Skip to Content
 

[TiGD] Theory Series #1: Decision Theory and Games

23 replies [Last post]
Zzzzz
Zzzzz's picture
Offline
Joined: 06/20/2008

Decision Theory and Games

Definition 1 (courtesy of wikipedia.org)
The study and application of mathematical techniques which provide a rational grounds for choosing between alternative courses of action in situation of varying degrees of uncertainty.

Definition 2 (courtesy of indiainfoline.com)
Is an interdisciplinary area of study, related to and of interest to practitioners in mathematics, statistics, economics, philosophy, management and psychology. It is concerned with how real decision-makers make decisions, and with how optimal decisions can be reached.

Commonly in Decision Theory, decisions are thought of as occurring under one of four conditions:

1. Certainty: The states of the game are known. But you have the problems of:
a) determining the outcome that will result from each action
b) deciding what benefits of those outcomes are
c) resolving tradeoffs with one action better achieves objective A, and another action better achieves objective B.

2. Risk: The states of the game are unknown, but the probabilities are known. For example, I don’t know if the having the longest road vp in Settlers will result in a win, but I know that 30% (I made this percentage up) of players that obtain the longest road vp, win Settlers.
So in addition to the problems associated with decisions under certainty, we must also evaluate probabilities, and combine information about probabilities and outcomes.

3. Uncertainty: The states of the game are unknown, and even their probabilities are unknown. There may be no previous knowledge about the effectiveness of some actions, for example. Now we have all the problems of decisions under risk, but we must also estimate probabilities somehow.

4. Conflict: The states of the game also consists of outside choices (what your opponents choose to do). Your opponents choices may be influenced by your choices (or their beliefs about your choices) and vice versa.
We have all the problems of decisions under uncertainty, and we have to think about strategy - how to respond to or influence someone else's actions. The study of decision under conflict is usually called game theory, and is fascinating, but we will cover that in another part of this Theory Series.

*****

The topic is Decision Theory? Why?

As a game designer, I feel that understanding Decision Theory will help me to improve my games. So I figure it was worth my time and effort to bring the discussion to the forum, in hopes to make us all decide how we handle decision making in our game designs.

For me, I think that the idea of Decision Theory is a core part of any good game. I think it can make or break a game. Can you imagine the game of PR where players have no reason to decide on which role to select?

I feel that without decisions, players would have no reason to “figure out” the best strategy, or better yet, worry about the outcome of an uncertain choice.

So a few questions to get us all thinking,

1) Do you as a designer think about decision making while designing your games?

2) How do you incorporate decisions making aspects into your games?

3) Would a game be a game without some type of decision making?

4) To what depth should a game designer understand theories that influence game design?

5) Was this a useful post or did I waste my time? ;) (just checking to see if I should continue the Theory Series)

Anonymous
[TiGD] Theory Series #1: Decision Theory and Games

Here are my 2 bits:

1) Do you as a designer think about decision making while designing your games?

Absolutely. This may be the most important consideration when developing a game. Sure, the theme, a certain mechanic or even a genre may come first, but drafting rules, playtesting and fine tuning all come down to creating the best game experience possible.

For me, that hinges on the type, frequency and importance of the decisions the players make over the course of the game.

2) How do you incorporate decisions making aspects into your games?

It depends on the genre/theme and the design goal I first set out, but I prefer developing multi-tier decisions. In other words, decisions that can have more than one effect, or impact the game in more than one way.

If you're dealing with a bidding game, as an example, not only do you need to gauge an item's relative value to you, but also the value to the other players. Passing on an item is a decision that impacts the game by virtue of allowing another player to grab the item. Bidding and winning the item impacts the game as well, but in another way.

I also like items that can be used in several ways. A good example of this would be the cards in certain card-driven wargames like Hannibal: Rome Vs. Carthage. A single card could be used to recruit more units, expand your political influence, activate stacks of units, move by sea transportation or played as an event -- having several decisions wrapped into a key component adds strategic depth, replay value and some good brainfood.

3) Would a game be a game without some type of decision making?

Games with literally no decisions generally come down strictly to probability and chance. These sorts of games are excellent for children, where everyone can play on equal footing, without skill disparity affecting the outcome. Reiner Knizia's Monkey Madness is an excellent example of this - it is a social exercise, I'd argue, and not technically a game.

4) To what depth should a game designer understand theories that influence game design?

I think a designer should be familiar with concepts and theories he would like to work with, but strong experience with these theories (based on gameplay experience or discussions like these) are just as valuable as reading about the nuances of a particular theory.

That said, I think the broader the experience and knowledge base (either through first hand gameplay experience or through research), the better off the designer will be, being able to leverage knowledge of several different ideas, fine tuning them to fit his particular needs within a particular design. And the more familiar you are with a concept or theory, the more likely that tweaks and changes to the theory will remain balanced and workable, rather than fall apart.

5) Was this a useful post or did I waste my time? ;) (just checking to see if I should continue the Theory Series)

Great topic for a thread. These are the sorts of forum questions and content I find most interesting and useful as a designer. Well thought out, great supplemental information, and the questions should trigger some interesting discussions from folks.

xantheman
Offline
Joined: 12/31/1969
[TiGD] Theory Series #1: Decision Theory and Games

I can not think of a single game that does not include some kind of decision making process. Without choices there is no point in playing a game with other people.

I often have the dilemma of trying to balance the number of choices that players have in my games. Giving to many choices can result in sluggish game play while not enough choices can make parts of the game boring or repetitive.

Balancing the usefullness of choices can also be tricky. Often I come up with an option that I think should be usefull but, the players do not try it out enough to learn how to use it properly.

Overall, I say that decision making is one of the most important aspects of game design. Balancing the number of options and the usefullness of options can be one of the most challenging parts of designing a game.

Xan Lynch
www.rentoys.com

Verseboy
Offline
Joined: 12/31/1969
[TiGD] Theory Series #1: Decision Theory and Games

Quote:
1) Do you as a designer think about decision making while designing your games?

I don't see how you can't think about it.

Quote:
2) How do you incorporate decisions making aspects into your games?
I also like items that can be used in several ways. A good example of this

Another example would be the produced goods in Puerto Rico. Do you trade them for money or ship them for victory points?

Quote:
3) Would a game be a game without some type of decision making?

The card "game" War comes to mind. There are absolutely no decisions to be made. It's an exercise for killing time. It is not a game.

That said, party games often don't involve decision making. That, however, is outside the scope of this thread.

Quote:
4) To what depth should a game designer understand theories that influence game design?

I don't think it's necessary that a designer be able to consciously articulate any of this, but I do think it's essential that at some core level decision theory informs all of the choices he makes in including this feature and excluding that one.

Quote:
5) Was this a useful post or did I waste my time? ;) (just checking to see if I should continue the Theory Series)

Not a waste at all. I hope you'll continue with it.

Steve

doho123
doho123's picture
Offline
Joined: 07/21/2008
[TiGD] Theory Series #1: Decision Theory and Games

Quote:
I can not think of a single game that does not include some kind of decision making process. Without choices there is no point in playing a game with other people.

There are many games that do not have a decision factor. And while you might not like games like this, many people DO enjoy them as a form of relaxation.

Some examples:

Many children's games come to mind, CandyLand being the prime example, but other more "adult" mass market games, such as Mouse Trap.

Bunco.

Slot Machines.

The card game, War.

Zzzzz
Zzzzz's picture
Offline
Joined: 06/20/2008
[TiGD] Theory Series #1: Decision Theory and Games

doho123 wrote:
Quote:
I can not think of a single game that does not include some kind of decision making process. Without choices there is no point in playing a game with other people.

There are many games that do not have a decision factor. And while you might not like games like this, many people DO enjoy them as a form of relaxation.

Some examples:

Many children's games come to mind, CandyLand being the prime example, but other more "adult" mass market games, such as Mouse Trap.

Bunco.

Slot Machines.

The card game, War.

Good I see people caught on to my tricky question #3. That is you can have a "game" (though many argue back and forth on this) that has no decision apsect.

So I will add that game designers that remove the decision process from a game, create what is typically called a Racing Game (race from point A to point B, that dont allow for player choices) Such is that case with games like Candyland or War (race to get rid of your cards first).

Finally this would mean that we as game designers should ask ourselves if we wish to design a game that contains decision making or not. Since it will help us to define the scope of our game design.

Anonymous
[TiGD] Theory Series #1: Decision Theory and Games

Quote:
create what is typically called a Racing Game (race from point A to point B, that dont allow for player choices)

I'd disagree with this, as written. A race condition in a game is merely a means of seeing who achieves Objective A first, but many such games contain important player choices and decisions. In fact, nearly all sports games modeled after racing include this race condition (as you would expect) but feature numerous decisions.

Ave Caesar, Formula De, Circus Maximus, Q-Jett, Arena Maximus, PitchCar... all these games feature a race condition, where Objective A is crossing an actual finish line, but all these games feature decisions. Some are simple decisions, like PitchCar (how hard to flick your car, what direction) while others are far more detailed like Formula De (what gear to shift to, how to approach a turn, whether to burn brakes/tires for positioning, if you should risk a collision, etc).

I think instead, the type of game described (Candyland, Monkey Madness) could be more appropriately defined as a Probability Race (to coin a phrase) since the winner of this particular race is based purely on mathematic probabilities, and is not influenced by decisions or player involvement.

Zzzzz
Zzzzz's picture
Offline
Joined: 06/20/2008
[TiGD] Theory Series #1: Decision Theory and Games

ynnen wrote:
Quote:
create what is typically called a Racing Game (race from point A to point B, that dont allow for player choices)

I'd disagree with this, as written. A race condition in a game is merely a means of seeing who achieves Objective A first, but many such games contain important player choices and decisions. In fact, nearly all sports games modeled after racing include this race condition (as you would expect) but feature numerous decisions.

...
..
...

I think instead, the type of game described (Candyland, Monkey Madness) could be more appropriately defined as a Probability Race (to coin a phrase) since the winner of this particular race is based purely on mathematic probabilities, and is not influenced by decisions or player involvement.

ynnen, I agree with you. I was too general in my statement. And I am happy you pointed it out since I dont want to go down the wrong path!

Instead I should have stated that games with no decision apect are typically consider a Racing game, though many Racing games also contain decision making.

And I think your term of "Probability Race" fits nicely.

Verseboy
Offline
Joined: 12/31/1969
[TiGD] Theory Series #1: Decision Theory and Games

Zzzzz wrote:
Good I see people caught on to my tricky question #3. That is you can have a "game" (though many argue back and forth on this) that has no decision aspect.

... this would mean that we as game designers should ask ourselves if we wish to design a game that contains decision making or not. Since it will help us to define the scope of our game design.

We can get into a lot of semantic juggling, here, but I think it would be useful to try to limit the discussion to the kind of games most people here are intent on designing, namely games with some strategic element and decision making. I would argue that slot machines, War, Bunco, Candyland, etcetera are more readily described as activities, even when they share some of the accoutrements of games. Even if in the broad sense, they are games, it doesn't mean any of us aspire to design anything similar. (I may be wrong, and I certainly don't mean to offend anyone here who is sitting on the next Candyland.)

I think it is possible to design a party game that removes most or all decision making. Party games serve to foster creativity and social interaction. That has been the main focus of the games I have worked on, though I have some strategy games in the hopper. However, I don't think it's a valid question for a designer to ask if he wants to design a game without decisions. For all practical purposes, if someone here wants to design a game, they are making an implicit decision that the game will have decisions. The only real question is how involved will those decisions be. I understand that exceptions exist. I'm one of them. Discussing the exceptions won't really help anyone, though.

That's my take anyway. I don't mean to come off as dogmatic. I was just feeling semi-articulate today.

Steve

Zzzzz
Zzzzz's picture
Offline
Joined: 06/20/2008
[TiGD] Theory Series #1: Decision Theory and Games

Verseboy wrote:
...

I'm one of them. Discussing the exceptions won't really help anyone, though.

That's my take anyway. I don't mean to come off as dogmatic. I was just feeling semi-articulate today.

Steve

I agree that we dont want to go to indepth on the expections, and the point is to get all of us thinking. Thinking about how and what we do when designing a game. At some level all of us consciously or unconsciously use many "hidden or unknown to us" theories when designing our games. Decision Theory is one such area.

So I guess I wanted to just get everyone talking about how decisions help create a "good" game. Or how they can cause a "bad" game. Obviously the amount (or lack of) decision making can influence the concept of Agonizing Decisions in a game. What else can decision making do to a game?

TruMobius
Offline
Joined: 12/31/1969
[TiGD] Theory Series #1: Decision Theory and Games

Decisions in my opnion reflect the depth of a game

obviously this line of thouht breeds the idea that a game with more decisions is arguably a deeper game than one that invloves little to no decision making.

While all that might be generaly true it doesn't take into account the types of decison made.

For instance, take a game that most here have played, Settlers. It can be said that there is a fair amount of decisoin making invloved in the game, from beging placment of houses to whether or not to spend the resouces not for a road or wait for that one you need to get a posssible development card. vs. Puerto Rico where your decisions you make (your job for instance) affect the types of decisions you can make, etc.

One coudle argue, qute well infact as would tend to agree, that Settlers is less deep than Puerto Rico. But in essence they are both games about trade and development and as such surly merit this compairison.

Ok now on to the point if one would compair Settlers to say Chess for instance which would come out as the more Deep basied on decision making. In my opinoin (and I can just feel the flak) I'd say that neather are more Deep than the other for the simple matter that they roughly involve the same amout of Decision making.

alright let the flak decend

zaiga
Offline
Joined: 12/31/1969
[TiGD] Theory Series #1: Decision Theory and Games

Hey Zzzzzzzz, interesting topic!

Quote:
1) Do you as a designer think about decision making while designing your games?

Yes, absolutely! It is in fact the number one thing I think about when designing games. At each decision point in the game I want to present the player with a handful of meaningful, interesting and distinct decisions.

I also find it important that the player knows what his options are, and that certain options are not obscured by layers of complexity, fiddly rules or poor layout. Basically I want the game to say to the player: "Hey dude, it's your turn. You can do X, Y or Z. What will it be?". The challenge of the game should be deciding what to do, not finding out what you actually can do!

Quote:
2) How do you incorporate decisions making aspects into your games?

A simple game model that makes it easy to create interesting decisions is the Resource ->VP model with a timeline. An good example of this model is Ticket to Ride. On your turn you can either take more resources (cards), or claim VPs by spending cards. When someone runs out of trains the game ends, which will happen sooner or later. Players cannot endlessly hoard resources and then end the game with a bang, because you can only spend six cards per turn at max. In a sense a turn is also a valuable resource in the game, because you only have a limited amount of them. This creates tension between deciding when to grab more resources and when to spend them to get VPs. Also, there's tension between mid term scoring opportunities (claiming track) and long term scoring opportunities (fulfilling tickets, longest route). Finally, there's tension because there's competition for limited resources, namely the tracks on the map.

So to sum it up, you can create interesting decisions by:
- Separating resources from VPs (or what other winning condition you have)
- Having a finite, ever diminishing number of turns
- Having short term, mid term and long term decisions with different pay-offs
- Having competition between players for limited resources

Quote:
3) Would a game be a game without some type of decision making?

Depends on your definition of "game". I think that yes, you can have a game without any real decision making. LCR, Chutes and Ladders, etc, qualify as games in my book, but I'm not really interested in designing such games.

Quote:
4) To what depth should a game designer understand theories that influence game design?

I think it's important to understand the fundamentals of theories that influence game design, at least if you take game design seriously. Understanding the theories makes it much easier to avoid certain pitfalls, and makes it easier to design a game in such a way that it delivers the experience you are looking for.

zaiga
Offline
Joined: 12/31/1969
[TiGD] Theory Series #1: Decision Theory and Games

TruMobius wrote:
Ok now on to the point if one would compair Settlers to say Chess for instance which would come out as the more Deep basied on decision making. In my opinoin (and I can just feel the flak) I'd say that neather are more Deep than the other for the simple matter that they roughly involve the same amout of Decision making.

alright let the flak decend

Flak, why? I don't disagree per say. It's very hard to measure the level of decision making in a game. Perhaps the decision making in Settlers is of a higher level than in Chess, because in Settlers there are many more factors to take in account, such as what your opponents are going to do (you have more opponents in Settlers than in Chess), the probabilities of the dice, the probability of opponents drawing or having drawn a certain action card, the various recources, the various ways of scoring points, etc.

However, even if you take all the right decisions in a game of Settlers you can still lose if the dice are against you, or because of player interaction. This doesn't happen in Chess. If you take all the right decisions in Chess, you will win. It's the same in other abstract 2 players games such as Tic-TacToe or Connect Four.

onew0rd
Offline
Joined: 12/31/1969
[TiGD] Theory Series #1: Decision Theory and Games

All games I like, at the basest levels are about 3 things which are all related:

1 Limited resources (game components, your turn, etc).
2 Decisions on how to spend them with limited knowledge (what will opponent do?, should I wait? is this the last one of these I get?)
3 Winner is the one who makes the best decisions and, oftentimes, who gets luckiest with their decisions or gameplay

I think all games I really like include all 3 of these things in differing doses. I feel if your game has these 3 attributes, it is well on it's way.

Scurra
Scurra's picture
Offline
Joined: 09/11/2008
[TiGD] Theory Series #1: Decision Theory and Games

zaiga wrote:
I don't disagree per se. It's very hard to measure the level of decision making in a game. Perhaps the decision making in Settlers is of a higher level than in Chess, because in Settlers there are many more factors to take in account, such as what your opponents are going to do (you have more opponents in Settlers than in Chess), the probabilities of the dice, the probability of opponents drawing or having drawn a certain action card, the various recources, the various ways of scoring points, etc.

One of the differences between Settlers and Chess is the depth of the decision tree. In Settlers, because there is so much unknown information (beginning with what the next dice roll will be) it is very difficult to create any sort of strategic plan. In Chess there is no hidden information at all, so the game is all about strategic planning. [That doesn't mean they don't both have considerable elements of skill involved such that the better player is more likely to win, but that's probably irrelevant!]

I guess I'm saying that the Settlers decision tree is broad but shallow, whereas the Chess decision tree is much narrower but hugely deeper.

Zzzzz
Zzzzz's picture
Offline
Joined: 06/20/2008
[TiGD] Theory Series #1: Decision Theory and Games

While reading the replies, I thought of the following questions:

6) What games would fall under the Decision Theory aspect of Certainty? For some reason I want to say Chess falls into this section, because most serious chess players, know all the states of a game. And most can determing the outcome of various moves (or actions) and understand what the benefits to each move they might make will be.

7) What games would fall under the aspect of Risk?

8) What games would fall under the aspect of Uncertainty?

9) What games would fall under the aspect of Conflict? I think this is where Settlers would fall. You have to assess what your opponents will do, or wont do. Many players are influenced by their opponents actions.

TruMobius
Offline
Joined: 12/31/1969
[TiGD] Theory Series #1: Decision Theory and Games

Under both Risk and uncertanty I think Poker would fall into
and for that matter most Card games that posses a "hand" would fall into Uncertanty

heck most games with "betting" would fall into both now that I think about it for that is some of their appeal (risk something you've earned earlier for the chance that you might improve your posistion)

good example (also falls into both categories): "game of thrones"

Zzzzz
Zzzzz's picture
Offline
Joined: 06/20/2008
[TiGD] Theory Series #1: Decision Theory and Games

Ok next item,

When designing a game, we want players to feel conflict and *worry* about making those hard choices we place in front of them.

So our game needs to have at least one optmial decision path that will yield the *best* possible payoff, based on the starting states of our game. And of course it will contain other sub-optmial paths that force the player into those great agonizing decisions.

This path is the optimal desicion path and can be determined by us, the game designer, by using Decision Tree(s).

A Decision Tree
Using the following image (symmetrical positions have been removed for the sake of simplicity and space) , let us start by looking at a decision tree for the simple game of Tic-Tac-Toe.

As we can see in the image, the initial state is an empty board. Player X starts the game by having the choice of placing an X on any open space.

At this point, player O has to decided where they wish to place their initial move. You can see that even the simple game of Tic-Tac-Toe can result in a long decision path (and I have only shown you the initial 3 levels of the tree).

Enough about that simple Tic-Tac-Toe game. But it served it purpose of getting you to think about the decision tree and I wanted to show a simple tree, since many of the games that we design here at BGDF are *more complex*. Most of the games designed here will include multiple decision trees, as a result of just how the initial state of the game begins for each player (think about how many different starting scenarios occur for Settlers!)

So why am I rambling on about this topic, as a designer you can use these decision trees to create optimal strategies by placing consequences in terms of VP earnings.

Think about the initial task in Settlers, placing your beginning settlements. The decision of where and the long term effect of where you place your initial settlements can truely alter your decisions for an optimal path to victory. As a result, as a designer we need to understand what paths to victory are possible from "every starting point*. Well for that matter we need to understand that there is *always* a path to victory, well unless you dont want any players to win at your games :)

A simple example of assigning long tern consequences, the simple concept of " Should I go to college?"

In this example it can be seen that the path A to B1 to C3 is the optimal path based on reaching the highest income earnings. The same is true for our game players, they want to assess and figure out the best payout path for achieving victory in our games.

So why decision trees?

1) They are useful tool for searching for holes in a game design. If you go through an excerise of creating the potential decision trees for your game, you might be suprised to see what it uncovers. Though I would agree this might be very tedious for some games.

2) They help you as a designer understand what goals and choices you are creating for your players.

3) It can help you determine if a game is "too easy". When looking at your games from a decision tree perspective, it can help you to determine if you are giving the players to many optimal paths to victory. Thus you may not be causing tension and those lovely agnoizing decisions.

Enjoy (hope this makes some sense I jsut tossed it together, a bit to quickly IMO but oh well!)

larienna
larienna's picture
Offline
Joined: 07/28/2008
[TiGD] Theory Series #1: Decision Theory and Games

Look, from what I have read about game theories, is that these theories are a way to convert the various decision and strategy selection in game into a mathematical fomula. These mathematical formulas would allow an automated process to easily make a decision, in other words develop computer AI.

As humans, we somewhat use the same concept but with common sense instead of complex mathematics. We use our judgement with external information, like probabilities, to know if this decision is better than another. The computer can`t do this. This is why we need to make maths with it so that he can "compute" a decision.

Of course, we can make errors in our evaluation, while the computer won`t make an error since the math say that it is the best choice.

So in the end, I think game theories are more useful for developping video game but they are not completly useless to board game.

Zzzzz
Zzzzz's picture
Offline
Joined: 06/20/2008
[TiGD] Theory Series #1: Decision Theory and Games

Larienna,

I agree that people who create computer games are more likely to use these theories, use them to create AI algorithms.

But many of these theories are the core of many subjects in real life. From economics, sociology and so on. And again I feel these theories are great tools for us designers to consider. I dont expect that many (if any) board game designers will truely use these type of theories to their fullest potential, but understanding these theories can help us all design better games, IMO.

With these theories, we can assess *every* aspect of our game, depth, optimal victory paths, maxmin (sum-zero, non-sum zero games), and so much more. And by doing so we can create more stable and solid games.

But I agree that actually carrying out any other these theoris to the *fullest* would be a rough and ugly chore for most of us.

Infernal
Offline
Joined: 12/31/1969
[TiGD] Theory Series #1: Decision Theory and Games

When designing a game I try to have the player make two types of descisions (not 2 descisions but types).

The first type is the Tactical descision. In this type there is a definite best strategy (but the best will change depending on the situation/context).

The second type is the hard descision. In this type there is no best strategy. In fact I try to make it so that all options in the descision tree for this chioce are equaly good (an nessesary) but also mutually exclusive and have some compromise(This is what make them "Hard" choices).

By having these 2 types and presenting them several times through out the game and letting them interact with eachother and them selves, it creates a dynaminc and varide gmae where each strategy is valid and the player is presented with meaningful choices.

Lor
Lor's picture
Offline
Joined: 12/31/1969
[TiGD] Theory Series #1: Decision Theory and Games

This forum is the college course in gaming I never took!

A Decision Tree superimposed upon a game is to me highly suspect. It might work well on a railroad game. What happens when all the paths to victory are exhausted? The gamers eventually know all the routes, the game gathers dust. On to the next. Well, I guess that's one way to keep the game economy on the move! ;-) Chris Crawford ("Balance of Power" pioneer) bemoaned this very situation in playing interactive video games. Always the same canned response, the same result, given a certain player action.

Okay, let me throw in my 2 bits on the topic questions to suggest an alternative mindset:

1) Do you as a designer think about decision making while designing your games?

Absolutely.

2) How do you incorporate decisions making aspects into your games?

I design tools (such as rules of movement, alternate actions and resulting gains or loss) into the game which allow players to exercise some command at every step. I encourage and reward risk-- usually.

Thought of some more examples...

The tilting Labyrinth marble game involves decisionmaking. Even Rock, Scissors, Paper involves decisionmaking, and at the split second scale too!

Dice games which allow you to split the die value involve decisionmaking. Any dice game I design uses that mechanic, as well as the expected special action on doubles. Never overlook the expected, the tried and true.

3) Would a game be a game without some type of decision making?

IMHO, no, merely an amusement which doesn't engage higher level cognitive function. Rube Goldberg type games fill this description. They fascinate, they may even inspire, but don't ask for your intervention. They are "low demand."

I think a lot of game builders lose this distinction. When you begin authoring draw cards which define the next move, propelling you here and there, you know you're on the wrong track. But if the card says you will gain certain resources IF you move here, and or lose a curse cast upon you IF you land there, IF you can stash the card and apply it just to get out jail... there's a bit of luck to the draw, which is like life, a useful model. But these are far more effective decision support mechanisms than Go Here, Do This Nasty Thing, Pick Up That Gnarly Griffin and see how another player reacts.

This is my bone to pick with prepared decision trees, although I can see where some plotting becomes unavoidable. I guess if they're artfully hidden the player makes the decision, steps forward and only when the hatchet falls does he get a glimmering he went down the wrong path... I hope another life is made available. Better I think to create interesting situation drawn from the game theme which players will encounter time and again and have available useful ways to cope. For me, that is far higher entertainment.

An engaging game demands, period! It asks of you to determine a path of action, it doesn't plot one for you, and I question the wisdom of attempting to plot all possible outcomes in this manner. This kind of knowledge representation STILL hasn't achieved real AI!

Rather, provide useful tools. The player has adopted the Game Need (game object), and will be making the hard decisions, and be seeking useful support (or "reward")-- and there should be many good ways to achieve a goal, not just one optimal path. As a player THAT potential would excite me.

Perhaps a ratio can be constructed which measures game decision requirement based on high demand, low demand, high support, and low support. It might vary even within a game. How well do you support your players when they make certain decisions? For me, that's effective plotting. It's a lot like parenting (and a quieter substitute!) As the age level rises, so does the demand on the player, and so support must become more sophisticated, less obvious.

4) To what depth should a game designer understand theories that influence game design?

Well, for me, it's an ongoing education. I would master them all to the point I know what's being applied to this or that design. I haven't played every game out there and certain gaps in my knowledge are downright embarrassing. But I know what I enjoy and that's one guide.

5) Was this a useful post or did I waste my time? ;) (just checking to see if I should continue the Theory Series)

Crucial!

Infernal
Offline
Joined: 12/31/1969
[TiGD] Theory Series #1: Decision Theory and Games

Quote:
A Decision Tree superimposed upon a game is to me highly suspect. It might work well on a railroad game. What happens when all the paths to victory are exhausted? The gamers eventually know all the routes, the game gathers dust. On to the next.

This would apply to a rigid decision tree that has no variability. If you add in factors that change the value of a particular branch or branches then that can rejuininate the game. Some of these factors may be hidden or random as well.
Considder this:
You are 5 points from winning the game. You have the choice of playing card "A" which may give you 5 points but if the other player has card "B" then they can block your card, leaving you with no points. Or you could play card "C" which would give you 3 points and then next turn you could play card "D" for 2 points winning you the game, but these can't be blocked as easy. Which do you do?

This is an example where hidden infomation (whatcards the other player has in their hand) which make the decision tree murky and uncertain. You could replace the cards with any decision (eg goods bought or sold in a trading game).

Having a decision tree does not nessesarily make a game stale after only a few plays. Infact a good player will know the decision tree and be able to adjust their strategy depending on the situation based on their knowledge of it, but a well designed tree will still make them have to think.

Lor
Lor's picture
Offline
Joined: 12/31/1969
[TiGD] Theory Series #1: Decision Theory and Games

[This would apply to a rigid decision tree that has no variability. If you add in factors that change the value of a particular branch or branches then that can rejuininate the game. Some of these factors may be hidden or random as well. ]

So it's less of a Decision Tree-- rather a Consequence Matrix! A certain effect arises from a specific cause. Deciding to play one card over another may present an unknown situation (although designer must know it and plot it). That feels better, a bit more like life.

Designing one of these to emulate spontaneity is tough work, and can only be done by playtesting with the widest variety of personalities and play syles. The situations that crop up under testing always reveal dilemmas you are unlikely think up in a vacuum. That's certainly happened to me.

Is there another way? Maybe. Design a game where you make up your own rules and value system, and so does every other player-- playing only under the most minimal of regulation and for an established goal. Is there anything out there which offers that degree of freedom? Does it use a playing board?

Syndicate content


forum | by Dr. Radut