I'm working on a game in the abstract puzzle genre and am at the point at which I'm basically calibrating it, meaning it is code complete but its gameplay is defined by a set of parameters for each level and I'm trying to nail down those parameters. I'm trying, obviously, to set things up such that the game starts out kind of easy but gets progressively harder.

So I'm running into two problems.

Problem (1) is balancing easiness of early levels with interestingness. Obviously, I want to start out easy and get hard, but if I start out too easy I worry that the user won't think the game is interesting enough to continue playing.

Is there any way you can tweak the design to make the game progress faster when the player is good? Like if you just put down pieces (or whatever you do in this game) you'll be playing level 1 for a while, but if you use your pieces efficiently to create a good combo, the points acquired from that will propel you directly to level 2 or 3. I think this would probably be best for a puzzle game.

The other thing that comes to mind is the arcade staple: even if the start of a first level of a shoot'em up is typically pretty easy to survive in, the scoring system is built in a fashion that makes the level hard to survive for a player looking to get a great score.

A few thoughts based on my own recent puzzle game:

If the game gets harder over time and the beginning is too easy for more advanced players:

• Allow the player to select what difficulty/speed to start on. The original Tetris had this.
• Use the slow/easy phase of game play more efficiently or with a different strategy than other game phases.
• End the game at a specific point in time.
• Give the player a way to score more points off of better play during the beginning phase.
• Judge the player based on some other metric than just raw points.

Explained:

My game starts out slowish and speeds up progressively over time and will end, in theory, after 20 minutes max. That's the point at which the game goes infinitely fast. The goal of the game then is to get the most points within that very limited frame of time. I think this mechanic works really well, as opposed to games that just keep going indefinitely. It's also better for short attention spans.

Because the time effectively stops in the 15-20 minute range, my game forces the player to cope with varying levels of speed in the course of one game. This neatly divides the game into phases as i experience it. I have to take advantage of the slow beginning period to build a reserve of collectible special objects that will be used to survive during the ending phase of the game (at least that's my current strategy). The point is that i actually have to plan my strategy for the whole game before i even start. I know that i need to store specials in the beginning, ramp up my point scoring multiplier and play efficiently in the mid-game, and use everything i've stored up to survive as long as possible into the end game. The end game is where i score the bulk of the game's points so i have to look good going into the end or i won't get a high score.

Another Idea:

My game also has an Awesomeness Rating. It's based on your time, hits, and score, averaged out to a percentage. This way you can play for 5 minutes on the worst difficulty level, get brutalized, and still score a high Awesomeness, while you can also play on normal difficulty for ~16 minutes and play pretty well and get a lower Awesomeness than on the harder difficulty level.

I gave all the difficulty levels an Awesomeness Bonus to compensate, too.If you play on the harder levels, you're Awesomeness gets a boost because it's just plain harder to survive as long as on the normal levels.

I think this link may help you with tuning your gameplay.

http://www.gamasutra.com/view/feature/6155/hot_failure_tuning_gameplay_with_.php

In this article, Chris raised a metric way to maximize the use of player's death or stuck in a level. I really love that way, hope it will also help you.