My attempt at analysing what aspects of games I like and dislike, and what design lessons can be learnt from this.

Before We Start

I’ll make references to the Bartle test of gamer psychology several times in this discussion. This is a pretty simple classification of gamers into four categories:

These categories seem fairly self explanatory, but if you’re not quite sure what one means refer back to the wikipedia article that has a nice summary with examples for all four. I also recommend you do the test at yourself.


Naturally the opinions below are just that, opinions. Some of them I think are fairly well justified that most people would agree with (e.g. Quick Time Events Suck) but many of the attributes I’ve selected as good things are largely just things I prefer in a game. Obviously if your results are significantly different you are likely to find yourself disagreeing on a few of these points.

My own Bartle result:

  • 73% Explorer
  • 47% Socializer
  • 47% Killer
  • 33% Achiever

With the disclaimers over, let’s get onto some real content. The points below are in no particular order and each point is prepended with the name of the game I’m using as an example for this particular point, if you haven’t played the example game it doesn’t really matter.

The Good

Rift - Custom Class System

Rift uses a strange skills tree system. Each class allows access to 6 different “souls”, any three of which may be equipped at one time, each soul opens up an entirely new skill tree. This essentially means that for each class there are 120 different sub classes! This system is great, it means that there are many different ways of pulling off each role in the game and players can come up with “custom” classes that suit their method of play. For example I played the rogue class with a soul that unlocks various teleport based dodge skills, a soul which unlocks various tanking skills (including “generate threat every time you teleport”) and a soul which gave me even more teleport based attack skills - suddenly I have a tanking character, built on a rogue class, which tanks by teleporting around the enemy and never getting hit. That class is nothing like how any normal tank classes work, which is what makes it cool, it’s my very own tank class.

I think this kind of thing is great. Players get oodles of choice, and get to build customised classes for exactly how they want to play, which keeps the game interesting for longer.

Guns Of Icarus - Teamwork


Hitman/Dishonored/Deus Ex - Stealth


Dystopia/Hawken - Large Time To Kill


Threadspace:Hyperbol - Sidegrades

Hah! I beat you because I’ve spent more time playing/microtransactions to buy a superior weapon. Isn’t that super fun? Yeah… straight upgrades suck.

I’ve never seen sidegrades done as well as in Threadspace:Hyperbol, every upgrade module gave is an exact sidegrade (i.e. exactly exchange N of attribute 1 for N of attribute 2), this leads to players becoming super specialised as people minmax their attributes to match their playstyle. Critically, there’s no one-true-build that people have to follow for certain roles, all changes are purely to make your attributes enhance your personal playstyle. This kind of thing is great, enhancing individual playstyles really increases the variety of things you see in a match, which keeps the game fresh and fun.

The Bad

Rift/Tera/Guild Wars 2/Knights Of The Old Republic - 5 Man MMO

MMOs are cool because they give you an entire universe to play with - thousands of other people who you’re playing with - it’s a unique experience in games when you know something truly huge, involving thousands of players, is happening. So why is it that most MMORPGs are games for 1-50 people? It seems to me to be entirely backward to go to all the effort of building all the complex infrastructure required to host a server with 10,000 people on it and then to only provide content which ~50 people can be involved in (usually even less people). Even ignoring the waste of effort on the part of the developers MMOs come with certain downsides for the player: overcrowded newbie zones, high levels gankers, lag, subscription fees - why put all these bad things into the game if you don’t really intend to exploit the advantages of an MMO? True MMOs focus on large groups of players working together, emergent gameplay and possibly market economy effects - think Planetside or Eve Online.

Rifts in rift were pretty good GW2 allows lots of people to complete certain missions together

Rift/Tera/Guild Wars 2/Knights Of The Old Republic - Filler Content


Fetch Quest Kill 10 Rats Escort Quest Kill 1 Big Bucket Of HP Singlehandedly Wipe Out That Cult Fetch Power Crystals To Improve Your Weapon

Rift/Tera/Guild Wars 2/Knights Of The Old Republic - Crafting Systems


Crafting systems are fine Level grinding gathering skills, then crafting skills is not.

ArmA/Day-Z - More Realistic == More Fun!

This makes no sense to me. The real world is a pretty boring place, why would you want to simulate it? Of course the natural suggestion is to have a perfect simulation of everything in the entire world and then you can carry out unreal situations (e.g. zombie outbreak) and see how it would “really” happen. We all know that’s bullshit though, the game isn’t going to be a perfectly realistic simulation. Also players will do things differently to how real people would, because most of the things we do in the real world are really boring. E.g. A realistic zombie outbreak would involve:

  • 7 hours staring at your character sleeping (each person taking shifts of 1 hour to stare idly out of the door window)
  • 1 hour looking for flammable material to build a campfire
  • 3 hours moving across the city at snails pace trying not to be noticed
  • 30 seconds being bitten in the neck by a zombie that jumps out of nowhere

Don’t get me wrong here, I really love super hard games, but making things more “realistic” (and I use the word realistic with some pretty huge reservations) is just lazy game design. There are certain kinds of hard which are fun to deal with and certain kinds of hard which are not - going for realistic is just like saying “Oh we can’t really be bothered to work out what the fun+difficulty+realistic things are, so we’ll just program whatever “realistic” things come to mind and hope it’s fun”.

Battlefield 3 (single player) - Quick Time Events


Does anyone actually need me to convince them that quick time events suck? This doesn’t even qualify as lazy game design, at least with the realism laziness it’s the designer thinking “we’ll stumble upon something fun (because it’s difficult) while implementing lots of stuff that’s realistic (and therefore difficult)”. I don’t know what the designer of a quick time event could be thinking, perhaps “players like pressing keys quickly, so we’ll require them to press some arbitrary keys to not fail, yeah that’ll be cool”.

Even worse than quick time events just being really pointless is that they’re immersion breaking, you’re watching a cutscene which is communicating to you the important story which drives the game and makes you care about the characters and then QUICK SUDDENLY PRESS E RIGHT NOW TO STOP YOURSELF FALLING OFF THE TRAIN. If you fail then you will have to rewatch a load of cutscene (which has now gone from important storyline to annoying movie you just wants to skip through to get back to the QTE you failed), if you succeed then you won’t be paying close attention to the rest of the cutscene because you’ll be too busy watching for more arbitrary QTEs.

A personal gripe I have with QTEs (on top of all the previous things) is that the key(s) they ask you to press is usually meaningless. You’re watching yourself falling off a train and suddenly it asks you to press E. Wait… what? We have established mechanics for climbing up things in games, you press W, not E.

blog comments powered by Disqus