CA's definite answer on the Loadgame "issue"

[Crazed Rabbit]

Do you think CA has hired the ex-Iraqi Information Minister to handle bug reports?

"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!

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

[Ciciocle]

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.

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.

[Mr Frost]

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! ...

Just because there is no threat , doesn't mean the AI does not percieve one {or the possability of one} .

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 .

[hrvojej]

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.

[Red Harvest]

"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!

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...

[oompalumpa]

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!!!

*calms down*

Ok... done with the rant, but still pissed.

[hoom]

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.

maybe those guys [us who don't find this an acceptable answer] should be doing something more useful...

like playing a different game

[SpencerH]

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.

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.

[Puzz3D]

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.

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.

[Swordsman]

Look into my eyes You are getting sleeepy. Verrrry sleeeepy. :

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!"

[Intrepid Sidekick]

Dear Hoom

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

I have to take you up on this. If you are going to take quotes from Creative Assembly at least use them in context!
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:

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..."

[player1]

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.

[Servius]

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.

[Kull]

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?

[Puzz3D]

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.

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.

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.

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.

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.

[Red Harvest]

It would be more accurate to say the complexity of the game got beyond the ability of CA's development team to handle.

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.

[JimBob]

What's actually happening is more subtle than the doomsayers would have you believe.

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.

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

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.

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..."

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.

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

[Kull]

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:

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.

[Magraev]

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).

[Catiline]

There is already a thread where this is being discussed at length. There's nothing to be gained from another. Thread closed.