Monopoly

[CompWiz]

Hmm, I just noticed this article on ticalc.org. Maybe you should try to keep this a fairly low-profile project, at least until it's finished. In fact, just posting about it on this website would probably be sufficient to keep Hasbro from noticeing it. Don't send ticalc.org any news about it until you release it.

[Timendus]

The second part is, if there's no feasible way around that, what types of compression could I use? (Keep in mind, speed isn't any sort of problem--it's plenty fast as it sits)

Well, if you have a large empty space in the center it's probably most efficient to use an SRL (selective runlength) encoding as I call it:

5,6,7,8,9,0,0,0,0,0,1,2,3,4,5,0,0,0,0,0,7,6,4,3,2,8

Can be stored as:

5,6,7,8,9,0,5,1,2,3,4,5,0,5,7,6,4,3,2,8

Using 0 as a run indicator and the next byte as a run length. Seeing that you will probably only have a few tile numbers, there are more efficient ways though.

[benryves]

I personally wouldn't use a conventional tilemap at all, rather using my own custom graphics system.
An easier way to break it down would be to treat the game world as 4 different tilemaps:

Code: Select all

+---+--------+
|   |    3   |
| 1 +----+---+
|   |    |   |
|   |    | 2 |
+---+----+   |
|   4    |   |
+--------+---+

Then again, standard RLE would compress the map pretty well.

[chickendude]

I was thinking about doing something like what you mentioned, but I'm not quite sure how to pull it off.
Actually, since it's a completely square board, I could have it so that once the tilemap reaches a 0 (the blank space tilenumber) that it displays 8 or 9 times.
OR! Since all of the tiles essentially go in order, from 1 to 80 something, I really don't need a tilemap. At all. I could still use my routine and stuff, but just display the tiles in order.
That eliminates about 150 bytes, plus the couple bits in the tilemap routine that can be deleted.

I'm going to look into that when I get home.

And Kirk Meyer was one of those pioneers of programming, a big shot. I'm not, so if they even find out about it I doubt they'll make a big fuss over it. If anything, I can rename it to something they won't recognize or think to check. But thank you for that link, I'll be sure to keep a look out for that.

But I'm still unsure about the computer player part. I plan to have it, I'm just not sure how to go about it. Maybe I could create a survey to find out what actual people would do if they were in such and such a situation, analyze that and put the averages into the computer's AI. I'll have to look at some simple stuff on artificial intelligence online I guess.

[CompWiz]

Hasbro's copyright states that their copyright lies with the game name, the four corner squares, and the cards. Therefore the prices and property names would have to be changed, the board layout would have to be changed, a new name found, and new rules found. By then no one would even know how to play.

so, better to just keep it from being noticed to begin with, at least until you release it. If they find out about it, changing the name wouldn't be enough.

As for the AI:

NEURAL NETWORK

[chickendude]

"NEURAL NETWORK"? Maybe it's some insider joke?

I assume that once it's out, it's out. I don't know how they found out about his version, but maybe I can release to a variety of smaller sites first and avoid ticalc.org. They'll probably be so unimpressed at my feeble attempt to recreate their game that they'll figure it's just a joke anyway.

Has anyone heard of Ghettoopoly? I wonder how Hasboro handled that

But I'll probably spend the rest of my time here at work reading up on AI systems.

[CompWiz]

Neural network for developing an AI is like this: Write a couple of Simple AI's that use numbers to determine their behavior. Run them against each other repeatedly with varying numbers(for behaviors) and see what numbers allow it to win more often. Maybe you could have the calculator do this in the background while it is waiting for user input while playing the game.

In fact, how about this: Have the calculator keep track of your playstyle, and match it up to some behavioral numbers, so it makes an AI duplicate of your playing style. Then while it's waiting for your input, it can be trying out how often different AI behaviors can beat your playstyle. Then the AI would get better and better at beating you as you played the game.

and yeah, I would assume that once it's out, it's out. The only reason they caught Kirk on it was because his wasn't out yet. If you just release it finished, they probably wouldn't do anything.

[chickendude]

Oh wow, that's insane. Maybe I won't take it quite to that extent, but maybe I'll develop a couple different AI styles and butt them against each other and see which ones tend to do the best.

Do any of you have experience in AI? Once I get the actual game written would you be willing to help out with that? I'd like to have a challenging AI, maybe not impossible, but challenging at least. That player profile thing...I like it I guess to do that I'd need to find some common attributes to record. Maybe have each person create a profile to play, and then in that profile would be the rank they've risen to (and the higher the rank, the more intelligent the computer)? Haha, this is becoming like some cult-type game. But it would be fun to program!

And this monopoly game has been open for nearly four years, if not more, and I've heard nothing from them (of course, I hadn't worked on it for over three and a half years, but still...) so I can't honestly say that I'm worried about it at all.

[CompWiz]

As for the player profile thing, I still think that it would be really neat if you could have the calculator make an AI that comes pretty close to duplicating your playstyle, then put varying ai's up against it to see what can beat the ai imitating you. There's plenty of idle time you can use for that when waiting for user input. Just run it in the background and have it check for a keypress in the loop. It can then get better and better able to beat you, as long as the program is open. The profiles idea is good though, because then if someone else plays on the same calculator, it won't mess up your ai model of the other person.

[chickendude]

Yea, I just need to come up with some values to keep track of for each player profile, like which properties they buy most often. And I'm sure the analysis would take hardly any time at all, maybe a split second. I could probably just run that right before the computer's turn.

What exactly do you have in mind for this "background analysis"? What will the program analyze?

[threefingeredguy]

Will it have linkplay? Maybe this was already asked.

[chickendude]

I can't say for certain, but once everything else has been added it's certainly a possibility. Although, if I manage to develop a decent AI, I don't feel linking would be as necessary. Still, I will give it serious thought once all the other major portions have been completed. I'm not going to guarantee it, but I will say that it's likely.

Hope that clears it up for you.

[tr1p1ea]

I dont think you would need too sophisticated AI, as most of the actions are random (dice rolls, cards etc). You would just need a few classes of diffferent types to represent different playing styles. Like types that buy up everything, types that dont, types that always want to put houses and hotels etc. Not saying its going to be simple, but you know .

[Liazon]

Are there strategic properties that one should own in order to win Monopoly?

And doesn't rent double if you own all of one color?

[kv83]

And doesn't rent double if you own all of one color?

Yes it does. But only if you have no houses on it