[TiGD] State Tracking

Without some kind of tracking, you are playing "Pretend", a sort of game without rules that most of us played as kids. The closest adult analogue that I can think of is Amber Diceless Role-Playing, but even that has character sheets to track certain information about what the character was like just after character creation.

Double blind chess. Both players simply remember the position on the board. 'Course, you have to be pretty good chess players to play it.

Fundamentally, every strategy game is what computer geeks call a "state machine". The game can be in any of a (usually huge) variety of states. Transitions between states result from player moves and random processes.

The question here is how much of the state is represented by the positions of game pieces, and how much of it must remembered by the players. In general, I prefer as much as possible to be represented with pieces.

-- Matt

How about "Who or what Am I?"
A game we used to play in the car on road trips. My dad would give the initials of the person and we would have to ask yes/no questions about the person/thing until we guessed it. It was a lot more fun than it sounds...at least when you're eight.

I guess if you made it into twenty questions, then one would have to keep track of the number of questions. But we didn't do that for some reason.

But this brings up a questions: Does this imply that if non-state tracking game exists, it is potentially infinite?!

What other properties must a non-state tracking game have?

-John

I think that a game is not a game if it cannot keep track of states.

I think the goal of this thread would be to find new ways of keeping track of states. For example, I once had the idea of making a card game that does not require a table to play ( allow to play in bus, car, etc). So theorically, you cannot keep track of any states by placing cards on a table. So the only way to keep track of states for a game with this restriction are:

- The card composition in your hand
- The card is upside or upside down
- The card is face up or face down in your hand.
- The position of the card in your hand ( 1st on left or right )

Here are some other restriction ideas. Let say a game which has a board and some glass bead of the same color ( for the same player ). Unlike tokens, beads cannot be fliped, rotated and cannot have different meaning since they are all the same. So what can keep track of states here :

- The position of the bead on the board
- The number of bead on the board, outside the board or in any other particular area.

It seems that most of the discussion is focused on the question of "can there be a game without state tracking", but I don't think that's really the most pertinent question for the kind of games that most of us are working on. To me, the opposite situation is the more likely one, where there is too much state tracking. This can lead to a game with too much information for players to process, or one in which playing the game becomes fiddly because every time you do something, you have to adjust the game state in a substantial way.

The metrics I typically look for is how many markers are required to notate each player's game state independent of the number of playing pieces, and how often the state markers need to be adjusted. The exception is the players' pieces that go on the board to mark territorial possession or whatever; it's a given that there might be a lot of those, or that they might move quite a bit. But if you have to mark, say, 3 production levels, 6 technology levels, 1 VP level, etc, it just becomes something to look out for.

I also find the question of clever state tracking mechanics to be an interesting one. In particular, are there ways to integrate state tracking so that it actually becomes a game mechanic unto itself? I think that's the ideal situation. One example that comes to mind is Knizia's Lord of the Rings. The various tracks on each board track the game state in that they measure your progress through the board, and indicate how far you are from completing it. But, they also have functionality, since most of the spaces either give you tokens or force you to lose cards or roll the die or whatever.

One system I've come up with that has a nice blend of state tracking and functionality is in my new Thirty Years War game. I wanted to have territories become "devastated" and have this reduce their production capacity. The way I initially wanted to do this was to give each territory a number indicating its base production, and to then add "devastation markers"; so, to get the production value, you subtract the number of markers from the base production. This was functional but not very elegant, and rather fiddly. The solution was to give each territory two types of features: cities and estates. Each devastation marker now covers up a city or estate, and the total production of the territory depends on the number of undevastated cities and estates. I made this of further interest by awarding VP bonuses to the players with the most cities or the most estates, so now there is also some incentive to preserve one or the other depending on which bonus you're pursuing.

-Jeff

It seems to me that the discussion may be getting a bit too broad in terms of what constitutes a game state. Games by definition are a form of competition and/or solitaire accomplishment, therefore if we define a state as any method of determining the winner, then even something as simple as peek-a-boo qualifies because discovering the adult grinning like a lunatic establishes in the infant a sense of accomplishment and "winning" the game. It's an interesting intellecutal exercise but does it accomplish anything in terms of provoking design ideas?

It's interesting that Ticket to Ride is one of the examples listed, because the given state-tracking mechanic, the train car, is not at all a way of determining the score, and is not a strong indicator of who may win the game. A player can have trains criss-crossing the country, but if he does not complete his assigned lines, they are virtually meaningless in the final tally. The score is kept with a nondescript stackable marker and the cards for the assigned lines are hidden from the other players.

So what do the trains represent? Obviously they represent constructed train lines, the immersive effect of which is part of their appeal. They also represent the development of a player's strategy over the course of a game. It's difficult to surmise a player's final score from his layout, but if somebody is making a beeline on a long route, then state tracking is defined as trying to divine a player's intended moves in the future and planning moves to either aid or impede that goal.

Look at Monopoly, destroyer of families. Is the state of the game determined by the amount of money held? The raw number of properties? Or is it the little houses and hotels that start to sprout from monopolized portions of the board? Those pieces indicate a state of advanced development and clue the players in to an additional (dis)advantage present in that area.

Look at Carcassone. Is there really any strong indication of how a dynamic tile-laying game is going? The endgame can mean a massive point swing when totalling open fields, or it can be a complete bust for the person trying to mount a comeback. Meeples do not represent areas you HAVE developed, as scored cities remove the meeple from the board, but areas that are BEING developed. The state tracking function of the meeple is therefore to show where future tile development may benefit you in the short term.

Look at Life. On second thought, don't look at Life as an example of gaming. Ever.

To this end, it might be good for people to keep in mind that games can have multiple objectives, and therefore, multiple methods of tracking the state of each objective. More on that in a minute.

So in a sense, the topic is too broad, and yet not broad enough. By that, I mean that any feedback the game provides the player could be considered a form of state tracking, but on the other hand, certain tracking methods will be specific to their own assigned values, which may be of varying priority. Again, more to come, as my epiphany sort of developed through the course of the post.

Settlers is a perfect example of what I'm talking about, and here's why: tracking resources only tells you about one aspect of the game: how much one player is able to impact the economy of the game with what he possesses. But there is also the matter of the positioning of his settlements, the roads he has in place, the number of soldiers played. In the Settlers of Canaan adaptation, there is also the question of who has stones in the Wall of Jerusalem.

Each of these methods tracks one aspect of the game: economy, board control, developement, etc, and each of them help contribute to yet another state tracker, the victory points. The brilliance of the Settlers model (also reflected in other Mayfair games I've played like Domaine) is that anyone can examine the board and calculate the number of points each player has, by checking each of these various states.

Now here's the epiphany I just had: not only is it important to have ways of tracking a player's progress, not only is it better the simpler you design it, but I think it is also important to prioritize what states have the most impact, and make them proportionally more or less visible. In Monopoly, houses and hotels are what will destroy bank accounts in the endgame, therefore they are the only pieces other than the event cards and player pieces that actually have a home on the board. In Ticket to Ride, the points circle the perimeter of the board (at least, they did in the Marklin version), and are small nubs, while the trains are attractively molded pieces that spider-web across the board. In Apples to Apples, not only do the green apples make it easier to tell the clues apart from the responses, but the visual differentiation really cues in a player to who has points sitting in front of them. Risk is a conquest game, so the armies that populate the different regions are the most prominent feature of the game.

So not only should we tie specific trackers to specific goals, not only should we keep it as simple as possible to monitor them, we should also prioritize and design the display of those trackers according to their impact on the game. If this proposal is correct, one would assume that if the trains were not such a key element of Ticket to Ride, I would be more inclined to actually change my gameplay design to make them more central and integral, rather than either making (for example) the point total more visually grabbing, or just leaving it there and having trains of little use just cluttering up the board.

:edit: jwarren's solution for his Thirty Years War game is another elegant example of this. Simplified tracking mechanics that are visually integrated into the gameplay based on their import.

Thoughts?