differences in strategy for normal, epic and marathon

[Divaythsarmour]

P.S. If it shows 2 processors, does that mean I have 2, or is it a Dual core processor?? I think Physically there is only 1

Did you upgrade the RAM yourself? Even if you didn't it would be easy enough to open it up and have a look. You might have to slide the optical drives out of the way to get a clearer view of the board.

 

[Divaythsarmour]

I was playing exclusively in Marathon. Recently I decided to try normal speed. Everything happens much more quickly at normal. It forces me to take more time with each turn. When you do make a bad decision, you find out much more quickly and it seems to be more devastating.

As far as combat goes, I've never really understood the explanation that the human has the advantage over the AI in Marathon. People say it and it seems to be commonly accepted knowledge. But I've never seen undisputable proof. Maybe that proof is out there somewhere.

 

[TheMeInTeam]

I was playing exclusively in Marathon. Recently I decided to try normal speed. Everything happens much more quickly at normal. It forces me to take more time with each turn. When you do make a bad decision, you find out much more quickly and it seems to be more devastating.

As far as combat goes, I've never really understood the explanation that the human has the advantage over the AI in Marathon. People say it and it seems to be commonly accepted knowledge. But I've never seen undisputable proof. Maybe that proof is out there somewhere.

OK. Here's some indisputable proof then:

1. All cities get the equivalent of military academies, though it works for workers too.

2. The AI unitprob is fixed. It will not take advantage of cheaper units by building them more frequently - it still opts to head for buildings at the precise rate.

3. The human can make up for a significant amount of disparity in terms of unit production due to difficulty - possibly all of it unless at deity - merely by adjusting and building more units.

4. The AI isn't good at planning ahead. Once at war, it trains far less units. This is my #1 gripe with marathon vs fast speeds (especially quick) - even as the amount of turns it takes units to move is reduced on faster speeds, you have to fight MORE units to capture the same number of cities. This is most ridiculous on quick. Here, add 2 longbows to every AI city you attack. Oh, and, they can go 3 techs to rifling in 15 turns or so. Better hurry up and kill those 15 extra longbows! But on marathon it's just the opposite. You fight LESS and move much, much faster while doing it.

None of the above is disputable. The human may choose not to take advantage of his free military academies in all cities, just as the AI doesn't. However, should he choose to do this, the AI is at a severe disadvantage. Also, tech leads last forever. Go ahead and take each city, heal 100%, take another city, heal 100%, and you're STILL going to take cities faster in game years/tech time than a charging stack could on quick or maybe even normal!

 

[Roller123]

Apparently Iraq was supposed to advance to space techs from 1991 to 2003, and having a couple of air carriers built.

 

[Roller123]

None of the above is disputable.

All of it is disputable.

1. Everyone building units faster is not giving human an advantage. It speeds up everyone. "Faster" means building a unit takes less time. For example replenish losses or war preparations.

2. Having AI build less units is not giving human an advantage. It makes the game more difficult for a builder human, since the AI will build more buildings and thus have better economy.

3. Not sure what it is supposed to mean. Are you saying that it is impossible to build a bigger stack than AI's on lets say Epic? And possible on Marathon?

4.As for the biggest gripe, it works both ways. It is more difficult to repel invasions and to defend against backstabbers. Units take forever to build. And no hope of invading army becoming obsolete too. Wars are more interesting on Marathon, mainly becausing they are decided by strength and tactic, not whose stuff obsoletes first. Or making a right peace treaty at the righ time and in 10 turns his army is junk.

It seems Marathon delivers just what it is supposed to deliver. An improved, more difficult builder game experience. Why would a warmonger play a marathon game is beyond me, making military argumentation slightly off the point i think( though certainly valid).

 

[SnowlyWhite]

tmit - your analysis is pretty ok, but you forget a key element: till rene, the cap on units is based on cash, not production. Obviously you can exploit the faster unit production, but you'll end up being broke; on high levels at least. Plus the ai checks you on military and you'll fail miserably far more turns. Even with caesar and already with col, fighting izzy(which was researching dr instead of feudalism), me whipping ch when I was taking a city, still I was broke.

pret. production was fine, research wasn't. Obviously I would've recovered, since I had 3 times the land of the other ai, but till recovering, 2 ais from the other continent declaring took me out. Praets are good, but not against cuirassiers...

leaving aside that in my limited exp. on normal, the defenses were clearly lower(also clearly not 33% lower, probably something around 25% lower). Though ok, when you start attacking, probably it's already the end - the problem is usually till you start your attack.

where I agree marathon is easy - if you have no imagination, play rome, build only praets, play pangeea, probably you can finish them game by building only praets... but normally you'll do that once-twice till you get bored. Probably with WE/cata spam too since even if they get to pikes, they don't spam the them.

 

[TheMeInTeam]

tmit - your analysis is pretty ok, but you forget a key element: till rene, the cap on units is based on cash, not production. Obviously you can exploit the faster unit production, but you'll end up being broke; on high levels at least. Plus the ai checks you on military and you'll fail miserably far more turns. Even with caesar and already with col, fighting izzy(which was researching dr instead of feudalism), me whipping ch when I was taking a city, still I was broke.

pret. production was fine, research wasn't. Obviously I would've recovered, since I had 3 times the land of the other ai, but till recovering, 2 ais from the other continent declaring took me out. Praets are good, but not against cuirassiers...

leaving aside that in my limited exp. on normal, the defenses were clearly lower(also clearly not 33% lower, probably something around 25% lower). Though ok, when you start attacking, probably it's already the end - the problem is usually till you start your attack.

where I agree marathon is easy - if you have no imagination, play rome, build only praets, play pangeea, probably you can finish them game by building only praets... but normally you'll do that once-twice till you get bored. Probably with WE/cata spam too since even if they get to pikes, they don't spam the them.

I was going with the 33% for comparing marathon to quick, which is kind of heresy because quick is awful .

I'd argue the cap on units expires long before the renaissance, and more in the period of currency/CoL, with a few cottage cities grown and markets, rest hammers. Granted, I've never attempted marathon deity, but I have played/beaten every other level marathon multiple times (except the really low ones).

I've actually played standard continents on EPIC and cleared my entire half of the world with just keshiks on immortal - tech extortion, tactics, and basic courthouse/empire management come into play.

My experience with marathon diplo mostly comes from mad's RPC's and HoF, but I don't find it TOO different from normal. You're going to fail war checks above monarch no matter what you do and on every speed. With reasonably high odds for declarations at cautious, odds are you'll get stomped even on normal speeds if you piss off the AIs/ignore diplo (though it's easier to defend). However, the AI is pretty well-manipulated. What I find marathon really forces is taking sides. On large maps the diplo hurts like hell, but on standard maps you can easily get 2-4 AIs to the point that they will never, ever declare on you. Outside of the aforementioned pangaea, this leaves long stretches of the game where you're immune to war (knowing the war, vassal, and AI mechanics and abusing them to hell really helps, but this is probably my single strongest play ability - much better than my micro and probably even better than my warmongering).

1. Everyone building units faster is not giving human an advantage. It speeds up everyone. "Faster" means building a unit takes less time. For example replenish losses or war preparations.

The argument is that the AI will build units at a fixed rate regardless of their flagrantly reduced cost. They ignore it entirely. Unless the human ignores it entirely too, it's an advantage for the human.

2. Having AI build less units is not giving human an advantage. It makes the game more difficult for a builder human, since the AI will build more buildings and thus have better economy.

A builder human IS arguably disadvantaged on marathon. However, no player who wins consistently on high levels is going to be called a "builder" or a "warmonger". Inability or lack of desire to take care of some aspect of civ IV is a hole in one's play. So, if one is bad at military prep/usage, marathon will expose this flaw more so than normal. Make no mistake though, it's a flaw. Show me one top notch player that ignores either his tech rate OR his military completely on a consistent basis.

3. Not sure what it is supposed to mean. Are you saying that it is impossible to build a bigger stack than AI's on lets say Epic? And possible on Marathon?

Of course I'm not saying that . You can build a bigger stack than the AI on quick. What I AM saying is that it's easier on marathon specifically, because even as units become cheaper, the AI continues building them at the same rate, not taking advantage of this bonus. We're basically comparing a 0% human and say 20% AI bonus on other speeds with a 50% human and 70% (or whatever) AI bonus on marathon, but the AI will still train the same # of units. It doesn't adjust, just as it doesn't on water maps or other maps and thus does more poorly.

4.As for the biggest gripe, it works both ways. It is more difficult to repel invasions and to defend against backstabbers. Units take forever to build. And no hope of invading army becoming obsolete too. Wars are more interesting on Marathon, mainly becausing they are decided by strength and tactic, not whose stuff obsoletes first. Or making a right peace treaty at the righ time and in 10 turns his army is junk.

I agree with this. I shouldn't have numbered it. It's just something that *really* irks me about quick, which is the speed I'd play otherwise. But that isn't really relevant to a discussion on marathon haha. Repelling invasions is easier if you are prepared on marathon, but if you AREN'T you're toast.

But all in all the argument is actually a simple one:

The AI was created and balanced based on a standard map and at normal speed. ANY variation to map, speed, and so forth hurts the AI because its programming is still fixed. It won't adjust, even doing ridiculous things like building a large military in always peace or having 3x your power but landing with 1/16th of its power for an intercontinental invasion. Marathon is just another example of this.

Of course, marathon barbs can still rape you, and bad starts with hard diplo or boxed in without resources are probably harder. But on average, the holes in AI adaptability make it weaker on marathon.

 

[monolith94]

I've tried marathon before, but I just can't stand it. It feels like tedium! Often, I find myself reading a book between the vast number of turns and only half-paying attention to the game.

 

[Divaythsarmour]

In order to provide indisputable proof, one would have to present a hypothesis/series of hypothesis and then test it/them.

What TMIT presents is a series of hypothesis only. Obviously it's based on experience. But his hypothesis was disputed. And the more people go back and forth on the various points, the more that it appears that "style" and "preference" come into play.

Maybe a test could be designed using world builder.

 

[TheMeInTeam]

In order to provide indisputable proof, one would have to present a hypothesis/series of hypothesis and then test it/them.

What TMIT presents is a series of hypothesis only. Obviously it's based on experience. But his hypothesis was disputed. And the more people go back and forth on the various points, the more that it appears that "style" and "preference" come into play.

Maybe a test could be designed using world builder.

Well, the fixed ibuildunitprob is a FACT. Whether you agree it's an advantage to the player or not is disputable, but disagreeing logically is pretty difficult...

Obviously marathon is the best speed for hall of fame, because cheaper unit costs and 3x moves increase the speed at which all empires develop across a game. If you're trying to win a space game (or any VC), you want to tech as quickly as possible, and you WANT the AIs to tech quickly also because that will speed up your finish via trades! This is because your opponent isn't truly the AIs anymore, so it makes more sense to have them help as much as possible so a global bonus makes even more sense .

 

[Divaythsarmour]

I suppose that level of expertise has a lot to do with it. For example, consider an expert level player who's familiar with facets of the programming and knows how to exploit or take advantage as such. That expert could say that there's an advantage in a given circumstance. But for people who don't have that knowledge, there's no advantage in the same cicumstance.

And even if a lower level player were to take your word for it and attempt to try a similar approach in taking advantage of a circumstance. There may be degrees of finesse required to succesfully do so that still remain outside of their grasp.

 

[TheMeInTeam]

I suppose that level of expertise has a lot to do with it. For example, consider an expert level player who's familiar with facets of the programming and knows how to exploit or take advantage as such. That expert could say that there's an advantage in a given circumstance. But for people who don't have that knowledge, there's no advantage in the same cicumstance.

And even if a lower level player were to take your word for it and attempt to try a similar approach in taking advantage of a circumstance. There may be degrees of finesse required to succesfully do so that still remain outside of their grasp.

That's true...in fact I play all the speeds somewhat regularly except quick (though I also play that). The adaptation is a bit harsh at first, but doing so definitely gives a clearer picture.

Knowing and TRULY exploiting knowledge of AIs from the XMLs ---> spreadsheets requires both that knowledge AND planning ahead strategically. My micro is always lagging behind my ability in that regard, which is why in some of my older emperor games my tech rate made me look like idiot empire but I'd still win . Of course even as it looks better now, it's still very average on immortal and subpar for deity, but the AI abuse remains constant .

 

[SnowlyWhite]

What TMIT presents is a series of hypothesis only. Obviously it's based on experience. But his hypothesis was disputed. And the more people go back and forth on the various points, the more that it appears that "style" and "preference" come into play.

as far as I'm concerned, it's definitely style only; I mean, late game, I'm researching ~12turns/tech on marathon - that'd be 4 on normal. 4 turns - 1 to airdrop, 3 of fight, and it's obsolete... err, no thanks(not that I war in modern too much since I seldom finish the game lately, but that was the thing that drove me to marathon in the beginning when I got the game and then I wasn't interested in changing).

after not playing civ since civ 1, 1st try was on quick - the game ended in 1h and I thought it was a perfect junk

 

[Roller123]

The argument is that the AI will build units at a fixed rate regardless of their flagrantly reduced cost.

Yes, however thats already point 2 of your argumentation. I cannot see how generally cheaper units by itself are a disadvantage, its a speed setting.

Well, the fixed ibuildunitprob is a FACT. Whether you agree it's an advantage to the player or not is disputable, but disagreeing logically is pretty difficult...

Agreed. Ultimately it can be reduced to the point that Marathon AI prefers economy over war. At this point, it instantly is becoming just another game setting modificator and all arguments from similar discussions are starting to apply. For example AggressiveAI more/less military. Or having a Marathon game full of Montezumas and epic games full of Gandhis, more/less military. Does adding Hatshepsut into the the game make the AI worse or the game easier? It doesnt, it just is playing different. Some people may find it more difficult, some more easy. Thats my gripe with a phrase "AI is disadvantaged on Marathon", which sounds like the AI is playing worse, while in reality it is playing more peacefully.

 

[TheMeInTeam]

Yes, however thats already point 2 of your argumentation. I cannot see how generally cheaper units by itself are a disadvantage, its a speed setting.


Agreed. Ultimately it can be reduced to the point that Marathon AI prefers economy over war. At this point, it instantly is becoming just another game setting modificator and all arguments from similar discussions are starting to apply. For example AggressiveAI more/less military. Or having a Marathon game full of Montezumas and epic games full of Gandhis, more/less military. Does adding Hatshepsut into the the game make the AI worse or the game easier? It doesnt, it just is playing different. Some people may find it more difficult, some more easy. Thats my gripe with a phrase "AI is disadvantaged on Marathon", which sounds like the AI is playing worse, while in reality it is playing more peacefully.

Haha. This is a really good post. I'm inclined to agree.

Well, technically speaking, marathon puts EVERYONE at an advantage, including humans and the AI. Cheaper units is just an advantage. But as you point it it IS just a different setting, akin to others.

I guess it is more technically correct to say that marathon opens up a very visible weakness in the AI to more plain sight, but this is true of most every setting possible.

 

[SnowlyWhite]

no, from seeing, it's easier on marathon to war, at least in start game, and in theory you can exploit it by constantly opening with a rush or things like that. In practice, probably axe rushing will become rather boring and you'll probably play more or less "normal". At least that's how I ended.

what's also definitely easier - you don't really care about ai cultural attempts; you have so many turns to raze one of it's cities it ain't even fun...

what's sucks most is having a neighbor aggresive or backstaber and wasting a gazzilion hammers on it and having to tech machinery in order to have something to defend and counterattack in medieval in order to avoid being totally pants down(you move 10 times better your troops, but your axes will still be pwnt by maces no matter how dumb the ai is) which makes getting something decent out of lib(bio or steel) abit harder.

 

[TheMeInTeam]

no, from seeing, it's easier on marathon to war, at least in start game, and in theory you can exploit it by constantly opening with a rush or things like that. In practice, probably axe rushing will become rather boring and you'll probably play more or less "normal". At least that's how I ended.

what's also definitely easier - you don't really care about ai cultural attempts; you have so many turns to raze one of it's cities it ain't even fun...

Ugh. I'm not looking forward to deity culture attempts. Going culture is suicide for AIs in my games when I'm at parity or better (I notice culture long before it's close), but if I'm behind by enough, it's like there's nothing I can do. Taking a non coastal city with rifles or even infantry vs the AI with tanks/arty/planes when you have to cross an ocean = no thanks.

 

[Shadowkx]

Well, the fixed ibuildunitprob is a FACT. Whether you agree it's an advantage to the player or not is disputable, but disagreeing logically is pretty difficult...

Unless the logic is flawed. While ibuildunitprob does not change with speed, ibuildunitprob is not the only factor when it comes to the AI build units. For example when they run of buildings to build ibuildunitprob is not really a factor... they build units. While some cities might build research/wealth/culture they usually build more units. Marathon is not so much a warmongers speed as it is an early rushers speed. The higher the difficulty and further into the game you get the less likely the AI has building in need of construction, so they are more likely to be building units.

If the AI is pursuing a war strategy then it will build more units as well. Basicly ibuildunitprob comes in when the AI has no pressing need to build something else so it needs to decide between a random building and units.