I think that it isn´t very sensefull to make the game too "symetric", it is round-based so i think that it would become a tower defense game, If you Choose the map creation options right, you can get a symetric or asymetric gamefeeling, for example: very much sytem connections with much of wormholes will create a hard, symetric and chaotic game. increasing the number of enemies would enlarge the difficult to conquer territory because there are more seeded empires who will grow much faster to a great fiend around you. You can influence the opponent behavior if you choose other victory terms. They will be more accessible for diplomaticy if it isn´t a kind of winning option... , take all victory options if you want to make your oponents hard in handling. I´ve played the game many hours before i get behind the techtree, so i think: it is large enough.

Void2258 wrote:
I would disagree with some of the AI comments. Keep in mind that processor tech has continued to improve, and most importantly we now have quite a few spare processors. You could run really good AI by running the AI on a dedicated (or even multiple dedicated) cores. The performance bottleneck still exists, but it is much higher than it's generally taken to be, if you just branch out and don't try to make the whole game single threaded.

Gosh THIS. ALL THE THIS.



In this day and age, single thread and 32-bit have no place anymore. People have decent computers, let them use them.

I would disagree with some of the AI comments. Keep in mind that processor tech has continued to improve, and most importantly we now have quite a few spare processors. You could run really good AI by running the AI on a dedicated (or even multiple dedicated) cores. The performance bottleneck still exists, but it is much higher than it's generally taken to be, if you just branch out and don't try to make the whole game single threaded.

Dawn of War had one of the best AI's i've ever seen, but that was because someone built a dedicated mod to do so. He gave the AI's playbook opening based off what actual players were doing at the higher levels and then had them go from there (among other more clever things).



The issue with cheating AI in a TBS vs an RTS is that it's very easy to get what they got from cheating. If I destroy the base of a cheating AI in an RTS then whatever. I've still got however many other AI's to kill. If I managed to take the planet of a cheating AI in a TBS i now have all the buildings they've built and maybe some tech WAAAY ahead of schedule because the AI got there much faster. Suddenly I've take then advantage and can now use it against them, so the moment you start winning the "advantage" the AI has disappears as you begin to leverage their own resources against them.

Nasarog wrote:
Once you figure out the Endless difficulty, it's not so bad. But yea, I prefer non-cheating AI.

It's always a little difficult to compete with a cheating AI, but yes, even fairly smart AI such as those in Supreme Commander can be conquered if you figure out how to funnel them effectively. Still, I understand why. It's basically impossible to make a "creative" AI, one capable of confusing and baiting like a human, but even a sufficiently well-done "reactionary" AI is pretty cool to see in motion. I always love when going full-naval in SupCom prompts the AI to start using submarines and torpedo bombers.

Sinnaj63 wrote:
That's the thing though. I'm fine with not winning, but I'm not fine with being overrun by AI ships at turn 10 on an Endless 1v1.



Though I guess it actually works fairly well if you think about it. When players start playing, at low difficulties, the AI doesn't get unfair advantages; in fact it gets unfair disadvantages even. Then as you get better, the playing field gets leveled, untill you start being able to compete with advantages AI at which point you can probably also start playing in Multiplayer. That said, while I'm a big fan of multiplayer and absolutely love it, setting up games and finding times where a sufficent number of players are avaible usually leads to it being much more limited than playing against AI which obviously works all the time because your Processor is always there.

Once you figure out the Endless difficulty, it's not so bad. But yea, I prefer non-cheating AI.

Nasarog wrote:
You need to ask yourself this - Why am I playing this game. Am I playing it to have fun? To win? To...?



If you want to play an AI that is different every time, create a player group and MP. There are demising returns to working on AI. Most studios won't spend the time/effort/resources to really do a solid AI.

That's the thing though. I'm fine with not winning, but I'm not fine with being overrun by AI ships at turn 10 on an Endless 1v1.



Though I guess it actually works fairly well if you think about it. When players start playing, at low difficulties, the AI doesn't get unfair advantages; in fact it gets unfair disadvantages even. Then as you get better, the playing field gets leveled, untill you start being able to compete with advantages AI at which point you can probably also start playing in Multiplayer. That said, while I'm a big fan of multiplayer and absolutely love it, setting up games and finding times where a sufficent number of players are avaible usually leads to it being much more limited than playing against AI which obviously works all the time because your Processor is always there.

Sinnaj63 wrote:
The thing is though, I feel like a lot more players would play(And especially finish)games on higher difficulty if the AI wasn't so frustrating to play against, being both dumb and OP and annoying.

You need to ask yourself this - Why am I playing this game. Am I playing it to have fun? To win? To...?



If you want to play an AI that is different every time, create a player group and MP. There are demising returns to working on AI. Most studios won't spend the time/effort/resources to really do a solid AI.

In many cases in Endless Legend and Endless Space, there will be an optimal science path for a given situation, faction and playstyle. This is something that we are okay with. However, we are considering ways of making lesser-used technologies more interesting through reduced technology costs or something (think of Search Party for instance).

Tdoggs wrote:
I love play balance, but endless legend wasn't play balanced and it was great. I am not sure the game will hold up in multi player and nor should it. It would be awesome to play a pilgrim faction just running way from the empire and their imperial guard. I enjoyed leading massive armies of cultist and necroage across the map and crushing my enemies.

This. This is what I believe. Fun trumps balance in most games. Call of Duty? Sure, balance that, it's all about multiplayer. Battlefield? Balance it, again, multiplayer. Forza Motorsport? Screw balance, focus on fun. Same goes for Endless Space.

I love play balance, but endless legend wasn't play balanced and it was great. I am not sure the game will hold up in multi player and nor should it. It would be awesome to play a pilgrim faction just running way from the empire and their imperial guard. I enjoyed leading massive armies of cultist and necroage across the map and crushing my enemies.



I agree the AIs should be tailored to the civilizations they play. I would love to play one over powered game where I chase the pilgrams around the map trying to finish them off as Rome/empire crumbles around me I believe an open and early mod community could make thing game great. I am not sure how the AIs could be changed and the focus should still be on civization management.

I realize that's not a nice title so allow me to explain and expand. Further there's a caveat here in that I haven't played either game in awhile, so if something is drastically off or been fixed, feel free to let me know.



Things Endless Space/Legend do right:



Interface- DEAR GOD I cannot praise this enough. I HATE the tedious micro that comes out in late game TBS games where you're just doing the same damn thing 100 times, and while that's still very much an issue in Endless, it is a LOT better, and early game is a breeze. Menu's aren't just aesthetically pleasing but actually extremely functional, and running a large empire quickly is a breeze. There might be a few things here and there for me to nitpick but it's obvious they've got this on lock.



Diplomacy- sorta. It's basically better than 99% of other TBS out there where diplomacy is tacked on and extremely minor. One of my best memories of endless space was of a game where the carver AI was running away, and I actually managed to broker alliances with allied players to put them in check. Once things were better they realistically unallied, and they did a decent job of dog piling the winner in a way that kept the game close, but didn't feel arbitrary or unfair. Endless legend adding a diplomacy resource and black market is a tremendous step towards finally making a TBS game that has a meaningful diplomatic layer, but there are issues I'll get into later.



Visuals- Games are gorgeous. No one is arguing that.



Wold building- plot/flavor/feel is always great.



Potential Variety- Lots of unique ways to play......but I say potential for a reason.



The problems:



They balance their games in such a way (or rather there's a lack of balance) that the actual variety is very very small, and there's larger issues that stem from that.



Customizing races- This is a neat feature. I know a lot of people like it. I personally though would love to actually play these games on multiplayer, and having everyone do custom races quickly becomes very unfun in a competitive setting. Not only do all diplomatic systems fall apart in human vs human (mostly), but the trait choices and race designs suddenly become VERY limited. "So just play with default races?" EXACTLY, and thus my second issue. They don't balance the default races very well at all. Some are freakish powerhouses while others barely even function. I'd love to be able to get a game going between my friends and I and know that we're on roughly even ground even if we all pick different races, but realistically this just hasn't been the case. They do take effort to balance them slightly, but in practice I've found its not nearly enough. Certain custom or race skills are just way too good or don't work nearly well enough to justify.



Potential solutions: Please make the base races balanced assuming no one ever touches the customization screen. It makes the game a lot more fun for those of us who don't want to deal with it or who might want to do PvP gameplay. Good AI's are very hard to make. Often companies do % resources bonuses which doesn't work very well since you just have to survive early and then take their resources (forcing military only strategies), and that's because making an adaptable AI that doesn't eat processing power is a very difficult task, so having a way to play against other humans for an increase in challenge is important. We can obviously do that now, but it's a very shallow experience because of how few viable stock races there are.



Variety in strategy- Expansion. Is. God. The only thing that changes between races (in both endless games) is just how fast you can expand, and in the case of cravers, what happens if you stop. Every single race you're going to want to get a new colony asap, and then just keep doing that is quickly as possible. This stems from a whole bunch of reasons (a LOT to do with the tech tree and flat bonus buildings), but mostly because no other playstyle is incentivized in the slightest (yes cultists, but even then you just find the first minor race you can, assimilate, and rush to the next one). This leads to a very large feeling of "i've done this before" when playing different races. "oh you mean i'm going to do everything I can to expand in the first 15 turns, then make sure i've got a military to defend it, while getting approval techs/buildings when needed, and then rinse and repeat until the game ends...yay". I would really like if there were in game reasons (not just racial affinities) that you might consider going tall instead of wide. The approval system technically does this, but the payoff for managing your approval so you can go as wide as possible is ALWAYS better than trying to do otherwise.



Potential solution- Some tech solutions (see below). Rework flat bonuses, rework approval, give better incentive to consider not expanding like a madman, make it so that some races (not just 1) don't actually scale perfectly when going wide, allow a way to keep up with production, etc, etc.



Variety in tech- there's a lot of techs, but ultimately your path boils down to a few key chokepoints with specific "must have" techs. You need approval techs, no question no option. You need exploration/expansion techs, no question no option. You need to pick a kind of military tech, and research it, no question no option (a binary influence mostly due to resources). What should be some of the biggest choices in the game actually feel the most formulaic. I stopped playing endless space when I realized I was choosing my entire tech tree at the start of the game. Not literally the whole thing, but by the time I was actually making choices about my technology I'd already run away with the game and was at the "cleanup" phase.



Potential solutions- 1. I've always wanted tech trees to vary more depending on your race, but i'll admit that's a very difficult goal. I would like better race specific techs though as some can be game changing while others are literally worthless 2. Be careful with flat bonus buildings, or give buildings/techs that only work if you've got VERY high approval (not just spam out colonies approval), but are balanced by giving out extremely good bonuses if you can meet that threshold. A wide player will skip them until end game and a tall player can grab them to keep up with the wide player. Do everything you can to give players options for playstyles from the tech tree alone (not just race selection). This branches into other areas as well, not just expansion.



Combat- This goes hand in hand with tech making it very shallow. Endless spaces combat, at the times I was playing it, sucked. Cards were broken in that barrier/heal/heal was the go to solution, and anything that didn't play well with that was basically worthless. All these cards and abilities that could be cool, but in practice were just glitz and glamour. I challenged myself to NOT use barrier/repair because it was more fun, but if I were playing human players I shouldn't expect the same, and it guts the entire system. Worse though is that there's not a whole lot to the underlying combat. Mixed with the tech issue again it was basically "pick a type of DPS, pick a type of defense, build a mobile equation". There's not a whole lot else to it. Resources could maybe limit your ability to get a specific weapon, but it was rare (especially with the "wide is god" problem), and the fact that the weapons themselves (lasers were the best) weren't super well balanced didn't help. Legend had a potentially cool system with movement and tactics and I remember a patch note saying we were going to get ZOC patched in to help make it deeper, but it never came. For the most part it just boils down to "spam your ranged unit + initiative" with very mild differences. It's way better than endless space (which felt awful) but still a shadow of what it could be if given proper care.



This also relates to the fact that "all races have the same units" syndrome. Again legend did this better, but for the most part races feel identical with maybe one unique unit that may or may not actually affect how combat plays out (how well does it help the "spam ranged" or "barrier/heal/heal lasers" strat). The tech/weapons further this issue.



Potential solutions- I don't think the card system from space is coming back, but if it does please do a better job of balancing it. Further though please add more....unique or conditional stuff. Lasers are lasers, no matter what ship they're on, and they do DPS. Ditto with missiles and mass drives. Give me EMP weapons that do no damage but have a chance of shutting down one of their systems. Give me tractor beams which only work on ships smaller than yours but hold a ship at a phase for an extra turn (something that would be good on big ships, but bad on small ones). Give me self destruct bombs so I can have a swarm race that builds tons of cheap small ships which get in close and explode to do damage. Give me actual options in execution that start from ship design (a tractor beam as I described would be worthless on a small ship, but possibly very good on a large one for example). Not just variations on my DPS calculation. I'd love even more reasons for each race to play different in combat (legend did a much better job at this for the record, and i'm super excited to see that approach hit space, even if it's not up to my standards)



TL;DR-



The current balance methods of Space/Legend lead to a "sweet spot" of unique play, that exists all too briefly, and then degenerates into the exact same strategy with very minor variations. By the mid to end game I feel like the unique flavor of my race has been lost and i'm just going through the motions to expand, buff my military, and win by whatever method I've chosen. Boosting science is the same as boosting industry, just with different buildings. Going wide is viable, going tall is not. Lasers work on all ships, with the only difference being how many I can fit. Give me more flavorful and unique choices that aren't just shifts in equations, but actual playstyles. Please identify the more unique options (Stuff like pilgrim evacuation or vaulters portal) and work to include more of it in a way that's viable, not a gimmick.





Edit-



I forgot to say that just from what i've read of the new hero system I think they're already somewhat aware of this. The old hero system had a very similar problem to what i'm talking about above, where it felt like a formality that was solved and didn't vary much from game to game. It was basically a test to see if you could properly convert your econ into more econ. The new system sounds VASTLY better and expands in all the ways i've been talking about above (not just "does this econ" but more "oh they gave me this, so I can now do that" sort of choices).

Nasarog wrote:
3 reasons:



The A.I. doesn't have an over arching personality - which it should. Necrophages should always be bent on conquest. Vaulters on scientific discovery. Drakken on diplomatic control. Those are just examples. The fact that the central A.I. personality can change from play thru to play thru (to increase replayability) makes certain of this.



The A.I. needs hardcoded coding for what tech it wants based on the situation it finds itself in which would be connected to a personality trait. That would make them more capable of handling human play.



The A.I. needs to learn how to wage war better.



Then the adaptive A.I. fills in the gaps. This would be linked to the affinities. Have a random affinity that would play it the way it's played now. If all of that was done, the game would be better, but.... People would complain that the A.I. is too predictable, and the min/maxers would quickly find the weaknesses of each A.I. variant, and an arms race begins between the dev and the players. That's all fine and good, except for Amplitude. Why? Simple, money. ALL OF THIS COSTS MONEY AND TIME,



Since only a small fraction of players would ever get to this part of the game anyways, it's probably not worth it for them. Just look at the global achievement stats to see who has done what. Using a number of 700K (from steamspy) and the achievements shows that 3.3% of people beat it on hard or higher. 1.7% of serious or higher. 1.0% on impossible, and 0.8% of endless. All together that's less than 7%. Less that 4,900 people of the player base, but the requirement to build and test so many different A.I. variants is probably substantial.



That's my 0.02 cents. Though, having said that, Amplitude is constantly working the A.I. I think that once the game is done and all of the core mechanics are in, they'll be able to do a proper A.I. pass and finish the game. Until then, the A.I. will grow "organically" or something like that.

The thing is though, I feel like a lot more players would play(And especially finish)games on higher difficulty if the AI wasn't so frustrating to play against, being both dumb and OP and annoying.

The-Cat-o-Nine-Tales wrote:


My remark was not concerned so much with how badly or well the AI plays, but with AI literally incapable of playing their faction.

3 reasons:



The A.I. doesn't have an over arching personality - which it should. Necrophages should always be bent on conquest. Vaulters on scientific discovery. Drakken on diplomatic control. Those are just examples. The fact that the central A.I. personality can change from play thru to play thru (to increase replayability) makes certain of this.



The A.I. needs hardcoded coding for what tech it wants based on the situation it finds itself in which would be connected to a personality trait. That would make them more capable of handling human play.



The A.I. needs to learn how to wage war better.



Then the adaptive A.I. fills in the gaps. This would be linked to the affinities. Have a random affinity that would play it the way it's played now. If all of that was done, the game would be better, but.... People would complain that the A.I. is too predictable, and the min/maxers would quickly find the weaknesses of each A.I. variant, and an arms race begins between the dev and the players. That's all fine and good, except for Amplitude. Why? Simple, money. ALL OF THIS COSTS MONEY AND TIME,



Since only a small fraction of players would ever get to this part of the game anyways, it's probably not worth it for them. Just look at the global achievement stats to see who has done what. Using a number of 700K (from steamspy) and the achievements shows that 3.3% of people beat it on hard or higher. 1.7% of serious or higher. 1.0% on impossible, and 0.8% of endless. All together that's less than 7%. Less that 4,900 people of the player base, but the requirement to build and test so many different A.I. variants is probably substantial.



That's my 0.02 cents. Though, having said that, Amplitude is constantly working the A.I. I think that once the game is done and all of the core mechanics are in, they'll be able to do a proper A.I. pass and finish the game. Until then, the A.I. will grow "organically" or something like that.

I seem to have sparked quite the discussion with my remark about AI.



For the record, I also accept that AI in 4X games is generally weak. Amplitude is at least taking the issues seriously and has been working on the Endless Legend AI quite a bit.

My remark was not concerned so much with how badly or well the AI plays, but with AI literally incapable of playing their faction. For example, as far as I know, the AI in Endless Legend is not capable of focusing on the specific goals of a quest, so they never receive quests from minor factions or ruins, and their main quest advances by itself. (If this has changed by now, I would hope a dev corrects me.) Similarly, I believe long ago a dev confirmed that they had originally planned a "Population Efficiency" mechanic that would result in diminishing returns if too many workers were assigned to the same FIDSI, but the AI did not handle it well, so it was scrapped (though of course, there likely were other reasons for scrapping it as well).



I'm all for differences in kind. I want factions that play differently because they have different abilities, not just because one gets +20% Industry and the other gets +20% Dust. I would love Necrophages that rely entirely on cadaver stockpiles and have to sacrifice population for the most powerful units, Cravers that leaves planets almost unusable, and Roving Clans/Pilgrims that never settle permanent cities and instead move through the territory of other empires. But at the same time, I feel that such differences in abilities would be wasted if the AI was not able to use them and instead received other bonuses to make sure it doesn't have to.

Yeah, even very adaptive AI like Supreme Commander could still be abused, and that was only with them having to manage economy and combat. Throw technology and diplomacy on top of that, and the AI will always be at a disadvantage when on equal footing. Obviously if we pump up their economy, that becomes a different story...

Eji1700 wrote:






This isn't totally accurate as a lot of games have gotten AI mods from dedicated players to better handle how players work. It doesn't matter if there's 1 million varied playstyles as long as the AI has at least one playstyle that's competitive. Maybe there's 50 ways to beat it, but right now with AI's being developed completely before launch we wind up in a situation where the players understand the game, and it's balance, vastly better than the AI because we've had more man hours with it than the devs after a month. The AI sits there and uses sub optimal outdated strategies with a resource multiplier and we use bare minimum competent tactics to win and then take all those extra resources, making the rest of the game even easier.

Shortly after release? Hardly.

Romeo wrote:
So, I know I'm going to get slapped silly for saying this, but balance is towards the bottom of my list of concerns. Normally balance happens by nerfing every strong point until everyone is the same. I realize that makes for the best competitive environment, but it makes for a very dull singleplayer environment. I prefer the Brood War system of balancing: If you can make everything overpowered, nothing becomes overpowered. Yes, having exploitable advantages for every race would likely make for a frustrating competitive multiplayer, but would still make the game much more enjoyable for the majority of players (Singleplayer and for-fun multiplayer).



I really don't want to see the game become generic and boring, like yourself. I firmly believe that identity and balance stand at opposition to one another. If they have to focus on one, I'd prefer identity.

That's not balance. Or rather it can be, but it's really really bad balance, especially for an asymmetric game. There's plenty of games with "how op can we make you feel" vibes across multiple genres which demonstrate how to properly balance something without making carbon copies (brood war is a balanced game and a good example). The goal really should be that each race has something that ONLY they can do, and do well, but that works out in the scheme of things due to generic options that can assist. Cravers/Automatons/harmony were a decent attempt at this in the first game.



Cravers could potentially do some flat out nasty teching/expansion if unrivaled, but all it takes is a decent military and just holding the line (not even winning) to slowly starve them out. This is both unique and fun ideally for both players. The cravers get to play in a style that no other race can replicate (sorta..poor numbers balacning and other issues lessen this) and fighting cravers can be solved in a way that every race has access too (general military might mixed with the fact that their inherent downside doesn't require full conquest, and in fact makes it less desirable).



Automatons allow you to do builds/expansion in a unique manner, but it doesn't much change how other players interact with you. It's a neat idea, but has moderately minor impact (you do overall the exact same thing, just in a slightly different way with one or two neat tricks).



Harmony I think is self explanatory given how heavy their race changes are.



It should be mentioned that someone can be the best "vanilla" race. Just solid all around and good at not being bad at stuff, but my issue with Endless games is that they feel balanced in exactly the way you're afraid of (everyone is very similar). Almost all the races feel like boring variations of that vanilla. I don't want someone who harvests 15% faster so I just focus food, or gets + 1 industry per pop so I just focus industry. I want a race that has hyper efficient population but capped growth, so I have to carefully pick what planets I grab or a race of bodysnatchers who can infiltrate a planet and take it over slowly without them knowing but is awful at real invasions. Unique playstyles with unique downsides, not small numbers shifts.

Ai will suck no matter what because 30k/100k/500k players will have many varied play styles, and the AI will not be able to account for them. That's why I am a proponent of a hybrid AI. A wee bit of hardcoding mixed in with adaptive AI. Then as the game grows, they go back in and tweak it here and there.

This isn't totally accurate as a lot of games have gotten AI mods from dedicated players to better handle how players work. It doesn't matter if there's 1 million varied playstyles as long as the AI has at least one playstyle that's competitive. Maybe there's 50 ways to beat it, but right now with AI's being developed completely before launch we wind up in a situation where the players understand the game, and it's balance, vastly better than the AI because we've had more man hours with it than the devs after a month. The AI sits there and uses sub optimal outdated strategies with a resource multiplier and we use bare minimum competent tactics to win and then take all those extra resources, making the rest of the game even easier.

Ai will suck no matter what because 30k/100k/500k players will have many varied play styles, and the AI will not be able to account for them. That's why I am a proponent of a hybrid AI. A wee bit of hardcoding mixed in with adaptive AI. Then as the game grows, they go back in and tweak it here and there.

So, I know I'm going to get slapped silly for saying this, but balance is towards the bottom of my list of concerns. Normally balance happens by nerfing every strong point until everyone is the same. I realize that makes for the best competitive environment, but it makes for a very dull singleplayer environment. I prefer the Brood War system of balancing: If you can make everything overpowered, nothing becomes overpowered. Yes, having exploitable advantages for every race would likely make for a frustrating competitive multiplayer, but would still make the game much more enjoyable for the majority of players (Singleplayer and for-fun multiplayer).



I really don't want to see the game become generic and boring, like yourself. I firmly believe that identity and balance stand at opposition to one another. If they have to focus on one, I'd prefer identity.

The-Cat-o-Nine-Tales wrote:
I have to agree that the balancing in Amplitude's games could be better. Just recently, I talked to a friend about the Era system of research in Endless Legend, and said to him that I could never agree to the opinion that it restricts choice, because the system would in theory allow for great freedom in picking exactly the technologies for your playstyle, but the balancing means that in the end, you always pick the same technologies. (We ended up having some disagreements over the distinction of a system and the items in it, but that's a different matter.)

In a system where expansion is king and siege defense is almost useless, nobody picks a Fortification building over the Mill Foundry. In a system where Long Range Kinetics deal almost as much as damage as long range missiles or melee kinetics, there's little incentive to pick the others, since killing your enemy quickly is crucial in a system where you can't have a designated tank.



More differences in kind would definitely help alleviate such issues. Unfortunately, many differences in kind also tend to make AI substantially more difficult to create.





On a related note, I would love to hear from a dev about their methods of balancing. Is there any system to how initial values are picked? How do you go about testing and adjusting?

I'd rather they just accept the AI will suck at launch, watch the player base for 3 weeks, and then patch both the balance, and add some playbooks for the AI.



As is most normal players don't care how good the AI is because it's "hard enough" while good players will be acing the game inside of a week (and that's without the usual custom race min/maxing). At least this way you might be able to see what better players are doing and upload some race specific playbooks rather than the % boosts we've all gotten used to hating.

I have to agree that the balancing in Amplitude's games could be better. Just recently, I talked to a friend about the Era system of research in Endless Legend, and said to him that I could never agree to the opinion that it restricts choice, because the system would in theory allow for great freedom in picking exactly the technologies for your playstyle, but the balancing means that in the end, you always pick the same technologies. (We ended up having some disagreements over the distinction of a system and the items in it, but that's a different matter.)

In a system where expansion is king and siege defense is almost useless, nobody picks a Fortification building over the Mill Foundry. In a system where Long Range Kinetics deal almost as much as damage as long range missiles or melee kinetics, there's little incentive to pick the others, since killing your enemy quickly is crucial in a system where you can't have a designated tank.



More differences in kind would definitely help alleviate such issues. Unfortunately, many differences in kind also tend to make AI substantially more difficult to create.





On a related note, I would love to hear from a dev about their methods of balancing. Is there any system to how initial values are picked? How do you go about testing and adjusting?