting with friends and family. Don't be surprised when they love your game. They want to support you and will do that by telling you the game is great and asking to play it again. This bias is why the next phase is so important. At this point the game is likely to change every time you play it, or you might be wheeling out something that sounded good on paper but when you play it is shockingly bad. Friends and family are great for this, as they have more patience with your misses than the gaming public. At this point honest friends are worth their weight in gold. Someone who can play your latest darling and say: 'Man! That was shit! I mean _really_ bad.' to your face is much more useful than someone who thinks it and doesn't say it or is just predisposed to liking it because it's one of yours.
The other big advantage of playing with other people is getting other ideas and strategies. You've played it a lot solo, it's great. It works really neatly. They you play it with Bob and he uses a different strategy, one you'd not considered. All of a sudden your great little game is broken :-( Yeay Bob! Much better you find this out before you invest thousands of pounds (dollars, euros, etc.) in the manufacture of the game - it's not too late to make changes at this point. Fix the problem and try again.
While you are still making a lot of changes time spent writing a great set of comprehensive rules is probably time wasted. Why invest time writing a set of rules with lots of diagrams explaining move 'A' if when you finally play that version, it turns out that 'A' is actually a bit weak and you have to replace it with move 'B' instead? While the rules are very fluid it's not worth the effort creating an awesome set of rules and hence you can't blind playtest it (more on this next time). So you have to teach it to your testers. Taught playtesting isn't the be all and end all, you have to taken into account the bias in the feedback you get and obviously without a set of written rules to learn from you're not going to get much in the way of rules feedback. But it serves a purpose, and it's very valuable if you are aware of its pitfalls.