Sid Meier's Talk at GDC

Well, not really Sid... Here are some notes that Matt Bowman took while at Sid's talk at this years GDC. I have to agree with a lot of Sid's stances. Some great observations here.

The winner paradox
Keep winning percentage abnormally high. In real life, not everyone wins. Only one of the 25 teams in the NBA can win the championship, but not so in games. The player is looking for a satisfactory conclusion. Developers should make sure players win big and win often. Skew the “winning percentage."

Reward vs Punishment.
Players like to find gold coins. Players are very inclined to accept anything you give them and think that’s the result of their own merit. If you punish bad behavior, the game is wrong, it’s cheating, players think. When there’s a negative consequence, it’s important that the player understand why that’s happening and how to avoid it next time. If you can emphasize that “next time” aspect, you increase replayability. Antime you can plant the seed of replayability, do it.

Also, the first 50 minutes have to be really cool. Let them know they’re on the right track. Cool stuff is happening and even cooler stuff will happen later on. In the first 50 minutes, you almost cannot reward them enough.

Difficulty Levels
I used to think we needed four difficulty levels. Apparently, I was wrong; we need nine. The reason we have these levels is to create a sense of progress. There always need to be more challenges. Desiring to make it to the next level enhances replayability.

You always want your player to feel they are above average. They are doing well, and they will probably do better.

The Unholy Alliance
This defines the relationship between the player and the designer. I’m going to pretend certain things; you’re going to pretend certain things, and together we’ll have fun. The terms of the alliance are as follows:

1. I’m good! (the player). Designers went off track with more and more realistic plane simulations: realistic crashes created the impression “I’m not good, my plane is on fire and I’m falling out of the sky.” Keeping them feeling good about themselves is an important part of the designer's role.

2. Suspension of Disbelief. The player needs to inhabit that character. That’s his part of the bargain. When my wife comes out of a movie and I ask what she thought of it, she says things like “I came out of that movie twice” or "I never got into it," or “never came out of it.” Players need to stay “in it” and designers need to help them.

3. Moral clarity. Imagine you are conquering cities. You have one town left to occupy, you outnumber them and can squash them in a second… but they’re defiant! It would be more realistic if they wanted to surrender and asked for pity. But that would put the player in a moral dilemma—you always want the player to feel he is making the right decision by continuing to play.

4. MAD: mutually assured destruction. The Cold War balance of powers is paralleled in the player-designer relationship. Players can play in all kinds of ways that destroys the experience. Designers can do the same. You need to keep the balance to maintain the suspension of disbelief.

5. Humor/ Style/ Music/ Atmosphere. If you start with lighthearted music and cartoony graphics, then all of a sudden heads start exploding, you’re not living up to the alliance. You don’t want to fool the player. To help the player maintain his suspension of disbelief, you need to consistently keep the player in your world.

Math 101
Often games give odds for the player’s chances of defeating a foe. For example, the Attacker Player has a 1.5 score and the opposing Barbarians have a 0.5. There’s a 3-to-1 chance that the Attacker will win. When polled, however, players thought they should always win when they have a 3-to-1 advantage, but weren’t surprised when they would lose if they were at a 1-to-3 disadvantage. Around 3-to-1 or 4-to-1 advantage, players expected to win every time. So we adjusted the computer to appease the player.

Players felt they could lose a 2-to-1 battle every now and then. But they had a problem if they lost a 20-to-10 battle. So we adjusted, and asked, “Now are you happy?” “Well kind of, but there's one more thing: I had a 2-to-1 battle and lost, which was fine (we went over that). But right after that, I had another 2-to-1 battle and lost again—how can that be!? The computer’s out to get me, obviously!” So we made sure that occurrence wouldn’t happen, and the player was happy.

When something happens that feels wrong (even if it isn’t), we start to lose the player’s suspension of disbelief.

My Bad (errors Meier made)

* Real-time Civilization. When we introduced real-time play with others, the player became an observer of other players. The game’s mantra was “It’s good to be king.” We needed to put the player at the center, so we took away real-time action.

* Rise and Fall. We thought, “Wouldn’t it be great if you are crumbling and just at the last minute before all is lost, you come in and save the situation and then rise again to an even greater status than before?” No. Players would reload at the cusp as things start to go badly. They would never experience the comeback, so Civilzation is a game of progress only.

* Don’t Randomize. Players want to be in control. Whenever anything random happens to the player, paranoia sets in. They feel the computer rolls that random number to make their life more difficult. Random events at a large level have to be treated very carefully. Low level randomization can help things seem more realistic, but if they are large events, players come up with worst possible explanation for it.

* Civilization Network. This network is built on the Facebook platform, and is social. We played with idea of being able to give gold to other players. They could trade or beg or give out of pity if they wanted. What we found was that nobody ever gave gold to anybody else.