Totalwar.com post
I find quite distressing as I guess it means no corrective measures against this feature.
PYJ99
Printable View
Totalwar.com post
I find quite distressing as I guess it means no corrective measures against this feature.
PYJ99
So they don't think it's a glitch, they think it's a part of the way the AI behaves over time. Part of that post is true though, I have had instences where the AI has laid siege to a city only to leave 3 turns later of their own volition. So the loadgame issue is a just the AI retreating to re-evaluate it's strategies, interesting.
hahahah, as if we should have expected anything better. pathetic.
I can now feel my unshakeable faith in CA actually starting to shake :( oh well
The fact is if you load a savegame the AI doesn't make the same decisions when you hit end turn that it would if you had played the turn without saving and reloading. In my opionion, proper design would be to have the game make the same decisions as if you played continuously rather than starting over with it's assessment of the entire strategic situation everytime you load a savegame because it seems to take the AI two turns to get all the decisions back in place.
The Shogun's post also seems to say they intend to squash anymore discussion on this issue.
CA say this is the AI breaks sieges on loading a game because on load it does a periodic assessment of its position, and decides its forces are better used elsewhere. This implies that the periodic assessment is a more thorough/better (and lengthy?) analysis of the AI's position, compared to the normal end turn.
Dedicated testing by others has shown than if you reload every turn, the AI never expands beyond its initial settlements.
Imagine if the periodic assessment DID occur every end turn - the AI would never expand.
So, we do not have a bug in the save game file, but we now a design flaw (bug?) in the periodic assessment by the AI, which gives worse results than if the AI just did its normal end-turn assessment.
It also sounds dumb that when you reload, the AI assesses that since it already sieged it's got to choose to do something else from a list of possibilities only to reassess later on that it's better to siege the same settlement. If that's the way it's designed, it's not realistic for an army to surround a settlement and then leave without trying to take it over. It should really think about whether finishing the siege might be a higher priority since it's always better to finish what you start.
For them to totally ignore other issues like getting protecture, lack of ships being built, etc is staggering really but not unexpected given the silence. Typical programmers attitude, they really have hidden behind the "it's a feature not a bug" excuse.
Oh and the attitude to calling people who complained about this issue as a "subset of players" was quite disturbing.
Excuse me but the last time I checked, EVERY player must load up a game at some point so won't EVERYONE get affected by this bug....oh sorry wait...feature. ??? :dizzy2:
Good-bye CA and Total War. I've had fun with STW and MTW but this is the end for me.
Cheers
Starkhorn
After a reload, my Julii campaign test clearly shows the AI breaking sieges on Syracuse and Athens when I hit end turn. If I hit end turn again, the AI reinstates the same two sieges. The AI didn't break the sieges because it was responding to other threats because there were no other threats. The sieging armies moved to places where nothing was happening and stood there. If I don't reload in this test and instead play continuously, the AI doesn't break those two sieges, and takes both cities two full turns before it does in the reloaded game.
If this is CA's idea of proper game design, it amounts to trick on consumers because virtually every gamer assumes that continuing a strategy game from a savegame doesn't alter the course of the game.
So it's a feature and not a bug ~;)
CBR
Well all I'll say is 9 times out of 10 the A.I. maintains the siege and assaults without reloading.
Reassements........ if I'm the Gauls and besieging Segesta and the Britons besiege Condate Redundom(sp) what chance does that army really have of breaking the siege ..........none well at least almost none.
Usually if there is no threat to the A.I. that they can reach in one and sometimes 2 turns they never leave.
But the main FACT I see is that NO siege is maintained after a reload.
Watch what you ask for, and some of yas got it an answer.
As far as I understand CA developers dodged the issue by saying that AI reasseses the situation after the load, and thus somtimes liftes the seiges.
But, since AI never does offensive action in turn after realod (including taking non-walled cities), that can only lead to reassessment algorithm that gets initiated after reload is buggy.
How elase to explain that AI factions will never (100%) expand if you only play 1 turn per session?
And this quote shows that there will be no patches for this issue in RTW.Quote:
However, we will take all of your comments and feedback into consideration with regards to other games in the TW series and any expansions. All of your comments were read (as always). So...
Pretty sad, since this stupid bug drasticly reduces preformance of AI for everyone who doesn't play the game for at least several hours (confirmed so many times).
If it was a feature, why did CA hide it for 6 months? Surely they could have gotten some PR value out of it such as, "Innovative AI which actually changes it's mind on reloads!".Quote:
Originally Posted by CBR
Great, a 'feature' that pretty much ends the game for me unless I leave the game running all the time. Nothing else explains why I didnt fight one defensive seige and why the AI didnt expand in the one campaign I played after switching from 1.0 to 1.2.
EDIT. I started a campaign yesterday as the Brits. I didnt save and it was pretty tough to get started. I finally took a city from the much larger gauls. They counterattacked and beseiged my new city. Unfortunately it was late and without thinking I saved and quit. Upon re-starting the gauls packed up and left. 2-3 turns later the gauls reappeared but I had re-equipped and was prepared for the festivities. With no further re-starts the Germans are now attacking me.
Well, it's been a while since I loaded RTW. From the outset, one of my biggest complaints was that the map never seemed to change. Factions stayed in just about the same places, very different from the dynamism present in MTW. Now I know this was a 'feature'. A pretty awful one, if you ask me.
I'm not one of those people who can sit for marathon gaming sessions. I also like to take command of all major battles (ie where I don't massively outnumber the enemy so it's not even a contest), so the number of turns played on any given day isn't that great. At the end, a save. I'd imagine most people work this way - so, given that, why on earth was such a 'feature' included? I'm quite certain MTW didn't suffer from the same problem...
It would be brilliant if CA actually announced they were going to fix this in the x-pack (it should be done anyway, IMO, but in an x-pack's better than nothing). For goodness sake, CA, put the dynamism and fluidity of map control present in MTW in RTW!
just stop playing it.
its not that good any way, i dont Like 98% of the game,
and the 2% i actualy like is the Idea behind it,
I prefer shogun :)
No don't be silly, this is a feature not a bug. THey can only fix bugs, not planned "features". :dizzy2: :dizzy2:Quote:
Originally Posted by Midnight
Cheers
Starkhorn
Oh, I'm pretty sure they will fix, .. er.. change that bug, I mean feature for the expansion pack, for the simple reason that they will be getting paid for it. No money in hot fixes. And apparently no guts over at SEGA to demand better for us.
Well, I liked it 98%, but awarenesss of this bug pretty ruined RtW for me.Quote:
Originally Posted by Shambles
I wonder if we could issue some sort of protest for whole disregard of the problem.
Like all unofficial forums chaning color to black for few days and adding problem in the news sectoion of their sites.
P.S.
Or it would be too risky (like CA staff not posting in such forums in the future)?
Dont get me wrong RTW had some real potential.
but i really dont like the feel of the game,
Or the layout of the campaign map and moovment with in it,
(I really love the way shogun did it)
I play shogun with 120 men in a Unit, So when i see A few units of hostiles aproaching In Shogun, Im afraid.
But in rome TW I dont feel any threat from any one apart from eliphants,
P.s
I used to like watching eliphants throwing My units in the air, but decided that this was a poor substitute for a decent game design.
Honestly though I love games of all kinds,
And Shogun Is 1 of my Very favs,
But where as Rome may be appealing to me, 1st thing i decide to play it again,
I get dissapointed and go to play Shogun.
But im off topic here,
So il Stop =)
ShambleS
:bow:
Who knows, maybe they really screwed up when they were figuring out how different groups people play the game. Maybe they thought only inexperienced newbies save and reload multiple times and have one turn sessions because they don't know how everything works yet. They keep losing battles or money and want to restart. So what they thought is hey let's mess with the AI so these people who've obviously never played a game like this before a break to catch up?
That of course would be wrong since both veterans and beginners equally have other priorities and attention span lengths.
If what they think is what they say, then they haven't read what we posted on that issue and haven't tested. In this case they are a bunch of liars and loosers.
If they think one thing and say another one, they are a bunch of liars and loosers.
That's it for my purchases of CA products.
Sorry I just had to vent off.
I haven't posted much since I was here but...I've now truly lost all hope of getting back into RTW soon. Like I said in another post if I want to play against static A.I I might as well play Battle of Middle Earth or Kingdoms Under Fire.
Complaining about this isn't going to change anything. We need to fight back by using public opinion to force a change in CA. This can be done by emailing review sites and computer magazines (CGW etc.) and complaining about the poor support we the users are getting from CA.
An expansion to RTW is going to be as poorly supported as RTW is today if things don't change.
:shame: Disgraceful. A game buster problem, being denied as a bug and instead treated as a feature. :no: :no: :no: Do you think CA has hired the ex-Iraqi Information Minister to handle bug reports?
Let's start with forum protest idea I proposed earlier?Quote:
Originally Posted by Turbo
player1 it is far from a bad idea.
But the problem is that at the same time we would have to refrain from actually discussing it as that would invariably lead to hot-headedness. So like CA we would have to consider the issue settled on our part that it is indeed a bug. No more discussion about it. And that should not impact our relation to the devs. Whenever they treat us with a visit we should be cordial and friendly and refrain from mentioning this (we wouldn't need to as it would be on the frontpage).
It would be childish of the devs to consider a silent protest an attack when we a as friendly as ever, just making certain that they understand that we don't agree with the choices made.
I just fear this can't be done as somebody will always voice their discontent.
Right now I actually think it would be better to start a campaign of advising potential buyers not to buy the game or any future products. After all this is a game. If it's not worth getting upset about, then it's not worth the effort to try to pick up attention of the devs in any other way either. The enjoyment I get for the money I spent is obviously not high on their priority list, and hence I don't see why their "feelings" or purses should be high on mine. And the one who programmed this assessment algorythm for the AI, since it's obviously a feature and not a bug, should be fired. Mistakes such as bugs happen, bad features are just bad any way you look at them.
ps.
Out of respect for this forum, I won't actively campaign against the game here, the staff need not worry about that.
"There are no bugs! We have driven the infidel bugs into the sea! All the bugs are dead! That is, that is....a feature! Yes, that is not a bug but a feature!Quote:
Do you think CA has hired the ex-Iraqi Information Minister to handle bug reports?
Indeed, it seems they have.
A sad day. It seems they pulled a Microsoft and advertised lag as a feature.~:(
Though, that really is unfair to Microsoft. They made several patches to AoK for balance and also to The Conquerors.
Crazed Rabbit
what a bunch of crap. they surely think we are fools. just try to play without fog of war (edit preferences file) and you will see that the AI lifts the siege every time, WITHOUT ANY THREAT! sometimes the AI lifts the siege, stays in place, and the next turn it starts the siege again. CA are simply sour loosers.Quote:
However, to describe this as a gamebreaker is a little unfair, and a misunderstanding of the game's AI. What's actually happening is more subtle than the doomsayers would have you believe. The AI does a regular reasssement about the best use for its forces. If one or more armies are engaged in a siege and it decides that there is a higher priority usage for them elsewhere it will lift the siege and use them somewhere else. The player might not understand why this is happening, but often the reasons for the AI acting this way may be covered by the fog of war - a Gallic siege might have to be lifted because the Gauls have suddenly come under attack by the Britons, for example, but the player can't 'see' this happening.
Quote:
Originally Posted by hrvojej
Currently, I sell CA's products for them. I've advised people to buy Rome and Medieval on several occasions when someone walked into my store asking for advice. Alot of these people take my advice. Am I doing them a disservice? I'm starting to rethink this practice now. (Maybe I should just stick to promoting Medieval)
P.S: I'm proud to say that I've managed to keep the same copy of Masters of Orion 3 on the shelf for the length of my employment. May the people who made that game find rich and rewarding careers doing something other than game design.
Just because there is no threat , doesn't mean the AI does not percieve one {or the possability of one} .Quote:
Originally Posted by Ciciocle
Have you ever had a conversation with someone whom took freindly comments as insults ? eg : me:"had a long day huh ?" moron:"you saying I look old , how dare you !"
How about people who get scared because their is a well dressed and polite/freindly Negro behind them .
That's how dumb humans can get and their brains are many orders of magitude more powerful than the computer the games' AI uses to "reason" .
I believe them when they say it is just "thinking" more , but for something as monumentally stupid as AI today , thinking more is a bad thing .
The fix would involve either removing the feature , or merely having the AI appily no more "thought" to the descision to continue seige or have brainfart that it would on an End Turn .
You are mistaken. Features cannot be amended by patches. They are set in stone. Just like the province ownership is if you reload.Quote:
Originally Posted by Mr Frost
Well written, I was looking for just such wording, but couldn't come up with anything. I was feeling creatively tapped by this whole issue...Quote:
Originally Posted by Crazed Rabbit
We need to design an appropriate smilie we can enter into our sigs. That would very quickly show the 'lay of the land' without offending anyones sensibilities.Quote:
Originally Posted by Kraxis
I must agree with the posters in this thread who said the response was extremely disaapointing. It is 100% obvious they have never even tested it at all. THE BUG IS THERE, ACKKNOWLEDGE IT YOU JERKS!!! :furious3: :furious3: :furious3: :furious3:
*calms down*
Ok... done with the rant, but still pissed.
Theres nothing wrong with having the AI review its actions on load & thus have reloaded games come out differently than a continuous game.
The justification given for that is adequate and quite acceptable.
The problem is the explanation of the seige lifting.
Its true that it would be reasonable to lift the odd seige if the AI considered there to be more pressing need for the troops to be doing something else.
The problem in the arguement lies in the provability that the AI (almost definitely) always lifts seiges then reinstates them on the next turn unless something actually changes that really requires the soldiers to do something else.
like playing a different gameQuote:
Originally Posted by CA
This may sound a little rude, but since i heard, tested and confirmed the existance of the primary-secondary attack bug, i pretty much believe everything from CA... That and their "fix" to the protectorate bug.
I mean, what kind of error is that??? And now this attitude!! I have to admit that it is not what i expected from the people who made medieval...
Why? Surely the AI should make the same calculations whether or not it is ending a turn that started from a save game or an ongoing game. The idea of the AI performing some 'xtra calculations' with a save game is nonsense. The most simple answer is that the info that was used for the calcs to trigger the seige/aggressive behaviour/etc has not been saved in the save game and the AI must start accumulating info from scratch.Quote:
Originally Posted by hoom
Although I must confess I removed RTW from my pc quite some time ago, with the intention of coming back to it when the 1.1 patch came out. This was of course asuming that the patch solved all the horrible problems which in my opinion were making the game far from enjoyable. But what CA have done now beggers belief. What however i fear they have done now is paint themselves into a corner, remincent of my mates four year old arguing with his dad the other day ~:handball: who had knocked something over playing in the house. In that they, after claiming that which is obviously (from reading the posts) a major bug is a "feature", cannot back track and say sorry folks we lied without further eroding the crumbling fan base and consumer backing of us the dedicated TW fans. I'm not sure of the solution to this but what really worries me is that if the TW brand is so blackened by this SEGA will choose not to comission any more in the total war series :(
Well, I'm usually not a big defender of CA, but I gotta be honest here, their defense is a good and logical one. That is, they have made the case that it's not a bug, but like ShellShock said, the problem is no longer a bug, but a bad decision-making process.
If it's true that if a player saves/loads each turn, the AI never expands, then there is something wrong with the deeper analysis the AI performs after loads. It's important to differentiate that, while this behavior by the AI may be dumb, it's not necessarily a bug. It may just be an overly-defensive tendency built into those more indepth reviews. If that's the case, it's likely the devs would not have seen this pattern during development for the bell curve reason CA mentioned. It may very well have taken the sheer number of games we've played to expose this tendency.
This may be another example of a negative consequence of too much complexity and/or too many variables. The more variables in an equation, the more difficult it is to ensure a favorable and outcome. RTW has a TON of variables the AI is supposed to take account of. This may just be an example of the cart getting ahead of the horse, of the complexity of an equation getting ahead of the AI's ability to make good decisions based on it.
There are two seriously scary aspects to the CA "it's not a bug" response - and I can't figure out which is worse:
1) They actually BELIEVE there isn't a problem, in which case they definitely won't spend resources to address it in the expansion pack.
2) The problem is so deeply entwined in the basic functioning of the AI (as Servius 1234 has suggested), that a fix is WAY beyond the kinds of changes typically planned for in an expansion pack - and therefore there's no budget to do it.
At this point we are reduced to praying that CA is simply lying to us now in hopes of dampening the furor until such time as the expansion pack - containing the fix - is released.
Why do I have this awful feeling that's not going to be the case? :worried:
There is a problem. There shouldn't be a need for the AI to ever review its decisions. I can't believe that the AI is somehow getting confused as the game goes along, and has to be reset periodically. It makes the best decision based on the data it has at the time. The AI is changing it's decision upon a reload, but the strategic situation hasn't changed. There is no reason for it to change its decision. So, which one is the right decision? The one it makes regarding sieges if you don't reload looks like the correct decision. The one it makes after a reload delays the taking of almost all cities currently under siege by at least 2 turns. That's not good, and has a significant negative impact on the progress of all the AI factions over the course of the campaign.Quote:
Originally Posted by hoom
I don't buy that argument at all. The AI should be doing its best analysis every turn. The turn after a load doesn't take any longer than the same turn without a load, so I think both evaluations are going to the same depth. You don't have to play a lot of games to see the problem, but you do have to turn fog of war off. You can readily see the problem on turn 4 of a Julii campaign, and possibly on turn one as well. The thing is nobody tested the savegames because who would have thought the AI wouldn't be able to properly continue the game from a reload? Nobody designs strategic games that way except apparently CA, and I don't think STW or MTW were designed this way either. According to CA, STW and MTW were made by a different design team which may be why so many nice features of those previous games didn't make it into RTW.Quote:
Originally Posted by Servius1234
I agree with you except for the last sentence. It would be more accurate to say the complexity of the game got beyond the ability of CA's development team to handle. It's obvious that all the variables are not being restored to their prior values upon a reload. If they were, the AI could easily come to the exact same decisions that it does in continuous play. Computers are excellent at doing that task, but humans are not.Quote:
Originally Posted by Servius1234
This game required much more development time than it got, but that probably would have bankrupted company. So, now CA has to scramble to justify the state of the game despite a sincere attempt to correct all the problems with the v1.2 patch effort. I'll bet they are completely strung out trying to get the add-on finished, but, if they continue to underestimate the developement time required to get things done, future releases are going to be plagued with the same plethora of issues. Far too many to be all caught by one patch.
Look into my eyes :uhoh2: You are getting sleeepy. Verrrry sleeeepy. ::dizzy2:
When you wake up, you will be convinced that it makes perfect sense for the AI to lift all sieges (not just one or two randomly). You will bow to the southeast 2.5 times daily in homage to the omnipotence of the all-seeing, all-knowing developers who do not err....
Of course, who among us haven't taken a car to the dealer with an ear-splitting noise coming from somewhere (your choice: engine, brakes, body, transmission) only to have the PhD mechanic say: "Noise, what noise? Oh, THAT. It's normal, they all sound like that... What did you say? Speak up-- I can't hear you over the NOISE!"
Yep. That's it in a nutshell. I reached this conclusion when the pri/sec bug was revealed. There is some really nice stuff in the game, but so much of it is left unfinished that it spoils the effect.Quote:
Originally Posted by Puzz3D
Doomsayer? I'm a doomsayer for demanding my money's worth, well in that case the Apocolypse is coming, duck, cover and stick your head between your knees.Quote:
What's actually happening is more subtle than the doomsayers would have you believe.
Between the time I save and the time I load know how many things change? 0, unless of course the computer gnomes are accessing my save files and launching a Sythian invasion of Italy.Quote:
It so happens that when a saved game is loaded, the AI does a reassessment of its position, decides what is relevant, and then acts
maybe those guys should be doing something more useful like sitting in that patch of trees for a few turns. Gotta let the city defenders rest, fair play and all that.Quote:
It's also a consequence of getting the AI to look after its interests during play. After all, many players review their overall position when reloading a game, send armies to new destinations, decide that a siege isn't worth pursuing, or simply reassign attack priorities in the light of their strategic reassessment. The decision was made that this was a reasonable and sensible thing for the AI to do as well: the loading of a game seemed like a good point to get the AI to sit back and go "hmmm... what's happening?... maybe those guys should be doing something more useful..."
There is one of two problems here:
1) There is a bug
2) The AI is built on a faulty premise which is *tada a bug.
So I say to you CA, you have kicked your core out. We are the ones who buy the games. You think joeblow who bought RTW is gonna buy The Next One: Total War, hell no, he's got his fun.
Anyway email game mags, sites, anyone. As money seems to be the drive
I suspect game mags get plenty of flame mail along the general theme of "mean old company "x" won't fix their lousy game so please tell the world how evil they are." And most of this, based on shrill-level, probably goes straight into the trash. If you really want their attention, what's needed is a short, respectful letter that points to evidence that a well known game company has perpetrated one fraud upon it's customers - and appears to be preparing to do so again:Quote:
Originally Posted by JimBob
The player community at game boards x, y, and z have all identified a game-breaking bug in RTW, and we have proven it's existence using a variety of different tests. Upon being informed of this, CA has gone on record - in writing - that it is not a bug. Since CA plans to release an expansion pack in the very near future, and all indications are that they don't plan to address the problem, we would hope you could investigate this further and warn your readers to avoid the game and the expansion pack. We refer you to links l, m, n, o, & p to back up our claims.
Respectfully yours, Joe So-and-so
This and only this has the potential to hit CA in the wallet and get them to deal with the issue.
Dear Hoom
I have to take you up on this. If you are going to take quotes from Creative Assembly at least use them in context!Quote:
quote from Hoom's posts
maybe those guys [us who don't find this an acceptable answer] should be doing something more useful...
like playing a different game
We did not refer to any human being in this fashion at all and you know it.
Who ever you are, you appear to be deliberately trying to stir up discontent and ill will. The full context of that statement is in reference to how the AI thinks on the campaign map and has nothing to do with how we feel about our customers.
Here is the actual text in context:
Quote:
The decision was made that this was a reasonable and sensible thing for the AI to do as well: the loading of a game seemed like a good point to get the AI to sit back and go "hmmm... what's happening?... maybe those guys should be doing something more useful..."
To say it's not a bug is either monumentally stoopid, an inability to face the truth or a straight lie.
Either way I foresee flagging sales for an expansion if it's not fixed. I'm certainly not buying it without making sure the bug is gone (along with the square coastlines and other smaller issues).
Well, It's not strange why there is some bashing.
While I'm against that, and more tolerant, it is still one of the consequences how CA staff handled save/load issue.
There is already a thread where this is being discussed at length. There's nothing to be gained from another. Thread closed.