Unofficial BTS 3.13 patch
- Thread starter Bhruic
- Start date
TheWilltoAct
I am observe
What is the difference between this patch and the fixes of yours that I have heard are incorporated into the HoF mod?
I can not see the problem with that. Why are people so paranoid about savegame compatability?!
All Bhruic has to do is say this version is not save game compatable. Then people who want to finish a current game can do so before installing the patch.
Why ignore what could be a major improvement to the game, just because players can not install it until they have finished their current game.
That is what I would do.
Simply because it is a stated objective of Bhruic to maintain save-game compatability (well at least to my knowledge). Solver did the same thing.
I agree with you - I wouldn't care less whether save games were compatible or not but in the context in which RJ made the suggestion it appeared to go against the point Bhruic had made almost right before.
Also, I'm assuming keeping the compatibility makes it far more flexible for players with different needs. For example, you can pick up a game from a point where you believe an AI was playing fairly poorly in 3.13 and try out the game from the savegame with the unofficial patch. It probably makes more work for Bhruic but at least then it will appear as more a patch - not a mod.
Minor Annoyance
Deity
Does that mean it totally ignores all tiles that it isn't working when deciding what to improve? I'm wondering if it was currently working a town with an unused workshop nearby and decided it needed more hammers, would it replace the town with a workshop rather than use the existing one? I'm hoping it's not that stupid.
Because it's extremely important? I'm don't want to have to end up dealing with a bunch of angry people when the next official patch comes out, and people find they can't load their old games with it, because of changed save game compatability.
Bh
No, I'm saying that it will only prioritize improving an unimproved tile if that tile is being worked by the city. Otherwise, it gives an unimproved tile the exact same weighting as an improved tile. However, it's important to note that when it's checking for the best improvement for a tile, it does make a comparison to how the tile currently exists (in terms of resource output). So a tile with a farm on it would end up with a lower valuation than a similar tile without a farm (where higher means more likely to be chosen to be improved). But an unworked plains workshop (at the -1
+1 
stage) would be valued the same as an unimproved plains hill (a rough comparison only, because they don't share improvement options).Bh
It doesn't sound like it's very important to most people. Starting a new game, when the next patch comes out, is hardly a life-altering decision. Dealing with angry people seems easy: tell them that your patch is NOT compatible with future official patches of Civ. If save-game compatibility is so important to them, then they can stick with your 1.11 version.
Roland Johansen
Deity
Bhruics proposed approach to tackle the problem would check the food/production/commerce output of the bare terrain each turn and because that doesn't change (sans global warming), it would have the same result each turn and wouldn't have to be saved in the savegame. Based on the output of that bare terrain, a city would be classified to be a normal, commerce, production or Great Person city.
But would that really work? Suppose we take the Great Person city into consideration. I'd assume you only aim to get one of those. What happens if the city changes hands, for example? The AI then no longer has a great person city? What if the great person city happens to be a border city and is at very high risk of being captured (and so most human players would not make it the GP farm)? Maybe part of the calculation you're talking about involves some sort of risk assessment but then that is not static for the length of the game, bringing me back to the original problem.
I guess I'm really just saying I don't know how you'd carry out the calculation. Surely there is great limitation in forcing the city label to depend soley on the terrain around it. Or would you argue it is good enough for an AI?
Another minor point - the appearance of new resources, whether due to a new tech or just a new source, could cause a city designation to flip as well right? I think most players would never change a city's designation based on a new resource if it has been significantly specialised alraedy.
jkp1187
Unindicted Co-Conspirator
Roland Johansen
Deity
It was not my suggestion, but Bhruics. It's clearly not as refined as the human decision making regarding choosing specialised cities. However, a human player also picks the city type mainly based on terrain around the city, so it's a good starting point.
It's not necessarily so that trying to let the AI behave more like a human will result in a better AI. The problem is that when you add layers of code to the AI, then those might conflict with oneanother in unexpected ways. The AI would do some stuff that a human player would never do just because it's an unforeseen result of the rules coded into the AI. It's always hard to predict those cases just simply because they seem so illogical and stupid to us.
In the case of city specialisation, it could very well be that after a lot of work implementing a true city specialisation AI into the present AI, it would appear that this AI is generally weaker than the present AI. Just because, inadvertently also some stupidity was introduced to the AI. Many things could go wrong. The AI could build too many cities of a certain type. Founding new cities could change the class of an old city resulting in a radical change of terrain improvements around that city. It's probably tricky to get it right.
The problem is that there is some pseudo city specialisation worked into the present code which lets the AI change the terrain around a city based on the setting of the emphasize buttons. Because these emphasize buttons are regularly changed, that will result in regularly changing terrain improvements around the city. I don't know why this was implemented this way, but I don't think it's a great idea to let the emphasize buttons determine the terrain improvements around a city. It doesn't result in city specialisation. It results in cities who don't know what they want.
Maybe it's better for the AI to just build generalist cities. It will maybe lose some 10% efficiency compared to a human player who builds smartly specialised cities but at least it is less likely to cause some unforeseen AI stupidity. The AI doesn't build specialised cities now and I don't think it's easy to get it to do so efficiently. I personally think that city specialisation doesn't help a lot anyway beyond the national wonder cities. It would help the AI if the presence of a national wonder would let the AI change the terrain improvements around a city and if the AI would place the national wonder in a city that would really benefit from it. But I guess such changes require a less deep level of programming compared to an AI which would try to specialise cities which don't have national wonders.
You both make good points, although I'm not sure how solid your 10% efficiency loss number is. Removing the emphasis would only be useful, I would think, if it caused a decrease in tile dis-improvements. If the AI continued to rip up improvements and replace them, I don't think it would be worth it. It's a pretty easy change to make, but I'm not sure how well many tests you'd need to determine if it was actually being effective.
Bh
Bh
Roland Johansen
Deity
It's rather hard to test the AI in civ4 to see whether some change is leading to improvement. Every game is so different that you can't compare AI-effectiveness from one game to the other. The only meaningful test would be to watch the AI play a game and to see whether it behaves as you expect it to. So a limited test would include:
1) make the code-changes.
2) take a game in an advanced stadium and see how the AI behaves by removing the fog of war.
But it's hard to get a definite judgement on whether it leads to improvements or not.
The AI isn't specialising cities at the moment, is it? It isn't creating specific terrain improvements around a city and it isn't limiting the buildings in a city to a certain subset which is the most useful for a specialised city. The theoretical loss is only there if it were specialising its cities now.
And yes, I do think that the benefits of specialisation in a cottage economy (as the AI is mostly running) is rather limited beyond the specialisation of national wonder cities. Of course, the specialisation of national wonder cities is clearly very valuable. It can result in cities which are maybe twice as effective as other cities (also because they're usually build on excellent terrain). But I think the specialisation of other non-national wonder cities isn't that valuable in a cottage economy. Many buildings are useful in every city because they give health or happiness benefits which are often needed in every large city.
By the way, are there other major causes that you can see in the code that would lead to terrain re-improvement (especially ripping up advanced cottages)?
I can see a potential problem which could arise when you remove the link between emphasize settings and the terrain improvements. Maybe, the AI uses the emphasize food setting to make a city grow. Maybe this is efficient at the start of a game. Maybe this will be used often by the AI to let its cities grow to optimal size. Maybe the standard settings aren't good enough for growth. I guess you can set them to be good enough, but maybe at present they are not. Maybe...
It's probably only clear what the effect will be when you can really dive deep into the code and can really understand what the effects of all of the emphasize settings are. I just know that they aren't really working very smart now as they will cause the AI to never settle for a best solution. I think that really substantially improving the improvement code is probably a pretty tough job.
Is the AI capable to see how many farms are needed to work the terrain surrounding a city? Will the AI increase the number of farms when needed to grow even without any emphasis settings?
Hi
I'm just asking but other than the "it is annoying because I wouldnt do it that way" reason of seeing the AI workers constantly change tile improvements. What is the problem with it to make it worth considering making big changes or any changes to the AI programing?
I mean yeah it would bug me to see two workers side by side one changing workshop grassland tile to farm and other changing grassland farm to workshop" but what exactly is game breaking about it? Game breaking as in crippling the AI to make it so non competitive that it loses games and you can beat it easy no matter what you do?
I mean it might be an argument if say while both those worker are busy doing reimprovemnts that result in zero net change for that city meanwhile on other side of its empire there is a 13 pop city with no tiles improved at all since the workers are in the firstc ity just doing busy work. But the AI seems to me to be very good at connecting cities and making sure that worked tiles get improvemnts maybe not imporvements we would pick or it doesnt always stick with them but still seems to do good enough that it is competive for players at warlord and noble and gets harder and harder to beat as you go up in difficulty.
Yeah ripping up towns and villages could be bad but again does it really cripple the AI? I mean even on settler AI still gets adantages in teching and upgrades and money so loss of a village or villages isnt as big a deal as it would be for human player.
And maybe all that changing is because AI is trying to be flexible constantly evaluating the game and making changes to fit and result is LOTS of changes. And it seems to be working well for AI. And if it misses optimum tile impovement ocasionaly it still has chance to get once in while and has bonuses to compensate for when it not.
And I think making it more likely to stay in a certain set of improvements runs risk of getting AI "stuck" in bad placemnt such as someone mentioned losing food sources so now city with workshops doesnt have enough food but doesnt changes because it decided that city is a production city so it gets workshops and thats that. Especially if there is going to be an attempt at city specialization.
I dont think even people who like specializing cities specialize EVERY city or NEVER changes a cities specialization as game changes. Which could end up with either AI's cities being stuck in a specialization or worse instead of city changing tile every turn it changes cities entire specialization every turn. Or it doesnt use its specialization and ends up building wonders in cottage spammed city while heavy mined city is building wall street or oxford and the big city with nothing but grassland farms is trying to build a tank.
Or lead to situations where players learn hey if you take AI cites with hills first it will never to decide to make the other cities production since its they already have a specialization and AI hates to change it or something.
It just seems that letting AI stay flexible and fluid seems to be working as far as making games a challenge and not a cakewalk and reworking AI in that kind of detail is something more for Civ5 or something than a patch since I dont really think its "broke" now.
To me it is like the city placment issue. Yeah it is annoying wihen city is one tile from coast or built too close to another city than I would like but there seems to be logic in AI's thinking to it and AI is competitive with it and working like intended so I would rather have AI competive but annoying rather than building cities where I like and improving them how I like and either need TONS of bonuses to stay competitive or end up where you can just walk all over them with no problem.
Just my 2 cents anways
Kaytie
I'm just asking but other than the "it is annoying because I wouldnt do it that way" reason of seeing the AI workers constantly change tile improvements. What is the problem with it to make it worth considering making big changes or any changes to the AI programing?
I mean yeah it would bug me to see two workers side by side one changing workshop grassland tile to farm and other changing grassland farm to workshop" but what exactly is game breaking about it? Game breaking as in crippling the AI to make it so non competitive that it loses games and you can beat it easy no matter what you do?
I mean it might be an argument if say while both those worker are busy doing reimprovemnts that result in zero net change for that city meanwhile on other side of its empire there is a 13 pop city with no tiles improved at all since the workers are in the firstc ity just doing busy work. But the AI seems to me to be very good at connecting cities and making sure that worked tiles get improvemnts maybe not imporvements we would pick or it doesnt always stick with them but still seems to do good enough that it is competive for players at warlord and noble and gets harder and harder to beat as you go up in difficulty.
Yeah ripping up towns and villages could be bad but again does it really cripple the AI? I mean even on settler AI still gets adantages in teching and upgrades and money so loss of a village or villages isnt as big a deal as it would be for human player.
And maybe all that changing is because AI is trying to be flexible constantly evaluating the game and making changes to fit and result is LOTS of changes. And it seems to be working well for AI. And if it misses optimum tile impovement ocasionaly it still has chance to get once in while and has bonuses to compensate for when it not.
And I think making it more likely to stay in a certain set of improvements runs risk of getting AI "stuck" in bad placemnt such as someone mentioned losing food sources so now city with workshops doesnt have enough food but doesnt changes because it decided that city is a production city so it gets workshops and thats that. Especially if there is going to be an attempt at city specialization.
I dont think even people who like specializing cities specialize EVERY city or NEVER changes a cities specialization as game changes. Which could end up with either AI's cities being stuck in a specialization or worse instead of city changing tile every turn it changes cities entire specialization every turn. Or it doesnt use its specialization and ends up building wonders in cottage spammed city while heavy mined city is building wall street or oxford and the big city with nothing but grassland farms is trying to build a tank.
Or lead to situations where players learn hey if you take AI cites with hills first it will never to decide to make the other cities production since its they already have a specialization and AI hates to change it or something.
It just seems that letting AI stay flexible and fluid seems to be working as far as making games a challenge and not a cakewalk and reworking AI in that kind of detail is something more for Civ5 or something than a patch since I dont really think its "broke" now.
To me it is like the city placment issue. Yeah it is annoying wihen city is one tile from coast or built too close to another city than I would like but there seems to be logic in AI's thinking to it and AI is competitive with it and working like intended so I would rather have AI competive but annoying rather than building cities where I like and improving them how I like and either need TONS of bonuses to stay competitive or end up where you can just walk all over them with no problem.
Just my 2 cents anways
Kaytie
Roland Johansen
Deity
The main problem is with re-improving cottages as you already mentioned. If there were no terrain improvements that would 'grow' over time, then there would not be a problem. Yes, the AI would have some workers do useless work and thus those workers would not have been needed. So some hammers are lost on workers and some gold on upkeep, but that's not a huge problem. The real problem is related to cottages.
A grassland cottage gives a net income of 1 commerce, a grassland town (under universal suffrage and free speech) gives 7 commerce and 1 hammer. The difference is staggering. If 20% of the AI's cottage tiles are subject to constant re-improvement, then the AI is effectively losing close to 20% of its commerce income, or at least a significant portion of its income.
You mention that this could be overcome by the difficulty bonuses given to the AI, but in general people like to play against an AI that is subject to similar rules than themselves and very high handicap bonuses for the AI are typically not appreciated very well by players. It starts to look as if the AI were playing a different game than the human player. It's much more elegant if the AI is capable of competing with only moderate handicap bonuses.
Next to that, it's not very attractive to see your opponent do some truly stupid things (like changing a village to a workshop and a workshop to a cottage in the same city at the same time). Especially not if that opponent is still competitive thanks to difficulty level bonuses.
I do agree with your point that changes in the AI code could lead to bigger problems than what they solve. It's not easy to improve the AI and a logical sounding improvement might lead to bigger stupidities by the AI. You really have to understand how the AI is 'thinking' before you decide to change it. It's not an easy job.
Stop the AI re-improving of terrain sounds easy, but you really need to find out why the AI is doing it and if it serves some purpose in other cases. I also talked about that in my previous post, so we agree on that. I just disagree that constant re-improvement of terrain isn't limiting the AI severly. I think it really hurts the AI and if it could be repaired in a smart way, then the AI could get a major boost in the modern age when they most need it.
How does the AI decide where to place national wonders right now? If I gift Literature to an AI, will it build the two Epics in the first two cities that finish their current builds, however badly suited they are, or will it make two cities (currently) high in production and GPP generation build them, or does it go by some other method?
Does the AI ever intentionally combine national wonders, will it purposefully combine Heroic Epic with West Point or Globe Theatre, or even with Ironworks or Red Cross?
Wodan
Deity
Yes.
The AI is already at a huge disadvantage in the modern era because the AI simply can't build city infrastructure and to specialize cities like a human can. That means that the relative increase in research speed dramatically favors the human.
Crank in the AI having has less commerce because he's losing Towns here and there, and this means the AI is losing even more ground in regard to research speed.
Wodan
Perhaps one (quick and dirty) solution would be to bar the AI from ripping up a cottage unless there is a resource underneath it?
