Hi
Im doing a modern pirate game where you plunder ships, run missions, build wealth, buys/upgrades ships and skills etc (think sid meiers pirates! but with more details). The thing is im interested in moving to multiplayer.

I cannot do concistent world (like a MMORPG) so you need to save the game and connect with a lobby to the same game with the same players (your friends) when you want to continue. Problem is how to pace and design such a game. If you play against each other, its hard to make it fun for players who get behind (in economy, levels etc) to continue playing in "the next session". Such a game is by nature long as you build up your wealth and trade for better ships etc.

Its not like : connect. Play a match of 1-2h. Declare a winner. Start a new match.
Its almost like diablo 2 with a long buildup to access more expensive/late stuff as the game progresses. In Diablo2, co-op is the natural solution, but i dont know if co-op works in such a game as I describe...

Any thought about this rather wage problem?
Thanks.
E

Here's an idea: when the players are sailing around at sea and meeting other ships, sometimes those other ship will be another player visiting their game. They can interact, trade, fight, whaterver, but when they separate the multi-player session is ended and both players go back to their respective worlds. A variation on that idea is giving players the ability to join another's single player game in it's entirety until one of them leaves. Of course for both of these implementations it would be best to have an option to disable it for players who would rather be left alone.

There is potential for griefing, but hey, they're pirates after all. Yarrrr!

One of the problems that plague many games is that of Positive Feedback Loops. These are effects where a player will take an action that through how that action influences the game mechanics that eventually leads back to improving that player's ability to take that same action in future (it also means that if you reduce an ability to take an action it further reduces the ability to take it in future). This leads to what is know as the Runaway Leader Syndrome.

Imagine this situation:

You have a Cart racing game with items that players can drop that hinders a player who hits it. However, these can't shoot forward, only backwards.

Now a player who is hit by one of these is slowed down which means the player in front get more a lead. However, as they don't shoot forward, this ends up in a runaway leader situation as a leader can take actions that increase their lead (by slowing other players), and as this lead give them access to these items first, the action leads to an increase in their ability to take that action (and also makes it harder for those behind to catch up).

This is why in the cart racing games they have weapons that attack those in front of you, and the players behind usually get them more frequently.

But this applies to other games too.

If, in your pirate game, you have a situation where better weapons are bought with gold, and these weapons help you to get gold, the this is a positive feedback loop and could (but not necessarily if there are other factors - see below) lead to a runaway leader situation.

The way to solve the positive feedback loop problem is with another feedback loop, that of a Negative Feedback Loop. These are loops where increasing a variable in the loop ultimately ends up reducing that variable, but if you reduce a variable, then ultimately it ends up increasing the variable.

Positive Feedback Loops are unstable as they force any change to increase in magnitude, where as negative feedback loops suppress change.

Now, positive feedback loops are not all bad. The best use for them in games is to bring about the end game. What you want is to have a positive feedback loop come to dominate when you want your game to end. This means the resolution quick and dramatic. What you don't want is for them to dominate too early other wise a winner of the game emerges early and the rest of the game is just following though the motions to reach the game over screen.

With negative feedback loops, these should dominate during most of the game to keep the differences between players relatively even. However a player who has the skills and strategy to keep their score (or whatever you use to measure success) high, then they will do better when the end game time comes.

I'll try to pre-empt questions that often get asked:

So, how do you determine if a loop is positive or negative?
This is fairly easy. In short, all you have to do is look at each influence in a loop and determine if it has a positive or negative effect (you don't need to know the strength, just the direction of the effect) on the next part of the loop. Negative feedback loops can contain both positive and negative effect. The deciding factor is if the number negative effects are or even.

If the number of negative effects is even or zero, then the loop is a positive feedback effect. If the number of negative effects are odd, then it is a negative feedback loop.

The real trick is determining what effects are in the loop, and the only help I can give here is to use some paper (or a drawing program) and write it out. It is a fair amount of work, but it can be worth it to get a balanced game.

The mathematics behind all this is that each influence is multiplied to the accumulated influence of the loop.

Technically, when multiplying you need to multiply the normalised magnitude of the effect. This means you treat the effect as a 1 dimensional vector with the magnitude between -1 and 1 (this is needed if you want to balance the power of the loops - which is useful for timing when they will dominate).


How do I fix a positive feedback Loop that is causing problems?
Again, in theory this is simple: Just ad another effect into the loop, and make it a negative effect.

As an example, think about the Gold and weapon problem I used above. With this you can add the effect that increased weapons means you slower, require more rare/expensive ammunition or are more likely to come to the attention of government warships which will hunt you down.

This causes a negative effect on the player who upgrades their weapons. Sure, They will find it easier to get loot, but they will find it harder and harder to keep that loot. They will be able to get ahead of the others, but the further ahead they get the harder it will be for them to get further ahead. This means that players trailing will get a chance to catch up either through improving their own strategies, of from the leaders making mistakes (or just having a bit of bad luck).

Another idea would be to make it that it is worth the trailing players to team up against the leaders (or take direct action against them). This would means that the leaders constantly have to watch out for loosing players backstabbing them. However, this creates an adversarial system (which could be what you want in your game).

If you look at the Cart Racing games, you will see that they use just this kind of effect to allow the trailing players to attack the leaders. This is to make the game so that at any point any player could win (but it takes a really good player to win constantly).

It is too complex
At first this can be a complex subject, and many will choose to ignore it and go by their gut instinct. Sure that can work and it often does, but this theory explains why it works or doesn't, and so allows you to use it to make sure it will work. However, you can practice with some simple loops to help you learn and understand how it all works. At, first, it can look complicated, but when you compare this to trying to figure out why you have an unbalanced game without this, then compared to that, this is actually quite simple and simplifies a lot of the design work.

It is too simple
Yes, this is only the basics of it.This is why it is simple. Some things that add complexity to this is multiple loops (for example with my pirate example if you included all the penalties, each of these could be on a separate loop and thus would not interfere/cancel with each other). Loops can have magnitudes where they might exist, but until a certain event or effect occurs they remain insignificant (or the reverse they are strong, but then go weak). You can have loops within loops. You can have effects change direction from positive to negative (or negative to positive), You can change the magnitude of effects (essentially increasing or decreasing the influence of that effect on a loop). You can have effects be part of several different loops at the same time (or even change between them).

Basically, there is a lot more complexity that can go on with this theory.