Anything worth doing is worth doing right.

But to hedge your bets, I would go with SETTLE FOR LESS first, then pull yourself up and dive into the STRUGGLE.

Say you want to make a jousting game with leveling up and mini-games.

Build up the main game play actions first like the actual jousting event.

From there you create the mini-games.

After that, you create the leveling system that gets attached to the rest of the game.

Basically, start with the most important aspect of the game; the parts that it needs to have for your idea.

Then, tack on the feature creep and bloat if you still have time and money.

Worst thing that happens is you already have to core game you wanted if the other features are too tough or break the experience.

Best of both worlds IMO

This could happen with anything creative.You get that one idea, rather its simple or ambitious, you reach a crossroads in your mind where you choose:

STRUGGLE
(Learn how to program this really difficult idea on your own or surfing the web for tutorials hoping you find the solution.Both taking a lot of your time and its not guaranteed you'll succeed. In the end its more rewarding and beneficial for the learning process.)
Or
SETTLE FOR LESS
(Make something close to but not as difficult as your idea.You won't feel accomplished but youll be satisfied.This is the easier but less rewarding path.)

Does this ever happen to you? what do you do when it happens.And what road do you usually choose?

Considering STRUGGLE

Making first game as 3d mmorpg is nono, jumping too high may hurt in landing part. If you go and fight a 7headed beast as level 1, you shall not prevail !

Considering SETTLE FOR LESS

A path that many walk, therefore hone your skills young one before venturing forward, the road your taking is a dark and dangerous place!

Simply said SETTLE FOR LESS but STRUGGLE for more.

Settle for struggle?

never settle.

Rome wasn't built in a day, and you gotta break a few eggs to make a REAL mayonnaise.

You learn how to do what you don't know how to do.

If its never been done, you invent your own method of doing it.

if you merely settle, your sub-par skills never improve, and all you can make are sub-par games.

And maintain focus:

All that matters is what you haven't figured out how to do yet.

Once you've figured out how to do everything, only then you can proceed to build with confidence.

There are really two different sides to the struggle option, and they're very different things:

1. You can have an ambitious goal that is beyond your current abilities and that will take a lot of time and hard work to achieve. You might also need the help of others to reach your goal.
2. You can have an overly ambitious goal that you will never satisfactorily achieve.

The first is definitely worthy of pursuit and will be an excellent learning experience -- although it's worth noting that it's often a helpful step to set smaller sub-goals (effectively settling by your definitions) in order to more easily achieve your goal by breaking it down into smaller projects you can tackle along the way.

The second is a waste of your time. Unfortunately it can be hard to identify without experience, but thankfully there's an easy method that doesn't necessarily rely on taking the word of others: by breaking your goal down into smaller sub-goals you can then test to see whether or not they are achievable, and will be much better positioned to judge whether or not it is worth investing a significant amount of time and effort into the project.

Either way, effectively settling for smaller goals is a helpful thing to do when working towards a larger project. Settling on the achievable doesn't have to be an end in itself.

by breaking your goal down into smaller sub-goals you can then test to see whether or not they are achievable, and will be much better positioned to judge whether or not it is worth investing a significant amount of time and effort into the project.

I'll second that.

When contemplating a new title, I'll start with a list of the major technologies / techniques / capabilities required that I don't know how to do off the top of my head.

Then comes research and rapid prototyping. Once all the unsolved parts of the puzzle are solved, the concept has been proven doable, and worth pursuing, at least from a technology point of view. Maybe not sales though. Usually a game idea must pass a fitness test from a marketability standpoint before I even consider the technological requirements or what technologies will need to be prototyped to make sure they're doable.