Take the 2-minute tour ×
Game Development Stack Exchange is a question and answer site for professional and independent game developers. It's 100% free, no registration required.

How is game difficulty typically tested or balanced so that there's a steady progression of difficulty; how do I ensure that no level is too difficult or too easy for its point in the game?

Are play testers of various skills brought in, or a random sampling of people? What considerations are/should be taken into account during the actual level design process?

share|improve this question

6 Answers 6

Two big helps in difficulty tuning.

Metrics

Tracking player data can go a long way to making difficult tuning as objective a process as possible. How long are people staying alive, where are the taking damage or getting killed, how many tries does a section of the level take. Getting good metric data can really help your team see where people are having issues in your game.

Kleenex Testers

The longer your QA staff plays the game the better they get at it, so it's good to have sessions where you grab new people who have never sat down with the game and have them go through it.

share|improve this answer
2  
Metrics - try heatmaps on where they die gamedev.stackexchange.com/questions/157/… –  Bryan Denny Jul 22 '10 at 12:46

Watch someone play but don't talk to them. By not talking to them, you can see things them do things that will make you pull you hair (You: the solution is obvious. What's wrong with you? #$@!).

Instead of saying your thoughts aloud, write down what you want to say. Use this as your basis to guide your game difficulty.

share|improve this answer

Ensure that you change the people playtesting your game. We shipped a title for the original PSX that was far too difficult (far too tight time limits) as we'd had all the same playtesters testing the game over the development period so they knew the levels backwards.

share|improve this answer

Some games adjust the difficulty based on the skill of player during the game.

For example every time a player dies at a level - difficulty decreases. Every time he finishes - difficulty increases.

Still this needs some testing to get the decreasing and increasing right.

This method makes the scoring a little less accurate and harder to compare with others but the game more enjoyable as you won't get totally stuck.

share|improve this answer
2  
I personally hate this technique and would recommend against it. Sometimes I'm in the mood for an easy game, sometimes I'm in the mood for a hard game. Autoadjusting difficulty not only deprives me of that choice, but sometimes ends up screwing me over based on the game designer's preconceptions of what's "hard" and what's "easy" - I have non-fond memories of getting obliterated in Privateer 2 because I was much better at making money than dogfighting. (Guess which part governed the global difficulty level.) –  ZorbaTHut Jul 21 '10 at 23:15
1  
The only kind of version that I find acceptable is if you die enough the game either recommends turning the difficulty down in a menu, or the game unlocks a lower difficulty level. Having the game dynamically change difficulty is frustrating. –  AttackingHobo Nov 7 '10 at 1:12
1  
I hate this too, but I'm not sure why the answer was downvoted, as it is a valid method for balancing. Thus it gets +1 from me. It could also be used purely during the testing stage alongside good metrics to automatically find the correct level to pitch levels at. –  Kylotan Apr 21 '11 at 23:54
    
I would suggest leaving the usual difficulty levels alone and also have an "automatic" difficulty level too. –  Calmarius Nov 20 '13 at 16:08

an idea I've seen mentioned elsewhere is to sit pairs of people in front of the application[game in this case]. They're more likely to reveal their thought processes to someone else watching and the 2nd person may ask questions without leading them like you or another team member might.

share|improve this answer

First a quality assurance pass is needed to make sure the tasks are reasonable and completable by your internal QA.

Now do a sweep on initial difficulty and/or level times, rinse and repeat until diminishing returns mean you need to move on.

The next step is to use all the other internal staff you can grab for a few play-thoughs. Be sure to record as much as possible for later analysis. Sweep, iterate, repeat.

Families are a good next line if your on a budget, after that you need to bring in focus groups or outsource testing.

Remember do small chunks, iterate and repeat.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.