Special conditions: USA found a pot of gold at the end of the rainbow: starts with bid 50!
Remember AI will fairly randomly spend this, ranging from all land-units to all navy or a mix.
Of course reloads are legal from the start of the game, so take your pick!
Overwhelmed by all the submissions it took me some time to review them all, but I was finally able to come up with an endresult! All those who contributed with advice on the competition I'm looking forward to seeing your participation in the next Game of the Month!
@ZjelcoP - Good challenges. I hope to get around to giving them a shot myself :)
My only suggestion is to start the challenge after the bid phase so that everyone starts from the same place. The main reason for this is that the AI's bid phase is really random and if you want to maximize your chance of winning the challenge then each player would probably keep restarting til the AI makes a really poor bid.
Hope to see your battle with your own creation Redrum!
I had been doubting on that point when thinking it up. But as I explained under special conditions, reloading from the start is the intentional special snack of this experimental challenge!
And don't worry, I haven't seen the AI make really poor bids, just very different ones.
Have to try it out to see what people think of it.
I would have to disagree about the AI not making poor bids. In the game I posted with the rules question, it bid 6 transports which is pretty bad. If I remember correctly (sorry no save), the first time I attempted the challenge it bid a carrier, fighters, a transport and a tank in Sinkiang, which is a really good bid.
This isn't necessarily a bad thing though. I suppose this is a good way to some feedback on the AI's bidding process. Players who are really competitive might be restarting once or twice during the first turn anyway to try and get better dice.
Just my input, I really think that LowLuck should always be used, at least for the experimental division. Some restarting for different bids is one thing, but doing for different dice is an ugly road to go down.
Although I'm not a huge fan of LL generally, people expressed an interest, so I went for it anyway. No tech, No SBR. +50 to USA on the bid. They used it for a battleship and carrier and a fighter in sz10. I just slogged through in one go, no reloads. I think LL currently works to the AI's disadvantage, since the HardAI doesn't really understand how strafing works and can be bloodied by the strafe mechanic pretty ruthlessly hehe.
Took me 15 rounds straight through to take Washington. Here is a gamesave in case anyone likes for reference...
Fun stuff though. Good old Revised still has it's charm.
I'm a little confused about what the experimental league is. Does this mean using the latest pre-release jar? Cause that is what I have been doing. I don't see much reason to backtrack. Whenever a new Jar is available, I'd suggest we use that one, so that Redrum can collect as much feedback on the current incarnation of the AI as possible.
Again to to return to the LL thing. I can't see any good argument for forcing the LL playstyle on our players. I know some people are very fond of LL, but the fact remains that Axis and Allies was not designed with LL in mind. It is basically a house rule, and a player innovation, and one which changes the gameplay dynamics substantially from the OOB system. I'm a definite partisan for dice, at least when it comes to the official boards. Some TripleA specific games may have been designed from the ground up with LL in mind, and in those instances it would make sense to have that be the default, but for the OOB boards I think its pretty boring. Every poll I've seen at A&A.org still suggests that LL players remain the minority, despite being fairly vocal in their preferences for this playstyle over dice. Its fine to accommodate people who enjoy different things, but I think if you restrict these challenges to LL only, you're probably going to have fewer players participating over the long haul.
What you gain from LL by controlling the potential TUV swings or battle reloads in any given game, is undercut by the fact that the computer doesn't really understand the LL playstyle. The AI is presently a a pretty lousy LL player, so forcing it to play this way instead of dice is like giving the Human a huge advantage over the machine. Until the AI understands how to strafe, and deal with multi-turn attacks, the Human will always kick its ass in LL, so if anything I would suggest Dice in the experimental league.
Just to be contrarian haha ;)
On the subject of the the +50 bid, and how USA can use it. This was probably the nuttiest one I saw... USA spent the entire pile on infantry, which would clearly have made this particular challenge a royal pain lol. AI bought 13 infantry and a 2 tanks.
Human Axis decided to pass on that one.
ps. New game below... Here USA bid the carrier 2 tanks 2 fighters 1 art.
In under the wire. One last match vs the Machine before the month is out haha! Good old revised
Again with USA +50, LL no tech, no sbr vs skynet, since the machine doesn't know how to retaliate yet hehe. (This allows for some interesting AAgun options against the AI, since you don't have to cover vs SBR like you would against a human.) Opted for a Germany focus this time with a Kriegsmarine expansion.
May I ask how many entries were there? I like the idea of having an experimental league and regular league, but I wonder if they could just be simplified to a dice league and an LL league. Winners would be posted from both. Or if the number of players entering isn't big enough one league might be enough.
@Black Elk - I think the idea of LL was to ensure a more fair competition, so the better strategy wins, not the better dice. This is obviously a huge debate for TripleA and strategy games in general, but in this format all the times a risky move doesn't pay off don't really matter if it gets made in turns 1-2 or 5-6, compared to Human vs. Human where it could cost the game.
With that being said, I also prefer dice (for most maps) and it would be great if the AI did factor LL into its decisions better.
Haha, Well 3 entries in total!
Not sure what kept others form joining in...
Over 2000 views on AxisandAllies.org and plenty comments on this forum.
Anyone willing to comment on this go ahead.
Time of year? Very nice weather over here..
Complicated? Think the basic challenge is pretty straight forward, but that is my view. Instructions unclear?
Just takes some time?
Well there has been quite some discussion on settings:
LL vs Dice.
StandardAI vs Latest AI
Bonusses or not.
Ofcourse it's impossible to please everyone, but I think all options are in the two challenges now, just not all combinations.
@Black Elk: Experimental means playing with latest prerelease jar, and often bonusses or whatever else we can think of to keep it interesting. Regular is meant to be as simple to play as possible (no fileswapping, strange settings, or harder AI.)
Plan to keep it going for a while and hope it picks up.
You could make an XML or TSVG file with the settings and putting it in the opening post, maybe that makes things more clear?(I thought it was straightforward as well). Maybe the discussion of unclear rules or illegal moves in the threads confused some people? Would you be allowed to post something about it in the lobby?
What if you made the regular challenge extremely straightforward, and have a choice between LL or dice (maybe all the other settings as well). Then the experimental requires downloading the latest AI, and requires LL. Though there could probably only be a single challenge until the number of participants increases.
I think that might be helpful, having a gamesave for a quick load.
Also, sourceforge was down for most of July, so that could have effected participation.
I tend to enjoy the more moderately scaled maps myself, since the AI can take a while sometimes when the map is large. I'm excited for the v3, v5, and Global challenges when those roll around.
I'm still interpreting this whole challenge thing as a "personal challenge" more than a competition or tournament, so I still don't see the need for LL. But that's just cause I'm a big hater probably hehehe. LL has its charms sometimes, but I'm a dice man at heart.
Maybe we could try to time it so that when the next stable (with all the AI updates) is released , that the "game of the month" for that month is one of the latest OOB games? I think the first month shows that controlling the bid conditions more narrowly might be helpful.
For a future v5 challenge, I'd suggest the Russian bomber bid or flat rate bous +5 to AI as pretty entertaining.
The regular challenge I tried to make extremely straight forward, not sure how to improve on that.
What I didn't want to do is to require people to download a save, might be too much or too complicated.
For Experimental i did plan on using saves at some point as it opens doors to all kinds of scenarios, possibly created by editing.
The regular challenge alternates between dice and LL. So LL this month.
I liked to set some sort of challenge for myself as well when playing the AI, however to me it seemed like more fun to compete against others. To have several options to let people choose from would make it impossible to compare results.
Now people have 2 options to choose from, future will tell if both are here to stay.
I sometimes prefer the bigger maps, allthough i agree the AI can get pretty slow there.
Elk: What is OOB and Hr-idea's?
Might go for a classic challenge too, played it recently and was surprised how well the AI played as Allies by setting up a pipeline to norway.