Positive vs. Negative Feedback

Well, all feedback needs to be handled delicately. Positive feedback unchecked becomes a "rich get richer" situation, and negative feedback creates a "poisoned in the first turn, die all game" situation. Yeah, I had to make that second comparison up, but you all know the feeling. Feedback loops, in both cases, skew the miportance toward the opening of the game, and almost always increase the first player advantage as well.

A feedback loop needs to be checked by the other players, either with another feedback mechanism (which can give you extremely strange results), or some way to remove the reinforcing mechanism through their own actions.

In general, I think that all feedbach should be able to be countered by appropriate action. The reinforcement of any effect, whether positive or negative, should only happen whenthe situation is allowed, for whatever reason, to go unchecked across turns. This raises the possability of tactical and strageic planning, and dramaitic decisions during the game, as players often choose between long-term or short term gains, or avoiding long-term consequences with painful but short-term actions.

How about in Puerto Rico, where your actions leading up to choosing Craftsman (building production buildings and selecting plantations such that you will produce goods) lead to producing goods as Craftsman, which will allow you to score points or make money later via Captain or Trader?

I think Puerto Rico, Caylus, and probably any strategy game on the market probably exemplifies the positive feedback thing, as well as just regular advancing of position. In fact, I'm finding it difficult to discern a difference - anything that advances your position on the board almost by definition does so in a way that will help you advance it further. I suppose a 'advance of position that is not positive feedback' (...neutral feedback?) could be something that costs you resources to get VPs in a 1-for shot. Like buying a big building in Puerto Rico, or a Prestige building in Caylus. That's the end of the line - you get points for it, but you can't use it later to advance your position further like you can with all the rest of the buildings in the game.

Hmm... on a more personal note, I'me thinking of my own games to see if I've got any positive (or negative) feedback in them. All For One has many actions which advance your position without making it easier to advance it any more. You can pick up stuff, but all it does is get you closer to completing a mission - it doesn't make doing anything else easier. Same with combat. The only exception is the Horse token, which allows you to move faster, so that's a tiny bit of positive feedback. I don't think there's any negative feedback. So how does thie differ from PR and Caylus? Well, I'd say it has to do with the weight of the game. A41 is intentionally a lighter game than PR and Caylus, so it's not a suprise I guess that it hasn't got the positive feedback.

My newer game, which I've been thinking a lot about lately (of course), I think has some positive feedback. In Terra Prime you have a spaceship which you can upgrade by buying stuff. There's also technology upgrades to make the stuff you've bought better. These upgrades make it easier to get more upgrades, by allowing you to more easily find and retrieve resources. I don't think it's so powerful as to create a rich-get-richer scenario, at least that hasbn't come up in playtesting. I must've done an OK job of balancing the upgrades :) I've tried to add some "neutral" feedback as well, analogous to the PR big buildings... fighting aliens for example takes time and resources, and gives you nothing in return except VPs and the possibility to get more VPs (by colonizing where the aliens were). Also, the hardest-to-get resource is used pretty much only for VPs, not upgrades. Like the Big buildings or Prestige buildings, these resources are delivered for big points, but don't help you otherwise.

This is a good example, but it's also an example of a slippery slope. And not just because ice is slippery (heh)! This sort of direct correlation between your benefits increasing based on the benefits you get sounds like it could easily lead to the rich-get-richer scenario mentioned above. Maybe the key to useing positive feedback is to have an indirect relationship... you do one action, which improves OTHER actions you take - not the same action. If doing the same action over and over makes that action better and better, then you have the glacier example. But if doing one action makes another action better then you have something more like PR and Caylus.

Those are my 2 cents on the subject. Good subject by the way! Very interesting concept, and a good way to look at our games. I think your thread has already taught me something about how to make a heavier game!

- Seth

One way to avoid a slippery slope from positive feedback is to limit opportunities to cash in; another way is to spread the wealth. In Caylus building a new building can be accomplished only by activating one of a few "builder" buildings, and not all such buildings exist at the start of the game; they themselves must be built before they can be used to build others. Furthermore activating a building costs an action, and actions are rare and precious, so you may need to use them all for other purposes. These effects limits the amount of building that can happen during a game, and that prevents the snowball effect. Also a new building can be used by anyone, so although it's a definite advantage to be the builder of a building, everyone can benefit to some degree by its existence. (There are exceptions such as the prestige buildings, as Seth noted.)

A good example of a posative feedback in a game is in Settlers of Catan (SoC). In this game you get resources from areas where you have settlements and cities. You then use these resources to build more buildings and therefore get more resources.

A good way to work out if a given loop is a negative or a posative feedback loop is to draw a diagram listing each component of the loop (as the rabbits and foxes above) and drawing a line between any for each direct effect between them (including the bi-directional one in the above example). Then work out weather an increase in the first has an increase (+) or decreaseing (-) effect on the second. Once you have doen this for all effects then you just count the number of (-). If the number is even then it is a positive feedback loop and if the number is an odd number then it is a negative feedback loop.

Posted by Hedge-o-Matic

A negative feedback is not like this. A negative feedback will give you a diminishing reture the greater the input. So it would effect the player(s) in the lead and help the player that got burned in the first round.

For example:
If you increase "A" it will provied an increase to "B".
If "B" is increased then "A" is decreased.

Call "A" Rabbits and "B" Foxes.

If the population of Rabbits goes up (as rabbits wont to do :D) then there will be more food for foxes and so their population will rise. Because ther are more Foxes eating Rabbits then their population will drop, and so on.

So putting it into a game: As the Rabbit player gets ahead in the game the Fox Player will be able to increase thier number and so catch up to the Rabbit Player.