In his article about vanity, Alex mentioned the importance of playtesting and acquiring useful information from playtests. In this article, I’ll give some examples from the recent open playtest of the next edition of Dungeons and Dragons and how the principles applied there carry over to strategy game design. While the goals in designing a roleplaying game aren’t the same as for making a board or card game, some of the lessons learned in the D&D Next playtest are valuable ones for all sorts of game design.
First, well designed games can (nearly) be played right out of the box. One of the key lessons that the design team learned was that players have certain pre-conceived notions about how a game ought to operate. In a strategy game, these ideas might be based on the game’s theme, its genre, or even its designer. Players might expect a game about empire building to reward controlling a large amount of territory on the board, or a game with colorful cards to contain a set collection mechanic, or a game by Stefan Feld to involve a menu of available actions based on rolling dice.
One of the best things a game designer can do, in turn, is to play up to those expectations. Listen to the feedback generated during the playtest to figure out what your players expect to be able to do in a game, and make sure the actual game play experience is consistent with those expectations. That leads into another important playtest lesson learned: mechanics should feel like tools for the players to use, not limitations on what they’re allowed to do.
In a previous discussion on the topic, I discussed Settlers of Catan as an example of a game that adheres to this precept very well. Settlers is a territory control game, so it makes sense that the mechanics enable the players to conquer territory and that the scoring rewards it. On the other hand, it’s not a war game, so there doesn’t need to be a way to invade other territories. Settlers is additionally a resource management game, so the trading mechanic is exactly what you might expect it to be. Plenty of games that also involve resource management don’t support–or expressly disallow–trading. Just because your game involves managing resources doesn’t mean you must include a trading mechanic. But your players might strongly expect one, and careful listening during the playtest is the best way to hear about it.
Another important goal for playtesting is to examine your game on multiple levels. On one level, the numbers need to work; the game needs to be mechanically balanced. To achieve numerical balance, the D&D Next playtest team used a “stress test,” in which the game was stripped of all its thematic trappings and run repeatedly to fix underlying math and smooth out rules contradictions. In the stress test mode of playtesting, the designer’s goal is to examine a single situation or mechanic out of the context of the rest of the game, try to break it, and make sure it stays intact.
On the higher level, a playtest can reveal how the players react to a game’s theme, and whether the theme fits the mechanics. (The idea of theme integrating with mechanics is a fascinating concept that Games Precipice will revisit later this year.) A game needs to have numerical balance to work, but a truly great game is one that elevates those numbers to mean something within the game. Tongiaki is a simple game with essentially one mechanic: when ships leave a port, there needs to be some sufficient (randomly decided) number of ships in the fleet for the ships to arrive safely at the next island. It’s fine to think of that number as a source of mathematical variance only, but it’s much more fun to think of it as experienced captains from different tribes working together to brave the currents.
Finally, the most important thing a playtest can reveal is to determine if the game is fun. Fun is the single most important quality a game can possess and the one that’s most difficult to assess simply with a rulebook and components. To that end, the D&D Next playtest team found that a mathematically perfect game is not the same as a good game. A designer can balance probabilities and effects on paper all day long, but if it’s not an enjoyable game to play–if the players aren’t having any fun–then the design isn’t any good. Of course, fun is subjective, and different groups of testers will have fun in different settings. Intelligent playtests allow you to narrow down your audience and pitch the game to the players who will like it most.
In conclusion, the D&D Next team was able to use a massive, iterative playtest to develop a version of the game that appealed to the broadest consensus of fans. Most designers probably can’t enlist thousands of testers to play every revision of their game and fill out detailed surveys about it. But the principles used and lessons learned can carry over for strategy game designers: figure out what your players expect to be able to do, and implement it (or have a good reason for not implementing it); tune the game to work on both a mathematical and a thematic level; and above all make sure the players are having fun.
Thanks! There are definitely places for both the “this game is fun because it is perfectly balanced” school of thought (I suspect a lot of chess players would make that argument) and the “this game is fun precisely because it ignores conventional numerical balance” idea (which is why Fluxx is so popular).
Your comment raises two great points about playtesting: first, the first draft of a game will probably not have perfect balance. Figure out if the players aren’t having fun because the game is imbalanced or if they’re enjoying themselves anyway. There is no way that Betrayal at House on the Hill left the Avalon Hill offices with anyone having any idea if Haunt 3 is as fair as Haunt 17 to both the heroes and the traitor. But that issue probably did not arise during playtesting, since Betrayal is a game whose fun factor derives from the experience and the interaction rather than precise balancing.
On the other hand, even the most beautiful and immersive game in the world will not get played if the “game” part of it is fundamentally flawed. Like you said, aesthetics are a way to enhance a game and make it memorable, not hide broken underlying mechanics.
A good modern game design will probably fall between the chess and the Fluxx ends of the spectrum, and the playtest is a great time to determine where a game best fits.
Great article! I sometimes feel balancing comes a close second to the fun factor myself, but then that can be used as an excuse when a badly designed game masks it’s short comings with glitzy packaging and an engaging theme.