Hi guys,

I would just like to know what you guys do to try and predict whether a game concept and initial design will lead to a fun game, before it is prototyped and play tested? Are there any key things that you think is essential to have, any paper/document based playtesting methods you employ to test out the gameplay etc? I have heard, and also believe that developing a game heavily involves testing and refining, but it would be great to have some sort of metrics during the start of the design phase to skip straight to a semi-decent design before a prototype is made. Looking forward to hearing from you, God bless

I personally always find it hard to predict. There are most likely a couple of guidelines you will be able to find in books that will tell you some best practice, but I personally think that you will find that out yourself over time when you have done some games and experienced first hand what works and what doesn't.

I usually start with something I think might be fun and ask people what they think of it, if I make a game like that, would you play it? One of the hardest factors is that people generally have a different taste in what they fun. I don't like soccer games, I don't enjoy them, but there are plenty who do.

What also helps is simply looking at other games, what made people want to play that game? what made it so special?

If you don't mind the read, I think The Art of Game Design: A Book of Lenses by Jesse Schell and A Theory of Fun by Raph Koster are a must for every designer out there.

Hope it helps

Just having played a ton of similar games, and analyzed relevant mechanics, you will often be able to recognize straight away that a design sucks.
Eliminating bad stuff at least gives you better chance of hitting something good when you start playtesting or prototyping.
I doubt there are many formal processes to prototyping. Just isolate key parts of the design as well as you can, and use a minimal prototype. You might be able to emulate some turn-based battle system with Legos or Post-It notes, or play a game of tag within Super Mario or Quake, whereas figuring out novel physics for a platformer like N requires a software prototype. Whatever works.

Yes, it's very easy to assure a game is fun before playtesting. Take an already existing fun game and clone it
Smarter ones might take an already existing fun game, analyze what parts of it are fun and what are unfun, then clone it while removing the unfun parts and strenghtening fun parts (Blizzard is the master of this approach). The trickier and riskier approach is to take two or more already existing games and combine and clone these (althrough here the fun is not so assured unless you know what you are doing quite well).

Of course "cloning" is just a word, in reality we design in a fuzzy state between "clone" and "original". The more in the "clone" part you are the more predictable the fun is, the more into the "original" part you are the more unknowns and risk for unfun game and more needs for playtesting.

The ultimate fully original approach basicly assure unfun and broken game, unless you are a genius (I know only one game that was designed 100% original in the last decades, it was Tetris, the rest were taking the best parts from already existing games).

Or to put it other way, if you borrow from an existing game, it was already playtested, so you don't need to do it again. The trick is we don't want to blantly copy it (exceptions), but to improve it, so we want to add originality as well. And we absolutely have to playtest the original parts.


Also, as Stroppy Katamari said, while it is not trivial to say if a given design will result in a fun game, it is almost always very easy to determine if a certain design will result in a very boring game. The bad designs are easy to spot, unfortunatelly it is not easy to distinguish a good from excellent designs without playtesting. And that's the problem, because good is usually not good enough

Near impossible to predict. Drastic changes can occur up to the very last minute.

E.g. Warcraft 3 was a very different game during its beta. It played like Starcraft. There was a major overhaul 3 weeks before release, which is the version we see today. I am pretty sure they couldn't predict the final gameplay right from the start.

How would you know if a book you were going to write would be entertaining? Typically there's a central theme where characters or other elements (maybe a particular gimmick the writer thought of) are gradually exposed to the reader. There's usually some sort of conflict or challenge to overcome. At some point you discover the results of that challenge and then maybe move on to the next one which is presumably either tougher to overcome or just a progression from the first.

Over all, I'd say that most games follow that sort of pattern in some way. It has some sort of theme, maybe a central mechanic and some characters you're introduced to. There's a challenge. Result. Repeat.

However, just like a book, it's hard to say if the various elements will go together well or how the author's presentation of them will be received until you have some sort of prototype to try out on an audience. Preferably someone that can take into account that it's an unfinished product.

If you have a target audience or genre you can do a survey of what is usually liked by that audience/players of that genre. Any other answers I gave, like give the game a strong coherent theme, would only be true for the audience segment I'm in.

How would you know if a book you were going to write would be entertaining? Typically there's a central theme

Heh we posted about the same thing at the same time. ;)

I think it's a combination of being easy to use, well designed interface, interesting environment and story, and solid game mechanics.

Do the opposite of: looking terrible, being very complex, not that interesting, and crappy combat.

When I have an idea for a game I don't think about it in terms of "will this be fun?" Instead I think about the mechanics and the core dynamic of the game and ask myself a series of questions: Will this be engaging (requiring skill and awareness)? What kind of potential situations could I put the player in? Are those situations immediately interesting to me? As well as a series of "if mechanic X does this, what would be all the possible consequences of that action?" As well as other questions pertaining to the specific idea.

If those questions lead to interesting answers showing that individual parts of the idea have potential than I think the game could be fun.

Much of it is kind of up to what you think would be fun and not having built anything to test the ideas it helps (at least for me) to break down the idea into more focused assessments to see if maybe there is potential.