Several years ago, players were very angry and 1-star rating the Monopoly iPhone game because they were POSITIVE the AI was cheating. The developer pleaded to them that there was no cheating and that the dice rolls were exactly that: entirely random. Still, people would not accept that answer even from the developer's mouth and continued to hate-rate the game into oblivion. Finally, the developer posted a screenshot of the actual code of the dice rolling part of the game, and even pointed out that since all the outrage had started they had updated the game on the sly and tilted the RNG in the player's favor to address everyones complaints. And despite all that people called the dev a liar and continued to 1-star the game because they were absolutely convinced the AI was cheating and the game was screwing them over. The moral of the story is that yes, you'll probably just have to let it go. Human nature is funny, and we have a tendency to remember with absolute vivid clarity anytime a game like this has a lucky roll that works in the AI's favor. Of course, there are probably just as many instances of the RNG working in the player's favor and getting them a miracle win, but for whatever reason people don't remember those good times and only remember those bad beats. There is a term for this but I can't remember what it's called. Anyway, short of completely rigging the game in the player's favor you'll most likely never convince some people that the game isn't out to get them, and even if you did rig it people would complain the game was too easy and that they always won. Pretty much a catch 22
Please, whatever you do, don't dumb down your AI! Having a strong non-cheating AI in a game is a rare find these days.
re: AI - The AI in this game IS quite difficult, but I don't feel like it's cheating. I'm actually more curious to hear more about how you coded the AI to play so well.
You can't counter it. It's a form of bias. Tons of people accused Puzzle Quest of cheating. Even after the developer posted the code that generated the board and explained how incredibly difficult it would be to program an AI that actually did cheat, people refused to believe it. On top of that is the issue of locus of control. There seems to be a preponderance of people these days with an external locus of control. They find it much easier to blame a "cheating" AI than to accept they didn't play very well.
Above, you meant that "AI did not cheat" right? Please edit. I don't mind AI cheats, frankly, but this game is too random (and I am too dumb) to say it gets any unfair advantage or not. (that's another reason why most of the time I prefer playing against human even I lost most of the time) Keep up the good work and I hope to see it keeps improving!
Cool, thanks for the advice. It makes sense, and I guess I'll just keep responding to reviews as they pop up and hope that at least a few of them will believe me. @slothwerks - I'll make the AI the focus of a future blog post and go more in-depth. Are you interested in the coding side of it, the design side of it, or both?
I'm specifically interested in the coding side - what types of algorithms you used and/or how you taught it to play. It seems really challenging given the complexity of how cards interact in this game.
The closest I've ever seen to an attempt at demonstrating that the AI didn't cheat was, I think, in Catan for iOS, where it'd keep track of the dice rolls and show you an actual chart with the stats of how many times each dice roll had come up. I don't think there's anything you can do on your end given that the whole game dynamics and mechanisms are different, but if it's something that can help, feel free to use it. Maybe a histogram of how many times each card showed up as "draftable" by each player or character would help? But, once again, I think this will only be valid after a lot of games, so that you have a lot of unbiased sample data. And to get there, you probably will need to like the game and not believe that it cheats. So that's the problem.
Ya, that's an interesting idea. I could certainly try something like that, and maybe even just the presence of it would help people believe there's no cheating. I'll give it some thought for the future. Thanks!
Okay, I'll write something up soon. Although I'll warn you there's really nothing earth-shattering going on. It's more or less giving the AI some factors to consider with some manually tuned weights plus some custom logic for the trickier cards. And then I kept tweaking it until it felt good.
Huh, so someone actually changed their rating from 3 stars to 5 stars because they decided to believe me that the AI is fair. Pretty cool of them to take the time to update their review.
Sadly nothing at this point. Once you've maxed all leaders and have all cards, the coins are useless to you UNTIL the wondrous devs add more leaders and cards.
I was just going to mention Catan. Stumbling across those statistics got me off the AI cheating bandwagon. But it really comes down to people. When RNG doesn't go our way (even though it's statistically accurate), we feel cheated. And you'll be unable to convince some people otherwise because some people don't necessarily care about a fair game or a strong opponent, but rather winning. On a side note. Can you make the game stop before the start of the 4th round when your deck is shown? Pretty please. ;-)
Do accounts sync cross device (eg: can I play my game on Android and iOS) if I own a copy on both plaforms? I like playing on my iPad at home, but I'd love to play on my Android phone at work during breaks. I'd definitely be willing to buy for both platforms if that's supported.
Yes, that is supported. You just have to log in with the same username/password on both devices. If you always play online, it works pretty seamlessly (although if you leave the app running and make progress on another device, you might have to restart the app to sync down the new progress). If you play offline a lot, check out this post: http://forums.toucharcade.com/showpost.php?p=4027440&postcount=205
