How does the direction of a game during creation affect its success?

This question has been bothering me for a while since I never managed to find a good answer for it. How exactly do all the direction choices when creating a new game affect its ratings, chances of winning an award, or how well it sells?

If I wanted to create a sumo puzzle game and put all my points into developing the game world, what will that affect, than say, focusing on a niche audience?

The direction seems to affect the success of the game very little. The conventional wisdom is that certain directions fit certain genres, much like certain genre combinations work well together. However,

• Unlike genre combinations, the game gives you no direct feedback on the quality of your direction/genre decisions.
• I have been unable to find any actual empirical evidence that specific direction/genre combos improve game rating or sales.
• The nature of the production system makes every game's development unique, and randomization is thrown in for good measure. This makes it almost impossible to effectively test.

Conclusion: the amount of points does seem to make a notable difference. The allocation of them may or may not. If it does make a difference, it's subtle enough that nobody has been able to effectively quantify it.

My experience playing the game has led me to believe that the direction points do not affect ratings, chances of winning awards, or sales... at least not directly.

It seems to me that direction points seem to play a factor (in combination with the genre/type) in how much of the 4 categories (Fun, Creativity, Graphics and Sound) get developed. For instance, I found allocating lots of points to the Innovation direction will cause your game to have a high Creativity score. I also notice high Graphics & Sound scores when I use many points on Polish. These correlations are not direct 1:1 relationships (Innovation does not simply affect Creativity, it probably also affects Fun) and as I've mentioned, I believe the genre/type also play a factor.

As sjohnston mentioned, a higher total number of points used seems to create higher scoring games. And as many other sources suggest, having balanced scores in the four areas (Fun, etc.) seems to help sales and chances of winning GOTY awards. This means that the direction points will not directly affect sales and awards but will affect game's scores which will affect sales and awards.

I would conclude that it would be a good strategy to distribute the direction points in all areas to be fairly balanced but use some common sense to maybe slightly stack a few areas (e.g. stacking Game World direction for Action RPG or stacking Realism for War games). I believe that this will lead to high, balanced game scores which will lead to higher sales, review scores, and chances for awards.

I always though it was kinda like mashing A and B while trying to catch a pokemon. It might help, it might not, but all I know is that I catch more of them when I do. Also, whenever I made an Action/Ninja game, the critics always wanted more polish, so I maxed it out and they put the game in the Hall of Fame.