Istanbul: Digital Edition

Istanbul: Digital Edition

View Stats:
flogae Feb 7, 2020 @ 3:15pm
Hard AI strategy
Been trying to beat two AI players on "hard" setting - succeeded once only in 10+ attempts.
Very clearly, the winning AI player completely ignores the Mosques and building up his handcart, purely going for trading goods and buying or trading rubies.
With this being an easy optimizing task for a computer, you almost can't beat him at it.

Remains to be tested how that strategy would work against human players, but unfortunately the online lobby is half dead most of times...

A pity, since it's a nice little game...
Last edited by flogae; Feb 7, 2020 @ 3:16pm
< >
Showing 1-8 of 8 comments
SteveMoffat7 Mar 5, 2020 @ 2:59pm 
of course that strategy works with human players too. Thatś why no one wants to play online beeing afraid to loose.
CalvinHobbes Mar 12, 2020 @ 8:20pm 
Interesting, I was stuck on trying to beat two AI players on "normal", I always seemed to come very close to winning but never won.

So I tried the strategy of ignoring the Mosques and handcart buildup - I won very easily the first time I tried this.

Fun game but disappointing that there appears to be an easy winning strategy which mostly entails ignoring a significant part of the game.
Anjovi May 20, 2020 @ 8:54pm 
Odd, I tend to find the mosque tiles to be very great in most circumstances. That said, I think a healthy combination of mosques and selling can make for viable strategies.

My favorites are:
red tile + the tea house/black market (can be a big boon for collecting rings if you're going for palace gems later on)
green tile + trading/palace
Blue tile is probably the most situational, though it's pretty handy all around adding up to 1 movement/activation to your chain of stuff...
yellow tile is about the same as blue but very awesome as well, adding essentially as many to your movement/action chain as you can afford.

Also, ♥♥♥♥ blocking the other players can work out pretty great, as many games seem to come down to a couple coins at the right/wrong time. Sometimes you can even deny them that space for that turn...though it's all about weighing out how much it benefits you/hinders them.

The key thing i feel is not worrying about getting gems right off the bat, but building a good little economy for future actions.

I feel pretty comfortable playing hard AI players.

I'd suggest making sure that you randomize the board as well, as this will definately force the AI to play at least a little differently from game to game.This is a big part of the appeal of this game to me.

*And don't forget cards! These can be a huge + and can determine wins/losses in certain games. It's also nice that you have access to the last discarded card, which can add some viable tactics.

Suppose if you wanted more strategy analysis in the game, you could check out this forum: https://boardgamegeek.com/forum/1442187/istanbul/strategy

Generally, board games will have decent amount of analysis and criticism by the folks who buy them as they tend to be bigger investments then their PC counterparts.
Last edited by Anjovi; May 20, 2020 @ 8:59pm
sjgf Jul 9, 2020 @ 10:38pm 
Originally posted by malcolmmackay224:
Interesting, I was stuck on trying to beat two AI players on "normal", I always seemed to come very close to winning but never won.

So I tried the strategy of ignoring the Mosques and handcart buildup - I won very easily the first time I tried this.

Fun game but disappointing that there appears to be an easy winning strategy which mostly entails ignoring a significant part of the game.

Hello. I just bought Istanbul because it looked like an interesting game. However your comment suggests to me that the game has a design flaw in it which means that each time the game is played all players should just follow the same strategy and therefore play the same way each game in order to win, which to me makes the game seem predictable, repetitious, and therefore boring.

So is there a best way to play Istanbul in order to win and is this best way the same way every game so that games end up being predictable and repetitious?
jimingolden Aug 11, 2020 @ 11:50am 
When you play random the board always changes and new strategies are considered. I enjoy playing with 4 AI's at the hard level with neutral assistants. i win about 20 to 25 percent of the time.
Corydoras Sep 21, 2020 @ 6:35pm 
Thanks for sharing these tips on how to beat Hard AI. I happened to love facing 4 Hard AIs too, in that way I can see what's the best moves (/actions) should be made, then I mimic their actions next time. I've learned and improved a lot this way. I successfully won yesterday it felt so good. :-)

I don't think there is a design flaw in game (as far as I aware). I especially love the AIs. They are unpredictable.
CascadeHush Jul 13, 2021 @ 3:12am 
Doesn't surprise me at all. This is a light euro and most people who buy the physical board game probably won't play it more than a dozen times. It probably doesn't matter that the game isn't well balanced, many aren't. It's a bigger problem when you buy a digital version and plan to play it dozens of times (I know I have with other digital versions of board games).

It makes me wonder if the developer worked out the strategy and then implemented it in the AI, or if they stumbled upon it whilst letting the AI run with various parameters.

I've often wondered how well most board games would stand up to a thorough rigorous brute force or learning AI to see what strategies are best. I suspect many of them won't do so well. Almost every board game I've played has some cards/spaces/actions/objectives that I know just are not worth anything.
Socratic Wisdom Dec 22, 2021 @ 7:54pm 
The game is not unbalanced.

The game was the Kennerspiel des Jahres Winner in 2014. That's the "gamer's" game of the year for that year, and it's designed by Rudiger Dorn, an accomplished designer. What are the odds that the game is unbalanced given these considerations?

What makes the game so great is that there is NOT one ultimate game-winning strategy. Optimal play will change every game, depending on the layout of the locations and opponent behavior. Certain combinations of locations will be favorable with any given layout while other won't.

Acquiring mosque tiles and extending your cart are very powerful and viable strategies, and I have seen the (hard) AI opponents win using them. In fact, I would say that in my experience the hard AI opponents often pursue them. The mosque tiles in particular are very powerful (I like them all a lot, except for yellow), and I almost always get at least one ruby pursuing them (I'd say I get 2-3 mosque tiles on average).

I respect everyone's opinion and grant that my experience might be different for some reason. Still, I feel the need to stand up for this great game!
< >
Showing 1-8 of 8 comments
Per page: 1530 50