Under rating threshold (hideshow)
I may try to add another level of difficulty. Not quite sure how to approach it yet, might require changing the way I wrote the AI. As of now it should try to protect against threats on the flag, but only immediate threats, The AI doesn't try to look more then 1 move ahead. Thanks for the feedback!
Under rating threshold (hideshow)
Hey ranoknas, the Piece movement being confusing is def something I thought about. Because the pieces level up in a way I wanted to add in some visual cues. So you have the 4 pieces types, soldier, centaur, dwarf, and minotaur. Then they have 3 levels which allow them to move 1 space, 2 spaces, or across the board in their own directions. Finally there is a visual cue for each level. Level 1 have no weapons, level 2 has a shield, and level 3 has a shield and weapon. Hope this might make it more clear.
Under rating threshold (hideshow)
The game was fairly well done, took about 90 minutes to beat every level. I wouldn't say it's very hard on hard though (Besides the fact that its extremely difficult to lay down an A.I. that could match wits with a human in the first place). Good job.
Under rating threshold (hideshow)
This game is actually quite brilliant. Great twist on the old chess. Would like to see it developed more. Graphically this could be awesome.
Under rating threshold (hideshow)
You have literally re-invented chess. you should really give this game idea to a company that makes board games and it may catch on and become just as famous as chess, risk, stratego, etc. good job man.
as for the A.I. that can always be tweeked or changed eventually. it doesnt matter that much what really matters is the gameplay itself.
Under rating threshold (hideshow)
So there are two things I noticed: one is that as has been mentioned (and is apparently a problem even on hard) the AI does not protect the flag, which as the term is even in the title you'd think would be mandatory (i.e. if the enemy threatens the flag you *must* defend it if you possibly can); the other being that piece layouts are randomized every game, which rather diminished the 'puzzle' aspect in my eyes (you'd need an extremely sophisticated generator indeed to make puzzles of a consistent difficulty every time, and it's hard to swallow that this is one as opposed to the standards for difficulty just being lowered.. which it sounds like it has in other comments, too). I am enjoying it, at any rate, and I'd like to see more, although if you could personally design the puzzles next time I'd really appreciate it.
Under rating threshold (hideshow)
Not sure where else to post this, the Hard AI has been improved quite a bit although I am not yet done improving it. Any feedback would be appreciated!
Under rating threshold (hideshow)
Interesting and entertaining, but one humungous flaw- It doesn't react to "check" on the flag. If you put it into check, it should immediately attack that piece.
Under rating threshold (hideshow)
AI loses when the flag is captured. So, if there's a powerful AI unit between a somewhat weak human unit and the flag... THE AI MOVES IT AWAY SO YOU CAN CAPTURE THE FLAG???
Under rating threshold (hideshow)
I'm still playing on easy, but I find the AI really stupid : I was threatening the flag, and the AI had a way to take my soldier with a dwarf, and instead chose to move another soldier. I hope that the other difficulties will be more challenging...
Under rating threshold (hideshow)
I will be trying to improve the AI this weekend and fixing the end game screen. I have a full time job and wife so that is why any updates have been slow. If I could improve the AI I will prob remove the Medium mode and replace it with Hard, while adding a Very Hard mode for the new AI.
To respond to skiracer33's comment, Easy mode would not go anywhere, there are plenty of people who are not great at Chess and I think Easy mode makes this fun and challenging for them.
All the feedback I have received on Kongregate has been awesome and I will be taking a lot of the advice to heart for a Check Flag 2 down the road.
Under rating threshold (hideshow)
Even on the highest difficulty level, the enemy AI almost always refuses to take a piece if its piece would be taken-- to "trade." Even if it would take two of my pieces for one of its own. Even if my vulnerable piece is about to take one of the AI's pieces. Which is silly.
So here's an idea. I'm not sure how easy it is to do this in Flash, but maybe assign each piece a "point value"? Then the AI can compare the point values of pieces (maybe adjust them if it's about to lose a piece anyways) and decide whether an exchange is worthwhile.
Additionally, the AI should ALWAYS protect the flag. As it is, I've been in this situation:
XMX
XEX
XFX
Where M is my piece (a shield dwarf that moves two spaces), E is an enemy piece, F is the enemy flag, and X is an empty space. And the AI does this:
EMX
XXX
XFX
Literally committing suicide. That shouldn't happen, ever.
Under rating threshold (hideshow)
Very nice game. Like everyone has said, the A.I. is not particularly smart, but I won't rate down for that because I would assume that writing the script for a good A.I. would be very difficult and take a long time.
Under rating threshold (hideshow)
Could use some low key background music, I know picking the right type of music is tricky, but it would help IMO. Just remember to be able to mute sound effects and music separately.
Under rating threshold (hideshow)
I conquered cities. I led armies to their inevitable fate. I saw men cry out in anguish as they were slain in the pursuit of their goals. I conquered those who had opposed me, and defended my honor as a general with the perseverance of my men and my own skills of leadership. But now, after all the bloodshed, I can finally clai-
"Congratulations, you beat easy, can you beat medium and hard?"
FFFFFFFFFFFFFFUUUUUUUUUUUUUUU-
Under rating threshold (hideshow)
Interesting game, but the AI is HORRIBLE. Several times I attacked with a piece (leaving it undefended) and it completely ignored it, allowing me to kill again with it. To implement a basic AI that doesn't suck this bad, do a short breadth first search with heuristic values. Give every piece a point value (based upon how useful the piece is) with the flag having a huge point value. The value of any state is then SUM(values of your pieces) - SUM(values of opponent's pieces). Then do a 2-3 move look-ahead using min-max search (you assume your opponent attempts to minimize your value and you attempt to maximize it).
Under rating threshold (hideshow)
While I love the idea, it is very difficult to tell at a glance which pieces can move where. The shield and weapon are neat touches, but they just don't stand out. I would use color instead - maybe the rank 2 pieces could be white with blue trim and black with orange trim, and the level 3 pieces could be white with green trim and black with red trim, in addition to gaining shields and weapons.
I am currently working on another game. This is an idea I would love to revisit at some point in the future and add multiplayer and more pieces and game modes to.
Under rating threshold (hideshow)
Dev: Have you seen chessvariants.org? It has hundreds of varieties of chesslike games. Could give you some new ideas for pieces, board, objective, etc. I wish there were more games like this on Kong. Thanks!
Under rating threshold (hideshow)
in reading your posts JoshuaJewell, I think the key fault is that the AI only looks 1 turn in advance (either attacking or defending). Most chess AI software looks 8 turns ahead. Granted, this game is a little simpler than chess, but it's very easy to fool the AI with moves like sacrifices etc to get its pieces out of position because it can't see that if it moves a piece, it's dead in 2 turns.... Otherwise a great game with a lot of potential!
Under rating threshold (hideshow)
Quite a brilliant game of chess, eh?
I haven't gotten to hard mode yet (And probably won't, It's a bit repetetive), but I have to say easy and medium weren't hard. I know you've addressed this, so I suppose this is already being fixed.
Under rating threshold (hideshow)
Even if I hated this game, I'd give it a 5 just because it's a new concept and not some duplicate with a small change. Too bad I can't rate this a 10.
Under rating threshold (hideshow)
Dwarf: "For the love of God soldier the flag is to your LEFT!", soldier: "It is not." Dwarf: "I can't fend of three minotaurs by myself! Just take one step to your left, PLEASE!" soldier: "Nope, don't feel like."
Under rating threshold (hideshow)
Love the pieces in this game. Wish there was an option to play on the largest board with the most pieces vs different levels of AI though, once you get to the final level there's not much more to do as it's pretty easy to beat the AI. Would love to see a version where you can set up a game against human opponents (two people on the same pc) too!
Under rating threshold (hideshow)
If you work on anything, work on the A.I. The graphics level is perfect for this game; anything else will only make it more confusing.
Under rating threshold (hideshow)
besides not thinking ahead, the main thing I'm noticing is that even(especially) on hard, the AI is way too protective of it's pieces. I had a shield minotaur threaten a basic dwarf, while the dwarf was protected by a shield human, and I had no back up after the minotaur attacked, but the AI moved his dwarf back anyway. you need to make the AI somehow understand what a good deal is, and that giving up a basic dwarf to kill a shield minotaur is acceptable. perhaps give the units point values, and have the AI try to get the mos points in addition to trying to protect the flag. Although that might require thinking two or three moves ahead anyway.