0
Declined

A.I. does not attack.

Галиев Рустем 4 anos atrás atualizado 4 anos atrás 10

Windows 8.1 64 bit, 8 Gm RAM, Intel(R) Core(TM) I7-4500 CPU1.80 GHz 2.40 GHz GPU: AMD Radeon R7 M265, Intel(R) HD Graphics Family. Version game fresh, 1.3.2f3


The official map of Border War A.I. almost never attack with the help of the flag, but the workers swept breaks. There in the middle of the river, which is fenced in the holy land.



If the map is more complicated, namely, between the rivals the empire with portals (without core or inhibitor), and the area is impossible to capture the (fenced in the holy land, the walls of rooms impassable stone) - the A.I. even digging a tunnel toward the Empire, and accordingly, never attack an opponent.



This is an obvious bug that certainly does not depend on my options and features of my laptop so programmed algorithms A.I. A bug, because on any map, which generally can be defeated, the bot must try to win.



Watched 3 times passivity A.I. (level - the master, the characters are different), on my map

"Kind word and a gun".

Here is a video - 70 minutes have passed, A.I. never had digged in the territory of the Empire.

War for the Overworld Launcher Log.txt

output_log.txt


Game Version:
Steam Public
Platform:
Windows
RUIM, NÃO ESTOU SATISFEITO

Of course, I do not agree. Bot not attack - a game against him - is meaningless (I did not do a map - a sandbox).

Satisfaction mark by Галиев Рустем 4 anos atrás
Declined

It's a known fact that the AI is a outdated (doesn't use all the tech available in the game) and the behaviour you're looking for/expecting here isn't programmed into the AI yet. With that in mind, there isn't actually a bug here as the AI is working as (currently) intended. An overhaul/update for the AI has been planned as a future major milestone patch (1.5) so i'll have to decline this ticket for now.


On an off topic note; If you put as much effort into making "proper" maps as you did into making gimmick maps, you'd probably be one of the people getting their maps showcased, you clearly have talent at map design so it's a shame you just seem to insist on using it in odd/weird ways that don't work out the way you want.

A.I. should try to win whenever possible (when the enemy core is not closed completely impassable stone - it is possible).


The bug can be different at different levels of detail specification. I am sure that the desire A.I. trying to victory - it is one of the basic principles of the game, to take into account when designing. So I completely disagree with you - there is an obvious flaw of the algorithm. When it is correct - is another question, but an issue - there. And do not blame me for what I do "not right" map. I do them for the exciting game, not to read the absurd accusations against anyone in the "dishonest creating maps", where the measure of dishonesty correlates with a measure of fantasy criticism.

P.S. I want to sign a pact with you, Andrew "Nutter" Jaggar, the mutual disregard. I do not write in your topics, you - in my? Is that possible? I'd rather wait for the day - the other, and wait for an answer from someone less biased.

Now it is not the recognition of a bug, the theme will be forgotten. And when upgrade 1.5 - the bug again.

Ok #1: There was clearly a translation miscommunication, because my offtopic comment was a compliment and you've misunderstood it to be an insult. I was saying that if you designed normal maps rather than gimmick maps they would probably be very good normal maps because you have a lot of talent. It was a compliment.

#2: I checked personally with our AI coder that this was a non-bug before declining the ticket. The AI is designed to work a specific way, that is it's intended behaviour. The fact it doesn't do something it was never programmed to do is not a bug. I am in no way biased in declining this ticket just because you made it (or whatever other reason you think i'm biased)

#3: It's my job to reply and deal with tickets so I'm not signing anything, but it's no accident that i'm answering a lot of your tickets personally after you made such an impression upon me with your map designing skills on Steam :)

I did not understand, whether the measures taken to ensure that in patch 1.5 A.I. on a map - attacked?
Or not, as my subject is rejected and will not be accepted if calculate future work with A.I.?

The AI is currently working as intended, that means this behaviour is not a bug. The AI will be improved in 1.5 and things like this will (probably) be added then. I have declined this ticket because the AI is following it's (currently limited) programming without fault, so there is no bug.

That doesn't mean the AI doesn't need updating, obviously it does, but this isn't a bug.

Just great. Intended no attack at A.I, that is, you're in a 1x1, but in fact play in the sandbox - and this is the norm. And nothing is done to remedy the situation will not be because some people simply reject the obvious mistake than to bring the bug to programmers.

The AI isn't programmed to deal with Sacred Earth in the way it was used in the Borderlands Map design, it cannot use the Outpost construct to cross the sacred earth/water, so it cannot reach the player to attack them. But that's it has never been "taught" to use that construct (as that construct didn't exist in it's current form when the AI was coded).


I've already told you that we recognise the problem, even if it isn't considered an actual bug, and that we're planning an overhaul of the AI (currently aimed to be patch 1.5) which should solve the issue. I'm declining the issue because it isn't a bug, not because there isn't a problem.


No amount of your complaining will change that.

+1

Hello Ontos.


Nutter asked me to chip in (as the QA supervisor)

I have read the this thread and have to agree with Nutter on this issue.

What you describe will already been improved in 1.5 there is no merit in raising a seperate ticket. That does not mean your problem will not be addressed as this is clearly one of our goals for the upcoming patches.


thank you for your report but nutters decision is correct from my point of view.

please be a little patient and the issue will be solved.

If the problem is known and will be solved - this is the main thing I wanted to know. Good.

Desenvolvido por UserEcho