Don't underestimate them, you'll play on random created maps with settings you proably won't be used to, I'd except good players to beat it. However, 1400-1600 players who still attack wastelands etc may not be better. You can always watch the games on the site, all of them are there, and judge the level for yourself.
My recommendation is that after the challenge is over, whichever bot wins, we set up a showmatch and we analyse whether or not the bot actually plays well. In addition to that, in one seasonal ladder where there will be no cards, the bot could play as a player, probably disguised as a normal player (program that it says gl hf in the beginning and gg in the end and have it make moves using random intervals of time) and after the season it could be announced that a specific account was a bot. I haven't seen any of the bots games so I'm curious if it'll have any chanec against a human being.
In all likelihood. I recommend looking at the debug output. - Stack Trace --> an error gets thrown - Simply nothing --> Probably an infinite loop, can check with System.err at beginning of each turn Or maybe there is another mistake in your code that leads to no output without anything crashing.
so pick_starting_region 10000 3 9 10 13 16 means your bot have to pick starting regions and 3,9,10,13,16 is the possible territories to pick. I recommend you print the bots output, use System.err.print() to do that.