Changing rules in a playtest

I change the rules. I try to do so in a way that won't hurt any particular player, but everyone playtesting knows that they're not there to win, so even that's not usually a problem.

Mind you there are rules changes I've made that only make sense if they're in place from the beginning of the game, so those certainly don't go into effect until the next play.

And there are also plenty of times I want to see the game play out with the existing rules, even though I know I'll change something next time. There are sometimes too many interactions I want to observe to not let it play through.

-- Matthew

As the others stated, I think it depends on the change, but for me it also depends on the problem.

First, I would decide how major the problem is and how much changing the rule would change the game. There are some rules that might be very easy to modify without breaking other parts of a game. While some changes can require a large rewriting. I would assess the impact and determine if altering the rule mid game or not was possible.

During a playtest session I also keep in mind three things :

1) Is a change needed because the players dont understand a rule or mechanic?

2) Is a change needed because the outcome or use of a rule/mechanic is not what I expected, wanted, or is WAY off base?

3) Is a change needed because I notice something in the game play, that I dont like, but goes unnoticed by the playtesters?

I am sure there are more then these 3 items that I use to consider rule changes, but these three I feel our the key during playtest sessions.

The first one needs to be fixed then and there, either explain it and create additional rule information that helps players understand, or you have to consider changing it. If rules/mechanics are not understood a game cant be played.

The second, might be the biggest problem. If your rules or mechanics dont seem to be working right, it might be hard to change these on the fly and keep a stable playtest going.

The third, is personal preference problem, after the playtest I would ask the playtesters opinion of what I noticed. Maybe the game is fine for others and not for me. So I would consider the audience I am writing the game for and if I feel the changes are needed, I would introduce them at a later playtest session. Once I had a chance to work the new changes in.

If you're testing something that will only be played once during that playtest session, then I agree with everyone who says "change on the fly". My games tend to be shorter than that, however, where the testers will try 3-5 times at a sitting. (I like games that inspire the "I'll get you next time" feeling, so if testers play once and have "had enough", I know the game doesn't work).

In case of shorter games, finish the game you're playing with the broken rule. It may be painful, but it's very important to see just how it's broken, and to not change too much.

For example, a card game that is in playtesting at LOG right had a session this week where everyone (myself included) decided that the "end-game" mechanism was totally broken for the number of players involved. That was the only major complaint, but there was also a minor issue with turn sequencing.

After we finished the first game (and yes, it was painful to do), I changed the turn-sequencing issue. Suddenly, the "end-game" problem didn't exist. In fact, the overwhelming opinion was that it should not be changed. This wasn't due to any genius on my part. I simply thought it would be easier to try the minor change first. Since then, the game is much improved, and it's received nothing but high marks for the end game.

So... my suggestion: during playtests, try to only make one change at a time, and do it between games. Of course, this isn't always feasible, but it gives you the best sense of what's really going on with the mechanic.