Login

Arantor · January 14th, 2014, 19:25

I took the day out today to visit the Science Museum, get a handle on the various sciences, to try and get a sense of how the tech tree should be handled, as well as reviewing how the different games have done their tech trees. I find it interesting how MOO stands basically unique in the way it approaches tech, not just with the quasi-random tech tree but other things.

So let's kick off with a quick recap of MOO before moving into the others, just so that we can compare notes.

MOO

MOO2

Civ II

SMAC

Endless Space

Galactic Civilizations (the first one)

I have also got Civ V, GalCiv 2, Armada 2526, and Sins of a Solar Empire to have a bash at and see how they do it. But I think the trends are clear and give me some serious direction on how I should be doing it.

1) This one tech at a time nonsense. This doesn't square even in the smaller scale of a flourishing civilisation, let alone a galactic empire. Not unless you specifically want to go down that road (a la MOO), so definitely keeping that aspect.

2) There is a fairly clear trend towards 'what techs are' rather than 'what techs do' - we see weapons, comms, propulsion etc coming up but only in Endless Space do we really see the tech tree named after what it does (warfare, expansion & exploration). I have to say, I do actually prefer Endless Space on this one. It just feels more correct to me, but at the end, it's primarily a name for a category of techs. But it does feel more suggestive of the types of techs that are likely to be found (which would fit better with Catwalk's suggestion of bringing together all the waste clean-up techs)

3) Colonisation/improvement of harsher environments does broadly come up in the different games (e.g. SMAC with xenofungus terraforming techs) but interestingly, ES has tech for pretty much all types of planets, even as I said desert, arid, along with the more exotic hostiles we all know and love (tundra, barren, lava, etc.) - there's also 'arctic' which is interesting. Certainly there's no requirement to drop the hostile colony tech! But I'm inclined to not have *too* many techs on the tree, though that does depend a lot on the number of variations on planet type; MOO has something like a dozen (14 offhand?) planetary types of which half or so are hostile, ES has fewer but all except terran require tech.

I have to say I'm not a fan of tech for tech's sake, so definitely not going to go overboard. I find it interesting that 'planetology' doesn't come up as a subject so much. Not sure what to make of that.

4) Universal translator. This is something that is unique to GalCiv as far as I know, where you have to research communication techs to be able to talk to other races. I have to say I like this. I love the idea that you could meet races you don't know how to talk to. The notion of entering a MoO style diplo screen and not knowing what to say, only getting growls or purrs or other random noises from the alien race rather than actual dialogue, and only able to offer 'hopefully positive sounding noises' or 'angry defiant noises' which might in themselves be right or wrong.

But I'm concerned about it being a tech you 'need' to have, in a way that we've never really entirely had in MoO before (unless you're a race that's a straight up warmonger and don't care to engage in diplo at any point). I also like the idea of introducing an empath race that doesn't have to research the tech (and presumably gets diplo bonuses)

5) I'm not a fan of the dependency style, all too often that seems to be a bout of 'research x, then y, then z'. Won't be having that. The randomised approach MOO has is far more engaging because it can screw you up, rather than having a comfortable framework to play against where you're following a list. I've also heard it said that you get into the 'todo list' mentality where you're just shown everything and it's like a microgame to see how many of the lights you can light up by the end of the game.

6) Definitely not a fan of the imprecise research goals of SMAC (leads to confusion over the tree)

7) I can understand why Civ/SMAC do the single tree and the rest don't, that's more a case of setting rather than game design, and that's cool. I find it interesting that 6 is common... MOO I and GalCiv I have 6 branches, ES has 4, MOO2 has 8 (and SMAC has 4 broad categories that all techs fall into)

I'm trying to get a handle on whether this should be changed in some way - either up or down - and whether some overall rejigging is required. Racial bias is quite strong, so whatever categories do turn up, they should probably reflect the different races to some degree (MOO certainly does this, with the birds relying on their natural agility and understanding of flight to buff propulsion whilst being rubbish at shields, or the Meklar being great at computers because they're mostly computers anyway but bad with the environment), which makes me wonder how much of the categorisation is really best placed with an understanding of the races in the universe or whether the tech categories are the primary consideration and then build the races around *that*. (I would argue that's what MOO did)

Needs more thought but I have plenty of ideas on how the propulsion and colonisation techs would come across wink

(This is more of a braindump -> brainstorm than anything else, please feel free to ignore me)

rho21 · January 15th, 2014, 09:20

Sword of the Stars is worth looking at as they've tried to do tech tree randomisation, like MoO.

And some opinions:

MoO does it so well because techs are generally not unique: they have alternatives or a better version further down the tree. This means that, although you are missing some important techs in one generation, you can usually act to mitigate the problem.

RefSteel · January 15th, 2014, 14:09

If you don't mind me spitballing here a bit too:

(January 14th, 2014, 19:25)Arantor Wrote: 1) This one tech at a time nonsense. This doesn't square even in the smaller scale of a flourishing civilisation, let alone a galactic empire. Not unless you specifically want to go down that road (a la MOO), so definitely keeping that aspect.

I agree that this is really important, not only for a realistic feel, but to create more complex choices and make it easier for the game to scale with empire size.

Quote:2) There is a fairly clear trend towards 'what techs are' rather than 'what techs do' - we see weapons, comms, propulsion etc coming up but only in Endless Space do we really see the tech tree named after what it does (warfare, expansion & exploration). I have to say, I do actually prefer Endless Space on this one. It just feels more correct to me, but at the end, it's primarily a name for a category of techs. But it does feel more suggestive of the types of techs that are likely to be found (which would fit better with Catwalk's suggestion of bringing together all the waste clean-up techs)

Just thinking about MoO techs though: Are Fusion Drives a warfare, expansion, or exploration tech? What about advanced range? Even Controlled Environment technology allows invasions as well as expansion to new worlds. (And expansion extends the range at which exploration is possible too...) Another thing to think about: One of the things that makes MoO's tech tree so interesting and complex is the interdependencies of different technologies. (I actually didn't realize just how much interdependency until Catwalk made some good points in the Imperium 40 discussion thread.) There are no hard-and-fast cross-tree prerequisites, but Robotic Controls become much more valuable as Industrial Tech advances (reducing the cost of building the new factories) and waste clean-up improves (increasing the new factories' value). This results in strategic choices with much greater depth, because you're not forced to pursue them in a particular order, but the need for each piece of the puzzle must be judged in each game. (There are several other interdependencies like this, but I think that's the strongest example, and the most central to the game's economy.)

Quote:3) Colonisation/improvement of harsher environments does broadly come up in the different games (e.g. SMAC with xenofungus terraforming techs) but interestingly, ES has tech for pretty much all types of planets, even as I said desert, arid, along with the more exotic hostiles we all know and love (tundra, barren, lava, etc.) - there's also 'arctic' which is interesting. Certainly there's no requirement to drop the hostile colony tech! But I'm inclined to not have *too* many techs on the tree, though that does depend a lot on the number of variations on planet type; MOO has something like a dozen (14 offhand?) planetary types of which half or so are hostile, ES has fewer but all except terran require tech.

Yeah, MoO has 7 habitable planets types, 6 hostile types, and no-planets asteroid fields for a total of 14. The importance of Controlled Environment technology is in part to prolong the expansion phase of the game, as Sirian once noted, but it also adds variety to the map. I've always liked the idea - in theory at least - of different races having different planetary preferences. For instance, speaking in MoO terms, one race might start on a Desert world and get the same size ranges on those those that Humans get on Terrans, with their preferred "habitable" environments running (say) Desert > Arid > Steppe > Minimal > Terran > Barren > Dead instead of the Humans' Terran > Jungle > Ocean > Arid > Steppe > Desert > Minimal. This desert race's tech tree possiblities would then include "Controlled Jungle Environment" and "Controlled Ocean Environment" techs in place of Barren and Tundra for humans. This might make things overly complex or reverse some of the usefulness that Sirian mentioned, but it's the sort of thing I'd really love to see.

Seven seems to be a magic number for us humans (the way we store short-term memory) so at least keeping the number of hostile planet techs required to six or less (for each race, if their preferences are different) is probably wise. (1. Habitable without tech; 2. Barren; 3. Tundra; etc.) Make the number too large, and the different types will start to run together. Sword of the Stars uses numeric hostlity values for each race, for example, and while this seems like it should be simpler or more intuitive, it makes the game less appealing to me. "-1000!" is a big, impressive number, but it doesn't have the emotional impact of "Inferno!" with images of active volcanoes and lava floes.

Quote:4) Universal translator. This is something that is unique to GalCiv as far as I know, where you have to research communication techs to be able to talk to other races. I have to say I like this. I love the idea that you could meet races you don't know how to talk to. The notion of entering a MoO style diplo screen and not knowing what to say, only getting growls or purrs or other random noises from the alien race rather than actual dialogue, and only able to offer 'hopefully positive sounding noises' or 'angry defiant noises' which might in themselves be right or wrong.

But I'm concerned about it being a tech you 'need' to have, in a way that we've never really entirely had in MoO before (unless you're a race that's a straight up warmonger and don't care to engage in diplo at any point). I also like the idea of introducing an empath race that doesn't have to research the tech (and presumably gets diplo bonuses)

For what it's worth, I think the way this would be handled in a tech tree like MoO's would be something like this:
Tier 1: Xeno Communications. Permits diplomacy deals including peace treaties, cash tribute, and trade agreements up to 10% of the smaller empire's production, plus a small boost to relations.
Tier 2: Universal Translator. Permits diplomacy deals as Xeno Communication, plus tech tributes, threats, non-aggression pacts, tech trading, and trade agreements up to 20% of the smaller empire's production, plus a slightly larger boost to relations.
Tier 3: Xeno Emulator Matrix. Permits diplomacy deals as Universal Translator, plus alliances, requests to break alliances with (or declare war upon) other races, and trade agreements up to 30% of the smaller empire's production, plus a significant boost to relations.
Every race's tree would be guaranteed at least one of these three techs, but not all three. (This is similar to the way Space Scanners work, for instance, but at tiers 1, 3, and 5).

Quote:7) I can understand why Civ/SMAC do the single tree and the rest don't, that's more a case of setting rather than game design, and that's cool. I find it interesting that 6 is common... MOO I and GalCiv I have 6 branches, ES has 4, MOO2 has 8 (and SMAC has 4 broad categories that all techs fall into)

I'm trying to get a handle on whether this should be changed in some way - either up or down - and whether some overall rejigging is required. Racial bias is quite strong, so whatever categories do turn up, they should probably reflect the different races to some degree (MOO certainly does this, with the birds relying on their natural agility and understanding of flight to buff propulsion whilst being rubbish at shields, or the Meklar being great at computers because they're mostly computers anyway but bad with the environment), which makes me wonder how much of the categorisation is really best placed with an understanding of the races in the universe or whether the tech categories are the primary consideration and then build the races around *that*. (I would argue that's what MOO did)

I agree about the way MoO's tech trees and racial biases were most likely built. As for the number of fields, this may depend on the intended scope of the game. Three is probably too few, and MoO2's eight is probably too many. (That magic number 7 again...) I could even imagine a game with enormous scope that had separate all-high-level tech trees that could be unlocked by other technology. (Any one of a set of techs perhaps, as with Space Scanners or my translator suggestions.)

Quote:Needs more thought but I have plenty of ideas on how the propulsion and colonisation techs would come across

(This is more of a braindump -> brainstorm than anything else, please feel free to ignore me)

Please also take my suggestions above as just spitballing/brainstorming. In particular, I suspect some of my "this would be so cool!" ideas could represent feature creep. Probably the most important thing is to pick some key ideas and build the core of the gameplay around them. (The top-level design of the tech tree of course would likely be one of these key ideas.) Then resist adding more stuff that doesn't directly serve the core concepts of the game. For great-sounding ideas that don't fit naturally into core gameplay, if the title is successful there are always expansions and DLC....

Arantor · January 17th, 2014, 11:28

Quote:- Big tech tree, lots of links and cross-links. http://chariot.nickersonm.com/ANY_TechTree.html

- A couple of expensive era-changing techs (fusion power and antimatter power) which are requirements for large swathes of the tech tree (the bits with yellow and purple background on that link).

- Each link has a percentage chance of appearing in your tree, dependent on race (some are 100% for all races). So if you miss out on the link to a tech from its standard pre-requisite, there's still a chance you'll get it later.

- If a tech is missing from your tree, there's a small chance of gaining access to it after winning a battle in which it is employed. It's also possible to do diplomatically. There will be no links out of the tech, however.

- There are individual translation techs for each other race; it's pretty cheap to be able to talk to them, rather more expensive to gain the tech to ally.

- A few techs are race-specific.

Big tech tree with cross dependencies is interesting. I like the idea of races having race specific tech, but also races having more or less chance to obtain a tech (sort of like the Psilons have more chance of any individual tech being in their tree than other races)

Definitely need to play SotS smile

Quote:The good:
- Each race won't always play out the same way.
- Different races do have a tendency towards one type of weapons, which gives them flavour.
- Era techs work pretty well to stop bee-lining to really powerful weapons being the best option.

The bad:
- Not knowing in advance which techs will be available sometimes results in pursuing expensive dead-ends. This means you will usually end up following the highest chance path and curse if you have to try a different one.
- While there are many different weapons lines, there's only one shield line, and only one armour line. If you miss out on the later varieties of those techs, your ships will be very weak.

I feel that both of these bad points detract from the strategy of the game. When you meet a warmonger who has invested heavily in missile tech, for example, you want to be able to counter this with point defences. If that option is missing from your tree, it doesn't leave you many choices to work around the shortcoming.

Each race playing out differently is a good thing, differentiates the races in a way we don't have right now.

I'm not a fan of the dead-end tech thing, that's what makes MoO great is that you have a line of techs and if you miss one you're in with a chance of picking up the next one so that any pain should be temporary with any luck...

Quote:Just thinking about MoO techs though: Are Fusion Drives a warfare, expansion, or exploration tech? What about advanced range? Even Controlled Environment technology allows invasions as well as expansion to new worlds. (And expansion extends the range at which exploration is possible too...) Another thing to think about: One of the things that makes MoO's tech tree so interesting and complex is the interdependencies of different technologies. (I actually didn't realize just how much interdependency until Catwalk made some good points in the Imperium 40 discussion thread.) There are no hard-and-fast cross-tree prerequisites, but Robotic Controls become much more valuable as Industrial Tech advances (reducing the cost of building the new factories) and waste clean-up improves (increasing the new factories' value). This results in strategic choices with much greater depth, because you're not forced to pursue them in a particular order, but the need for each piece of the puzzle must be judged in each game. (There are several other interdependencies like this, but I think that's the strongest example, and the most central to the game's economy.)

I deliberately didn't cover that level of interdependence in the tech trees - because that's both a component of any good tech tree and also an implicit factor in all the games we've seen on the subject. No point having big pointed sticks if you can't get anywhere to use them (or can't use them in time). The main criteria I was using for reference is whether or not we have distinct tech pre-requisites to get from tech to tech; MoO doesn't require you having controlled barren as a requirement for controlled tundra, which is not a requirement for controlled dead, etc. Some games as observed physically have dependencies, but often applied interdependencies as well.

Interestingly, the interdependency angle isn't lost in the likes of CivII, you still need techs across the board to make it work - you can't shoot down one branch of the tech tree indefinitely (but whereas in MoO it's a balancing mechanism by way of that implied interdependence, it's explicit in CivII)

To answer the other point: I believe drive techs would come under 'expansion and exploration' if pressed for categorisation, they're not primarily a warfare tool, though they are a component of warfare, but moreso with expansion and exploration, if that makes sense.

But you make a great argument for keeping techs categorised as what they are, rather than what they do.

Quote:Yeah, MoO has 7 habitable planets types, 6 hostile types, and no-planets asteroid fields for a total of 14. The importance of Controlled Environment technology is in part to prolong the expansion phase of the game, as Sirian once noted, but it also adds variety to the map. I've always liked the idea - in theory at least - of different races having different planetary preferences. For instance, speaking in MoO terms, one race might start on a Desert world and get the same size ranges on those those that Humans get on Terrans, with their preferred "habitable" environments running (say) Desert > Arid > Steppe > Minimal > Terran > Barren > Dead instead of the Humans' Terran > Jungle > Ocean > Arid > Steppe > Desert > Minimal. This desert race's tech tree possiblities would then include "Controlled Jungle Environment" and "Controlled Ocean Environment" techs in place of Barren and Tundra for humans. This might make things overly complex or reverse some of the usefulness that Sirian mentioned, but it's the sort of thing I'd really love to see.

I did actually sort of come to the same conclusion but for a different reason. I was thinking about the Trilarians from MoO 2, which are a sort of galactic jellyfish, but I thought... what if they were more conventional jellyfish? As in: thriving on an ocean environment, finding a terran environment tolerable but not ideal, and treating desert as a hostile.

Certainly this would emphasise race as being a choice and give them more character but I am sort of concerned of the complexity. But I'm always drawn back to changing up the relationship of everyone vs Silicoids; we all know you gotta play the rocks differently and I'm intrigued by the idea of having other races that you have to play differently too - but without introducing unnecessary complexity or introducing RP mechanics (as that was something that raised concern)

Quote:Seven seems to be a magic number for us humans (the way we store short-term memory) so at least keeping the number of hostile planet techs required to six or less (for each race, if their preferences are different) is probably wise. (1. Habitable without tech; 2. Barren; 3. Tundra; etc.) Make the number too large, and the different types will start to run together. Sword of the Stars uses numeric hostlity values for each race, for example, and while this seems like it should be simpler or more intuitive, it makes the game less appealing to me. "-1000!" is a big, impressive number, but it doesn't have the emotional impact of "Inferno!" with images of active volcanoes and lava floes.

7 works well for a variety of things (though, realistically, 7 +/- 2 is the target, not a hard 7)

You're right though - the races are running together except the rocks. And the whole 'hostility numbers' thing is not something I want to entertain. Aside from the absolute lack of panache, it's not significantly better than having the names we have now. As for 'Inferno!' that was one of the things commented upon by Arne, that the planet screen should show some sense of what the planet is.

Quote:For what it's worth, I think the way this would be handled in a tech tree like MoO's would be something like this:
Tier 1: Xeno Communications. Permits diplomacy deals including peace treaties, cash tribute, and trade agreements up to 10% of the smaller empire's production, plus a small boost to relations.
Tier 2: Universal Translator. Permits diplomacy deals as Xeno Communication, plus tech tributes, threats, non-aggression pacts, tech trading, and trade agreements up to 20% of the smaller empire's production, plus a slightly larger boost to relations.
Tier 3: Xeno Emulator Matrix. Permits diplomacy deals as Universal Translator, plus alliances, requests to break alliances with (or declare war upon) other races, and trade agreements up to 30% of the smaller empire's production, plus a significant boost to relations.
Every race's tree would be guaranteed at least one of these three techs, but not all three. (This is similar to the way Space Scanners work, for instance, but at tiers 1, 3, and 5).

That works for me smile

Where it gets interesting... do both sides need to have tier 1 to engage in diplomacy, or does it only require one of the two sides doing so?

Mind you, the idea of 'communicating' without actual ability to do so intrigues me; more than one sci-fi story has tackled the idea of two races at war because they couldn't communicate with each other (most influentially for me, Ender's Game) and accidentally triggering a war seems interesting.

Quote:I agree about the way MoO's tech trees and racial biases were most likely built. As for the number of fields, this may depend on the intended scope of the game. Three is probably too few, and MoO2's eight is probably too many. (That magic number 7 again...) I could even imagine a game with enormous scope that had separate all-high-level tech trees that could be unlocked by other technology. (Any one of a set of techs perhaps, as with Space Scanners or my translator suggestions.)

Well, 6 is definitely in scope of 7 +/- 2 in a way that ES/SMAC with their 4 categories aren't. 7 itself intrigues me, there is something very appealing about having something based on 7 - it's less regular than 6 visually, conceptually etc.

The problem with enormous scope is that already most people here tend towards smaller games because they're a bigger challenge (due to earlier expansion pressures), while you'd need to be on the larger galaxies to really have the time to actually get to those later techs. That could, though, be an interesting scenario - have a small galaxy and keep things from boiling over long enough to obtain a research victory. That and the fact that huge galaxy games tend to drag because glassing the entire galaxy a system or two at a time becomes a real drag.

Quote:Please also take my suggestions above as just spitballing/brainstorming. In particular, I suspect some of my "this would be so cool!" ideas could represent feature creep. Probably the most important thing is to pick some key ideas and build the core of the gameplay around them. (The top-level design of the tech tree of course would likely be one of these key ideas.) Then resist adding more stuff that doesn't directly serve the core concepts of the game. For great-sounding ideas that don't fit naturally into core gameplay, if the title is successful there are always expansions and DLC....

That's where I'm trying to get to... I recognise that the tech tree is possibly one of the most important aspects of the game, and will potentially colour everything. I'd argue it's easier to balance the races after the fact than it is to rebalance the tech tree.

I'm just brainstorming here too, because there's a lot of ideas here, a lot of potential to do a great many things and feature creep is always a concern, but so too is getting the right feel. I'm still floating between a couple of things, but I'm definitely narrowing down what I want to make and all of this still comes into it smile

The trick is keeping something I'd like to play as well as what other people would like to play; already I've noted the resistance to 'having to RP certain things' (because that would appeal to me personally), even though slightly oddly that's almost exactly what must be done to successfully play the rocks. I wonder if that puts people off playing the rocks.

RefSteel · January 17th, 2014, 21:50

(January 17th, 2014, 11:28)Arantor Wrote: You're right though - the races are running together except the rocks.

I know others have said that, but I actually disagree: I find that every MoO race has a particular flavor, and though the Silis are the most extreme, I don't play any two races in the same way as one another. For what it's worth.

Quote:Where it gets interesting... do both sides need to have tier 1 to engage in diplomacy, or does it only require one of the two sides doing so?

I think the best way for this to work would be for diplomacy to require that only one side have the relevant technology, and most AIs (i.e. those without major diplomatic focus) should assign it a very low priority. This means AIs that benefit a lot from diplomatic deals (e.g. MoO Humans) or just really want to have them because of in-game personality will be able to make deals with the player and other AIs, but players can't just ignore the tech and count on the AIs getting it instead. If both sides need the tech, I think that will just lead to player frustration ("Still waiting for the Meklar to get around to teching their stupid translator") or gamesmanship. ("I might as well put that tech off anyway for now; the AI doesn't usually get it until like T40-45.")

Quote:Well, 6 is definitely in scope of 7 +/- 2 in a way that ES/SMAC with their 4 categories aren't. 7 itself intrigues me, there is something very appealing about having something based on 7 - it's less regular than 6 visually, conceptually etc.

I don't think of 7 as the perfect number actually, but as a pretty good upper limit. It is claimed that human beings can keep no more than 7 discreet units at a time in their short-term memories (e.g. a 7-digit phone number) - though we can improve on that by mentally linking things into sets and the like, thereby reducing the number of "discreet units" involved in the memory. (I've memorized the 15-digit number I have to dial to get my brother internationally, but it's not a 15-digit number when I call it up: It's basically five "words" of between two and four "syllables" apiece. YMMV.) It therefore seems to me that if you have more than seven categories of (e.g.) tech to consider, it becomes difficult to really keep track of them all at once.

Quote:The problem with enormous scope is that already most people here tend towards smaller games because they're a bigger challenge (due to earlier expansion pressures)

Yeah; creating a game with enormous scope would not be easy, and creating a good one is nigh-impossible. Just because I can imagine something doesn't mean it can (or should!) be done.

Quote:The trick is keeping something I'd like to play as well as what other people would like to play; already I've noted the resistance to 'having to RP certain things' (because that would appeal to me personally), even though slightly oddly that's almost exactly what must be done to successfully play the rocks. I wonder if that puts people off playing the rocks.

I actually love role-playing in a strategy game ... which is part of the reason I hate the feeling of being forced to roleplay one in a specific way. If you consider the 'coid design to be one that forces players to RP, and you want to do that more, I'd say try to stick to the model of the Silis. The trick is this: Don't directly penalize players for "playing the wrong way." Instead, design the race so that playing the "right way" is especially effective for them. I'd say giving Warmonger Race a morale penalty if they've been at peace for too long will feel like "forcing" to the player and be unfun. But if Warmonger Battleguys are economically weak without war, and can make major (including economic) gains through war, towards whose prosecution they hold advantages, they can encourage playing to their strength but still leave the player the option to play them peacefully, accepting the higher difficulty this means from game start. The problem comes in no small part from making things worse once the game is already underway just because of a race (or the game as a whole; see Civ5) being played against the grain of its intended RP.

Arantor · January 18th, 2014, 19:36

Quote:I know others have said that, but I actually disagree: I find that every MoO race has a particular flavor, and though the Silis are the most extreme, I don't play any two races in the same way as one another. For what it's worth.

That's the thing... do you play them strategically differently? You might RP them differently, you might react to different events differently because of the rules of them game but do you physically change the strategy to play the same way?

Consider: supposing you're a Human player with a very awkward diplomatic situation - everyone's against you almost as badly as you might be as the Mrrshan. Unlikely, but it serves to illustrate my point: would you play those Humans differently towards winning than you would the Mrrshan?

I have to be honest, I'm going more off my experience and I've admitted I don't win the harder games, but possibly because I'm playing all the races except the rocks much the same way... I kind of figured others would do.

All the comments I've seen have indicated that everyone plays the rocks differently because of their starting conditions mean it is most favourable to do so.

Quote:I actually love role-playing in a strategy game ... which is part of the reason I hate the feeling of being forced to roleplay one in a specific way. If you consider the 'coid design to be one that forces players to RP, and you want to do that more, I'd say try to stick to the model of the Silis. The trick is this: Don't directly penalize players for "playing the wrong way." Instead, design the race so that playing the "right way" is especially effective for them.

I think we've been trying to argue both sides of the same coin - because the model of the rocks is exactly what I have in mind. If you play the rocks the way they were intended, it's especially effective. Play them the same as any other race and you're going to find it significantly harder, penalising you for more conventional strategies of growing colonies.

Quote:I'd say giving Warmonger Race a morale penalty if they've been at peace for too long will feel like "forcing" to the player and be unfun. But if Warmonger Battleguys are economically weak without war, and can make major (including economic) gains through war, towards whose prosecution they hold advantages, they can encourage playing to their strength but still leave the player the option to play them peacefully, accepting the higher difficulty this means from game start

I can see where you're coming from, but are they not just doing the same thing: forcing the player to play a certain way? The difference is really just whether it's a carrot or a stick. Though the carrot would be more effective. (I'm a programmer, semantics are a big part of my psych makeup :P)

But yes, I can see that being important.

Quote:I don't think of 7 as the perfect number actually, but as a pretty good upper limit. It is claimed that human beings can keep no more than 7 discreet units at a time in their short-term memories (e.g. a 7-digit phone number) - though we can improve on that by mentally linking things into sets and the like, thereby reducing the number of "discreet units" involved in the memory. (I've memorized the 15-digit number I have to dial to get my brother internationally, but it's not a 15-digit number when I call it up: It's basically five "words" of between two and four "syllables" apiece. YMMV.) It therefore seems to me that if you have more than seven categories of (e.g.) tech to consider, it becomes difficult to really keep track of them all at once.

It's not strictly the perfect number since YMMV as ever, however George A. Miller's work The Psychology of Communication, 1967, covers a decent amount of this stuff and he arrived at 7 +/- 2 as being the most we could normally handle - but your example of making it a collection of sets fits just fine: you're creating discrete units (chunking it), and thus you're dealing with 5 chunks of content which fits into 7 +/- 2.

I am amazed, actually, that more people that don't study psychology to some degree because it has so much bearing on our lives that most people don't quite get. I brought this to a debate a few months ago when no-one could quite put their finger on why something was perceived as complex to use - it's because the items aren't presented in groups that fit in that sort of limit. (I'm no psych major but I've found it very useful to have a broad overview of some things like this.)

Quote:I think the best way for this to work would be for diplomacy to require that only one side have the relevant technology, and most AIs (i.e. those without major diplomatic focus) should assign it a very low priority. This means AIs that benefit a lot from diplomatic deals (e.g. MoO Humans) or just really want to have them because of in-game personality will be able to make deals with the player and other AIs, but players can't just ignore the tech and count on the AIs getting it instead. If both sides need the tech, I think that will just lead to player frustration ("Still waiting for the Meklar to get around to teching their stupid translator") or gamesmanship. ("I might as well put that tech off anyway for now; the AI doesn't usually get it until like T40-45.")

That works from a gameplay standpoint. I still think it could be interesting to have diplomatic tussles before either side has the tech, though I'm beginning to accept I might have to let that one go in the name of good gameplay. Especially if for whatever reason neither side manages to get the earlier one and has to wait a tier or two before communicating. Could be interesting in smaller galaxies.

Quote:Yeah; creating a game with enormous scope would not be easy, and creating a good one is nigh-impossible. Just because I can imagine something doesn't mean it can (or should!) be done.

That's one of the challenges I'm hitting with trying to review the tech tree as a whole... even starting from the MoO tech tree and trying to judge where these should fit seems to suggest making the tree longer - but that's only any use if you're in a with a real chance of getting to the higher tiers anyway.

Hmm. I've been wrapped up in other stuff the last few days, but I think it's time I really got down to working on this rather than trying to fit it in around everything else.

RefSteel

(January 18th, 2014, 19:36)Arantor Wrote: That's the thing... do you play them strategically differently?

Yes. Every race has unique strategies that work better for it than for any other; sometimes there's partial overlap between two specific races, but the overlap is never complete. Learning which strategies work best for different races and in different situations is one of the keys to improving one's game. (Of course there are strategic goals that are common to all races, but that comes with the nature of a 4X game).

Quote:Consider: supposing you're a Human player with a very awkward diplomatic situation - everyone's against you almost as badly as you might be as the Mrrshan. Unlikely, but it serves to illustrate my point: would you play those Humans differently towards winning than you would the Mrrshan?

So there are three ways this can happen: One is that you've already been playing the Humans very differently than you would have played the Mrrshans, taking advantage of your extra diplomatic leeway. The second is that you've already effectively won (thereby making everyone hate you for your power) or lost the game, and it doesn't really matter what strategy you follow anymore. The third is due to weird random factors or save-file edits. In the latter case, first of all you're saying, "Would you play the Humans differently from the Mrrshans if one of the key ways in which they differ from the Mrrshans were changed so that they're exactly like the Mrrshans instead?" But second of all, the answer is still an emphatic yes! (Those two are extremely different races!)

The exact strategy I would follow would depend on the game situation, but as the Humans, you can bet I'd be maxing trade wherever possible (which costs the Humans ~nothing, unlike other races) and counting on the gravitational pull of the default relations to help bring things back to normal, taking advantage of the Humans' good planetology abilities to strengthen my empire (and maybe their shield excellence if the situation called for it defensively) in the meantime. You can bet I'd be planning my defenses, establishing no more than minimal trade, and relying on the Mrrshan attack bonus and Weapons excellence to defend myself and make gains through combat as the cats.

Quote:I think we've been trying to argue both sides of the same coin - because the model of the rocks is exactly what I have in mind. If you play the rocks the way they were intended, it's especially effective. Play them the same as any other race and you're going to find it significantly harder, penalising you for more conventional strategies of growing colonies.

I think you're right - I think we're likely in agreement on most of this stuff and just making semantic arguments. I would say the rocks don't "penalize" the player for using a non-standard Silicoid strategy, but I think it's just because we're using "penalize" differently.

Catwalk · January 22nd, 2014, 17:32

I think it sounds like this moving far away from purity of form and into feature creep territory. I'm not saying you have to make this game like Master of Orion, but I think you should think carefully before adding a lot of features. Taken one at a time they may make sense, taken together they overwhelm the player and distract from the strategy aspect.

As for races, I don't think it's a question of rewarding or punishing. It's a matter of accuracy. If you reward or punish a race based on whether or not they're going to war, you will see players using all kinds of loopholes to get bonuses / avoid punishment. If you create circumstances where warfare becomes more profitable for some races than others (easy to do), you let the player make an accurate assessment of what his options are. If you create fuzzy variables, you frustrate the player. I find that strategy games shine when all aspects of the game are easily and intuitively understood, while putting them all together and deciding on the best strategy is the challenge. When elements themselves are clouded and fuzzy, I find myself losing interest.

Arantor · January 30th, 2014, 00:06

Yeah, I've been mulling that over an awful lot lately (when real life has given me leave to spend time on it, yay for crises to have to deal with!)

So, mulling over the tech plenty has lead to some very interesting ideas. I don't know how far I'm thinking yet but there's definite food for thought. I'm aware of the haunt of feature creep but I don't see how to approach this than either 'starting from a working MOO implementation and gradually tweaking' or 'pulling apart from the get-go and using that as a common ground for discussion'.

OK, so I gently suggested 7 rather than 6 branches, I'm still thinking that, the standard 6 from MOO plus Sociology. Partly because it covers the diplomacy techs, partly because it made me think of a bunch of other things.

First up: concerns over rebalancing the waste reduction/pollution clean-up tech. I get why it would be consolidated under one branch of the tree but that screws over any race that's poor in that branch (since they get hit badly both ways). Now, I got to thinking about our own planetary issues, our own growth issues, and what I realised is that there is a different way we can approach the waste clean-up type thing. Remember all the campaigns at the start of the 1990s about going green and stuff? Essentially a sociological change to get us to recycle, followed by encouraging reduced packaging etc. which is a sociologically motivated change. Or programs to help us deal with having more people in less space (since 'space' is an issue people have to deal with) with an attendant construction cost to build buildings that can hold more people. (Certain amount of handwaving of theme will be needed to fit the mechanics) There's also scope for 'training programs' or whatever to get better production - in short, it's a source of routes for races that (in MOO terms) would be poor in computers and/or construction to be able to make up the slack. Races that don't care, of course, don't have to research it.

Secondly, and more broadly: terraforming. I don't actually like it thematically. Look at sci-fi in general: how many times do we see 'terraforming while there's a colony already in place' vs 'terraforming from offworld'? What I'm thinking is that it wouldn't be infeasible to add a new tech to the weapons tree perhaps named the Genesis Device or similar, that gets added to a ship as a special, which terraforms a planet from space. This has other potential consequences - what if it were used offensively?

Which brings me to my third point: per race planetary bias. I never understood why - other than for balance purposes - all the starting races in MOO all got Terran 100 starting planets. Even if we go down the progenitor-race deal, it still doesn't sit right that they would all evolve on similar environments. The bears would likely do better coming from a jungle world, the cats from a steppe environment, perhaps even the rocks from an inferno-esque world. Now this is the one place I'm quite worried about it becoming not so much 'fuzzy' but perhaps overcomplicated. What if each race had a planetary type that it thrived on? Perhaps certain kinds of planetary types would be more or less hostile to others. Rocks, we know, don't care. But I can't imagine bears coping so well on the deserts, or cats in ocean worlds(!)

That gets complicated to remember who likes what but perhaps it's just as simple as 'this planet can hold x felines' as the felines would see it and 'this planet can hold x rocks' for when the rocks could see it if that makes sense. But that is certainly getting into the realms of fuzzy and intractable logic. I'm still trying to figure this out in my head. But thematically, the notion of each race trying not just to make each planet liveable but aspiring to be like their home worlds (be that biodomes or something else)... I could even see the rocks getting in on the terraforming action by converting planets to infernos where all that lovely magma action gets them going.

I've also been thinking about a shakeup of races - dropping the shapeshifters as we have them for a race of telepaths. They'd take the place of the spymaster race, but they'd have some more leverage during diplo (not sure what I'd do with the Humans yet, but the telepaths wouldn't get trade bonuses only diplo... all a bit theoretical!), but as a balancing measure - they wouldn't research the xenodiplo techs, they'd automatically be able to speak to other races, but not get the perks of the higher techs. IOW, their advantage is in early diplo and as other races get the xenodiplo techs, that advantage is nerfed over time (like other races, really)

Originally I was all for MOO on steroids but I know I wouldn't be happy with that. It needs to find its own identity and that does mean changing up even core mechanics (e.g. rethinking how terraforming works, sociology tech, but we need a common reference to be able to discuss potential mechanics) but as ever I think the *real* proof in the pudding is what happens when trying this stuff. It is all well and good to explore the theoretical side and what 'feels' like it might fit.

Still haven't managed to convince the other folks to give me the nod to go ahead properly (or any assets) so I think for now I'm just going to play it as a hobby project, building it in my home environment of PHP (I can do Unity but I can do PHP that much faster) and that would mean being able to drop *something* on to a server where I could let people play with it under controlled conditions. Potentially even multiplayer.

Arantor · February 5th, 2014, 21:27

Been doing some more of the numbers in how MoO does things, have to say that the more I outright study it looking for some understanding, the more I find myself simultaneously thinking how ingenious it is, and how straightforward it is.

It takes some time staring at the complete tech tree to really understand its subtlety. One can't help but admire it.

We have a tech tree arranged in 6 branches, with 50 theoretical levels (though no branch, not even weapons, has a tech in every level, not even weapons which looks like it has a gap at level 3 and at level 49, though level 36's Death Ray is an exception of sorts too), these levels are arranged in 10 tiers of 5 levels each.

Tiers have two principle factors from what I can see... firstly, if you're in a situation where you only have one tech in a tech tree due to the limited research list (LRL), that's going to be your choice for that tier. Secondly, and I think I'm right in this: when you get the choice of picks for tech (excepting the first round which is always confined to first tier, levels 2-5 only), your choice is upper-bounded by the range of what's in your current tier and what's in your next tier. For example, playing as Psilons (and thus 75% chance of any individual tech in the LRL), I researched ECM 2, then BC 3 (both tier 2) and the furthest choice I had was ECM 3 and BC 4, which are tier 3 techs. I had ECM 3 and BC 4 as choices before researching BC 3 because I was still in tier 2 after researching ECM 2. (Note: the level you have in a tech actually doesn't seem to play into this, since I'm currently level 13 in computing, but the highest tech levelled item I have in Computing is BC 3 at level 10, so it's still basing from what I can see more on what tier I'm in rather than what level I am, and leaves level to be a factor for miniaturisation)

I'm not sure what the mechanic for lower bounding is, same game I'd researched Warp 3 drives (tier 3) but still had the as-yet-unresearched range 4 tech from the first tier as a choice. I'm guessing the lower limit is 2 tiers below whatever your current tier is.

But the subtlety doesn't quite end there... far from it actually. If you plot it all out on a spreadsheet you start to see patterns emerging, assuming you ignore the level 1 techs. IIT techs are always third in the tier, for the 8 tiers they appear. Reduced Waste appears every other tier, and is always the last in its tier. ECM is always second of five, BC is always last of five, the space scanners are always third of five in their tier.

There's more, of course, planetary shields appear every other tier, second in their tier. Terraforming is always 6 levels apart, meaning that tier 5 has no + terraforming (but has hostile conversion), warp drives from warp 2 onwards are also 6 levels apart, meaning that they skip tier 7 (instead leaving us with only High Energy Focus)...

Weapons is the most interesting because if you skim the tiers, there is no point at which you have competing techs. There's always variety in that you have tiers with MIRV missiles/ground/particle weaponry/regular missiles and lasers, or tiers with ground/4x firing/particle/missiles/regular beam, there's never a point where you have directly competing techs in the weapons tree... and outside of that the only time directly competing techs show up is in propulsion tier 1 (range 4 vs range 5), planetology tier 2 (tundra vs dead env) and planetology tier 3 (inferno vs toxic)

There is a point to all of this, though... I've been running the numbers and I think I have a way to make my game function broadly the same (and thus capture what's really important) - I've already brought up the 7-branch tech concept, but I thought about going more down that road. Instead of 6 branches in 10x5 tiers, I was actually looking at 7 branches in 7 tiers of 7 levels of tech.

If we're already used to the idea of being constrained to see only a short distance ahead of where we are now, as represented by the tiers already, the idea of fewer broader tiers is not so far fetched, though instead of showing the entire next tier as an option, I'd possibly limit it to 'the next 7 from the current highest tech'; in MoO you can potentially see 5-9 techs ahead of the one you have as the highest level in a field, and interestingly enough more often that not in the above mentioned setup, you end up seeing 7-8 rather than 9 anyway.

Also for terraforming, I've sat and plotted out my races broadly in terms of what they are and what environments they would find hostile, and something interesting happened. I found myself creating race profiles where I was consciously not putting 'Radiated last' into the list and I found myself not looking at a list of 7 non-hostile + 6 hostile + none for 14, but a list of 16 instead (I added Mountainous and Arctic somewhere along the line)... where each race has 9 non-hostile picks and 7 hostile picks. This means of course that terraforming/colonisation for the more extravagant types potentially becomes a one-per-tier deal. But I'm not sure about that, especially if I want to go down the road of "Improve Planetary Atmosphere" type techs, a la MoO 2's terraforming worked, where each successive use brings it closer to what works for that race. But I'm concerned that this is too complex for the player to deal with (and no it's not what I'd call feature creep, it's about deconstructing a core mechanic and replacing it with a different one)

More thought required.

Navigation

Extra Menu

About us