Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Micro102]

What about the ability to tweak the nation after nationgen? If I could iron out the oddities and add some custom spells then I think not only would it make nationgen more enjoyable, but some actual nations that could be shared with others could come out of it.

I wonder if it's possible to design a tool to easily let people make their own sprites too. Run the tool, alt-click the sprite you want to edit, it pops up in an editor, you edit  and save it, and it overwrites the proper file.

[E. Albright]

There's no tool to do that, but it probably wouldn't be terribly difficult to code one (note: I'm not volunteering :p). We had everyone write fluff for NG nations in the round 4.10 game (http://www.bay12forums.com/smf/index.php?topic=144007.msg5680284), and that was fun IMO; not only did we get to exercise our creativity, we were forced to fit it to NG's random nonsense rather than letting ourselves excessive creative freedom.

SpriteGen got a UI overhaul in the last release, and while it's not as freeform as what you're asking for, it's easier to use than it was.

[Cruxador]

Assassins can't get magic paths. However, #assassin can crop up on mages. That's always been a thing, and no provisions are made to exclude paths that are most readily exploitable owing to cheesy battle summons - it would be bizarre to do so when skelispam amulets are still a thing in vanilla. I suppose I could make assassin-mages be cap-only, but it would be rather odd to eliminate them entirely given the vanilla examples of mage-assassins.

Yeah, wasn't trying to say "fix this specific thing that happened", just wondering about general process since I don't really follow it closely. It came up in discussion elsewhere whether it's better to give everyone three nations to choose from, meaning some can have three good options and some can have three bad ones, or to let everyone pick from a shared list, meaning that nobody will get totally screwed out of good choices by randomness, but unavoidably conferring an advantage based on timezone and availability. Since the underlying problem is heterogeneity of balance, that's why I asked if there had been anything done in this regard.

Overall, you hit the crux of where NG (or any procedural content) falls short: the tradeoff between cohesion and variety is the key one, and finding a balance can be hard. I've tried over the last several years to shift the balance in the direction of cohesion, but it's hard to produce a large number of procedural nations because at a certain point increasing cohesion requires increasing semantic depth. In this context, that goes from a mechanical problem (producing varied units) to a natural language processing problem (articulating sufficient coherent thematic and semantic relationships between the units and the nation), and that's a rather harder one.

I think NationGen is pretty good at this, honestly. It establishes core conceits in race and mechanics, and then iterates and diverges from there in systematic ways, which is, I reckon, the most you can ask for.

It may be more cohesive than it was when you tried it (depending on how long it had been), but it's fair to say that it would be worthwhile for us to work on national cohesion at a greater depth than mechanical and visual thematic cohesion, even if it's meaningfully difficult to expand in that direction.

What, like constructing a dynamic narrative for the descriptions? Sounds like a big job. If you do intend to do something like this, I think a lot can be done with a relatively small amount of random text strings, allowing people to connect the dots on their own. OSR type tabletop games which have big tables for random setting elements would be a good source for inspiration there.

What about the ability to tweak the nation after nationgen? If I could iron out the oddities and add some custom spells then I think not only would it make nationgen more enjoyable, but some actual nations that could be shared with others could come out of it.

I wonder if it's possible to design a tool to easily let people make their own sprites too. Run the tool, alt-click the sprite you want to edit, it pops up in an editor, you edit  and save it, and it overwrites the proper file.

Editing descriptions in the mod files that nationgen creates is very easy. It should be very easy (though I'm also not volunteering) to make those editable in an application, if you wanted to, but there's not really any point. The amount of technical knowledge required is opening the file in a text editor and using ctrl+f to find the unit name, then looking where it says "descr". Sprites can also be replaced just by replacing the file, and you can make a new one with spritegen or pretty much any image editor. You can also make these changes (or any that only change images and strings, as far as I'm aware) after a game has already started without causing any problems. At least in PBEM, I never tried it on direct connect.

There's no tool to do that, but it probably wouldn't be terribly difficult to code one (note: I'm not volunteering :p). We had everyone write fluff for NG nations in the round 4.10 game (http://www.bay12forums.com/smf/index.php?topic=144007.msg5680284), and that was fun IMO; not only did we get to exercise our creativity, we were forced to fit it to NG's random nonsense rather than letting ourselves excessive creative freedom.

Seems like it would be good to encourage that kind of thing for nationgen games as a general rule. Only caveat I can think of is if someone wants to be sneaky by obfuscating how a unit works, making your weak things look strong, your strong things look weak, implying a unit has special abilities (for which there aren't icons) when it doesn't, or removing reference to abilities it does have. Since this level of deception is possible, but tedious and inconvenient, to circumvent, naturally there should be a gentleman's agreement not to do such things.

[E. Albright]

It came up in discussion elsewhere whether it's better to give everyone three nations to choose from, meaning some can have three good options and some can have three bad ones, or to let everyone pick from a shared list, meaning that nobody will get totally screwed out of good choices by randomness, but unavoidably conferring an advantage based on timezone and availability.

The general balances here IMO are to make decisions more collectively and more measuredly. If each person gets three nations, let another player ban one of the three. That cuts down on the worst cheese, even though you may end up with two tepid choices. If it's all from one pool, don't make it a free-for-all, make it a draft with random order. My preferred option is to combine these two: each player picks two nations from a reasonably large pool via a round-robin draft, then bans one of another player's nations. There's an argument in favor of doing the banning pass first and then letting the player pick from their two drafted nations. There's also one in favor of doing a first banning pass so the cheesiest cheese is rooted out ahead AND the remaining cheese being homogenized by a banning pass afterwards as well.

Spoiler: That seems clear enough to me, but just in case, an example (click to show/hide)

Players 1,2,3; Nations 1-30 -> [0,0,0];[1...30]
Player 1 bans Nation 6  -> [0,0,0];[1...5,7...30]
P2 bans N16  -> [0,0,0];[1...5,7...15,17...30]
P3 bans N26  -> [0,0,0];[1...5,7...15,17...25,27...30]
Random order of 3,1,2 is selected
P3 picks N1  -> [0,0,1];[2...5,7...15,17...25,27...30]
P1 picks N11  -> [11,0,1];[2...5,7...10,12...15,17...25,27...30]
P2 picks N21  -> [11,21,1];[2...5,7...10,12...15,17...20,22...25,27...30]
P2 picks N5  -> [11,21:5,1];[2...4,7...10,12...15,17...20,22...25,27...30]
P1 picks N15  -> [11:15,21:5,1];[2...4,7...10,12...14,17...20,22...25,27...30]
P3 picks N25  -> [11:15,21:5,1:25];[2...4,7...10,12...14,17...20,22...24,27...30]
(Either stop here and players pick which they want, or... every Player P now bans one of Player P-1's nations in any order)
P1 bans N25 -> [11:15,21:5,1]
P2 bans N11 -> [15,21:5,1]
P3 bans N21 -> [15,5,1]

Seems like it would be good to encourage that kind of thing for nationgen games as a general rule. Only caveat I can think of is if someone wants to be sneaky by obfuscating how a unit works, making your weak things look strong, your strong things look weak, implying a unit has special abilities (for which there aren't icons) when it doesn't, or removing reference to abilities it does have. Since this level of deception is possible, but tedious and inconvenient, to circumvent, naturally there should be a gentleman's agreement not to do such things.

The best solution for this IMO is to forbid/limit renaming unit types and then just make sure everyone has the original seeds and Advanced Description files. If you are renaming units, though, then agreeing ahead of time to always be unambiguously descritpive would be best, yes.

[Cruxador]

It came up in discussion elsewhere whether it's better to give everyone three nations to choose from, meaning some can have three good options and some can have three bad ones, or to let everyone pick from a shared list, meaning that nobody will get totally screwed out of good choices by randomness, but unavoidably conferring an advantage based on timezone and availability.

The general balances here IMO are to make decisions more collectively and more measuredly. If each person gets three nations, let another player ban one of the three. That cuts down on the worst cheese, even though you may end up with two tepid choices. If it's all from one pool, don't make it a free-for-all, make it a draft with random order. My preferred option is to combine these two: each player picks two nations from a reasonably large pool via a round-robin draft, then bans one of another player's nations. There's an argument in favor of doing the banning pass first and then letting the player pick from their two drafted nations. There's also one in favor of doing a first banning pass so the cheesiest cheese is rooted out ahead AND the remaining cheese being homogenized by a banning pass afterwards as well.

Spoiler: That seems clear enough to me, but just in case, an example (click to show/hide)

Players 1,2,3; Nations 1-30 -> [0,0,0];[1...30]
Player 1 bans Nation 6  -> [0,0,0];[1...5,7...30]
P2 bans N16  -> [0,0,0];[1...5,7...15,17...30]
P3 bans N26  -> [0,0,0];[1...5,7...15,17...25,27...30]
Random order of 3,1,2 is selected
P3 picks N1  -> [0,0,1];[2...5,7...15,17...25,27...30]
P1 picks N11  -> [11,0,1];[2...5,7...10,12...15,17...25,27...30]
P2 picks N21  -> [11,21,1];[2...5,7...10,12...15,17...20,22...25,27...30]
P2 picks N5  -> [11,21:5,1];[2...4,7...10,12...15,17...20,22...25,27...30]
P1 picks N15  -> [11:15,21:5,1];[2...4,7...10,12...14,17...20,22...25,27...30]
P3 picks N25  -> [11:15,21:5,1:25];[2...4,7...10,12...14,17...20,22...24,27...30]
(Either stop here and players pick which they want, or... every Player P now bans one of Player P-1's nations in any order)
P1 bans N25 -> [11:15,21:5,1]
P2 bans N11 -> [15,21:5,1]
P3 bans N21 -> [15,5,1]

These methods which increase the complexity are a good idea with respect to the factors mentioned, but have the potential to cause other problems. First of all, they require a non-trivial degree of coordination, the complete lack of which is why I often see the first-come-first-served approach treated as a default. Although it can work well for some groups, others may find their interest in the game greatly impeded by the necessity to undergo tedious bureaucratic measures before even getting to play. After all, how many people forgo diplomacy entirely (or nearly so) because it's not a part of the game that they enjoy?

Furthermore, I think the issue of tepid choices that you touch on is very significant. If someone has a nation that they're excited to play, and then are banned from doing so, where does their enthusiasm for the game go?

Obviously, the exact point on the spectrum that you want to aim at depends on the group you play with, but I think these systems sacrifice too much of the fun and uniqueness of the experience for the sake of the balance; in order to maintain appeal with a wider array of players, improvements to balance should prioritize methods with minimal impact on player enthusiasm.

Seems like it would be good to encourage that kind of thing for nationgen games as a general rule. Only caveat I can think of is if someone wants to be sneaky by obfuscating how a unit works, making your weak things look strong, your strong things look weak, implying a unit has special abilities (for which there aren't icons) when it doesn't, or removing reference to abilities it does have. Since this level of deception is possible, but tedious and inconvenient, to circumvent, naturally there should be a gentleman's agreement not to do such things.

The best solution for this IMO is to forbid/limit renaming unit types and then just make sure everyone has the original seeds and Advanced Description files. If you are renaming units, though, then agreeing ahead of time to always be unambiguously descritpive would be best, yes.

I think names are less recognizable than sprites. If you're just glancing at battle summaries this isn't the case, but when watching a battle proper, the sprites are the thing you see and you've got to examine to see the names, at which point you'll see the stats anyway. There's no perfect solution, and I'm not concerned about the very obvious cases of abuse anyway, since those are easy to notice and as long as you say not to, nobody is going to do something too cheeky like making their units all identical. It's more like those small deceptions that are a bit borderline that will be an inevitable temptation. Of course, to a certain degree it's an acceptable hazard I suppose.

[E. Albright]

Furthermore, I think the issue of tepid choices that you touch on is very significant. If someone has a nation that they're excited to play, and then are banned from doing so, where does their enthusiasm for the game go?

This comes down to a matter of expectations. If they go into the game knowing there will be drafts and/or bans, they should realistically not be excited to play a specific nation upon seeing the 20/50/100/etc that are in the pool; they should be looking to find a set of nations they'd like to play. This is no worse than being excited to play a particular nation (vanilla, NG, or whatever) but then having someone else pick it before you can; if you can't maintain your enthusiasm at that point, I'd worry you'll not maintain enthusiasm for long enough to finish a game unless you're winning. That's not an entirely fair assumption, but it's not an entirely unfair one either.

YMMV, but drafting processes can be fun in and of themselves if you're into that sort of thing. Admittedly, I've played to completion two or three MP NG games where everyone took whatever trash a random hash of their name or somesuch gave them, so I may have more fascination with exploring mechanics and/or squeezing blood from stone than the average player.

BTW, did you ever do anything with the African (I think?) sprites you mentioned thinking about fiddling with for NG a year or three ago?

[EuchreJack]

I enjoyed the idea of our three player disciple team blind NG.  Not knowing what you're getting was fun, and balanced somewhat by three-nation teams.

[Cruxador]

Well, I got a specific balance complaint that is probably worth considering: Blood nations are consistently too strong, in part because script considers a B1 to be equivalent to (for example an F1 mage, and prices them according to the value of the latter. Just repeating what I heard so I don't know if this is true of the new version or old info. I did just confirm that you can get S1 for 45g, which sequani stargazers also get, but on a nation where that's arguably the main feature since elder druids were nerfed in the transition to dom4, although of course boarspam is also important.

Also, it looks like the capital generated sites are all using the same graphics, on my end; all a picture of volcanoes. I didn't test this thoroughly.

Furthermore, I think the issue of tepid choices that you touch on is very significant. If someone has a nation that they're excited to play, and then are banned from doing so, where does their enthusiasm for the game go?

This comes down to a matter of expectations. If they go into the game knowing there will be drafts and/or bans, they should realistically not be excited to play a specific nation upon seeing the 20/50/100/etc that are in the pool; they should be looking to find a set of nations they'd like to play. This is no worse than being excited to play a particular nation (vanilla, NG, or whatever) but then having someone else pick it before you can; if you can't maintain your enthusiasm at that point, I'd worry you'll not maintain enthusiasm for long enough to finish a game unless you're winning. That's not an entirely fair assumption, but it's not an entirely unfair one either.

Although it's not entirely unfair, it doesn't take into account the feeling of investment. When you see a cool nation and get excited by it, you're getting invested in that, so when you're told you can't play it after all, that's a problem. On the other hand, when you've been playing the game for a week or two, by that point you're invested even if you start losing.

BTW, did you ever do anything with the African (I think?) sprites you mentioned thinking about fiddling with for NG a year or three ago?

Nah, not yet. It wound up on the backburner in favor of other projects. I'll get around to it eventually I'm sure, but no ETA. I do have some other stuff that I could separate into its layers for you, although I still don't have a very comprehensive understanding of how the format is supposed to work.

[MCreeper]

Tiruin wants to play too (not sure about the preffered mode  ). So do we actually start something? NG game for 6 (or however much there is) players maybe?

[MCreeper]

In hindsight, i'm rather grateful that none of dom5 games there had any damned golems in them (or i just didn't pay attention?  Don't think they did anything important, anyways). In place where i play now, most games end with a torrential spam of golems from everyone capable of making golems. If game draws out long enough, someone gets their hands on blatantly exploity golem with ring of returning and horror harmonica. And before and during golem spam, everyone and their dog spams sea kings. Those always come with exactly same equipment or none at all. It's frankly revolting.

[Sean Mirrsen]

Welcome to "competitive" multiplayer, I guess. Esports and competitive-minded design has ruined plenty a good strategy game, because people keep trying to find a "meta" to exploit. It really makes me wish for more games with randomized elements in them, where you can't reliably hit any given "meta" and have to conform to what you're dealt. Like, say, Sword of the Stars and its semi-randomized research tree.

[Cruxador]

This is an artifact of either non-standard game settings or players that are consistently not very aggressive. Maybe very excessive numbers of throne points required. The typical meta in both Dominions 4 and 5 has the game ending far before the point where SCs dominate, most of the time. It hasn't been since Dom3 that the game was routinely like that on default settings (or nearly so).

[MCreeper]

This is an artifact of either non-standard game settings or players that are consistently not very aggressive. Maybe very excessive numbers of throne points required. The typical meta in both Dominions 4 and 5 has the game ending far before the point where SCs dominate, most of the time. It hasn't been since Dom3 that the game was routinely like that on default settings (or nearly so).

For some reason i thought we were playing on standart half+1 throne, but i actually checked ongoing games from out most frequent host now and lo! 9 out of 12 by 1's all over. Which did lead to hilarious results in one case - 112 turns long 4 team disciples game, no team fully eliminated yet (although ours is on a slow but steady way to it since beggining), no end in sight. So yeah, slogs supreme. Besides those, there is 80 turn "should have ended early" game with "one 3, two 2, five 1, 8 for win" thrones. Not really SC domination, but there already were at least 5 golems - two from me-MA Rlyeh, three from Tien Chi. And in other game, stupidly overfed MA Ermor  with a golem on turn 45. Along with enchantment 6. Yep.    Gang up on Ermor fast and don't get distracted by "free" castles that Jotunheim loser left behind.

[E. Albright]

Re: the development of predictable metas in Dominions, NationGen in the past has had pretensions (and even some dev work) towards randomized new spells and items. It's something that was dropped to the wayside, though - unsurprisingly, given how hard of a problem doing that right is.

[Cruxador]

Re: the development of predictable metas in Dominions, NationGen in the past has had pretensions (and even some dev work) towards randomized new spells and items. It's something that was dropped to the wayside, though - unsurprisingly, given how hard of a problem doing that right is.

I think it depends on what counts as "right". I would think that making some random summonable units and the spells to summon them would be a pretty straighforward extension of what the engine already does, even if it's only touching a tiny portion of what random spells could potentially do. But even for vanilla nations, national summons are often the most defining spells.