My notes. Sid's point of view is very much from his lifetime of creating Civilisation, so some bits of his experience might not work with an entirely different game type. It was a nice, avuncular, gentle talk, quite rule-centric. Lots of useful bits and pieces within a talk that covers a lot of areas in quite a loose structure. I'll put some pics up later.
Psychology of game design, everything you know is wrong.
I was going to call it people are funny, gamers are goofy, everything you know is questionable. But I thought this would get more interaction. The premise is that game play is a psychological experience. When I design a game, a lot of my games are based on historical topics. Railroads, pirates. And the way I’d generate these would be to make it realistic, true, historical, the more piratey it is, the better.
What I found in taking that approach is I found that what I thought I did was wrong. I hadn’t taken into account what happens in the player's head. Things became a lot clearer. by acknowledging that game play is a psychological experience, we can save time in areas. We end up with a better game. As we explore this idea we're going to run into a few psychological concepts: egomania. paranoia. delusion. self destructive behaviour.
If you play Civ you are an egomaniac. Ha-ha! Basically. Build a Civ to stand the test of time? “o yeah I can do that” you say. That’s egomania.
The first area is “the winner paradox”. In real life you don't always win. There are 28 teams in the NFL, but only one gets to win the superbowl. 25 teams in the NBA. At the end of the day there's only one team holding the Trophy of NBA Championness. But in the world of games you pretty much always win. The psychological phenomena is that the player doesn't object. This is true of other forms of entertainment: Sherlock Holmes always solves the mystery. Rambo comes out okay. It's kind of fundamental to entertainment: the player is looking for a satisfactory conclusion, and this often includes winning.
Some of the tools that we have at our disposal include reward and punishment. There's an interesting dichotomy in how these are deployed in games. When you reward the player, the player.. you know, gladly accepts the bit of gold you just rewarded her with. They don't really question it. They accept it gladly and feel that it was their own clever play that earned them a cool reward. On the other hand if something bad happens, it's really important to be very careful with the setbacks that the player experiences. That they understand why it happened, and how to prevent it the next time. Every time you plant that next-time seed, you're well on your way to replayability. Any opportunity that you have to plant this seed is important.
In terms of rewards, another key thing is The First 15 Minutes. They have to be really compelling, really fun, a foreshadowing of al the cool stuff that happens later in the game. You can almost not reward the player enough in the early stages of the game.
This doesn't negate the value of difficulty levels. A number of years ago I gave a talk on difficulty levels, and made the point that there were 4 difficulty levels: introduction, casual, experienced, expert. I was wrong about that. Apparently we need NINE difficulty levels (Civ example). It points out that this idea of progress and advancement is really rewarding for the player. Mastery and move on. So you want your player to feel that they're above average.
: I missed out on the uncanny valley, and I never trademarked Interesting Decisions, but the Unholy Alliance, I’m going to get this one. It defines the relationship between player and designer. I'm going to pretend certain things, you're going to too, and together we'll have a great experience. One thing we pretend as designers is that the player is good. You're really good. We want you to feel good about yourself while you're playing. This went off the tracks with Flight Simulators: early on they were accessible and easy to play. Then we got to every iteration where they went more complex, more realistic.. and pretty soon the player went from I'm Good to I'm Confused. My plane is falling out of the sky. The fun went out of it. Keep your player feeling good about themselves. Alliance
The player's role in this: they need to suspend their disbelief. Take on the mantle of king, or pirate, or whatever it is. That's part of the bargain. I think in some ways those of us that are old-time designers have a bit of an advantage because we worked in the good old days of 16bit graphics, we really had to work hard to get the player to believe they were building an empire or whatever and that helps a lot in the challenge of getting the player to suspend their disbelief.
It's more satisfying to win against bad Genghis Khan if there was some kind of moral cloud over the actions you were taking (example of bringing up women & children in GK's village). Mutually Assured Destruction is another thing to think about: who remembers the Cold War? We didn't blow each other up because we knew the other side also had nuclear weapons. The player can destroy the game experience any time they like: they can cheat, quit, play wrong. As a game designer, we can mess up the game as well: lose the thread of fantasy... at Microprose years ago, there was an adventure game, you adventured down to the castle to see the king, and when you got to the king it was revealed to you that the king wasn't the good guy, but the Bad Guy, and you had to go back to the beginning to do it "right". That was us messing up the game. "Imagine the look on the player's face when all this work they've done was in vain!" No: they're going to think, I just wasted 8 hours on this stupid game.
Another part of this alliance is in the area of style. Things that convey what the spirit of your game is. If you have light-hearted music and graphics, and heads start exploding, you're not giving the player what you promised them. You're fooling the player. If I’m playing a game that starts of happy and cartoony, and terrible things happen, you lose your suspension of disbelief, and the player turns off the game and moves on. Humour, style, music, activity, keep them consistent.
Probably where it became most clear to me that player psychology had nothing to do with rational thought was putting together the battle system in Civ rev. in Civ rev we'd show you before a battle what the odds were. As a mathematician this is a 3 to 1 battle [pic] but players feel they're going to win the battle. “I had this battle, it was 3:1 and I LOST!” The player would say, “no no you don't understand. 3 is big, 1 is small. I had the big number, I should have won!”
So we adjusted our system to make the battles more like what the players expected. This time the player had 1, and the AI had the big gigantic 3. And lo, the player won. And I said “doesn't that feel wrong to you”? And the player said, “no? Not at all? I had tactics”! So we made more adjustments. “Now are you happy with the way combat works”? And it's around 3:1 or 4:1 where players pretty much expect to win every time.
We can live with that. So we asked: anything else? Are you okay losing a 2:1 battle every now an then? “Yes. But I had a 20:10 battle. I had ten more than them! I lost!” I say, “isn't that 2 to 1?” “NO! that’s 20 to ten! Whole different odds!” So we adjusted again. You have ten more. That's a lot. So now.. are you happy?
“So I had this 2:1 battle. I lost. That was okay. But after that I had another 2:1 battle and I lost again, how can this be??? The computers' out to get me!”
So we take into account the results of previous battles when we do battle calc. We don't do this just to make the player happy, but when the battles start to feel wrong, the player starts to lose the suspension of disbelief, the player will get distracted. Something really clear: the interaction between logic, science and math.. and psychology... it's counterintuitive. Including the psychological part of things in game design is going to mean running into these counterintuitive things. This can be to your advantage, as you'll see.
Times when I really screwed up: my brain was too logical and scientific, and I hadn't thought about psychological aspects. The first one was to make Civ a real-time game. Inspired by SimCity, you're watching something grow... what we found with the real-time version of Civ was the player really became an observer. Watching what was going on. Just an observer. The mantra was "it's good to be king" but when we made it a turnbased game suddenly the player was the star of the game.
Another bad thing I did in original Civ was to include this idea of rise & fall. You build a civilisation and it crumbles... and then it comes back. Wouldn't that be awesome? No. At the cusp of crumble, most players would reload a savegame, and would never experience the glorious re-rise we had in mind for them. Civ is about the rise, rise and more rise. and not about the fall.
Another bad: Civ tech tree. Players want to be in control. They don't like randomisation so much. Any kind of randomness needs to be treated with a lot of care. Great natural disasters. Wouldn't that be cool? Plagues. Volcanoes. Randomly. No. It wouldn't be cool. Again when something random happens, PARANOIA strikes. The computer did that just to make your life more difficult just when you're about to win. Randomness at a low-level helps with replayability and variety, but be careful with it at a significant level.
Civ Network. Civ on Facebook. People playing on an interesting schedule. Co-op game play. This is a fun world to take Civ into. And one idea we played with was the idea of gifting gold to other players. Wouldn't this be neat? What we found was that no one ever gave gold to anyone else. The thought never crossed their mind. That was kind of sad.
One of the things I promised in the blurb was to how to save millions of dollars in game design. AAA games on a shoestring: really use the player's imagination as a very key tool. What's happening in the player's mind is always more compelling, more dynamic, than what [they're seeing]. You don't have to literally show everything in the game to make it happen. The player can imagine what's going on. You don't have to create all those assets!
One example from Civ Rev : you're sent a caravan containing 7 dancing bears. If the player WANTS to believe something, the player will believe it. Yes, foreign leaders should send me 7 dancing bears. The bears are dancing in your mind, because you agree that you deserve those bears. But there are no bears. We didn't create model bears: we didn't need to. We just put up a text box.
Take advantage of what the player already knows. Pirates. People bring a lot of their own pirate knowledge to the game. Here's a swordsman with a black curly moustache. Okay, we know he's evil and we don't need to know his back-story. So focus on what's new, really impressive, and hard to sell in the game: take advantage of what they player will automatically provide through their imagination.
AI is a place where psychology comes into play very strongly. In conflict situations, your psychological juices are flowing. Players project a lot onto the AI. I see AI as one element of creating an overall experience for a player: in a single player game, the key experience is the experience the player is having, and AI is a tool to craft that experience. There's another POV that the AI should act like another player: I don't subscribe to that that makes me shy away from that. In my experience, if an AI has human characteristics, I want it to surprise me and be tricky and clever. When AI does that, most players feel like it's being dumb, because there's a fine line between clever and dumb, and the player is likely to think the AI is dumb. If the AI is brilliant, the player might think the AI cheated.
I approach AI in making it a metric, a solid competition against which you play, but against which you feel your advancement as you go along. The AI is kind of predictable, kind of solid and steady, and you're getting better and better against it. Feedback and validation in single player is really important as the AI is kind of your friend at that time. Players like that, someone understands them.
Self-destructive behaviour. Ways that the player can damage their experience: one tool we have is ways to protect players from themselves. I see players saving a game before a battle, playing the game, losing, reloading the game, trying again until they succeed. So in Civ Rev we made it so that you can't do this, a savegame would have a marker in it meaning the outcome would always be the same, ha-ha! If you allow too much load/save behaviour, you kind of ruin the [play of the game].
Too many choices: customisations, settings, options, they're all key but there's been a bit of a tendency to open up all sorts of game play decisions to the player, and to me that's wrong. We're as designers here to make the game. Be careful with what you open up to them. Cheat codes are another damaging thing. Civ II: there on the main menu was CHEAT. What's this? You can bring in tanks in the middle ages? I don't disagree that that's cool - tanks in the middle ages! - but .. on the main menu? Can't we just bury it a bit? But the cheating idea led directly to the idea of modding. Modding is good.
Another psychological skill we need to develop as designers is "listen to the player". Game design is an iterative process. build a game over time. if you do that, you're going to spend a lot of time talking to players. But it's not THAT useful to take literally what you're players are saying to you. There are different kinds of ways players react: one is to provide solutions. "Take this out". Occasionally that's useful, but a lot of times this doesn't take into account stuff that might break other parts of the game. But it is useful to drill down to what is *motivating* that solution. What's unsatisfying? Another type of comment is an emotion. "I'm frustrated. It doesn't feel right." That's our job to figure out what's causing that emotion. The last thing important here is knowing individual personalities. Some people just might not like your type of game. Some people are too generous and will always say nice things.
What is the point of all this, you might be asking. I want to wrap it with this idea of An Epic Journey. There are so many genres, so many styles.. but they can all be encapsulated in this idea of an epic journey. How do we use psychology to make our journey more epic?
- Interesting Decisions. This is core. These are basic fundamental pieces of your game. The journey is the process of exploring these decisions. These are very powerful decisions.
- Learning and Progress. Fundamental to the journey. You're constantly progressing: it's fundamental but it's easy to lose track of this! You can't reward and acknowledge and reflect enough in a game. "What value has this last hour of playing been to me?"
- One More Turn: another way of making the journey epic. The idea that the player is leaning forward, anticipating things coming later. You can foreshadow ideas. Let them know cool stuff is happening soon.
- Replayability. If you create a fantastic journey, and the player reaches the end, they realise this is only one segment: epic cubed.
So know you know everything.