Slashdot Mirror
Microsoft Packs Up, Moves To Catan
Klytus writes "According to a press release posted at GamingReport.com, Catan GmbH and Mayfair Games have made a deal with Microsoft for the creation of the 'Catan Online Game', based on the popular Settlers of Catan board game. From the post: 'Available on Microsoft's Premium Game Service, the Catan Online Game will be accessible via both the MSN Zone and MSN Messenger. There will be multi-player online and download variants, each with rich graphics and a sophisticated AI.' Now we can play Catan endlessly through MSN." This deal is in addition to Capcom's various versions of Catan for console and PC in Japan.
it's news because Settlers of Catan is one of the more popular board games out there (i.e., it's got geek appeal). It's the third highest rated game at boardgamegeek.com.
I recently picked it up at one of those christmas/calendar stores that was closing for 25% off. Haven't played it yet though...
The game I'm waiting for an online version of is Stratego. Found a java version a while back that was buggy, but I wish zone/yahoo/someone would license it (or at least put up a nice ripoff-ware version)
Who would I have to kill to get them to do a version for x-box live?
This is exactly the kind of game that live needs, FPS and sports are all fine and good, but what harm could a little variety do?
On Wall Street they say "buy low, sell high" On the pad we say, "buy high, sell high" Isn't that somehow better?
lawsuits are what happens when you break someones copyright.
Settlers of Catan is quite a fun game. The only problem is that the games can be very long. When my family (6) plays, the games average 4 hours. So we don't play it very often because it takes up the whole day.
First of all, there IS a settlers game for windows out there, since more than 4 years. It's was done by funatics development in strong cooperation with Klaus Teuber (the creator of catan). There's also a version of this game for the PS1 (which does not use the same AI) and some more versions for the Settlers Card Game and for Settlers in Space (i actually don't know if it's called exactly like this in english). I've done the AI for the PC Version, but many ideas for it where also contributed by Teuber. I'm not going to do an ad - i've stopped working for that company several years ago ;)
:).
... (once more i guess that anyone in the game industry has heard or said this sometimes).
It was a hard projekt to do, but i wouldn't blame it so much on rules which are to difficult to play for AI, but more on the usual factors which happen a lot in game development. Like me not beeing very experienced at this time, very tight schedules (first project of a new company - you're not allowed to miss many milestones in such a situation) and some more problems which would probably sound familiar to anyone in this industry (like doing a peer-to-peer network solution and still trying to synchronice the game and the bots).
The basic high-level strategies are mostly getting the longest road, getting the most knights, or winning by building the most settlements and cities. Then there are a rules how to complete this goals and to check if they can be completed. At the lowest level are the ratings for individual building positions, which are mostly influenced by the amount of resources you can reach by them, by harbours or other special points (in other scenarios than the basic one) and by the danger of someone else reaching them first (or the need to reach them to block someone else). A set of rules for which resources are most useful for the player is needed (different strategies need different resouce distributions and certainly it's also important too see what others do already own, to find out which resources will be rare).
The setting of the raiders is easy - check the resourceneeds and if you want your bot to be emotional then make sure to remember who angered you lately and prefer this player and play some additional samples just to make sure he knows the bot is pissed off
One of the most complex parts is the trading. We did several approaches to this and i guess if we'd had more time we would have tried some more. Transfering the communication of a table game to computers is close to impossible. You just won't get that feeling when playing against bots. It's not so much which resources the bot needs, but more of a problem like when to stop trading or when to push the player some more in hope to get another resource. Some problems: In real life people talk at the same time (which sounds horrible in a computer game - you have to do it serial) or stop trading by just looking someone into the eyes or at theire gestures. These are elements which can't be simulated in a computer game (actually we did insert some shortcuts for "stop bothering me" and "make faster" in the first patch).
Besides the basic rules we inserted some more scenarios which all had different game rules and needed other strategies, but once the basic catan ai did work it was not difficult to adapt it. To make things a little harder for the poor AI we also gave each bot a different personality which had to take care that his reactions (and the animations) did fit those of a few friends sitting around a table.
After the projekt i was not too happy with the results of the ai. It was a fun game for someone new to catan, but experienced players could beat the ai without much problems. If i just would have had a little more time
An interesting fun fact: Teuber did himself complete a whole AI which did not play bad just using an excel sheet! You typed in the dice result and your move in a column and using some heavy magic the sheet displayed the answering move of the bots in another column. Most impressing thing i've ever seen done in excel.