28 November, 2009
There is often a divide between perception and reality, especially in online games. For individuals, perception is reality. The way players perceive your game will be reality to them, despite whatever models you have that demonstrate that the player’s perceptions are wrong. People don’t like to hear that they are wrong even in the best of times, and when they feel your system just screwed over their enjoyment, they’re going to be even less impressed.
I want to take a look at some reasons why I think that player perceptions vary so much with reality.
Yes, sometimes developers are wrong
Of course, it goes without saying that sometimes developer perception doesn’t match reality, either. Sometimes a carefully designed and implemented system has a subtle bug that makes that system do something unexpected. Or, perhaps the developers take some shortcuts while testing and didn’t know that some previous step along the way wasn’t working as intended. This is especially true of systems with a random element, where the behavior isn’t entirely predictable, where proper testing may be beyond the resources available to an independent game developer.
For this post, I’ll do the developer thing and assume the developer is right and the player perceptions are not matching reality. In reality, always double-check your work.
A mind is a terrible thing
If you read much about the studies of the human mind, you’ll find that for the most part, people are terrible at some pretty basic things. For example, most people are terrible at understanding probability. The classic example is how an event has a “one in a million chance of happening to someone today”, it’ll happen about 6,700 times according to current population estimates; not exactly as rare as it might first sound. People are also have a hard time grasping very large numbers. how many people do you know who could continue this series: million, billion, trillion, …. Probably engineers and game developers are the few that might be able do it without reference.
We also have a lot of perceptive “blind spots”, also known as biases. One of the most famous is confirmation bias, where most people tend to notice data that confirms their own opinions more readily than data which contracts it. (I believe this is the current basis of the U.S. political system.) A variation is when you form a conclusion based on a few data points then focus more on data that supports this conclusion. If you have had bad experiences with drivers from a specific state or country, for example, you’re more likely to notice when you have a problem on the road with someone else from that same state or country.
There is also the plain old personal bias, where someone willingly adjusts data (or simply lies with statistics) to support their position.
Perception in action
The reason I’m writing about this topic is a rather entertaining post by Melmoth over at Killed in a Smiling Accident about developers plotting ways to annoy players. (Note that despite having a mustache, I don’t often twirl it and laugh manically.) The issue at hand is a few monsters in LotRO that have speed-affecting effects that last for a while. Melmoth wonders if the evil developers didn’t put this in the game intentionally to slow players down. To be honest, I’ve found that these abilities are really annoying, too.
Note that Melmoth is a fairly intelligent person (or script able to write intelligent and creative posts and comments). Like many players, I don’t think Melmoth is stupid or unperceptive, I think that he does not look at mechanics from a game designer’s point of view; not surprising since he’s not a game designer as far as I can tell. I do highly recommend reading Killed in a Smiling Accident. The authors there can write entertaining stuff (or, at the very least, the scripts are showing us what life will be like once the robotic overlords take over the world and enslave us all).
I left a comment on the site giving a few possible design-related reasons why designers my intentionally put a movement slowing ability right as a monster dies. Not being one of the designers, I can’t say how accurate these are. I won’t go into details about the design justification in this post, either. I want to focus on the perception Melmoth had that these abilities seemed to happen frequently near the end of combat.
Going to the dogs
So, the first step to see if developers are truly diabolical is to do some research. A lot of this is going to be LotRO-specific. I won’t go into a lot of detail about the game mechanics, so if you just want to see the results go to the next heading. LotRO players and game designers might find this interesting, though. If you want to know more about the game mechanics, there are plenty of guides online.
I took a level 26 minstrel (happen to be the character I had parked in the location) outside the town of Esteldin in the North Downs. There I fought Wargs just to the west of Esteldin, which were approximately the same level as my character. This is not a scientific study, exactly, but I tried a few things.
First, I fought them normally. For a solo Minstrel, this means going to War-speech and nuking the crap out of the enemy, finishing up with a few weapon strikes at the end. Often my initial attacks would stun the enemy, allowing me to finish it entirely without getting close. I noticed that most of the time they didn’t get the run speed debuff off on me. They did apply a wound-based damage over time and a fear-based damage over time on me often. It seemed that I’d get one of these debuffs in a fight, near the end.
Second, I prolonged the fight by not using my big attacks to start, meaning that there was no chance for the enemy to be stunned. Obviously a lot more enemies closed to melee range. Not surprisingly, I found that I would get one or two DoTs or debuffs most fights.
Finally, I engaged the enemy and simply turned my back to it while not being in War-speech. I healed myself when necessary, but I let the monster just beat on me. Obviously I got a lot of the debuffs and DoTs, but here’s the interesting part: I got the run speed debuff even when the enemy had nearly no health missing, meaning it was not near the end of the combat.
What I believe is happening is that the Warg has a random chance each hit to do a debuff or DoT. If the random chance is selected, it chooses from one of the possible afflictions and applies it to the character. I’m not sure if the chance to trigger an affliction is dependent on health left or not. At the very least, the code is not designed to only put he affliction on near the end.
The debuff only lasted about a minute, less than the two minutes that Melmoth wrote (but still rather annoying). Now, this is just one enemy in a game full of enemies that do this, so I can’t guarantee that all monsters act exactly the same. But, it provides some fuel for our discussion.
Why it appears to be at the end
So, why was Melmoth sure that these afflictions were only applied at the end of a combat? One key element here is that the Warg debuff requires the enemy to be within melee distance. If Melmoth is playing a character that has ranged attacks, the monster will only be within range near the end of the combat, and thus can only apply the debuff during that time.
Another possibility is related to perception. Melmoth noticed that the debuff lasts longer than combat with the enemy does. He notices this more when the debuff is applied near the end of the combat. If the Warg has a uniform chance to apply the debuff during the combat, then the debuff will be applied during the last half of combat (nearer the end) half of the time. The half the time the debuff is applied near the end are noticed more due to confirmation bias.
And, to show that it’s not just Melmoth’s fault: it could be an (unintended) consequence of some other mechanic. Perhaps enemies get an increased chance to trigger special attacks as they get more wounded, or as time progresses in combat. Perhaps enemies will do special attacks more frequently in response to player special attacks, and Melmoth’s character has more special attacks at the end.
What a developer should do
As I said above, perception is reality. Even if the fault lies with Melmoth, a good developer should take the player’s perceptions into consideration as the design evolves. Let’s assume that this is not a desired perception, what can a developer do to change it? The important thing to do is to decide what the design goals are for these debuffs. I’ll assume the primary reason is to add an element of unpredictability to the encounter by giving an affliction that changes the encounter. The run speed debuff makes it harder to run away from an enemy, so the player has to come up with other ways to survive a potentially fatal encounter, or start running early.
The first option is to give players a way to counteract it. If I hate the run speed debuff, I should have some way to avoid it. There is a way to remove the wound-based debuff in the form of potions, but it costs money and inventory space. Some classes can avoid the debuff by killing from a distance and/or having an ability to heal wounds, but not all classes have this option. I actually tried to counteract this by increasing my character’s wound defense, but I found this to have very little effect in practice.
The second option is to make sure the code does not have a system that makes these affects appear more frequently at the end of the encounter. This only serves to annoy the classes that do not enjoy powerful ranged attacks or the ability to heal. In fact, it might be better to increase the chance that the effect happens at the beginning so that it can have a meaningful impact on the encounter itself. The debuff should not last significantly past the end of the encounter.
Again, the important thing here is to make sure the design supports the design goal.
So, what do you think? How can this type of system be made more fun for players? Or is it something that will always be an annoyance?