next »

[TheClintHennesy]

Hey guys. Serious question here.

We've been working on this game for around 3 months or so now. One problem was always difficulty scaling.

When we get newer people to try our game- it's always TOO DIFFICULT for them. But it's also very difficult for us devs to gauge what is hard and what is easy because we play the game SO MUCH times- it's just all too easy.

I want to know what are your solutions for this. :-(

Hopefully it will help others as well who are having a hard time with it.

[ThemsAllTook]

This is a very common problem. Different tuning techniques will apply to different game genres.

Most recently, I've been working on level design in a puzzle game. A technique that I used to gauge difficulty is to consciously try to put my foreknowledge of the puzzle solution aside, and simply make observations about what I see in the room I've built. Starting from the basic stuff like "I see that my goal is here; this is in my way, and to get past it I need to do this other thing first". Parsing my levels this way helps me notice when I have one that's way more complicated or subtle than the other things near it, so that I can redesign it to be gentler.

Playtester feedback takes some work to interpret into something you can use. Keep a clear goal in mind for how you want the game to feel. How many times is it OK for a player to fail at something before figuring it out? Do they always get to see clearly why they failed? In the ideal case, every failure should give players a clear understanding of what they did to cause it and how to avoid it next time, and it should never be an arbitrary punishment by the game that they couldn't control themselves.

Levels that are fun for the developer are often most appropriately placed at the very end of the game. You might design the last level first. Since you've skipped straight past the learning phase in your own game by inventing the mechanics and therefore understanding them before you've had a chance to play with them, your default skill level is going to be a lot higher than someone seeing your game for the first time. To design earlier levels, look through the lens of a first-timer and introduce concepts at a measured pace - give some room to learn and master one thing before moving on to the next. With a gentle buildup of concepts and mechanics, you can guide the player to the level of knowledge you started at by the time they reach the later levels where the real fun lies. Not to say that early levels can't be fun and challenging, but some constraints in the amount of understanding they require would usually be a positive thing.

I'd try to give some more specific advice if I understood better how your game plays. I had a peek at the devlog, and it sounds like a pretty neat concept. Hoping to see some more detail there sometime soon.

[TheClintHennesy]

This is a very common problem. Different tuning techniques will apply to different game genres.

Most recently, I've been working on level design in a puzzle game. A technique that I used to gauge difficulty is to consciously try to put my foreknowledge of the puzzle solution aside, and simply make observations about what I see in the room I've built. Starting from the basic stuff like "I see that my goal is here; this is in my way, and to get past it I need to do this other thing first". Parsing my levels this way helps me notice when I have one that's way more complicated or subtle than the other things near it, so that I can redesign it to be gentler.

Playtester feedback takes some work to interpret into something you can use. Keep a clear goal in mind for how you want the game to feel. How many times is it OK for a player to fail at something before figuring it out? Do they always get to see clearly why they failed? In the ideal case, every failure should give players a clear understanding of what they did to cause it and how to avoid it next time, and it should never be an arbitrary punishment by the game that they couldn't control themselves.

Levels that are fun for the developer are often most appropriately placed at the very end of the game. You might design the last level first. Since you've skipped straight past the learning phase in your own game by inventing the mechanics and therefore understanding them before you've had a chance to play with them, your default skill level is going to be a lot higher than someone seeing your game for the first time. To design earlier levels, look through the lens of a first-timer and introduce concepts at a measured pace - give some room to learn and master one thing before moving on to the next. With a gentle buildup of concepts and mechanics, you can guide the player to the level of knowledge you started at by the time they reach the later levels where the real fun lies. Not to say that early levels can't be fun and challenging, but some constraints in the amount of understanding they require would usually be a positive thing.

I'd try to give some more specific advice if I understood better how your game plays. I had a peek at the devlog, and it sounds like a pretty neat concept. Hoping to see some more detail there sometime soon.

Hey Man. I really appreciate the feedback.

I'm gonna try to pull our game designer into this forum to read what you said (haha.)

So if I understand correctly- Does this mean that majority of the "tutorial" levels have to very very easy?

We've always sort of been conflicted with 2 philosophies. Since we're working with a demo- we want to keep the first 15-20 minutes as exciting as possible. We want to show that there's a LOT more to do upcoming that level. The problem is is keeping those 15-20 minutes exciting but not frustrating.

There were also times where we tried to "pad down" difficulty by adding more content with very similar encounters. What ended up happening was it was too easy it became a chore for the players to do the content.

What we're trying to understand as well is how do games keep repeating content enjoyable. Hmm... I'm trying to think of ways to really open this up here in TIGSource cuz we have a lot of struggles developing the game that I think if we could show and ask people about it, we might really make the game 10x better. haha.

Again. Thanks so much. >.<

I really appreciate the response.

[MSThalamus]

I'm not sure how popular this suggestion may be, but if you're referring to combat difficulty (and assuming it's a single player game), you could do what they (eventually) did with Diablo III: have difficulty levels, and allow the player to switch on the fly, though I think a difficulty slider rather than three discrete levels might be better. Implementation here is to a large extent just a value applied to the combat math at game time. A slider could also be a good solution if you want the game to be friendly to people who have never played the genre before but to still be interesting to veterans. This is how I'm planning to approach it.

If you're referring to puzzles, as ThemsAllTook mentioned... that's a lot harder to gauge, and there's no straight math that could be applied to puzzles (in general) to make them easier or harder. My solution is to make the puzzles completely optional. The player gets a cool bonus for completing them, but there's no harm in skipping them. My game is a story-driven ARPG, though, not a puzzle game, so that solution obviously wouldn't help someone making a puzzle game.

Hope this is somewhat helpful.

[Morse]

Another way would be to find a single a or a couple of persons to playtest it, considerably a noob to the game, but not to the gaming in general, someone who plays games. Let them play 3 times, Record the time spent on each level he/she played.
Do the same for yourself and compare the timing, to get an average ratio coefficient between the pro and noob plays. Do a simple survey with the noob(s), just to get more details about what he/she found hard, frustrating, easy, interesting etc.

That way you'll get an idea of how hard the level is based on your own playtest.