Level flow and progression?

[Atomic_Sock]

Hello. I'm a pure newbie on these boards, but I had a question, and was wondering what everyone's take was. I'm working on a casual wordgame, and I'm having a little difficulty nailing down a proper level progression.

The level progression is point-based, and right now the problem is that it ramps up too quickly. Does anyone know of any good formulae or rules of thumb to apply when considering level progression and flow? Any advice would be appreciated.

[Coyote]

The most popular one you find nowadays that ramps up but doesn't go out of control is:

[Points Needed to Reach Current Level] + [Current Level x Constant]

So if it's a 100 point constant, you get:

Level 1: 0 points
Level 2: 100 points
Level 3: 300 points
Level 4: 600 points
Level 5: 1000 points
Level 6: 1500 points
Level 7: 2100 points
Level 8: 2800 points

And so forth. Unlike linear, progression slows, but it not anything nearly as bad as an X-Squared progression.

[svero]

Well I use to try to base my level progressions starting with hardest. Ie.. I'd setup the hardest level I wanted.. the maximum difficulty and skill I expected a player to attain and then try to graduate that back to the easiest skill level interpolating it. But that's not an entirely good strategy as it turns out because the easiest skill level might be too easy, if it's geared towards a player getting a feel for the controls etc... and that can make the early part of the game a little dull. Nowdays I like to do a similar thing but try to staircase the difficulty (imagine a graph that looks like sloped stairs going up to increasing difficulty instead of a smooth slope) a bit and combine that with player testing. So instead of doing a gradual jump from level to level I'd split the levels into sets of x levels and do gradual increases within sets with bigger increases jumping from set to set.

But all that is just general rules of thumb to get a feel for the overall game layout. And that style of tweaking really works best when you have a clear definition of what difficulty is. In some games you can just increase the speed a little every level and that leaves you open to interpolation for tweaking, but in other games it's a lot trickier to judge. Is puzzle a harder than b? Does adding 10% more q's make a 10% harder game? Not clear. So from that base I think you need to go in and do a lot of playtesting and tweaking. It's a starting point though. The staircasing difficulty can also help you mitigate the difficulty ramping up too quickly.

[Atomic_Sock]

The formula that coyote posted is very much the sort of thing that I'm looking for. I initially made the attempt using your method, svero, but I found that the end was too hard, the beginning too easy, and the middle part where it was about right didn't last long enough. Then we tried going with an X-squared progression, and by the end of the game, it's out of control.

Anyway, thanks to those who have already responded. I appreciate the advice.

[Huge]

I think you will find that that formula will end up like x-squared as your level gets higher. Keep the levels smallish (as shown here) and you should be ok.
Still better than 2^x, where you must "do it all again" to go up 1 level.

[jefferytitan]

I don't have any solid answers for you, but if you're looking for a function to correlate the level and the score you should need to finish it, a few suggestions are:

Cubic (shifted up and right). The inflexion point in the middle means that you spend plenty of time in the middle zone. However the difficulty changes quickly both at the beginning and at the end.

Logarithmic. Starts quickly, slows down.

Sigmoid (s shaped). Starts slowly, finishes slowly, changes quickly in the middle.