MMO Development Lesson #28
Perception is reality. When it comes to game balance, what really counts is how it feels while playing the game, not what the numbers tell you. It may be perfectly mechanically balanced–you have all the numbers, you’ve made all the calculations, and it is balanced. But, if that’s not how it feels when you’re playing the game, it isn’t balanced and needs to be changed. This is one of many reasons that it’s 100% vital that you play your game as a player does (a lesson unto itself).
Totally agree on this one. Back when I worked on Day of Defeat for Valve, we used to have huge player discussions/arguments in our forums about gun accuracy and shot dispersal patterns. What people FELT the guns did versus what they actually did were not always in alignment.
One thing it did prompt me to do though was go back and do some real testing to get a better feel for just what the guns actually did. You can see the results online at http://oregonstate.edu/~holtt/range/ where I made a “rifle range” map, and proceeded to test fire every gun at varying ranges.
What it proved is that certain guns were as good as others. But it did little to nothing to change how people felt about them. One interesting side effect is that I think the “technician” players who are out to maximize used it to get the best out of their weapons. It let you look at the Thompson for example and say, “Head shot at 10 feet, good odds. at 20 feet, not as good. 50 feet, forget it”
This is true in everything, not just games. As a manager working in several industries the most important thing you can know is that “perception is reality.” The sooner you realize that, the more successful you can be as a manager. That’s because you find out how important setting and example is, and you know what people think is true because of how they perceive it.
One important thing to know is that there are people that will fail to be close to perceiving some form of reality and will always try to “take you to China.” You can’t be distracted by these individuals, because you could take them to steak dinner and they would tell you that their chicken was overcooked.
I’ve been saying this for years to my friends, and they all think I’m nuts. “Perception is more important than reality. Does it matter more that a car actually is red or that the majority of people see the color and agree that the color is red? If you are a car designer trying to sell red cars, the latter always wins.”
I always wanted to try some mind nerfing. Like modify just the sound that a gun makes in a gun, and add a release note saying “Modified performance of ” then stand back and see what the reaction was. “So much more accurate now!” or “OMG they nerfed it SO BAD!” etc. My thought was you could just deepen the note a bit or up the volume to make it seem more powerful, or the reverse to nerf it.
Arg follow up post.
Here’s actually one story from Day of Defeat where perception was actually based on fact. After we ported the game to Source, a lot of people (or loud minority) complained it didn’t feel right, shot accuracy wasn’t as good, etc. Ends up that in fact in the HL1 (goldsrc) based version, the code was doing a weighted shot cone - where more bullets hit towards the center of the shot location than the edges. However in the Source version, the code got changed unintentionally to a completely random system where the bullet patterns were totally random with no weighting. Some enterprising players actually made a modification (server side) to implement the weighted shot cone.
So in this case, perception (and a lot of bitching on the forums about nerfing) actually was based on reality, and for a long time it was simply ignored as being in the player’s heads. So my advice is to not totally ignore what people are perceiving unless you implicitly know it is a fact they are “wrong” (in quotes).
Writeup on the shot cone thing and some replacement code (with comments FFS!) I wrote at http://developer.valvesoftware.com/wiki/CShotManipulator
Yep. Even if you’re sure players are wrong, you have to listen to what they’re saying and check it out for yourselves. There are times when players are entirely right about the facts and not just the perception, You may realize that something is looping too many times, an unexpected calculation is being run, a table in the database has been modified, or something else is different than you (the developer) thought.
Great advice; however, playing one’s own game is hard enough as it is unfortunately
The best way to burn out is to work all day and then play that game all night.
That’s why as you say, listening to your actual players is hugely important. You just gotta know when to, and not constantly be chasing the last thing that everyone said on your forums. We have ample examples of both ends of the spectrum, from blind-eye ignorance to too much pandering.
“Perception is reality.” - you couldn’t have said this any more accurately.
Would it really help for developers to play the game? A quick trip to any game’s forum makes it clear that devs only play the uber unbalanced classes
The more options there are for character generation/development, the less chance that a developer, even a team of developers, can play the game as the players do. Imagine trying to do that in EQ2, where there are 24 classes, then Achievement Ability builds that modify those. It would be difficult, imo, for a developer to really test out how more than perhaps 4-6 classes operate in solo/group/raid situations well enough to have a grasp on the classes ‘perceived reality’.
Playing the game as players do doesn’t mean “legitimately play the game as every option.” It means you have to play the game with no developer commands to know how everything works in practice. If an issue specifically comes up that you haven’t personally tested “as a player,” you have at least the following options:
1) Find another dev that plays that class and talk to them. 2) Observe how it plays out while grouped (you can generally tell things are screwed up when grouped with someone). 3) Use your hax0r dev commands and specifically test the perceived problem. 4) Talk to trusted class players within the community.
The real idea is that playing your game as a player gives you a much better idea of the state of things than just reading the forums or /feedback.
To Darniaq’s point: Yeah, it is hard to play the game as a player. At the very least, you need to play it while at work. If your company doesn’t let you play your game at work at least several hours a week, it’s a dumbass company that doesn’t know what’s good for it.
What you really need is to work at one company that is developing 2 MMOs… then make a friend on the other dev team, you play his game and have him play yours.
Completely agree. I know there is one company out there that some of the classes are very screwed up and I can almost bet that not one of the devs has played those classes to the max level, and it shows! I think it should be a requirement that the developers actually play the game they are making.
I’m not asking for them to play every option, but a set of typical ones would be nice. That way perhaps some significant imbalances would have been seen before it hit the servers and caused an uproar in one class community or another. Just ask someone who plays a high level ranger in EQ2 about arrows (NB: I don’t play one, but I have heard plenty about it). If one class is being abandoned or betrayed (if the game has that mechanism) to another class more than the opposite direction, the developers should be asking themselves why that is, and test out the classes for themselves.