The Science of Player Motivation

Almost every game relies heavily on extrinsic rewards. You accomplish a task and the game gives you a reward (sometimes). These rewards work very well in a simple game like Bejeweled or Peggle — Many will play Bejeweled for hours trying to beat their high scores, and many would wonder if there had been no high score table, would most have bothered?

External rewards are only effective when tasks are mechanical -- in any task requiring "rudimentary cognitive skills"  performance is reduced when rewards are introduced.

The Puzzle of Motivation — Dan Pink

A Quick Definition of Rewards.

Extrinsic Rewards

XP, points, loot, gear, achievements — things that we give the player as a reward for finishing a task.

Intrinsic Rewards

Pride, happiness, useful skills — things a person gets from a task that isn't explicitly tangible or quantifiable.

Here's the thing about extrinsic reward systems: they don't usually scale linearly upwards with game complexity. Bribing players with points in Bejeweled is very easy, and many will happily to go along with it.
Bribing them with points in Halo might not work nearly as well. Why? Well, because the player's expectation is rooted more deeply in skill and execution in Halo. It's not exactly a thinking man's game, but it's a lot more mental and physically demanding of effort than Bejeweled, and so that behavior is a reward in and of itself.

The problem is, once a task reaches a certain level of complexity, it has to be intrinsically rewarding for someone to want to do it. This could be linked to why it's hard for many people to learn programming or math — neither are rocket science, but they're hard enough that if a person has no intrinsic motivation, they won't have the motivation to proceed far.

Massively Multiplayer tiered Rewards.

MMOs are a good example where a lot of reward systems are in place concurrently. Almost all quests and tasks have very clear instructions about what is required. Completing it is a purely mechanical process, and the player's willingness to do it is amplified by the reward.

Then consider any quest that requires some creative thought or where what is required isn't immediately clear. That extra bit of thought that would be satisfying in any other situation suddenly becomes frustrating. It may seem to impede your progress towards the next reward because complex problem solving requires intrinsic investment. When players do complete it, they may generally have spent more time and were less engaged during.

Creativity doesn't need to be required for it to be used. Many puzzles can be solved by brute force, but this tends to be a tedious and time-consuming approach, and people derive pleasure from solving them by shorter, more elegant methods.

It's easy to write a Sudoku solver, but some people still enjoy doing them by hand. Computers play Chess by brute force and memorization; humans play Chess by creative thought and memorization. It is the balance of the two that creates engaging experiences.

Hard to define Games.

Sandbox games provide a vast window into discovering Intrinsic Value. One of the best examples of this is Sim City — there's no real goal to the game, and you don't necessarily win, but it's still a very rewarding game to play even though there aren't many explicit rewards.
The reward itself is in building and maintaining your own city.  The game gives feedback along the way, but not with badges or push notifications, but through the behavior of your citizens and the outcome of your fiscal and political decisions.

Minecraft is another example of this. Hundreds of thousands have created sprawling worlds with intricate subway systems, massive architecture, and hugely popular multiplayer minigames.
There are hundreds of website for Minecraft servers, textures, music packs, etc, and even tool to create maps of your world.

And all of this is in a game that basically just lets you place and remove cubes — that's it. It's barely even really a game, it exists as a medium that allows people to play how they choose and the cooperative aspect really helps.

As an industry, we sometimes define games too narrowly. Sim City, Minecraft, Tamagatchi, and The Sims barely even classify as games in the traditional sense (there's usually no end, concrete story, or progression of levels), but they've also all sold millions of copies collectively, and people love them.

We might have to throw away some very basic assumptions about what a game should be, and start build games with meaning.