I read about this in a book a while back, and some digging around the net turned up a GameDev.net article about it.

The short version is that the intuitive way to approach balance is to make everything "fair". If something is overpowered, nerf it. If it is underpowered, buff it. The alternative presented here is that something is imbalanced when you have no other real choices, and that maintaining balance is about keeping the number of real choices high.

An example of this would be whether or not to use JP/HMG combos back in the days of NS 1.04. The intuitive approach is to say infinite jetpack fuel & HMGs were imbalanced because nothing could beat it. The alternative is to say it was imbalanced because marines had no other real choices when picking the best way to win, and that aliens had no choices at all on how to fight back.

What comes about through this subtle change in thinking is that there are now many ways to fix the imbalance apart from nerfing the JP/HMG combo. Now you can start thinking about how to give aliens a viable counter. You can think about why other marine tactics seem so unattractive in comparison. The NS devs fixed this issue by fixing the bug that caused your jetpack refuel time to be affected by your framerate, and then set the jetpack to recharge too slowly to allow long flights. Perhaps there were other ways to address this and similar issues?

What about the eternal debate on the imbalance created by player skill? Is it simply that skilled players can too easily dominate an entire enemy team? Or is it rather that the enemy team has been given no choices by the developers when it comes to fighting back? How else could we approach the issue of player imbalance without attempting to "nerf" skill?

Good post - IMHO, I like the 'More Choices' approach better. But here is an interesting challenge, how do you increase both sides options without necessarily increasing UWE's workload of models, animations, and so forth? What kind of approach allows a lot of choice, without too much hard labor?

What popped to mind for myself would be the use of RPG style character building when it comes to marines and kharaa - in fact, I could have sworn there was a thread where it was suggested the Kharaa could tweak themselves within a range with regards to their chamber evolving upgrades. What if marines did something similar? Small percentage increases to parts of the playing experience which is guided by the different tech branches a Commander researches; so instead of NS1's weapon research level 1 which gives +% damage, what if in NS2, marines were given a choice of putting a point into an area for their character, in the case of weapons research, +% damage of specific weapon or clip size of specific weapon or reload time of specific weapon or effective range or guided projectile ... and so on. That would allow for greater variation but not necessarily (could maybe add a color effect to the muzzle flash or whatever is easy breezy) greater number of animations/models/textures or what have you, right? I imagine with enough options, there really would be no right choice, just choice - though that would probably not stop some know-it-alls from telling you their opinion of what would be the best stuff to put points into, I would just hope there would be differing opinions, thus displaying greater variation of the right tactics.

Thanks again for the good post, it sparked that idea for me.

About the above idea, there's the consideration that (some? many?) players need to know exactly how many bullets it takes to kill someone, exactly how many bites it takes to kill someone, etc. If players can tweak some of the values then it'll create a degree of uncertainty. Now I don't see that as necessarily bad, but others will.
(I saw this attitude presented in response to the 'buff/debuff' suggestion for new/advanced players respectively.)

Trust me if they added that, players will make a damage / armor chart fairly fast, displaying and showing players exactly how much damage, amount of hits to kill, and armor protection. NS already has one made, why wouldn't they for NS2?

How will you tell whether a player has one upgrade instead of another?

I suppose it'd work the same way as the Kharaa though.

There's already been like a bajillion topics on game balance and how to fix it (which ironically seem to repeat the same ideas over and over).. By 'balancing' the game so that lesser skilled players will be at par with the higher skilled ones will only hurt the more experienced gamers and those who sacrifice more time and effort.

I think It's safe to say that there is no way of really balancing it at this point in the games development.

The RPG idea would create too much randomness in the game as Harimau somewhat stated. Knowing how many hits/bites/etc you needed to kill something was crucial in NS1 (parasite kills ftw) and I'm sure It'd be the same in NS2.

This is why I created a new topic instead of replying to one of the other balance threads. Many, if not all of them, seemed to revolving around the notion of making the game arbitrarily "fair", which does not lead to a truly balanced or rewarding game.

Keep in mind that balancing via choices doesn't have to mean RPG-like character development. A player's choice can be as simple as which path through a map to take, or what part of an enemy to shoot at. As long as there is more than one meaningful decision a player can make to further his own goals, then there is at least some balance. When a squad of normally-played vanilla marines focus fire on a 2-hive onos to take it down faster, they have chosen to do so from a variety of other options, and in doing so may well have saved their own lives.

When that same squad of marines comes up against an extremely well-played 2-hive fade, it doesn't matter what they shoot at, or even if they shoot at all, or even if they run away. No matter what they do at that point, they die. No matter what they do for the rest of the game, they lose. At this point, the marine team no longer have any meaningful choices to make because they all lead to the same result. All they can do is pray for someone who can aim really well at very fast moving targets to join their team, as nothing in their arsenal helps them deal with that situation if they can't do it on their own.

I personally prefer the random aspect of games, it makes it more fun, varietal, and not so... predictable. I understand from a strategic stand point it makes it harder, but why is harder frown upon? If anything should make you strive even that much more. Such as scaling a mountain instead of some hill in your backyard. But anyways for my own style, I don't always like to know every single possible thing that can, will or should happen. I rather take it as it comes and deal with it the best I can with what I have, and what I know at that given situation. Makes the game a lot more fun then executing predictable commands and outcomes. Again.. a matter of opinion.

Probably also worth pointing out that knowing exactly how many hits are required to kill someone is only useful in NS because it makes a difference. That's easy enough to change when you're making a whole new game.

Ok, I'm not targeting anyone in particular here and this might sound rather straightforward and/or narrow minded, but from my own personal experience I find that the reason people complain about 'balance' is because they don't know how to play the game properly. You just gave two perfect examples of situations where it seems like nothing can be done, but in reality theres actually perfect counter measures that can be taken.

If there is a very good fade player and a small group of marines (say 2-3) all those marines have to do is hold their ground and wait for the fade to attack them. Yes, this may seem hard to do for someone who wants to run around and kill things, but it works. Get the fade in a small enclosed space and It's boned. For some reason players seem to assume that 'whatever I do I can't win' or 'the fade has too much hp, theres no point shooting at it'. Things like that make the team lose.

The other scenario being 'WTB pubgod' - again, if the players are thinking that of course they will lose. All they need to do is sneak a marine or two near a hive and build a PG or a siege. Then again, if the game has been going for something like 20mins+ and they have been losing the whole time It's their own fault for playing badly in the beginning.

The marine and kharaa teams are very different not only in looks and game play but also in strategies.. When people figure out how they should be playing, only then will balance become something of the past (hence why It's too early to be solving the problem at this point in time). There is nothing you can do about balancing the skill gap between players anyway.. It happens in every aspect of life. Work, sport, recreation, games. Unfortunately there will always be someone better than you and there's nothing you can do about it.

Unless you introduce random damage to NS it will always make a difference. In a multiplayer game you need to know how many hits it will take to kill someone. If they can trake more than normal, there needs to be some way of telling how (e.g. for Hive 2 cara you can just count how many Hives are active to get an indication of how many shots).

Now, players won't always know precisely how many hits an enemy will take, but after playing for a while they will get a feel of how long they need to focus fire on an enemy to take it down. If you start messing with those values without giving the player some way of seeing to what extent they have changed, the game won't be as fun. You'll get situations like a 200HP Fade blinking in to kill a single Marine building a PG and have the Fade die because the Marine had upgraded bullets or upgraded armour. Normally this would be an acceptable gambit on the part of the Fade (obviously assuming that the Marine hasn't got leet aiming skillz), but because of the unpredictable element a display of skill and awareness is not rewarded but instead punished.

This post helped spawn another idea for me, what about ways to paint targets, something with just a bit more emphasis than parasite but something that pretty much screams to a team "Focus your fire here!" so that a team works together better ... hmm, something like, where the more team mates you have targeting a particular player, the more the user interface shows it, either by Kharaa mind stimuli or Marine HUD - basicly the game lets us know actively when we are assisting our team better, than just finding out about it later in the event of a target's death ala TF2.

On the subject of the character tweaking aka RPG elements, why are you all assuming the worst possible stat boosts? It doesn't even have to be related to weapon or armor if need be, that is just one possible avenue for improvement, there are lots of other places in a game experience that can be incrementally improved, that is proven just by the way chambers work, which make improvements to HUD, speed, stealth, and so on. It doesn't have to be game breaking, in fact, I thought I was specifically encouraging that it not be game breaking, only add variety so it allows a diversity of ways in which to assist one's team without taxing the resources of the Devs.

Also, why all the talk about being able to chart and predict exactly how much of X to kill Y. Seriously? These days when I play NS1, there is pretty much only one path to take, doing anything else is seen as "noob" and brutally pointed out as such, often with much swearing and F4'ing, and that is very much tied into that other players can predict exactly how that will effect the outcome of the match. And you want that for NS2? Its not like I was suggesting the infuriating criticals of TF2, I was suggesting the research side of the game being tied even further into giving players small incremental enhancements that they could gear towards their play style. Not everyone plays the same way or even wants to, but if there is one common theme in Natural Selection, its that we love playing together as a team, and good team's have roles that player's adopt to help out - that makes me think of RPGs, but it doesn't have to break anything. In fact, a little unpredictability, a range of options, IMHO can only improve the game's playing experience if my current NS1 experiences are anything to go by.

Improved options also extend to the map themselves, I really hope maps are a lot more dynamic this time around, so that its not just a choice of path A, B, or C but an additional choice of scrapping A for a D, or changing C into an E. Utilization of the enviroment could be very key to improved chances of working together as a team, which I think is a goal even better than just winning. Often in games of NS1 or TF2, both sides seem to appreciate it when there is a challenge of an opponent who strives to attack, defend and counter, because even if they lose, they go, we did really well together as a team. That's encouraging and speaks more to a balanced game than any amount of predictability.

Dynamic > Predictable.

Even in NS1 you dont always know EXACTLY how many bullets it will take all the time. Take a alien that has cara vr one that has regen or redemp. Sometimes it will take more or less to kill em. If you dont know how much damage someone can take or dish, then that will just make people a little more conservative and that fade might not rush that PG (just incase they got that upgrade). Also it would make higher level aliens a bit more of a risk (do to the more uncertan nature of the game). Now one could argue thats a bad thing or a good thing, but it is diffrent.

Randomization makes strategic gameplay harder the unfair way. You need some certain relatively certain facts before you can really start creating tactics. Of course risk control is important, but the risk should come from your enemy actions, not from some random numbers created on server. For example leaving your base open for baserush is a strategy I can live with. I take the risk that the aliens see the opportunity and rush the base. If they do, I'm in trouble, if they don't I gain some advantage on the field. Then again, a skulk crit hitting my baseguard out isn't fair since I was prepared for the attack, saw it coming and still the luck factor allowed them to win with a predictable plan. Make the enemy actions and you closely related and don't make scouting too effective and you've got nice variable match going on even if you don't have any random stuff going on.

Starcraft has absolutely brilliant risk control and strategy system based on enemy actions. I recommend checking some english commentaries on pro matches on youtube if anyone wants to see how the variable gameplay goes without randomization.

I found your post very interesting. If options exist, but aren't widely known, then perhaps it could be because they are not immediately obvious. In this case, it could well be that the fault lies not with imbalance, but with the training & hinting mechanisms in the game.

That being said, whether you really have no options, or you simply think you don't have any, you still end up with the same ultimate result: The player stops having an interactive experience, they get bored, and they stop playing your game. Balance and fairness aside, this is very bad for any multiplayer game.


This phrase right here is a perfect example of extreme game imbalance, despite having nothing to do with fairness.

I agree, to an extent: Too much randomess kills any sense of influence over the game's outcome. At this point all choices become irrelevant, and we're full circle back into an imbalanced situation. Check out the new Mario Kart on the Wii for an example of this.

That being said, I'm not convinced random effects are completely bad.

Er, who said anything about randomization? Just curious, because the idea I put forward was about enemy or friendly making decisions about their character, similar to picking out their equipment, just extending that same idea to include expanded tech/bio trees for the Marines and Kharaa.

What there seems to be the problem with it is predictability, which I take exception to because a predictable game is a dull game. A good opponent is anything but predictable, both sides should have the chances at a large number of options, in map control & manipulation (infestations, buildings, resources, sub goals, welds, etc) and in character development (evolves, research, equipment loadouts, etc). Greater range, not greater randomization.

Personally, I can't stand randomization, its a bane of my character's existence in things like TF2 criticals or D&D low levels, I do everything I can to avoid it. Give me Dynamic game play though, that kicks ass, I feel like my game is only limited by my own creativity with the system then.

Seconded. The last ~6 posts or so have been in regards to randomisation (a la TF2?). But that isn't what we were getting at.

What I initially introduced was the idea of 'unpredictability' and how certain gamers (eg. Crispy, Bacillus) would be against it; despite there actually still being a level of predictability associated with CanadianWolverine's idea: because rather than a certain number of bites/hits, instead you'd have a range; and you'd probably play around the maximum number of hits required to kill someone based on the highest possible armour/health values they could have. In that sense there still is predictability, but it's not a 'this will die in exactly this many hits' situation.

Initially, the two conflicting viewpoints were possessed by those who supported an unpredictable/variable/dynamic system, and those who supported a predictable/certain/static system - an argument which I thought was good, and keeping on topic. (Admittedly I'm leaning more towards the former.) But then we got off-track with the 'randomisation' argument.

Also, I like CanadianWolverine's idea of the HUD indicating who was being targeted by your fellow teammates, so that you could focus fire on specific enemies.

He's probably talking about things that are redundant in the game, like OCs (without a supporting Gorge) before JPs, or maybe he's complaining about 'opening moves'.

The OCs case is basically a case of the game being broken and OCs not being worth the res spent on them. In short they need to be redesigned.

If it's what order you put your buildings down in base (i.e. the equivalent of chess's 'opening move'), then there are standards. Building an IP and an Armoury first is a standard opening move if you don't plan to relocate. The IP respawns your players and the armoury is vital to unlocking new technology.

But after this, there are plenty of options. You could drop an early shotgun, depending on the map you might drop an early welder to block off some vents nearby the Marine Start. Often people will research Hand Grenades or get the lengthy Advanced Armoury research going; the first allowing you to keep your resource towers alive much more easily, the latter giving you a mid-game advantage versus higher lifeforms (Adv Armoury leading to both HMG/GL and HA/JP via the Proto Lab).

The point is without clear examples you (the Original Poster) don't make a good case.

[Edit] I can't believe Firefox kept this entire post cached for me while the forums were down for maintenance! \o/

If you have no way to tell the range of possibilities then it is unpredictable. That is what I'm against.

E.g. by looking at the scoreboard and keeping track of a players kills/deaths, you can calculate when a player who has stayed Skulk for the whole round will be ready to drop a Hive, because it can be deduced from the scoreboard. What you can't deduce is if a player has been using personal resources/experience to statbuild their character and you have no way of being able to tell how much of a threat they pose, as per this example.



Now, there are instances where you can call someone's bluff. E.g. sometimes if I have a shotgun and a Fade is doing the rounds I'll keep my pistol out so the Fade doesn't know if I have a Shotgun or an LMG. They might assume that because they don't see or hear any Shotguns they can get away with rushing a few vanilla marines for a kill or 2. The point is if you just keep your pistol out you are more vulnerable to attack from any alien that takes you by surprise, so there is risk involved. There's no risk involved with invisible statbuilding, so it just becomes unfair.


---
One of your crucial arguments is that there should be more possible ways to counter; more choice. But if you dilute the game too much with too much choice you actually reduce counters because any commander-led decision on what to give their team may nerf some players who have statted in one direction and buff others who statted in the direction the commander has anticipated. Anyone unlocking a teamstat needs to know that it is taking the team in the right direction, they need to know that it is a hard counter that is in their interests. If you only have soft counters that don't really have a huge effect then it makes the FPS side more important and brings NS closer to deathmatch, reducing strategy.

---
Finally, another counter-argument to the character-building idea is that it distances good and bad players because the unlockables will help the best get better, while the less able depend on a vanilla or lower-tech loadout. Upgrades that benefit the whole team are better because they keep less able players on a more equal footing with their teammates. We already see how stacking, via RFK (res for kills), allows one team to get the upper hand because of more team resources for the commander to use on tech or more Fades who are less likely to die. Statbuilding is a bad idea for NS imo because it reduces teammate interdependance and encourages/rewards ramboing.

Nice post Crispy, though some of your responses to me give me the impression we're talking about the similar things and just labelling them differently.

That right there is exactly the message I'm trying to convey. Redundancy is imbalance. Under the philosophy of balancing via choice, OCs are not in balance with the rest of the game because it doesn't make sense to use them.

This is perfectly fine. Your options are "set up a base" or "relocate". Relocating is a higher risk but can lead to a higher reward. I would say that these two options are in balance with each other.

This is a good example of well-balanced gameplay. Plenty of meaningful choices to make.

There don't need to be a lot of choices, you just need to be able to make choices that matter. These choices can be as simple as fighting or fleeing. In the marine vs fade example, a single marine can't out-gun or out-run a fade, so in that situation the marine has no decisions to make.

My argument is that you can more accurately detect and fix game imbalance by looking at these situations and deciding what to keep as is and what to change. I am trying to argue for the use of a particular philosophy when tackling the issue of game imbalance in NS2, and I am doing this by drawing on examples from NS1. My goal is not to argue for or against imbalance in NS1, but that is hard to avoid.

Oh well... I might have misinterpreted the "random aspects". I just pointed out that I disagree with the fact that being unpredictable via random aspects is good for the tactics. I must admit it was an unnecessarily long post for a little point.

As for the skill and its counters:

Another long post with a little bit of ideas in the end coming up.

I quite agree that the player skill is too dominat in public games, but then again I feel people are adapting very little.

So let's say I see during the first minute that my team is going to be trouble vs the 2nd hive fades (btw, any team is, even good aimers have difficulty pushing against 2 hives unless they've got protolab in use). I go for the early game build involving a lot of shotguns and welders, marines packed together, early upgrades and such. Make sure people move quickly and push the early game advantage to the max. The aim is to take out enemy enemy nodes, effectively delaying the fades and the 2nd hive, before the top players can do much harm as a skulk. Then again, don't allow the 2nd hive to go up. If it seems its going up, make another plan that makes the fading routine a lot more difficult --> Secure the 3rd hive location and get ha. With ha you should be pretty comfortable fighting off the fades at least.

If you're able to play that simple plan somewhat well, you should never be facing those dreaded 2nd hive fades in a bad situation. At worst you should be fighting them off on some of your phase gates where you've got a fresh supply of marines and then again while pushing to the hives with heavy armor.

Obiviously that takes some cordination and tactical knowledge of you and your team, but since you're getting outskilled, you have beat them in some ways of gameplay.

The problem is that people can't pull off things like that. The game is too confusing on public. Now, if you could get people to the level where they know how to stick together, weld each other, plant mines on phase gates and follow the basic orders, we could start working out on effective tactics. As long as people just wander off, blast at some critters and die without any greater meaning, its going to be difficult to counter those skilled players.

I guess the bottom lines are

1. Make the game less confusing. I don't know how big percentage of players is really willing to follow the waypoints, but right now most people are too confused to even notice them, not to speak of forming a group and moving to the waypoint quickly. I guess better autohelp, better means of communication and such would be big use. Pretty much anything goes, as long as you can get people to understand the basics of the game.

2. Give people more information about the game. After you've got some basic knowledge combined with the fact that your communication and ideas might change the course of the game, its interesting to start figuring things out by yourself. There could be some actual development in gameplay, instead of the people doing their routine without really being part of the bigger picture. This could be done for example by some end-game statistics. The fact that aliens got 900 res out of the nodes in 10 minutes might lead to the conclusion that next game you must be playing attention to the alien resoucres.

At that point you've hopefully given people the means to combat higher skilled players and thus bringing some balance without simply 'nerfing' the skill.

I think that the bottom line boils down to content. How much content are the developers willing to put into a game?

Generally the formula is

content = weapons/equipment/alienlifeforms/toilet seats etc...
more content = better game
better content = better game

There is a choice there that you aren't taking into account: to stick together or to spread out. Like what someone said earlier about keeping someone at base to build or having the commander do it so that more marines can be on the field at once. When a Commander drops an RT for a lone Marine he chooses to risk it being lost before it gets built. Alternatively, he can either order another person there as backup, or he can scan the area so the builder can check their map for movement while building.

1 Marine versus 1 Fade will always lose, but a normal Marine doesn't cost 50 res. I can't remember how much an HMG and level 3 HA costs (15 + 15 + cost of upgrades?), but it seems fairly balanced that it should cost so much.

I'll reiterate Revenge's point: the design philosophy isn't about balancing numbers ("normal marine doesn't cost 50 res"); but to provide more choices to 'balance' the game. If there's a lone marine against a fade, as you say he will always lose. It's balanced at the strategic level (through resource costs), but not at the tactical level (that marine has no choices, is helpless and left for dead). That marine doesn't stand a chance fighting or even running, and that makes for a bad experience and a (mis?)conception on that player's part of imbalance.

Also, I'm not too sure whether CanadianWolverine was suggesting a "grind-heavy" minibuff (that would encourage rambos); so I'll leave it up to him to defend and clarify.

NS isn't the game with the most choises, but some of you are looking at it without seeing the big picture. I think I run out of choises at the point where the 2nd hive aliens have taken out my last pg outpost and I'm down to 1 rt. At that point I deserve to lose.

Crispy summed it up already, but here it goes anyway. Its the marine's own fault he was alone.
This choise was to go out alone. It's a risk, which didn't pay off at that point. Now, if you had managed to sneak a pg, get into some spawncamping position or something similar the choise would have paid out most likely. Now you lost that gamble, respawn and make a new decision.

My decision might be to get sg, welder, and 2 marines. At that point we can outlast almost any lone fades. Also, I'd move quickly, risking to die on some skulk ambushes, since 3 marines isn't going to last against a lerk and a fade and so on. I still think getting people somehow to the level where we can form those 3 people groups effectively is quite crucial. Right now its very likely that at least one of the three wanders off, decides to stay in an useless area for a minute or something else nasty, making a good decision using a lot less viable.

Er, what grinding? I wasn't suggesting grinding at all, unless the research a Commander does is considered grinding. You know how right now in NS1 a commander researches weapons and armor upgrade levels, well, just take that, split it up a bit so that it turns out a bit like Deus Ex, where a player goes "Oh, I can use upgrade type A or upgrade type B or upgrade type C with this upgrade, since column B suits my play style, I'll go with that small boost there". Why, that's a lot like how chambers work for a Kharaa currently in NS1, that seems to work out most of the time, why wouldn't it work out for the Marines to expand their tech tree? *shrug* One thing I can think of the RPG like character building would not be favored would be because it could make both sides similar if implemented poorly, so take it with a grain of salt. Oh, I guess if RFK (which I personally detest) was in and affected research and authorizations of equipment, then that could be a grind situation.

Oh well, it was just a thought, some brain storming on the topic, trying to see a horizon with expanded options that accommodates all the different play styles that are suited towards different roles in a squad or pack.