Dumb AI ? Or Brilliant Posters

[Johnus]

I just tried the 1815 scenario as France on the hardest level and got absolutely nowhere. I have been playing wargames since not long after they were first commercially published by Avalon Hill.

I am not surprised, however, that someone playing a scenario as the defensive side (difficulty level unspecified) might have little trouble beating the AI. I don't, however, see WITP characterized as a bad wargame on this basis.

I have no stake in Slitherine, Firepower, The Lordz or Matrix. But I must say that I have never seen a wargame take this type of unfair and unwarranted abuse. This treatment began even before release (see the thread over at the wargamer) and has continued here.

Those that are too sophisticated for this "beer and pretzels" wargame should not buy it. But please layoff with the attacks; the glare of your brilliance is blinding me.

[Erik Rutins]

The 1815 scenario is one of the toughest ones. I have no doubt that some folks are able to beat the AI up at the normal level, hopefully the harder levels will provide more of a challenge. In any case, Johan is taking another look to see what can be improved.

We don't mind criticism, as long as it's constructive.

[Smirfy]

No, harm to your opinion but I got this game for a "beer and pretzels" experience of the Napoleonic wars. The 1815 scenario on the hardest level is difficult but I still managed a stalemate with ease though I did not really pay £45 for a 4 turn scenario with most of the map blacked out and options pointless.[;)] Whilst CEaW is fun and would recomend it, this game while well presented lacks the immersion of its predecessor. I could live with no proper campaign in Africa and DDAY being uncontested because a couple of houserules (summer only invaision and not before 44) kept the game interesting because Ostfront was fun. The entire docile Grand Armee swaning round the cafe's of Paris has got me vexed as to which house rule to employ to make my purchase fun!

[killroyishere]

The porblem is they don't make AI's for the smart/intelligent they always make them for the dumb and untrained. Any experienced wargamer of years is going to beat the majority of AI's out there even on their hardest difficulty the first time. I am glad to see beer & pretzels wargames appearing though because I am tired of playing week long games vs AI's that come nowhere close to beating me or even challenging me. Some beer & pretzels games like Warlords III and Warlords IV (after the 1.5 patch) are still pretty challenging and hard to beat sometimes. (has a lot to do with placement). That's what we need more of though imho is more been & pretzels wargames. The Tin Soldiers series is another that is quite enjoyable with a challenging ai.

[Adam Parker]

No it really is a case of PC processors not being able to do the math well. There are plenty of AI's that are hard to beat. Because they cheat.
 
I'll add another good AI to my list of one's I respect in my gaming career:
 
1. SSG's Battles In Normandy (very historical outcomes but only if played against a defending German AI - as the Allied attacker it was hopeless.
 
2. Avalon Hill's Flight Commander 2 (highly successful because purely based on aspect and angle geometry - pure math).
 
3. Koios Works' "Athena" found in its Tin Soldiers series (tenacious all round - but taken outside this series, poorly performing).

[Erik Rutins]

Adam,

There are quite a few good non-cheating AIs. With that said, no non-cheating AI that can be made on the budget of a commercial wargame will ever be able to beat a good wargamer. It can give him an interesting game though and perhaps beat him the first time or two and it may beat less skilled wargamers regularly, but in the long run the multi-player options are there for the real challenge.

Regards,

- Erik

[IainMcNeil]

Performance is really not an issue - the way AI works for a game like this does not work like chess where tehre are limited options and the AI can test all of them. Programing and testing time are the real issues. Small companies dont have the resources to do the testing thats required for these huge games.

[Cunctator]

First of all I hope that you'll forgive my poor english, but I'm trying to clarify my point of view.
Answering to Mr. Johnnie, I cannot understand the reason why he is trying to defend someone who was not attacked at all.
As a beta tester of this game I received a copy of it one week only before the release, so I was not able to test anything.
Considering myself an after-market beta tester I underlined some defects of the AI in my limited (two campaing games only so far) experience of this game.
I reported my opinions trying to be useful for a future development of the game.
I did not intend to prove to be a military genius or a brilliant player (unforunately I'm not so gifted).
I only noticed that AI does not use overhelming forces to put obstacles to my conquests, having played at two different level of difficulty (normal and moderate).
I only noticed that tons of coalition troops remain stationary around their capital city.
I only asked for a more agressive AI, just for having more fun, playing this game.
Mr. Johnnie I strongly believe that my contribution may be useful while yours is only useless sarcasm.
Greetings
C.

[IainMcNeil]

The game has been beta tested by 150 people over the course of many months. The last batch were added to test the copy protection system in the last week.

[Phatguy]

I'm going to chime in with my 1.3 cents. Getting AI to be very challenging is nearly impossible. Unless you give the AI cheating powers.People complain about AI all the time but those that complain the most don't do anything about it but whine more and more. Please. 95% of  games out there are going to have poor AI. I know it , you should know it. It has been like that for years. Its  not going to change for your favorite game just because you wish it. If you have a problem go make your own ***** game! Sell it and see what people say about YOUR AI.

Sorry to vent but this whole "dumb AI" things is getting on my nerves.. Its like beating a dead horse..old

Oh, Erik R. You sent me a pm. I tried replying but your folder is full. Tried sending you a e-mail but have no clue if you received it

[Arsan]

I disagree. [:-]
Its not impossible. Its not easy, that's for sure. But it's possible to do it. Just look at Stardock games, Panther games, SSG games...
There are some games with good AI's like those, and many others with not so good but reasonable AI's.
The "get over it or make your f**** game and stop complaining" is not a good answer for the PC wargame industry future, as a great majority of players are AI players.
If we AI players really "get over it", it will mean and big drop of sales.

[IainMcNeil]

AI is hard to get right but that doesn't mean we wont do our best to improve it. As mentioned elsewhere we need feedback on where its going wrong as we knew the AI was not perfect (it never is) but we did not have such big issues in Beta.

[Phatguy]

ORIGINAL: Arsan

I disagree. [:-]
Its not impossible. Its not easy, that's for sure. But it's possible to do it. Just look at Stardock games, Panther games, SSG games...
There are some games with good AI's like those, and many others with not so good but reasonable AI's.
The "get over it or make your f**** game and stop complaining" is not a good answer for the PC wargame industry future, as a great majority of players are AI players.
If we AI players really "get over it", it will mean and big drop of sales.

I said nearly impossible not impossible. a tactical or even operational game will have better AI imho because code can be written for a specific time frame/battle/operation. Writing code for a strategic game is much tougher. Not impossible but tougher because it has to deal with more choices. Of course the expertise of the coders is important too and the size of the team. Having a dedicated group of AI coders makes a better game, but most wargames dont have EA sized budgets. I liken most wargame companies to "Ma and Pa and extended clan" type of operations. You will get a reasonable AI out of most wargames out there but it wont be what everyone is clamoring for. Maybe i was a bit harsh in my criticism but that seems to be the recurring theme in every single wargame release.Lets face it, we wargamers are not the run of the mill Wal-Mart crowd. We expect our games to be good but inevitably, the ai always rears it's mediocre head. My opinion is if the AI bothers you, play against someone if possible, or set the game to hard, or make your own. Screaming, ranting and raving will not improve the AI.It might clean up some of it a bit. I dont think the programmers plan on releasing craptastic AI in a niche product where sales are measured in hundreds to thousands. They do the best they can in the limited circumstances they have.
Well enough of MY rant and rave... I am now going to pop a couple of pain pills for my heart and spleen and play LOTRO

[Espejo]

I hope you get the aI up to the rask
 
Its possible to do non cheating good SIs
 
- Stardock: GlaCiv2
- Even sC2 WaW + PDE have good AIs right now.
 
I would like that normal should be a balanced game like a game between human vs human.
 
Personally I hate it if I have to give the computer 10x  the units to be slightly challenging.
 
This game is directed at wargamers so I expect at least a certain competence of the AI
 
I dont see it as fun If I beat in the first game the AI on normal and afterwards on hard.
 
The artwork etc is beautiful but to say hey the AI isnt so challenging because we wanted not to frustrate beginners is quite of weak
 

[firepowerjohan]

ORIGINAL: Sombra

I hope you get the aI up to the rask

Its possible to do non cheating good SIs

- Stardock: GlaCiv2
- Even sC2 WaW + PDE have good AIs right now.

I would like that normal should be a balanced game like a game between human vs human.

Personally I hate it if I have to give the computer 10x  the units to be slightly challenging.

This game is directed at wargamers so I expect at least a certain competence of the AI

I dont see it as fun If I beat in the first game the AI on normal and afterwards on hard.

The artwork etc is beautiful but to say hey the AI isnt so challenging because we wanted not to frustrate beginners is quite of weak

The AI is being worked on, will certainly be fixed and some problems occured due to other improvements made that is why they passed by undetected.

We do not make the default difficulty harder by adding any hidden bonuses or advantages or allowing the AI to cheat by using its own rules. So normal difficulty in CNAW that means on equal terms as like if playing a human.

[tran505]

Actually I take the other side of the argument. We are all studly mental ubermench whose finely tuned intellect (honed by years of wargaming) far surpasses that of the general population. What chance does even a brilliang AI have against our awesomeness?

- P