Apolyton University Mod (Thread II)

[nbarclay]

Originally posted by Nor Me
The potential problem of altering Rome's upgrade path is whether Longbowmen are built instead of Legionaries in significant numbers.
The AI might then build few Legionaries for offence while building them instead of Pikemen for defence.

With the upgrade path broken, Medieval Infantry would still be available for offensive use. They just wouldn't come into existence through upgrades.

[Nor Me]

Originally posted by alexman
We could increase the cost of Map Making (which is very high on the AI priorities anyway) to keep the total beakers of the Ancient Age constant.

This could have a bigger effect than the other change because Map Making can be critical if you are alone on an island.

Overall, I'm not sure if making the AI research a less useful tech so the human can't sell it is sensible. If the change can be shown to make the AI research more diverse techs then it would be a good idea but it might mean they just all research the same techs in a different order. I've been beaten to Mathematics before 3000BC on Emperor before now so it doesn't always pay off.

[Nor Me]

Originally posted by nbarclay
With the upgrade path broken, Medieval Infantry would still be available for offensive use. They just wouldn't come into existence through upgrades.

How do you break the upgrade path just for Rome while still allowing them to build Med Inf?

Other civs need:
Warrior->Swordsman->Med Inf

Rome needs:
Warrior-> Legionary

If Legionary->Med Inf then Rome can only build Legionaries if they can't build Med Inf. If Med Inf-> Legionary then Rome cannot build Med Inf.

So the only way to do it is to remove the warrior upgrade. That doesn't sound like a good idea.

[nbarclay]

What you would have is Warrior->Legionary->Guerilla and Medieval Infantry->Guerilla. So warrior upgrades would (for better or for worse) stop at legionaries until guerillas become available. (Same thing for the Celts with Gallic Swordsmen if the Gallic Swordsman to Medieval Infantry upgrade path is broken.)

On farther thought, Roman AIs are probably more potent (especially against other AIs) if they can upgrade to medieval infantry. But the question of which unit is better is less than entirely clear-cut.

[Nor Me]

Originally posted by nbarclay
What you would have is Warrior->Legionary->Guerilla and Medieval Infantry->Guerilla. So warrior upgrades would (for better or for worse) stop at legionaries until guerillas become available. (Same thing for the Celts with Gallic Swordsmen if the Gallic Swordsman to Medieval Infantry upgrade path is broken.)

But then no civ can upgrade Warrior->Med Inf! I've been talking about removing the option to build Med Inf for the Celts and possibly Rome.

[nbarclay]

If you remove the ability to build medieval infantry at all, that also removes the Warrior->Med Inf upgrade ability. I think the case for removing the upgrade path is better than the case for removing the ability to build medieval infantry at all, especially in the case of Gallic Swordsmen where an "upgrade" actually involves a loss of shield value.

[nbarclay]

I'd like to propose that starting now (preferably with AU 207), changes that are untested and do not have a clear consensus behind them should be confined to separate, "experimental" branches of the AU Mod rather than incorporated into the mainstream mod. Such an approach is often used in Free Software development, allowing people who want to experiment with new, untested versions to do so while ensuring relatively mature and stable releases for people who don't want to be guinea pigs for every new idea that comes along.

The down side to that is that it would mean two AU Mod releases, not one, in cases where experimental features that do not have a clear consensus behind them are involved. But it would resolve a lot of the inherent conflict between Alex's "We can always change it back after the next AU game" attitude and my preference to play with a rules set that I know won't offer nasty surprises. Both the experimentalist constituancy and the conservative constituancy can have versions of the Mod that fit their priorities.

Following this philosophy, there should be a 1.17 version of the Mod without the changes in cost to Mathematics and Map Making and a 1.17e experimental version with those changes. That way, Alex and others who want to experiment with the cost changes can do so without forcing me and others who feel as I do to either come along for the ride or do without the things we like in the AU Mod (or perhaps customize the rules individually for our own copies of the game). AU games would be released for both the regular version and the experimental one.

I'm sorry if this makes me look like a wet blanket, but think where the Linux operating system would (and wouldn't) be if the only releases available were experimental ones. Sooner or later, a product needs to have stable releases that people can count on not to introduce more problems than they solve if it wants to move beyond having only fellow tinkerers as "customers."

Nathan

[nbarclay]

By the way, I view changes to the cost of Mathematics and Map Making as full of potential for unintended (or perhaps intended but not universally desired) consequences. Will the AIs still make Map Making an ultra-high priority? How will efforts to trade for Map Making be affected? If Map Making is still a high priority, will the increased cost make it easier for human players to gain an advantage elsewhere on the tech tree while the AIs are bogged down? If Map Making is ruined as an AI priority, will human players be able to get a fairly reliable advantage researching that on a 40-turn basis? If the change is too effective in luring AIs over to Mathematics, would that give human players a significantly better chance of winning the race to Writing?

Nathan

[alexman]

Nathan, I already know the answers to these questions. Otherwise I would have not proposed the change.

As for removing the Med. Infantry from the upgrade path for the Romans, well, you can't do that without removing it for everyone.

I wish people would argue a bit less about things they don't have as much knowledge about.

[Nor Me]

Originally posted by alexman
Nathan, I already know the answers to these questions. Otherwise I would have not proposed the change.
...
I wish people would argue a bit less about things they don't have as much knowledge about.

Perhaps you could argue your side of the argument using these facts rather than leaving it to other people .

Originally posted by alexman
As for removing the Med. Infantry from the upgrade path for the Romans, well, you can't do that without removing it for everyone.

I've spent a couple of posts trying to explain this above but I'll try again. Upgrade paths are not civ specific and can only converge not diverge. Currently it's:

Warrior-> Swordsman-> Gallic Swordsman -> Legionary-> Immortal-> Medieval Inf-> Guerilla.

You could have:
Swordsman-> Med Inf-> Guerilla
and
Gallic Swordsman-> Legionary-> Immortal-> Guerilla

But what would warriors upgrade to?

[nbarclay]

Originally posted by alexman
Nathan, I already know the answers to these questions. Otherwise I would have not proposed the change.

As for removing the Med. Infantry from the upgrade path for the Romans, well, you can't do that without removing it for everyone.

I wish people would argue a bit less about things they don't have as much knowledge about.

The chain is currently Swordsman->Gallic Swordsman->Legionary->Immortals->Medieval Infantry. What makes pulling Legionaries out of the chain any more difficult than pulling Gallic Swordsmen out? And what would interfere with pulling either or both out and rearranging the chain accordingly? (For that matter, do you have any idea why Firaxis arranged the UUs together in a chain like that in the first place?)

Regarding Mathematics and Map Making (and mod-making in general), part of the scientific method is that you share your methodology, not just your conclusions. When you fail to show how you've done adequate testing and/or theoretical analysis to address possible concerns about side effects, you leave the rest of us no way of knowing what, if anything, you've done in that regard. You might have all sorts of testing and/or theoretical analysis to back up the idea in your own mind, but how are the rest of us to know that? So while I'm sure it's frustrating when people "argue...about things they don't have as much knowledge about," I contend that the proper solution is to provide adequate knowledge in support of proposed changes instead of simply expecting us to accept on blind faith that you haven't overlooked anything important.

[nbarclay]

Okay, so the reason for the Swordsman->Gallic Swordsman->Legionary->Immortals bit is because that's necessary for warriors to be able to upgrade to all of them? (Thereby answering my question above about why Firaxis arranged that weird chain?)

[nbarclay]

I ran a quick debug-mode test (Emperor, huge map, 16 civs) to see how the change to the cost of Map Making would affect efforts to trade for it, and the effects weren't as bad as I feared they would be. Whether Map Making is cost 12 or cost 16, if you're the first civ with Code of Laws, you can get a straight-up trade on a 16-civ huge map once two civs know Map Making. If you're the second civ with Code of Laws that an AI meets, though, it may make a difference. I needed 13 contacts with other civs that knew Map Making before Germany would go for an even swap with the cost set at 16 when Germany knew another civ with Code of Laws. With the cost at 12, Spain went for an even swap when I had only eight other contacts. (I didn't think to lock in the civ I'd be trading with.) So if you're first with something an AI with Map Making wants, the change may not make a huge difference other than affecting your ability to get something else in addition to Map Making from the civ, while if you're not, you could find yourself having to shell out quite a bit extra for Map Making compared with the default rules. (Of course this assumes you research and trade rather than extorting techs.)

Nathan

[Yahweh Sabaoth]

This is a highly theological thread, isn't it?

Well, I thought I'd just jump in here with a comment that may be totally inappropriate for the AU mod, but I think it might well serve it.

Has anyone considered making the nationality of the guerilla unit hidden, as a privateer's is?

I know that this could radically alter the way the game is played, and that is not necessarily the point of the AU mod. However, it would be somewhat more "lifelike," and hence, you all being intelligent people, I thought I'd throw the idea out there and see what sort of feedback you have.

Well, keep up the good work everyone. Your arguments enhance Apolyton as a whole.

[Dominae]

Thanks for the test Nathan.

Yahweh, the idea of giving the Guerilla hidden nationality has been proposed a couple of times already, and was rejected pretty fast. Not only would it change the game significantly, any human player could exploit the heck out of having hidden nationality land units. I personally would cut my Rubber supply and just draft Guerillas until I had enough of them to simply dicate where and when the AI would move (peaceful AIs, that is). That's not the kind of thing we want happening in our games, right?


Dominae

[Yahweh Sabaoth]

Originally posted by Dominae
Yahweh, the idea of giving the Guerilla hidden nationality has been proposed a couple of times already, and was rejected pretty fast. Not only would it change the game significantly, any human player could exploit the heck out of having hidden nationality land units. I personally would cut my Rubber supply and just draft Guerillas until I had enough of them to simply dicate where and when the AI would move (peaceful AIs, that is). That's not the kind of thing we want happening in our games, right?

Well, no. But if one could figure out a way so that the AI used it significantly as well... then...

Perhaps it's not appropriate for the AU mod. But it would be nice to see in some mod, some where. Perhaps in a "real world" mod?

[nbarclay]

Dominae, in what you did watching the AIs in debug mode, did they still go for Map Making quickly in spite of the higher cost, or did the cost push them to go after other of Writing's children first? (You might want to preface your answer with a spoiler alert for people who want to be surprised at how the change affects AU 207, but I view the information as important in weighing the merits of the change.) It seems to me that the ability to research something else while the AIs focus on Map Making is one of the few things that make peaceful research a viable alternative to tech extortion on Emperor level in the ancient era. And I'm a strong opponent of any change that pushes people significantly farther toward the dark side.

[Dominae]

Nathan, I used the PTW 1.14 scenario for debugging, so I cannot tell you anything about the change.


Dominae

[Konquest02]

For guerrilla, is there a way to give it 2 movement only on mountains and jungle? Or maybe treat jungle and mountains as roads, which I'm sure is feasable. It would give them 3 moves on mountains, allowing fast attacks out of nowhere, just as the real guerilla.

Food for tought...

--Kon--

[Yahweh Sabaoth]

Originally posted by Konquest02
For guerrilla, is there a way to give it 2 movement only on mountains and jungle? Or maybe treat jungle and mountains as roads, which I'm sure is feasable. It would give them 3 moves on mountains, allowing fast attacks out of nowhere, just as the real guerilla.

Yes, very good comments.

It would also be awesome to see a "contra" unit (not that I'm a fan of the contras in real life... don't get me wrong...) that would represent government-trained "counterinsurgency" (re: mercenary) forces, with similar stats to the guerilla, but slightly better perhaps.

[badams52]

Kon and Yah,

Sounds like changes just for the sake of changes and not much to do with making the AI more formidable. My understanding of the AU mod is to remove as many human exploits/advantages and to improve AI mistakes by way of modding.

[edit: If you can show how modifying the guerrilla unit in such a way would be advantageous to the AI, then we'd be in business.]

[Yahweh Sabaoth]

Originally posted by badams52
Sounds like changes just for the sake of changes and not much to do with making the AI more formidable. My understanding of the AU mod is to remove as many human exploits/advantages and to improve AI mistakes by way of modding.

You're probably right.

Still, someday, when I have a little confidence in my mod-making ability, I'm going to make a "dirty third world" MOD that will include these changes.

If only C3 and PTW included the concept of client states... sigh...

[badams52]

edit: double post

...Now where did my post go...I know I posted it, but I can't seem to find it. ...Oh here it is on the next page..

[Konquest02]

My proposed change to guerilla (treat jungle and mountain as roads) is to have a better shot at realism. I have not tested it (yet) in any way in Civ. My argument is that guerilla is a form of fighting that takes advantage of the terrain, setting traps to the enemy and then fleeing before anyone could catch them.

Would taht work in civ?

--Kon--

[Nor Me]

Konquest02, recently Dominae suggested giving guerillas the ability to treat mountains etc. as grassland.
The problem is that they have one movement anyway.

Giving guerillas all terrain as roads would make them too good compared to cavalry.

Making them faster in mountains/jungle than grassland seems illogical.

Giving all open terrain a movement cost of 4, giving all fast units ignore movement cost in each open terrain and giving guerillas a movement of 4 and ignore movement cost in all other terrain would mean that they only move 1 square in the open, 2 going through forest/mountain/jungle/hills and 12 on road. In the age of rail, the latter might not be a unbalancing. I haven't thought of any side effects yet.

But I still don't think it could be justified.

[Jaybe]

Guerilla: 6/6/2, ignore cost of hills, mtn, jungle & forest.
I have had much experience with this and it works well. I heartily recommend it for anyone's personal mod.

[nbarclay]

1) Giving guerillas faster movement is too much change for too little reason for the AU Mod. One of the goals of the AU Mod is not to change gameplay significantly from the stock game without good reasons.

2) Guerillas' fast attacks "out of nowhere" are a matter of tactical (i.e. short-range) movement, not strategic (i.e. long-range) movement. Or to put it another way, they're quick, not fast. Guerillas attack, do their damage, and then try to get away before before their targets can get organized enough to come after them. (And they may very well have nasty traps ready for their pursuers.) But tell them to travel a hundred miles and they can't do it particularly faster than conventional infantry. The way I see it, guerillas' tactical quickness is part of their attack and defense ability (and where the rationalization for their zero-range bombard ability in the AU Mod comes from), not something that should be reflected in a faster movement rate.

Someone proposed giving guerillas a ZOC ability as well, and that would be another way of implementing a form of "quick attack out of nowhere" ability. I wouldn't have a problem with that from a perspective of realism.

Nathan

[Konquest02]

I like your proposed changes Nathan, but mostly, I like your arguments. I agree that my change might be a little too drastic for the AU mod and I understand it. In fact, I was unaware of the zero-range bombard for guerillas and I like it. The idea of ZOC is good as well and it shows better the guerilla ability I was trying to show.

Thanks for your explainations!

--Kon--

[Dominae]

As of 1.16, Guerillas have ZOC in the AU mod.


Dominae

[aaglo]

To the true nature of berserker, it should have defence dropped to 1 and hitpoints increased with 1