Did they purposely weaken everybody's favorite strategies?

[apotheoser]

Again Bubbayeti, there weren't certain strategieS that were useful, there was one direct path that was clearly better than everything else to a huge degree. Civ is not supposed to be about finding the One Strategy and then applying it every game, also having the AI follow the One Strategy.

Unfortunately, there is always an optimal strategy. It maybe that in some situations which strategy is best changes, but there will always be one course or another that produces superior results.

 

[Pantastic]

Apotheoser, like I keep saying, the problem is that there was one strategy that was the optimal strategy virtually every game. The reason they changed things was that there was one course that pretty much universally produced superior results, there was no "or another". The point is that you're supposed to find an optimal strategy for a given situation, not that you're supposed to find the One Strategy and blindly apply it every game.

 

[Watiggi]

I got used to the chopping strategy and found that it was becoming essential to start that way (unless I was purposefully after a religion start or something).

I am amazed as to how much they nerfed chopping. I'm talking about the amount you get and how quickly it decreases the further it's away from the city. I haven't played a new game yet, so I don't know the effects of the changes at that point.

After I got used to it though, I found it interesting as chopping was no longer the best solution right there and then (industrial era, conquered city with lots of forests. I also had access to lumber mils). I found myself building lumber mils instead of the automatic 'no brainer' of chopping - which I ordinarily would do.

I don't mind the change.

Watiggi

 

[Sir Janus]

Unfortunately, there is always an optimal strategy. It maybe that in some situations which strategy is best changes, but there will always be one course or another that produces superior results.

The fact that there is one optimal strategy for a particular situation is not unfortunate -- it is what game playing is all about. The idea is to try find out what that optimal strategy is for the particular situation and try to implement it.

Did I just state something obvious??? What are the alternatives - try to shoehorn one particular strategy to the current circumstance?

What are you really trying to say?

 

[Watiggi]

The theme for civ4 was about having to make choices. They wanted to remove the repetative strategy making in the game and put people in a position whereby they have to make decisions.

If a strategy starts to evolve to be surperior to the others, then the decision making starts to disapear.

What they are attempting to do (whether successful or not, time will tell) is rebalance the decision making so that it isn't so simple to choose one over the other.

...keeps the game interesting I think.

 

[scooter]

When everyone uses ONE particular strategy almost all of the time, I think thats a very good sign of an overpowered area. There should be variety based on circumstances, not just one right solution.

 

[zeeter]

It's funny - just about everyone seems to agree here except for two people. We're all saying pretty much the same thing.

Anyway, to add to the original question, the programmers undoubtably spent the majority of their original programming time working on graphics and the AI. Those are the two hardest things to program. The basic rule set probably took no time at all.

So with all of the talk about the chopping strategy the programmers looked at what they had. The choices were to change the AI that took months to program and will take more months to playtest after the change, or change the ruleset that will take more time to compile than to change and will require only game balancing testing. It's a pretty easy decision.

 

[apotheoser]

The fact that there is one optimal strategy for a particular situation is not unfortunate -- it is what game playing is all about. The idea is to try find out what that optimal strategy is for the particular situation and try to implement it.

Did I just state something obvious??? What are the alternatives - try to shoehorn one particular strategy to the current circumstance?

What are you really trying to say?

It is unfortunate because it means that you do not make choices as a player. Trying to "find out what the optimal strategy is" is not complicated. Are you saying that the idea is only learning to play? What happens when you've learned the game, are you saying you only do the same thing every time?

There is another alternative - having a game where there are a few varied, different strategies, and you can pursue them without knowing that an equally wise and skilled opponent will have an advantage as soon as you pick that strategy (for example Cultural victory as a strategy is worthless in multiplayer.) The game could be more like chess.

This is the problem. Would you ever actually say "oh gosh, I'm so glad I picked Creative/Expansive over Financial"? No. Financial is still quite obviously far more powerful than either of those traits. Even if you are going for a Cultural victory, Financial is better than Creative; perhaps Creative should be +5% or 10% culture production, with a minimum of 1 or 2. Smilarly, Expansive is almost always a weak trait - it should be something like +1 food on squares that already make 2, perhaps.

 

[zeeter]

This is the problem. Would you ever actually say "oh gosh, I'm so glad I picked Creative/Expansive over Financial"? No. Financial is still quite obviously far more powerful than either of those traits. Even if you are going for a Cultural victory, Financial is better than Creative; perhaps Creative should be +5% or 10% culture production, with a minimum of 1 or 2. Smilarly, Expansive is almost always a weak trait - it should be something like +1 food on squares that already make 2, perhaps.

I believe that everyone's point here is that you can pick any civ to run and be able to win with it using various strategies. Certainly if you're going to pick financial and you find a winning system you have the choice to always pick financial. Most people like to change things up every now and then by playing a different civ.

 

[apotheoser]

In multiplayer if my opponents* pick Financial and I do not, I am going to lose. End of story.

The fact that I can beat up on a bunch of ******ed AI players without picking Financial is meaningless.


*equally skilled opponents, assuming no obstacles that can't be overcome (for example, starting alone on an island with poor resources.)

 

[Garand]

I think much of the problem is that people, by nature, want to win rather than lose a game. However, to me Civ is less about achieving a victory and rather the journey I take throughout the game. I never enter a game with a set strategy, and I never use the strategies others devise. To me, boiling a game down to numbers and science to dominate and win every time defeats the purpose of the game.

Instead, I just play, then as I expand and meet neighbors, I develop my country and begin formulating plans on how to get more resources, or build up my military, or defend my empire, or to catch up in tech, or whatever I think needs to be done. Each game should be unique and interesting, not just the same old techniques employed on a different map. However, some people will always be concerned with winning and will always find "the best way" regardless of how much you try to balance.

And I agree with the earlier post about making the AI more active rather than nerfing the overall ingredients of a strategy. It would be amazing if they could create an adaptive AI that watched how players played games and then adjusted accordingly. If a human player chop rushes, the AI chop rushes back. If the human player builds the Kremlin and buys his way to victory, the AI makes an effort to build the Kremlin first or deprive the human player of it. I guess we're a few years away from that, though.

 

[DrewBledsoe]

I.

And I agree with the earlier post about making the AI more active rather than nerfing the overall ingredients of a strategy. It would be amazing if they could create an adaptive AI that watched how players played games and then adjusted accordingly. If a human player chop rushes, the AI chop rushes back. If the human player builds the Kremlin and buys his way to victory, the AI makes an effort to build the Kremlin first or deprive the human player of it. I guess we're a few years away from that, though.

Agreed completely, but that kind of AI (what we think of as being a simulated intelligence) would need more number crunching power than is yet available.
Humans have the huge advantage of being able to learn, remember and apply decisions to situations they've encountered before, without the need to blindly plough down all the wrong dead end streets, just to get to the right one.

Simpleist example I can think of, think of a number, any number. Now tell me if its odd or even..You did it instantly right? You didn't have to divide it by 2 , and then check if the answer is an integer, or chop the last digit and examine it for 1,3,5,7,9 did you . You just knew. Well even the best Chess playing AI computers are still just huge incredibly fast numer crunchers..Kasparov lost to Deep Blue, but claimed the computer made a move "with human intentions", and accused the Deep Blue team of human master lvl intervention in the backrooms....(And as an ex pretty high lvl player myself I have to agree with him)

Anyways, sorry I don't know how it got onto this ramble, lest to say I dearly hope the time will soon come when AI actually means Artificial Intelligence..instead of just speed, size and speed again of number processing.

Its late , I think I'll go now....

 

[Sir Janus]

Trying to "find out what the optimal strategy is" is not complicated. Are you saying that the idea is only learning to play? What happens when you've learned the game, are you saying you only do the same thing every time?

No - I am saying that you should have to do something very different with different settings. Knowing that you have to choose financial, have to chop, have to cottage spam in every game is NOT making choices. That is why they have to tweak things to try to rebalance the game. Obviously they cannot tweak things too much in a patch as it would take many months of game testing to check that a mod does not totally unbalance things in other ways.

There needs to be many different viable ways to play, otherwise people will get bored very quickly and go and play chess instead. Ideally you should still be being surprised by new things, still be trying out a new strategy, etc when Civ5 is announced.

 

[Watiggi]

It's a bit like playing a Warcraft kind of game where one side is far more powerful than the other. No guessing what side you will play when you want to win or simply get ahead.

The idea of balance is so that there are different ways, different paths to finishing it. If there was only one way, it would loose it's re-playability fast.

 

[Watiggi]

The problem with AI isn't really the processing power. Computers are getting faster all the time. If it is still a problem, then it will change as computers advance.

What the real problem with developing a genuine AI is that no-one knows the 'algorithm' for an automatic learning system that can grow to 'infinite' complexity as new information comes in. It's fine to make a system that can 'understand' complex information, but that system will only know that information. To make an AI that has that infinite depth that intelligence has, as seen in puppies, horses or babies, is currently not possible. We simply don't know how to make such a holistic algorithm (which it has to be) that is capable of taking in and reorganising the ever increasing amounts of information, and then put that infomation in a way that resembles learning - just as the brain does. We just don't know how to deal with that complexity yet.

Simply put, that automatic learning with infinite increasing depth that nature just simply relies on for it to become intelligence still remains unknown. As a result we are left to number crunching.

Once we have found that magic learning algorithm, then we will know what sort of processing power we will need and then be able to develop the appropriate system for it.

The day that algorithm is made, it will be a holistic system whereby it will just take in what it senses and will then use that information to 'wire itself up'. Essentially it will be a universal learning algorithm. 'Universal' and 'learning' must be one and the same, if they are not, then the system will have a hard limit to what it can learn, thus it will only imitate learning. It's 'easy' to make a voice recognition system and to have it 'learn' voices. But that is where it stops. A true learning system wont care about the information it's given and what output results from it, it will just 'learn' it - regardless of what 'it' is. That is, it will just wire itself up properly in such a way that it resembles that same something we sense in puppies or kittens that we don't sense in computers, machines or software right now. Learning is just an algorithm - a process, just like any other. We just don't know it yet nor how to apply that learning 'feel' to a system that continually grows in complexitiy. When we do, you'll see a lot of fear in peoples eyes.

Watiggi