Marinecraft, or observations on thinking about game design and balancing
Komatik
Join Date: 2013-02-14 Member: 183057Members
Join Date: 2013-02-14 Member: 183057Members
Just some random notes from someone who stumbled onto the forums and knows jack about NS2.
I play Magic and Street Fighter, and used to be a Starcraft 2 fan until Blizzard ruined the game. Some observations, most of them through a lens of Starcraft because it's what I know and because people here are probably less emotionally invested in SC2 than in NS2:
Blizzard balances Starcraft by spreadsheet and statistics, and seems to be clueless, without a good understanding of what their game is about or how to develop it. The game has several deep issues like the way the economy works on a fundamental level (3 base cap), and the lack of meaningful terrain leading all maps to be the same because only choke points matter. The rest could as well be flat ground. Meanwhile Blizzard designs RTS units like MOBA heroes. Why is this a problem?
Because people don't experience spreadsheets. People experience individual games and gut feelings. Those may eventually converge to what the spreadsheet says, but a fun game that makes sense trumps statistical balance every day of the week.
Games also have certain fundamental things to them - for example, RTSes are about troop and infrastructure production and movement. I like to take as an example a sub-game of Starcraft. Let's call it MarineCraft. Marinecraft has Command Centers, Barracks, Gas Extractors, perhaps even the Factory and Starport. Unit-wise, we have the common worker, the Marine with a gun, a dropship and perhaps a Siege Tank that deals AoE damage. Just things that make other things and things that move and make stuff go boom.
It's a very, very simple game, but when you play it it lays done some emergent ground rules about things like troop movement - cutting off reinforcements, defender's advantage due to travel time and so on (stuff that not very surprisingly almost directly mirrors The Art of War) and the tension between focusing resources on economy, tech and army as broad strategic foci. (NS2 seems to lack the "army" focus on a superficial level - can anyone clarify if there's an equivalent?). These emergent rules create basic expectations for how the game is supposed to work - who has the advantage and what should theoretically work against some line of play and so on. People often say "there are no 'Rules of how RTS should work' so this random thing is fine" because none have been explicitly stated. These people are wrong.
As a brilliant example, the ancient Chinese board game of Go/Igo/Baduk/Weiqi has something like 3-4 rules on stone placement and that's it. The entirety of the gane's other hard and fast rules are emergent properties of this small handful of explicitly stated ones.
There was a time in SC2's life that Terran players struggled a lot against both Zerg and Protoss. Most were very, very frustrated with Terran vs. Protoss matchup simply because it didn't make sense. Terran was disadvantaged against both factions at the time, but playing vs. Zerg was apparently tolerable because Zerg didn't grossly violate the basic rules of Marinecraft. Protoss, on the other hand, did. They had all manner of silliness like frontloaded production and no travel times that frankly kill Marinecraft, bury it and spit on it's grave.
There has also been some talk about economy in terms of resource points, respawning and the like. How gross is the disparity, and what is it's nature? In SC2, for example, the Zerg economy has the ability to grow drastically faster than the Terran (and to a lesser extent Protoss) one because the Zerg are able to use all their production slots to make workers. This ability is completely broken if left unchecked, because Zerg can just reach optimal economy and let the income crush the opponent with little recourse.
It results in Terran having to pressure Zerg early lest the economy run rampant and Terran be drowned by sheer numbers. It used to be that Zerg had to invest production slots into units if the opponent did some light pressure. The mere act of making the units made the pressure worthless in that it realistically couldn't do damage, but resulted in a situation where both sides made army, econ and teched at the same time, instead of Terran doing all of that a little while Zerg maxed out on econ and then just went nuts, so the game continued.
People celebrated the matchup as the best and most glorious thing in the game - varied, relatively balanced and all around interesting. The Zerg had a distinct advantage late game, but they had to fight through a very even early and midgame to reach it, and it was reached late. Terran lategame was much weaker, but Terran had time to compensate and adequately prepare. Many great battles were fought against the Zerg leviathan.
Then Blizzard did something utterly ridiculous - they made the Queen, Zerg's only unit that doesn't require larvae - their universal production slots - better. The Queens also happened to be a production structure, provide vision via creep, be good antiair and a capable healer. They happened to have low anti-ground range and be bulky, which made them inconvenient in ground defense. Some Zerg players cried about having to make anything but workers (the inbuganity!) and thus the Queen's range was buffed. A numbers change only, surely? No. It was a qualitative difference, a difference in kind.
Suddenly, Zerg didn't have to spend larvae to defend early game pressure. They just made Queens and Drones, and were safe from anything but the most severe all-in attacks the Terran could make. Naturally, the economy went nuts and Zerg had more resources in the midgame, completely killing Terran midgame potential. This allowed them to tech faster because less army was needed. 17-minute Hive timings being greedy were replaced by 12-minute ones being utterly safe. Given Terran's need to have lots of expensive infrastructure to try holding their own in the Zerg-favoured late game, you can imagine what happened.
The above is important to note because most Zerg units are fine. Blizzard simply broke the economy, and now there's too much stuff to handle. It's ridiculously important to realize what your game is about. Blizzard apparently didn't grok that the beating heart of Starcraft is the economy, screwed it up even worse than it was and ruined their game.
Another interesting tidbit is that some people genuinely like difficulty - they love stuff being hard to do in and of itself. The one fault in these people is that they typically want everything to be difficult to do for others, too, even if these other people may not enjoy it. Street Fighter example, people wanting execution requirements to be sky high across the board.
Starcraft example, Terran having to micro more than the Zerg, all other things being equal. Zerg may not have needed micro, but they used to need to outplay the Terran opponent in those other fields to win, and the same was true of a decently-microing Terran player. If he didn't outplay the Zerg player, he wouldn't win. In Magic, some decks are just a lot harder to play than others while not being necessarily more powerful. But there's an enjoyment to doing that, and I like the possibility of doing that while not disadvantaging myself in absolute terms. Proper play may be more difficult, but not inordinately so, and can compete well. Same with Street Fighter. There's a ton of characters, some weaker, some stronger. Some harder to play, some easier. As long as there's choice, it's fine. When there isn't, well. Not so much.
Which leads to the next points - imbalance makes itself felt very differently in different kinds of games. A 6-4 matchup in a fighting game is not a very big deal - you feel it, sure, but it's not torture. 6-4 in Starcraft can feel completely hopeless in a way the SF match never could, unless the SF matchup was something like 7-3 or worse - something where you need luck or an opponent screwup to hope to win.
It's more imperative to get SC2 with few factions and matchups balanced relatively tightly than Street Fighter with 40 characters and less severe-feeling imbalance. One reason this is the case is because SF doesn't have much of an economy and has reads - few options are universally right, and there's not much of a slippery slope. Numbers can't drown you most of the time.
So, given that NS2 is basically RTS + a shooter where player skill matters a ton: get the economy and RTS-level Marinecraft equivalent about right, and the rest follows to a relatively great extent. If you feel you have to outplay the opponent a lot and success isn't logically rewarded - say the Terran push-out that gets deflected by Zerg's impenetrable-for-free defences despite an attack being the logical response to greedy play. It makes you just not care. If, on the other hand, it's harder for you to push and harass, but the rewards for success are at least the same as for the opponent, the game is skewed but makes sense. You can actually try stuff. You get properly rewarded for proper action. This is good. If the econ is fine but higher lifeforms are monsters, for example, it makes decent flavour sense and means they can still be individually outplayed. If the numbers mean outengaging the Alien players means jack, well...
Lastly, how about freeing up the matchups? There's a certain attraction to "being the Alien player" or "the great Marine team" that gets diluted by competitive games being both sides as both factions. I'd want to see Alien and Marine mirrors. It'd also allow more specialization and thus higher skill.
[/rant]
I play Magic and Street Fighter, and used to be a Starcraft 2 fan until Blizzard ruined the game. Some observations, most of them through a lens of Starcraft because it's what I know and because people here are probably less emotionally invested in SC2 than in NS2:
Blizzard balances Starcraft by spreadsheet and statistics, and seems to be clueless, without a good understanding of what their game is about or how to develop it. The game has several deep issues like the way the economy works on a fundamental level (3 base cap), and the lack of meaningful terrain leading all maps to be the same because only choke points matter. The rest could as well be flat ground. Meanwhile Blizzard designs RTS units like MOBA heroes. Why is this a problem?
Because people don't experience spreadsheets. People experience individual games and gut feelings. Those may eventually converge to what the spreadsheet says, but a fun game that makes sense trumps statistical balance every day of the week.
Games also have certain fundamental things to them - for example, RTSes are about troop and infrastructure production and movement. I like to take as an example a sub-game of Starcraft. Let's call it MarineCraft. Marinecraft has Command Centers, Barracks, Gas Extractors, perhaps even the Factory and Starport. Unit-wise, we have the common worker, the Marine with a gun, a dropship and perhaps a Siege Tank that deals AoE damage. Just things that make other things and things that move and make stuff go boom.
It's a very, very simple game, but when you play it it lays done some emergent ground rules about things like troop movement - cutting off reinforcements, defender's advantage due to travel time and so on (stuff that not very surprisingly almost directly mirrors The Art of War) and the tension between focusing resources on economy, tech and army as broad strategic foci. (NS2 seems to lack the "army" focus on a superficial level - can anyone clarify if there's an equivalent?). These emergent rules create basic expectations for how the game is supposed to work - who has the advantage and what should theoretically work against some line of play and so on. People often say "there are no 'Rules of how RTS should work' so this random thing is fine" because none have been explicitly stated. These people are wrong.
As a brilliant example, the ancient Chinese board game of Go/Igo/Baduk/Weiqi has something like 3-4 rules on stone placement and that's it. The entirety of the gane's other hard and fast rules are emergent properties of this small handful of explicitly stated ones.
There was a time in SC2's life that Terran players struggled a lot against both Zerg and Protoss. Most were very, very frustrated with Terran vs. Protoss matchup simply because it didn't make sense. Terran was disadvantaged against both factions at the time, but playing vs. Zerg was apparently tolerable because Zerg didn't grossly violate the basic rules of Marinecraft. Protoss, on the other hand, did. They had all manner of silliness like frontloaded production and no travel times that frankly kill Marinecraft, bury it and spit on it's grave.
There has also been some talk about economy in terms of resource points, respawning and the like. How gross is the disparity, and what is it's nature? In SC2, for example, the Zerg economy has the ability to grow drastically faster than the Terran (and to a lesser extent Protoss) one because the Zerg are able to use all their production slots to make workers. This ability is completely broken if left unchecked, because Zerg can just reach optimal economy and let the income crush the opponent with little recourse.
It results in Terran having to pressure Zerg early lest the economy run rampant and Terran be drowned by sheer numbers. It used to be that Zerg had to invest production slots into units if the opponent did some light pressure. The mere act of making the units made the pressure worthless in that it realistically couldn't do damage, but resulted in a situation where both sides made army, econ and teched at the same time, instead of Terran doing all of that a little while Zerg maxed out on econ and then just went nuts, so the game continued.
People celebrated the matchup as the best and most glorious thing in the game - varied, relatively balanced and all around interesting. The Zerg had a distinct advantage late game, but they had to fight through a very even early and midgame to reach it, and it was reached late. Terran lategame was much weaker, but Terran had time to compensate and adequately prepare. Many great battles were fought against the Zerg leviathan.
Then Blizzard did something utterly ridiculous - they made the Queen, Zerg's only unit that doesn't require larvae - their universal production slots - better. The Queens also happened to be a production structure, provide vision via creep, be good antiair and a capable healer. They happened to have low anti-ground range and be bulky, which made them inconvenient in ground defense. Some Zerg players cried about having to make anything but workers (the inbuganity!) and thus the Queen's range was buffed. A numbers change only, surely? No. It was a qualitative difference, a difference in kind.
Suddenly, Zerg didn't have to spend larvae to defend early game pressure. They just made Queens and Drones, and were safe from anything but the most severe all-in attacks the Terran could make. Naturally, the economy went nuts and Zerg had more resources in the midgame, completely killing Terran midgame potential. This allowed them to tech faster because less army was needed. 17-minute Hive timings being greedy were replaced by 12-minute ones being utterly safe. Given Terran's need to have lots of expensive infrastructure to try holding their own in the Zerg-favoured late game, you can imagine what happened.
The above is important to note because most Zerg units are fine. Blizzard simply broke the economy, and now there's too much stuff to handle. It's ridiculously important to realize what your game is about. Blizzard apparently didn't grok that the beating heart of Starcraft is the economy, screwed it up even worse than it was and ruined their game.
Another interesting tidbit is that some people genuinely like difficulty - they love stuff being hard to do in and of itself. The one fault in these people is that they typically want everything to be difficult to do for others, too, even if these other people may not enjoy it. Street Fighter example, people wanting execution requirements to be sky high across the board.
Starcraft example, Terran having to micro more than the Zerg, all other things being equal. Zerg may not have needed micro, but they used to need to outplay the Terran opponent in those other fields to win, and the same was true of a decently-microing Terran player. If he didn't outplay the Zerg player, he wouldn't win. In Magic, some decks are just a lot harder to play than others while not being necessarily more powerful. But there's an enjoyment to doing that, and I like the possibility of doing that while not disadvantaging myself in absolute terms. Proper play may be more difficult, but not inordinately so, and can compete well. Same with Street Fighter. There's a ton of characters, some weaker, some stronger. Some harder to play, some easier. As long as there's choice, it's fine. When there isn't, well. Not so much.
Which leads to the next points - imbalance makes itself felt very differently in different kinds of games. A 6-4 matchup in a fighting game is not a very big deal - you feel it, sure, but it's not torture. 6-4 in Starcraft can feel completely hopeless in a way the SF match never could, unless the SF matchup was something like 7-3 or worse - something where you need luck or an opponent screwup to hope to win.
It's more imperative to get SC2 with few factions and matchups balanced relatively tightly than Street Fighter with 40 characters and less severe-feeling imbalance. One reason this is the case is because SF doesn't have much of an economy and has reads - few options are universally right, and there's not much of a slippery slope. Numbers can't drown you most of the time.
So, given that NS2 is basically RTS + a shooter where player skill matters a ton: get the economy and RTS-level Marinecraft equivalent about right, and the rest follows to a relatively great extent. If you feel you have to outplay the opponent a lot and success isn't logically rewarded - say the Terran push-out that gets deflected by Zerg's impenetrable-for-free defences despite an attack being the logical response to greedy play. It makes you just not care. If, on the other hand, it's harder for you to push and harass, but the rewards for success are at least the same as for the opponent, the game is skewed but makes sense. You can actually try stuff. You get properly rewarded for proper action. This is good. If the econ is fine but higher lifeforms are monsters, for example, it makes decent flavour sense and means they can still be individually outplayed. If the numbers mean outengaging the Alien players means jack, well...
Lastly, how about freeing up the matchups? There's a certain attraction to "being the Alien player" or "the great Marine team" that gets diluted by competitive games being both sides as both factions. I'd want to see Alien and Marine mirrors. It'd also allow more specialization and thus higher skill.
[/rant]
Comments
NS2 works at constant supply, so army is mainly constituted of equipment (weapons, jetpacks and exo) on the marine side and lifeforms on the alien side. There's little tension between army and tech-economy because there's a special resource pool for army that you can't trade for tech or economy (personal res).
Personally I pretty much dislike creep in sc2, and find interesting that it had the same kind of problems in ns2.
- We had bonus and malus on infestation during the beta, it makes the bonus team camp on creep and the malus one avoid engaging on creep and trivialize strategy. Bonus always feels like a malus when you exit creep.
- We had infestation giving vision during the beta. It was very bad. It trivialize scouting and positioning.
- Makes a lot of boring army vs static structure fights. This specially bad in ns2 because it's actually real humans that have to kill the structures, but it's also quite boring in sc2.
- Allows to spam static defenses, more boring units vs structure fights.
- Doesn't have any depth, the topology of the infestation is dictated by the map structure, in NS2 it also hurt map design as every path to every natural RTs from each alien starting position need (or used to?) to be relatively flat to allow cyst placement.
This is a significant point since this is why people tend to gravitate to the larger 24 player servers. It's also why balance on those servers is better.
Here is a simple example. On NS2stats I set up the following filter. I selected only build 239, only public games, and only KingKahuna servers 1-3. I also went to the mod section and UNselected combat mode to make sure none of those games influence the totals. Of the ~900 games played in build 239 on those servers, the aliens won ~52%.
52%
Is it any wonder why people like playing on larger servers?
However, the other important factor is that larger servers lower the significance of skill differences in players. You'll have a much better chance to have a even spread of people from all skill levels when you have 12 players a side as compared to if you only have 6 players a side.
Marinecraft thinking isn't so much about balancing the game as a whole than making sure the people who make the game try to spot the unwritten rules of the game and ensure they aren't broken too egregiously. The other part is having an economy that makes sense. In SC2, for example, the endgame in Terran vs. Zerg is lots of clunky infrastructure making narrow counters vs. less expensive more flexible infrastructure making pure threats. This does not make sense on a qualitative level. Just one change - say, Terran being able to make threats too - could make it OK.
The rest is comparatively irrelevant, and steers away from the strictness of RTS, and the dominance of pure economy into the territory of player skill, where it might actually be, ammo permitting, possible for a Marine to duel an Onos and come out the victor.
The point is assuring that if you can do that, the rewards are proper. You can make the Onos into a god of war, but have it respect the rules of war.
Marines scale much better with player sizes in general compared to Aliens. Spawning scales better, DPS scales better, building scales better.
I don't doubt that the game is more balanced (at least in win-rate terms) with larger servers, but I think it's more from the scaling factors than some sort of better skill balance (1 or 2 really good players in a large server will still throw the game to their teams favor).
I don't think that's so different, tempest dragoon can kill infinite number of marines, a normal dragoon, not so much. There's huge difference between units effectiveness depending on who controls it. There also difference between unit effectiveness in a single game as the player will not dedicate the same amount of apm to each, sure units are a bit more homogeneous and reliable but I don't think that a very important difference.
I could also play a thousand game against a good player and win none, rts side might not be strong enough even in rts games.
I don't think that marines scale better overall, but that they scale better in the component that is one of the most important with regards to balance; spawning.
I would say they did. When you remove the one and only actual limiter to a mechanic that makes economy broken in a game about economy, in the process turning a balanced matchup universally praised as the best or second best in the game into something dreadfully dull (to the point where most people preferred to watch TvP which was in a horrible state at the time) and remove most indications of player skill in it, I'd say you've ruined something.
Add to that Blizzard's insane idea of asymmetric balancing in time instead of just method - that is, they deliberately make different factions just plain stronger at different points in time, which leads to very scripted-feeling gameplay. The steps of a game of SC2: WoL are so well defined most of the time you don't even need to watch the match. This when the clearly proper way to leave the determination of strength at a point in time to be decided on the build level instead of the faction level, and just making the methods of achieving ends asymmetric.
In short, Blizzard did something even worse than making their game unnecessarily more imbalanced - they made it less fun, on average, for both the player and the spectator. I don't see why I should mince my words - much of the stuff is plain as day, and I don't hold the teamliquid culture of hushing any notion that imbalance might exist - I've experienced and enjoyed civil, vocal and harshly to-the-point discussion on matters of balance and design in the fighting game and Magic communities, and that frankness is one of the most refreshing aspects of them.
I don't understand you. Are you saying that hard counters are required for an RTS? If so, I have to disagree. I prefer an RTS with almost no hard counters and a wide variety of soft counters.
The best way to solve this problem imo is to introduce hard counters, such that I am assured that no matter how skilled the zergling player is, he will never beat a Colossus. However, FPS players don't like this because they prefer the outcome of a battle to depend at least partially on skill, such that there is always a chance that the zergling can overpower the Colossus.
I kind-of agree with you. I think you're right, that a good FPS team will beat a bad FPS team no matter how well they execute on the RTS aspects. HOWEVER:
a) I don't think it's absolute. Say we're rating FPS skill on a one-to-five scale. A five team will beat a three team, but a four team could lose to a three team if the three team does a much better job at the strategic elements.
b) Between teams of roughly equal FPS skill (less than a full point difference) then strategic play will determine the outcome.
c) That's not so bad.
d) Better matchmaking would make for more games where better strategic play was important.
I'd rather see better matchmaking than hard counters.
I don't know for sure that SC2 is ruined though, so I'll take that with a grain of salt. I haven't played enough SC2 to make any comments one way or another. Marine rushes are still fun at my level, and I still win half of the time I guess.
Not sure how this applies to NS2. Were you providing a cautionary tale to UWE? ie, Don't make a Queen unit for NS2?
Too late, we have gorges!
It's not wrong, and it's a good thing. That's why the economy is the prime focus - it's ruled by hard numbers a lot more than actual engagements, where it's possible for a lone Marine to kill an Onos if he's really good. As it should be. Same stuff in SC2 where Banelings counter Marines, but MarineKing's Marines go even or even counter Banelings. Same in SF where you can win a near-hopeless fight with Zangief against Sagat because you play well. But that needs to be enabled. What's the point of being able to duel Onos (Onoi?) to death if they respawn faster than you do? Or, if a fully equipped Marine and the better Onos cost the same but the alien just gains more resources and you again drown under pure numbers in a way skill can barely hope to compensate?
Of course, this still assumes there's some asymmetry - like, say, with Ryu and Akuma in Street Fighter. Very similar characters, but Akuma is undoubtedly the better one overall in every single game that I know of. Yet it's not a straight upgrade, there's a bunch of things Ryu can't do but Akuma can, but also some that Ryu can and Akuma can't. Edges the Ryu player can exploit to overcome adversity.
You winning half the time is Blizzard's matchmaking doing it's job properly, nothing else. You could play the most hideously gimped pile of roach poop and you'd still win half the time. The ruination stuff largely applies to Master League/Professional level games, though there's a separate issue of a huge learning curve rise at Diamond/Low Master for Terran.
The point was part cautionary tale, part just simple advice. No use in repeating past mistakes, when a proper holistic frame of thinking helps make a game that's more fun, where imbalance should be less severe, and most importantly feel less bullshitty. But sure, if they want one side to have a clearly better economy, make absolutely goddamn sure as mustard-with-chili that the economy can be pressured by the other side. Otherwise you're in for some miserable time.
Also, in NS2 economy and engagements are strongly related, such that winning engagements is a prerequisite to expanding your economy. In SC2, it would be like having the only way to increase your crystal and vespene gas flow is by capturing more expansions, instead of building more workers.
I stopped reading there.