I has been observed recently that the gaming world, and the 360 in particular, is overwhelmed by first person shooters. Why should this be the case?

I think the factor here that is often overlooked is the impact of processing power on game development. When I started gaming a shooter, or a first person game of any kind (as we understand it today), was impossible. To fully understand what I mean you should look up games like '3D Monster Maze' or 'Nightmare Park'. From there we moved to arcade games, with static screens and seemingly unlimited levels. Once power improved it was possible to introduce side scrolling and then orthogonal projection, the first 3D games.

We are at a stage where 3D has become possible, but is still demanding. Ever wondered why so many are set in space or alien worlds (no plants) or why the Master Chief wears a mask (no facial animation). We see the worlds becoming more lush and green and draw distances increasing, but there is still a long way to go before games are lifelike.

The unfortunate part is that visuals are the easy part, relative to AI, so while the worlds become more lifelike it will be a while before we can convincingly interact with game characters in more sophisticated ways than just killing them.

The development of multiplayer gaming is an important one.  Without a guiding human intelligence computer games can feel lifeless and inflexible.  Humans can adapt and refine their tactics leading to constantly evolving patterns of play.  Computer oppponents do not do this very well, even the more sophisticated expert systems.  At first sight it would seem that the AI drone (bot) is dead.  But this is not the whole picture.

Imagine a world populated solely by human characters.  Who is going to be willing to play the part of the sleepy guard you garotte stealthily, or the cowardly barman you rob?  Who is going to want to follow the orders that will (realistically) get them killed everytime?  We all want to play the starring roles, the prime movers, and no matter how different our tastes there will be plenty of unfilled roles left over.

We must improve game AI to the point where the bots are (superficially) indistinguishable from the players, so they can blend in easily.  The boring jobs will get done and we can spend our time despatching cannon-fodder...

...until we meet another 'hero'.

This is really just a particular example of a more general question of whether any single index scoring system can truly summarise something as complex as quality.  We should first look at how we (each) determine the quality of a game.

Let us pretend for a moment that the idea of 'goodness' of a game can be broken down into a finite number of aspects or qualities.  Perhaps we may have graphics, sound, controls, gameplay etc.  Even if we were then capable of giving each a score that truly reflected how good that aspect was, we would still need to agree how (relatively) important each aspect was to the overall score of the game. I know that I rate the quality of the AI more highly than many others do, rendering our weightings different.  Even then we would need to agree how to combine these weighted scores so as to give us one score.  This is also not simple, as one weak aspect may cancel out many strong features i.e. fantastic graphics but repetitive gameplay.

The truth is that we all do this complex process subconciously.  The final score we give reflects not only our feelings about the game, but also our perspective like our personal weighting of factors.  How closely you should trust a score depends upon how closely you share the 'bias' of the reviewer.

The best bet? Use the score as a (very) rough guide and then read the review and pick out the aspects that matter to you.  Learn from others but make up your own mind.