Re: Quick bits of feedback after about 100 hours of play
 

<font size="2" face="sans-serif, arial, verdana">I guess that if my experience and Kriss's comment about it being possible aren't correct regarding my commander being targeted, it is possible that it was luck.

I have no problem losing commanders to anything other than them being targeted by the computer specifically for the obvious reason that I can't do that either.

I'll keep my eyes open and if it happens often again, I'll keep the files. I still wonder why Kriss did mention that it was possible then I saw it 3x in a few hours of play but ONLY vs. Pythium.


<font size="2" face="sans-serif, arial, verdana">I agree 150% that it sometimes helps when they keep pursuing regardless of the situation. I'm not talking about making the game easier. I'm talking about how absurd it looks for 50 units to be chasing one horse and bypassing 50 other units who will surely kill them eventually if they don't dispatch them.

I don't like that sort of odd behavior as it breaks my suspension of disbelief and that is the most important thing to me in any game.

For those happy with the lack of switching, the game is working as it should for them. I'm putting my vote in for better switching in a few situations to make the men act more logically.

<font size="2" face="sans-serif, arial, verdana">I agree entirely as long as the options are there in the editor. I'd be the first to pitch in to help do it! http://forum.shrapnelgames.com/images/icons/icon7.gif

[ January 29, 2004, 11:53: Message edited by: diamondspider ]

Re: Quick bits of feedback after about 100 hours of play
 

delete me

[ January 29, 2004, 11:26: Message edited by: diamondspider ]

Re: Quick bits of feedback after about 100 hours of play
 

<font size="2" face="sans-serif, arial, verdana">>Don´t use this with Wish. http://forum.shrapnelgames.com/images/icons/tongue.gif

Re: Quick bits of feedback after about 100 hours of play
 

I did come up with an alturnative possiblility for what happened in terms of my commander being targetting now that I'm more familiar with how the AI works.

I think that the Xbows were told to "fire rearmost" and what the AI does is select a random squad in back and concentrates fire on it. Even if several units are in back (and it seems that the commander counts as a squad although it is possible that the commanders have a lower chance of being picked over other "normal" squads).

So, while my commanders were, indeed, singled out in these cases, that doesn't mean that it was purposefully targeting them.

It is probably yet another example of the "sticking" problem in this game in that once the AI picks a target it keeps shooting at or attacking it until is is dead or broken. This usually works fine, but there are cases in about 1 in 5 battles where something obviously odd happens due to this general approach.

That is my current theory in regards to this and my suggestion would be to have it both be a lower probablility of choosing a commander over a squad (again, probably already works that way) then to have a chance of switching off of it since with 100s of men in the way, there is no way they'd be able to "lock on" to a commander in all of the fracas.

[ January 30, 2004, 07:00: Message edited by: diamondspider ]

Re: Quick bits of feedback after about 100 hours of play
 

That's what we were saying about positioning http://forum.shrapnelgames.com/images/icons/icon12.gif And for the most part it will probably keep firing on a squad/commander/large creature until it either dies or routs then switches targets (from my experiance).

There is also a % that units will actually be able to fire at the 'rearmost' and not 'mostly rear'. So while putting your commanders as the rearmost may seem like a good idea, it is in practice not. Unless you want to take the chance that a flier or archer to actually target the rear.

Re: Quick bits of feedback after about 100 hours of play
 

<font size="2" face="sans-serif, arial, verdana">Right. So this thread is about suggestions. I suggest that the switching be improved in this game such that a single crossbowman can't select a commander and "stick" on him for many turns in the face of there being another 100 targets in the way.

It looks ridiculous and like cheating no matter the case of being unlucky. Again I don't want the game to be any easier, just to make some sense. I should have kept that one battle, it is comical to see the one guy targeting the most important single commander over and over when a mass of 100s is attacking and obsuring his view http://forum.shrapnelgames.com/images/icons/icon12.gif

Seems that the switching issue (and the random bugs) is the biggest problems with Dom2. That is my feedback at this point at least http://forum.shrapnelgames.com/images/icons/icon7.gif

[ January 30, 2004, 07:23: Message edited by: diamondspider ]

Re: Quick bits of feedback after about 100 hours of play
 

Well I can say personally I didn't know the in's and out's of the AI as apparently you do after just 100 hours of play. I actually see new things from it from time to time and I've been playing the game for much more than a 100 hours. The fact that a unit gets 'stuck' on another particular unit doesn't seem all that 'unrealistic' or 'unbalancing' to me at all since I can't reproduce it at will. Especially if it is a commander. I find other things more of an issue that have little or nothing to do with the computer getting stuck rarely on the same target throughout a combat. Especially if it's already been addressed and acknowledged as something they may or may not be able to fix. But of course it is opinion that guides feedback. And my opinion it's hardly considered cheating, since it doesn't use it all the time, to it's obvious advantage or even to 'win the game', but by a semingly random occurance.

[ January 30, 2004, 08:55: Message edited by: Zen ]

Re: Quick bits of feedback after about 100 hours of play
 

<font size="2" face="sans-serif, arial, verdana">Well all depend of what you mean by forever. In numerous case ( antiquities battle mainly ) troups going into pursuit never come back to the battlefield until long after battle end.

We have account of pursuit for several hours to even days.

In fact, this even was a part of battle strategy and at a point Hannibal (IIRC) specifically deploy his small cavalry in front of the heaviest Mauritanian one in the hope they will rout and be pursued, effectively deprieving his opponent of one of his better weapon, leaving footmen fight between them. His tactic work and the ennemy cavalry doesnt show up until next day.

I think that the current AI could use some cleaning but the problem is not nearly as bad or illogical as it can seem.

BTW I am totally against a specific fire at commander, from a gamebalance and historical point of view. It is simply impossible to distinguish the commander from the unit in the battlefield in real life. All you can do is target the rear area and hope for a lucky shot. I am speaking of unit commander and not HQ wich tend to be more identifiable but usually out of harm way.

The current target rear and target archer commands are enough, it is up to you to protect your valuable unit by carefully setting them up... For the lone XBow targetting your commander I think it could be the AI unit commander equipped with magical weapons at they seems to target more specifically their counterpart... wich I fairly understand due to their "magical" power.

I think we need to remember that Dominions worl is more antic ( ancient romans, greek, and so ) than modern age thus the current way the battle work ( give order to unit before hand and pray, having no power to alter battle after it start ) is absolutely right for thsoes times.

[ January 30, 2004, 09:36: Message edited by: IKerensky ]

Re: Quick bits of feedback after about 100 hours of play
 

<font size="2" face="sans-serif, arial, verdana">Obviously I do, since all my Posts you see are flames.

<font size="2" face="sans-serif, arial, verdana">Well since you claimed the AI was *cheating* (as you put it) I'd assume you are not providing honest feedback on what you know, but your opinion about something you may or may not understand. I don't claim to understand the game in it's entirety, and even in part the Develeopers have said the same themselves. So I find it particularly hard that you suddenly know the secrets to the AI in a 100 hours, unless you are reading the code and can say.

It's hard to come to a discussion claiming "cheating" and "this is what is WRONG" when you in fact don't know if it's wrong, right, bad luck, circumstance, or in fact cheating. And blanket statements like "The AI this" are in fact very misleading.

I'd hope you'll find as you play with the game more and discover more about it, how you think it works, and how you were wrong about any number of things, you'd have more appreciation for that, but not everyone does.

Re: Quick bits of feedback after about 100 hours of play
 

<font size="2" face="sans-serif, arial, verdana">This isn't quite correct. According to the devs, there's a chance that a unit will pick the most rearmost squad. If this chance fails then the next closest squad is checked, then the next, and so on.

<font size="2" face="sans-serif, arial, verdana">It could also be that his commander was one of the only mounted units, or units with a size greater than 23 on the battlefield, which would make him a candidate for both 'target largest', and 'target cavalry'.