CS Slingshot - Is it overpowered?

[dsplaisted]

I'm not convinced that the Civil Service slingshot is the one right strategy in all games. Maybe it is in the majority of them, but I think in Epic 7, going Oracle -> Metal Casting -> Colossus was just as powerful. Of course that's in the special case of a water map, and perhaps the Colossus slingshot is something you would also like to avoid.

I believe the fastest three finishes were Compromise in 1214 AD (!), Uberfish in 1553 AD, and me in 1595 AD. Compromise used the Colossus slingshot and as far as I can tell never researched civil service. Uberfish used the civil service slingshot, but in a mirror game (which he didn't finish), he used Pyramids and the Great Library for similar effect. I used the Colossus slingshot and didn't research Civil Service till after 1000 AD.

Of course in most games civil service the one right choice. I would support watering it down somehow, but I would prefer that at first it be handled with event-specific variant rules so that we can test what works best. But Sirian is really the expert and boss here so of course I'll be happy with whatever he decides.

[Sirian]

<chuckles for a moment at a spelling typo right before the words "careful choice of language" >

Hey, I said careful CHOICE of language, not careful typing and editing.

I'd like to know exactly what you consider to be the problem with the Pyramids. The wonder is expensive enough that the opportunity costs are real and serious (especially without Stone), unlike that of the Oracle.

That's my best argument for why the Pyramids are OK, but it's not holding up in the face of incoming data. The Pyramids themselves are OK, but early Representation provides too much goodness to food-rich civs if leveraged to the max.

The Pyramids' chief benefit is using it to enable early Representation. Early UniSuff enables cashrushing, but that has been reined in already and comes without any shield bonuses since one won't have mature Towns yet. Early Police State was not even used by people in Epic Six, an ALWAYS WAR game. (I used it, but we'll get to that later). Hereditary is reachable by other means.

I don't buy that the Pyramids-Library combo is the problem. Four flood plains instead of four grass can do for you the same thing the Great Library can: let you feed two extra specialists. No, it's something else. It's anywhere that the math overlaps too heavily. Caste+Scientists gives you the same rush, if you've got the food on hand. Doubling the output of every specialist you employ, for a strat that is already viable without this boost, is too much.


See, a lot of things going on in the design are great by themselves, but turn to poison when overlapped. Sometimes ideas started out one way, and were fine, but got changed and created new side effects. Sometimes the side effects are a lesser poison than a problem that was removed with the changes, but that doesn't mean that "better than before" equates to be "good enough now".


All this would fall by the wayside if sooooo were right about players tiring of perfecting the uberstrat and exploring alternatives. Unfortunately, that's not reality. Reality is Blake's logic of "people accepting that some starts DEMAND the CSS."

My problem with wonder-based gambits and AI exploits is that they vary too little from map to map. You open a can of whoopass, and rather than being something that shifts dramatically from game to game, with lots of nuance, it is mass-produced by the factory, no (or little) assembly required. The Epics bear my personality stamp, and I lean too much toward craving newness to be content designing events that feature perfecting the execution of a strategy that is fully mapped before the game opens, chosen from a pull-down menu of three or four uberstrat options on turn one.

A strategy of building up a number of cities is more compelling. It varies a lot from map to map, with serious disparities between dotmaps in most cases, but viable paths for more than one option. I think that's the strongest part of civ, interaction with the map (which varies) whereas the tech tree is the weakest part of civ (being always the same).

Too much of the game, at the moment, is capital-centric, where most games become "OCC+" through the middle game, with your ubercity being the driving force amidst a playfield where you can only have a few cities in the early game before expansion constraints block your progress.


Ice Age is my favorite map script. Why do you think that is? (Anybody.)

I'm certain it's the best one I made, even though I am quite proud of Highlands, Pangaea, Archipelago, Great Plains, and Team Battleground.


- Sirian

[Sirian]

I'm not convinced that the Civil Service slingshot is the one right strategy in all games.

I appreciate that, but that is not the benchmark. Nor does it matter that there are a handful of other paths that are equally unbalanced. (That only means we need to correct several problem areas, in addition to this one).

Blake's pithy line about strong food starts "demanding" the CSS is the end of the road on debating whether or not there is a problem here.


We're already trying some variant rules to this effect. However, this is only to further feel out the issue, since any variant rule added to most games is better placed in the tournament rules, with events allowed to use variant rules to ENABLE the overpowered options where they are intended to be part of the playfield for a specific event.

Anybody with a particular interest in this topic should find time to play Adventure Thirteen. Data from that will further our discussions, I expect.


- Sirian

[darrelljs]

Can we ban certain things for certain events only? Sometimes building the Pyramids and running a specialist economy is a fun way to go, but not necessarily the optimal way to go. I'd hate to have that option taken away all the time...

Darrell

[Kodii]

I think it can be intertwined into the variants. For example, I think Adventure Thirteen's purpose is to experiment with eliminating the Pyramids and the Civil Service Slingshot. I mean, an archipelago map where our capital cannot build wonders, and our cities must build walls first? I'm not being spoilerish, just stating the variants

[Kylearan]

Hi,

All this would fall by the wayside if sooooo were right about players tiring of perfecting the uberstrat and exploring alternatives. [...]
My problem with wonder-based gambits and AI exploits is that they vary too little from map to map. [...] The Epics bear my personality stamp, and I lean too much toward craving newness

In a sense, your highly creative scenarios which differ greatly from each other are part of the problem, ironically. If in a given game so much feels different from last game, why should the player try out a new strategy? Seeing how old, known strategies work in the new variant environment can be interesting enough already, so the incentive to walk unusual paths often isn't high enough.

For example, Epic 6 had been my first real attempt at Always War, and so I felt happy to see if a CS slingshot would work in that environment at all, or if the player has to devote too much resources to the war to do it. I didn't do it because the situation demanded the CSS; I did it because the combination of AW plus CSS was new territory for me (note that I do not play private games anymore, and never played many, to try out scenario/strategy combinations).

I absolutely love variants. I'm variant scum, that's the prime reason why I'm here at RB. But sometimes I find it hard to resist the urge to try out old strategies in new environments, to see how they work there. But doing just that has convinced me by now that the CSS is indeed broken, because it works so well so often, regardless of restricting variant rules. I'd be happy to see it ruled out, maybe by using the beaker cap Blake has proposed - by a rule and not by a mod, if possible.


Regarding the Pyramids, I can't say much about their potentially exploitative powers. I have built them several times, sometimes with the Great Library afterwards, but never had tried them in conjunction with a highly focused specialist economy. I might have a couple of additional specialists if I am in representation, but never focused on irrigating more than usual to run more specialists.

There are a few other things I like to see adressed though, but we'll talk about that when Adventure 13 closes.

-Kylearan

[Sirian]

In a sense, your highly creative scenarios which differ greatly from each other are part of the problem, ironically. If in a given game so much feels different from last game, why should the player try out a new strategy?

Everybody thinking that in Epic Three lost, and big time, to your expert analysis and divination of the most effective opening. Sometimes the old faithful becomes a trap.

You're right, of course. Purely for purposes of having fun, people not already disenchanted with a particular strat can actually stretch its shelf life by applying it in new circumstances. But on the flip side, there's now enough map variety and other forms of variety (traits, UUs, resource distribution, lay of the land with religions, etc) that one could make that argument even in the absence of variants. ... How will it play with a bunch of Silks at the start instead of a Gold? What if you have Monty on your doorstep. Etc. At some point you have to stop and say, "I've seen this strat enough times. What am I MISSING because it keeps getting in the way?"


I'm looking forward to discussions about Adv-11. That's not too far away, now.


- Sirian

[Swiss Pauli]

FWIW, if you're looking for opinions on this matter, I'd support solving the CSS (and Pyramids if necessary) via RB tournament rules, rather than by actual game rules modification. And for the CSS, I think the best option is just to disallow the Oracle or Great Prophet shortcuts to CS.

I'd second this, but rather than outright bans, I'd favour a 'no Great Wonder without its resource' rule. This simply feels right to me: how would you build the Pyramids without stone, the Oracle without marble? It would make the Great Lighthouse relatively more attractive, but this is a quite situational Wonder anyway.

This would also stop the Oracle-GP-CS Lightbulb, and limit the opportunity of a Pyramids-GE-Great Library combo.

[Rowain]

Just one comment on Sulla's game:


You have (with Lib and Academie) ~70 beakers without CS so using Scientists before Oracle and building the Academie earlier I'm quite sure you would have been able to research CS yourself very fast anyway. Yes the CS would have come a few turns later but you could have used the shields for additional settlers or an Army.
I'm quite sure without the Oracle you would have reached Bureaucracy and the 100+ beakers still at this time == far too early for the AI to stand a chance against you.

What this shows is not so much the overpoweredness of the Oracle (or GP) but that of Bureaucracy which comes far to early and is far to easy to reach and that has its origin in the easy crab of CoL via Priesthood

[uberfish]

Too much of the game, at the moment, is capital-centric, where most games become "OCC+" through the middle game, with your ubercity being the driving force amidst a playfield where you can only have a few cities in the early game before expansion constraints block your progress.

The game encourages the player to build SSCs, but great wonders are not the cause of that as only Great Library and Temple of Artemis directly enhance a SSC. Unless I turned it into a military production centre, my capital is usually doing the lion's share of the research even in games where I don't build any of them. Rather, the extra resources granted to capital cities, along with the long headstart it gets on development and cottage growth given a Worker first build make the city very strong. And because the game favours city specialization and rewards building a SSC with the mini-wonders of academy-Oxford that you always want to apply to the best city regardless of when you get them.

If you have a start like Sulla's, and are playing with the goal of a peaceful space race, the capital is so powerful that the optimal strategy is always to get academy-bureaucracy-Oxford as soon as practical, whether or not Oracle-CS is allowed or not. So restricting CSS doesn't actually introduce any strategic variety in that case.

The only way to de-emphasize the capital relative to the other cities is to start with a sub-par capital. People will still build a SSC somewhere, but there will be variety in the positioning and it won't have the long headstart the original capital does, and if they want to run bureaucracy there they have to pay the not insignificant palace move cost. About the only times I've seen a decentralized economy be competitive with a SSC is with great lighthouse or pyramids in play.

It's combining pyramids-rep with a SSC that makes pyramids-rep-GL a tier 1 strategy. I think if the ability to build the SSC was restricted for pyramids-rep somehow, it would be more balanced.

Blake's pithy line about strong food starts "demanding" the CSS is the end of the road on debating whether or not there is a problem here.

His conclusion is mistaken, and I'm unlikely to play oracle-CS at RB again even were it not to be restricted - because I've already concluded that the best result I can hope for is to come in second to a more aggressive strategy.

I am a little concerned that restrictions will end up pushing the game into a 'one right strategy' of a different sort.