AI testers wanted

[Lord Zimoa]

We're looking for a few people who had strong views on the AI to help out with Beta testing of the patch. Anyone interested?

Go and post here:

http://www.slitherine.com/forum/viewtopic.php?t=8619


Thanks,

Tim aka LZ

[firepowerjohan]

We should not post it here, the beta patch is only for the Slitherine CNAW version.

[Lord Zimoa]

For those who have the Slitherine version or knows someone who plays the Slitherine version. Any help is appreciated.[:)]

Tim aka LZ

[balto]

It will probably take a few patches.  For the first patch there is no need to re-think major parts of rules or think of ways to tweek the individual units strengths/weaknesses.  Keep it simple -  just do a few simple things to reduce Russia's initial impact into Austria and change the initial setup of the doomed French navy off the coast of Gibraltar.   See my post for some real simple changes.
 
Try those easy solutions, and then see how that goes.  If that does not work, then start getting fancy with your patches. 
 
Good luck to all.

[pzgndr]

2. The AI still doesn't recruit any Generals

I note this tester's feedback comment on the Slitherine forum. With all due respect, this particular AI issue really needs reconsideration, yes? I mean, the "C" in CNAW and CEAW is for Commanders. If the AI will not use them then that's a pretty significant handicap against human players who do purchase and use them. For balance, one could choose to not use commanders against the unarmed computer opponent, but what's the point of that? Along with other AI improvements, this rather important feature really should get implemented, in both games.

[firepowerjohan]

Leaders and Artillery units are high risk units that are suited for certain situation. Using them excessively or randomly would be a poor value for money so unless they are part of a good grand strategy they are not worth it to use. If they are used properly they are effective, otherwise they cost a lot more than they are worth.

They would need to be in the correct spot every turn to be worthwhile using because with a game having light cavalry moving 8 hexes and heavy cavalry moving 6 hexes you need only so many units to cover every angle. Even in (my) human vs human games, it is difficulty not making mistakes leaving a exposed leader or artillery unit in the wrong place sometimes and surely it woudl be a larger problem for an AI.

I could cheat and make leaders or artillery cheaper for AI to build making them worthwhile in an artificial way but so far I have chosen the path of minimised cheating. It is all in the difficulty level instead. France receive (from next patch!) even more manpower and even more manpower on high difficulty level which for these cases makes Artillery and leaders less obviously beneficial.

Then there is the cosmetical side of it also, seing the nice enemy leader face [;)] oh well [:)]

[pzgndr]

Appreciate the insights Johan.  I've wondered about the possibility of AI mods for each side, but you're suggesting that even if more leaders and artillery are added to a scenario there's no guarantee the AI will use them well?  More cheating is not the answer.  Understand AI programming is challenging but I would encourage further development.  Eventually it would be nice to see the AI recruit and use leaders.   


Update:

• AI now use leaders
• AI recruits more leaders on higher difficulty levels
• AI now build some artillery

There you go! Thank you. [8D]