The Military Academy

[Catt]

Clearly the aggregate advantage or disadvantage grows when additional, even minute, elements are added to an equation. That is straightforward.

There are a number of different ways the MA could arguably “place builders at a building disadvantage” if one first accepts the proposition that pre-building or even certain other attributes fall within a “builder’s sphere” (which I wouldn't, but that's a different subject) – i.e., an MA-owner could build high-shield armies for disbanding elsewhere (does that make it a builder’s toy rather than a warmonger toy?); the MA offers an additional culture point per turn (and so will the Pentagon); etc.

I am not advocating ignoring the additional, ancillary effects of game features or AU mod changes – but I think it’s easy to miss the forest for the trees. From the viewpoint of someone who will play the AU Mod on only limited occassions, the fact that the decision to make the MA available to all at MT could depend on the fact that it offered an additional pre-build opportunity (or offered additional high-shield builds for disbanding, or offered increased cultural production) just comes across as surreal.

At its simplest level, I am no doubt simply disagreeing with the weights you must be ascribing to the MA itself and to the ancillary effects an available MA offers (principally the pre-build). But I’d argue that if such ancillary effects can tip the balance on a close decision about a more substantive feature, then the AU Mod would be better off looking for ways to negate the ancillary effects.

Catt

[Theseus]

I think Catt just said that Nathan's point wasn't important enough to swing a decision.

I think so, mind you, but I could be wrong.

J/K

I agree... I haven't gotten an MGL as of the Modern Age in a Sumerian game I'm playing.

No Armies for me... and no Mil Academy either. Woe and death.

So be it. Them's the breaks. It's called a "challenge." And using the Mil Academy and Armies as pre-builds or Armies as shield transporters is the least of the issue.

Now, again, that is the player side of the issue, and that's all I have to say about that. ("Life is like a box of RNG" )

I still have not seen nor heard of an AI C3C Army, and that IS a critical issue (either for the patch or the AU Mod).

[Catt]

Originally posted by Theseus
I think Catt just said that Nathan's point wasn't important enough to swing a decision.

I think so, mind you, but I could be wrong.

That sums it up nicely .

The only additional intention was advocating making the underlying concept behind that point applicable to more than just the pre-build / MA issue.

Catt

[ducki]

Originally posted by Catt


That seems to me to be a much smaller advantage relating to the availability of the MA than the actual ability to build armies (and would also work only in favor of the human player). How is this pre-build ability any less 'fair' than the ability to build armies offered to the (lucky) warmonger?

Catt

While the AI may not be able to use the army as a pre-build, wouldn't it be an awfully handy way for them to not get screwed over so badly on a cascade? As consolation prizes go, that one isn't too shabby. Or does the AI even have logic to minimize shield-loss in a cascade?

Edit: Just to be clear, I don't think pre-builds are a good enough reason to make the change. I think the other reasons given before(including prior AU versions) stand on their own. On the other hand, if the AI is smart enough to a)build this and b)use these shield-sinks to save a missed wonder, that might be worth considering a bonus, but still not anywhere close to the main argument for the change.

[nbarclay]

Catt, maybe the fact that I'm getting better at manipulating prebuilds is prejudicing me, but I view the prebuild issue as a lot bigger deal than just a more convenient and a little bit more efficient way of transporting shields (which can be done perfectly well with conventional units) or one extra culture point per turn (two factoring in the Pentegon if the player builds three armies to make it available). I think you're trying to trivialize the issue.

There are at least three times after Metallurgy when having two prebuilds going at once can be very useful. The most important is the early modern era, where the techs for SETI and the Internet are back to back and the United Nations is on the same tier of the tech tree as SETI. Consider the implications of that in a close game (and especially one where the player is just a little behind) where early modern wonder cascades are a real threat. Losing SETI is unlikely to be a game-breaker, but losing the Internet could be in a closely contested space race (especially if the player's closest space race rival gets it instead), and losing the United Nations could fairly easily cost a player the game if the AIs don't like him. So on admittedly very rare occasions, the stakes involved in that second prebuild can be surprisingly high.

Another place where multiple prebuilds can be useful is around the ToE/Hoover time frame. Under the standard rules, using one prebuild for ToE and a second to get Hoover sooner is essentially just a matter of a few extra turns of power-plant-enhanced production (for every city on the continent that doesn't already have a coal plant). But if we move Longevity to Sanitation and leave ToE where it is, the ability to have two prebuilds at once in that timeframe would get a lot more important to reduce the risk of losing either ToE or Longevity. Again, that's especially an issue in a close game.

Finally, depending on a player's research path, some optional wonder-enabling medieval techs may not have been discovered as the medieval era winds down and Universal Suffrage approaches. In my current game (the one that sparked my rethinking), I have a palace prebuild left over from not getting as many medieval wonders as I wanted and there are enough shields in it that using it for Universal Suffrage is the only option that would not involve a serious waste of shields. But that means I have nothing available as a prebuild for Shakespeare's Theater (on the assumption that one of the AIs will research the necessary tech in time for a prebuild to be useful). Losing that wonder wouldn't be a huge deal (aside from the fact that it produces four times the culture of the Military Academy and Pentegon combined), but what if it were Smith's Trading Company that hung in the balance?

The idea of players' potentially losing wonder races without the ability to build the Military Academy or armies where they could have won with that ability rubs me very seriously the wrong way. Granted, it probably wouldn't happen all that often, but the times when it could happen are also the times when losing a wonder race would be the biggest deal for the player.

Also note one other thing. A palace prebuild cannot be used in a nation's capital, while a Military Academy prebuild can. (Army prebuilds, of course, depend on where the Military Academy is ultimately located.) The implications of that are especially serious if a player wants to go after a one-city cultural victory in the capital (admittedly rare in epic games, but I finally did it once recently). And, for that matter, it can make a difference on a personal level if a player just wants his capital to be a leading cultural center.

Nathan

[Catt]

Originally posted by ducki


While the AI may not be able to use the army as a pre-build, wouldn't it be an awfully handy way for them to not get screwed over so badly on a cascade? As consolation prizes go, that one isn't too shabby. Or does the AI even have logic to minimize shield-loss in a cascade?

I can't be certain, but my impression is that the AI will start wonder builds when enabling techs become available. A wonder cascade will never dislodge an AI Military Academy build because the MA is a small wonder and therefore not lost if another civ completes it first. Same thing for an army. A wonder cascade does however seem to cause the AI to change the wonder-building city to the new wonder or improvement (i.e., not implicating the MA or an army build unless no other wonders are available and then the MA might be the "switched to" improvement). Whether or not the AI would switch a failed wonder build ("broken cascade") to an army is open to question as far as I am concerned, but it would obviously require, at the very least, that the wonder build had been happening in the MA city. I'd be willing to venture that the incidence of an AI civ being able and willing to switch to an army build (or MA build), with a lots of saved shields, following a broken cascade is very rare, and certainly rare enough to make the possibility less compelling a reason to contemplate a change to the MA than the pre-build ability enabled by the MA.

Catt

[nbarclay]

By the way, just to make it crystal clear, if I thought the victorious army requirement was actually good for the game, I certaionly would not regard the prebuild issue as sufficient reason to justify a change. But as best I can tell, even defenders of the requirement resisted change more because they didn't view the issue as big enough to warrant a rules change than because they saw the requirement itself as clearly beneficial. Under such conditions, it doesn't take anywhere near as much to tip the balance toward the view that a change is justified.

Nathan

[Catt]

Originally posted by nbarclay
I think you're trying to trivialize the issue.

I'm not (consciously) trying to trivialize it at all. If the real value of the MA, or a substantial but subordinate value of the MA, is as a pre-build, I'd argue that something should be done about either (1) pre-building in general; or (2) the timing of the available pre-build and how the timing and availability might be made available to MA-owners and non-owners alike. My own view is that the pre-build value of the MA is very small (present, yes, but very small).

If the stakes are very high related to any given wonder race -- meaning a pre-build would be materially valuable -- then it only highlights the choices made (and luck) earlier in the game. If one views the game as likely to depend on reaching Hoover's, the UN, or the Internet first, then I think one must take action, well before the moment arrives, to increase the odds -- that action might consist of any number of actions. In other words, I am having trouble acknowledging that a 400-shield pre-build is a game-breaker whether it comes in the mid-Medieval Age or the Modern Age (no matter how rarely the "gamebreaker" tag might be labeled on such a pre-build). If my view is not shared among AU Mod players, then it strikes me as more appropriate to institute some artificial pre-build small wonders -- somehting without value other than as a pre-build rather than something that might serve as a pre-build but might also serve to produce armies.

Finally, depending on a player's research path, some optional wonder-enabling medieval techs may not have been discovered as the medieval era winds down and Universal Suffrage approaches. In my current game (the one that sparked my rethinking), I have a palace prebuild left over from not getting as many medieval wonders as I wanted and there are enough shields in it that using it for Universal Suffrage is the only option that would not involve a serious waste of shields. But that means I have nothing available as a prebuild for Shakespeare's Theater (on the assumption that one of the AIs will research the necessary tech in time for a prebuild to be useful). Losing that wonder wouldn't be a huge deal (aside from the fact that it produces four times the culture of the Military Academy and Pentegon combined), but what if it were Smith's Trading Company that hung in the balance?

The above fact pattern would be something to consider when considering wheteher early warmongering, and an MA ability, was worthwhile. Whether it is Smith's, Shakespeare's, Sun Tzu's, Sistine's or any other, should the potential worth of an MA pre-build really drive the discussion of whether the MA, with its inate abilities, should be available to all? IMHO, the solution to the issue above is to enable one (or 2, or 5, or 9) artificial prebuilds available at the same time as the MA.

The idea of players' potentially losing wonder races without the ability to build the Military Academy or armies where they could have won with that ability rubs me very seriously the wrong way.

I'd still fallback on the view that if it is important enough to rub one the wrong way, it is important enough to alter one's play to avoid, or to argue for AU Mod changes which address that subset of the issues involved.

The implications of that are especially serious if a player wants to go after a one-city cultural victory in the capital (admittedly rare in epic games, but I finally did it once recently). And, for that matter, it can make a difference on a personal level if a player just wants his capital to be a leading cultural center.

Again, if one wants it, one must play for it.

All this is IMHO -- I recognize that there will be different views. I also recognize that my arguments might be worth discounting (I'll play stock more than the AU Mod). But I like the dsicussion of game mechanics, I like this discussion of the MA and its role and how its role might serve as a broader game view, and I like Nathan's rigorous and thoughtful arguments on the other side.

Catt

[Theseus]

Is it just me, or is no else really bothered by the lack of AI C3C Armies? Especially given the new bonuses, and the limitation of uses for MGLs, this seems to me a significant issue.

I'm actually starting to bend on the idea of a builder Mil Academy (i.e., Army-winning-less), but regardless, in the meantime, just as we are all fantastically pleased to see a fortified AI defensive unit in a fort on a chokepoint, shouldn't we also be clamoring for AI Armies? Hmmm...

[pauli]

i'm waiting for a working c3c version of c3mt - then i'll be able to see how wars go stock, and compare it to results if i, say, give the ai two or three empty armies.

i've never seen an ai army. i'm quite curious.

[ducki]

Originally posted by Theseus
Is it just me, or is no else really bothered by the lack of AI C3C Armies? Especially given the new bonuses, and the limitation of uses for MGLs, this seems to me a significant issue.

You must have missed me making this point due to my blathering and rambling.
I think it may be heighted because in previous AUs, we implemented this change specifically to get more AI armies.
Now without it, I have yet to see one. Instead of fearing an AI army, it's a novelty. I'd actually even cheer the AI on if I actually saw one. But I haven't and I haven't seen many game reports suggesting it's common, or even just not very rare.

I'm actually starting to bend on the idea of a builder Mil Academy (i.e., Army-winning-less), but regardless, in the meantime, just as we are all fantastically pleased to see a fortified AI defensive unit in a fort on a chokepoint, shouldn't we also be clamoring for AI Armies? Hmmm...

I'm already in a test game on this change, hopefully I'll see some armies. Even so, I realize that a sample of 1 isn't really worth much unless the results are astounding. We'll see.
The AI should definitely not have to jump through hoops to get armies and the human is far better at generating MGLs than the AI. (I also don't think peaceniks should be unable to build armies either, but that's a separate issue.)

[Catt]

Originally posted by ducki

You must have missed me making this point due to my blathering and rambling.
I think it may be heighted because in previous AUs, we implemented this change specifically to get more AI armies.

I play stock rules far more often than the AU Mod and I regularly saw AI armies in PTW. I still think I've seen one in C3C, but can't be certain -- seems like others are pretty convinced that their incidence has either declined notcieably or the AI is actually not using the buggers (my own sample size is pretty small).

Catt

[alexman]

OK, I'm not sure if this proves anything, but I started a debug game with all the AI civs starting with a leader and several Warriors.

What did all do? They rushed a Warrior!

[Tall Stranger]

That's really sad, actually. (OK, it's hilarious as well! )

Is there anything we can do to trick the AI into building armies more frequently? Add features to armies that humans can't/ won't use? I don't know how the flags for armies work, but can things like "Enslaves units" or "lethal sea bombard" be applied to the army unit WITHOUT giving it those real capabilities? We know, for example, that Marine + berzerk armies can't be amphibious (despite Theseus' pleas ). Is is therefore also safe to assume that, even though the army unit may have Enslave, as long as the units inside the army don't, the player won't be able to use this ability?

[vulture]

How many cities did you give the AI? You can't actually build an army if you have just one city. I guess you need 4 (or whatever the number of cities per army is supposed to be...) I know this because I recently went to great length to get a MGL in a OCC, and was then horrified to discover that I couldn't build an army with it.

[vulture]

BTW, in C3C, how did they use an MGL to build a warrior? Or did you give them SGL's - but then how would you expect them to build an army (never mind the number of cities)?

[alexman]

Good point about the number of cities, but I reduced the number of required cities to 1.

Also, I believe MGLs in C3C can still rush things, they just can't rush Wonders.

[ducki]

MGLs can rush anything except Great Wonders - they can still rush Small Wonders, according to my experiences and the docs.
SGLs are the only ones that can rush GWs, but, IIRC, they can also rush anything else.

How about this for a different test, alexman?
Change the Palace to spit out MGLs like the SoZ does ACs, but less frequently. That way you'd have a steady stream of MGLs for the AI to waste.
(I think I saw that you can do this, but I may be mistaken.)

[vulture]

As always, alexman is one step ahead of the game...

You'd think that the AI would be programmed to do nothing with MGLs apart from build armies.

[MrWhereItsAt]

Originally posted by ducki
Change the Palace to spit out MGLs like the SoZ does ACs, but less frequently. That way you'd have a steady stream of MGLs for the AI to waste.
(I think I saw that you can do this, but I may be mistaken.)

Sounds neat, but how would you stop yourself getting the same MGLs from your own Palace?

And it would take some work... even if the AI did build armies with them, it might be difficult to find the balance of 'enough MGLs that the AI uses them for Armies' with 'so many armies there is no chance for the SIngle Player'.

[ducki]

It's only an idea for a test, MrWIA.
No balance necessary, since it's just to see what sort of situations arise, how "wasteful" the AI is willing to be with the MGLs, etc. Not for a real game, but just for a further test like alexman's above.

A starting leader went straight to a warrior, but if the AI doesn't "think" it needs more units, will it rush a temple? An aqueduct?

The SoZ-style MGL-spurting Palace would allow alexman to see how the AI uses a freebie leader at any stage of the game and with regularity.

I shudder to think of trying to balance something like that for actual gameplay.

[alexman]

Yes, that was a great idea ducki.
I actually did that, and the AI used them all to rush Warriors and Settlers.

I'm sure the AI builds Armies sometimes, because people have reported seeing them sometimes in C3C, but it certainly doesn't build them often enough.

[ducki]

Just pulling this to the front page to be close to it's brother, the Re-balancing Armies thread.

[Thoth]

Has anyone given thought to Player1's idea of changing the MA to produce an army every 15 turns instead of allowing their construction? It would remove the prebuild issue as well as guaranteeing the AI has armies to play with.

Combined with the removal of the "Victorious Army" requiement the changes will feel different from stock, but should help out the AI significantly in the mid-late game.

[ducki]

Personally, I think that's a bit too heavily weighted in favor of the human, with the caveat that I haven't tried it. As expensive as armies are to build, is 15 turns "realistic" in terms of how often you normally would build one? Plus you get to use that city's production for something to actually put in the army.

Just random thoughts about the player1 idea, nothing concrete.

[lockstep]

Regarding the idea of an army-spawning Military Academy: Every 15 turns seems way too often for me, given the current power of armies - I'd say every 30 or even 40 turns. Furthermore, testing results from korn469 indicate that army spawning ignores the cities-needed-per-army limit (whatever you set it in the editor). OTOH, this suggestion would at least force the AI to use armies (that is, if the AI doesn't disband them for their shield value, but I refuse to believe that for now).

Sidenote: I'm still for a Military Academy that doesn't require a victorious army, even if we don't change anything else in the AU mod. nbarclay's '400-shield-pre-build' argument is yet another reason for this particular change. But as for now, we should wait for possible changes to armies and the Mil Ac in upcoming patches.

[alexman]

Well, the 1.22 patch has done nothing to encourage the AI to build Armies, and the timing of the next patch is uncertain, if there even will be one.

This shortcoming of the AI is the most serious in the game, IMHO. Even Sid-level AI stands no chance without Armies against a human with inferior units but with Armies.

The power of armies in C3C along with the inability of MGLs to rush Great Wonders make it virtually a no-brainer for humans to use an early MGL for an Army, so the PTW stretegic decision that might have denied builders the Military Academy is no longer there anyway.

I say we should remove the victorious Army requirement from the Military Academy.

[bongo]

Originally posted by alexman
I say we should remove the victorious Army requirement from the Military Academy.

Would it help? I read some other thread where some guys tried to *encourage* the AI to use armies. They did get some success when they let the palace spit out an army every n turn and the AI had LARGE nubmers of units to put in it. Even then the 2-3 first armies just stayed at home, defending the capital.

[ducki]

That seems a bit far from stock(edit:referring to boingo's suggestion); I'd rather not start out there. If we end up there after trying less radical things, so be it.

Removing the Victorious Army is conservative, and we don't really have enough evidence to decide whether it will be effective of not.

I second alexman's suggestion.

[lockstep]

I second it also (or would that be 'third'?).

This change isn't that radical and has already been part of former AU mod versions.