The balance debate revisited

245

Comments

  • tankefugltankefugl One Script To Rule Them All... Trondheim, Norway Join Date: 2002-11-14 Member: 8641Members, Retired Developer, NS1 Playtester, Constellation, NS2 Playtester, Squad Five Blue
    Hehe, Savant -- you're arguing that there is a resistance for changes that helps larger servers, I'm saying there isn't. It's a bit pointless to go on arguing that there is while attributing the statements of previous development members to the current development team. So let's end that debate, shall we? Consider it like you've "won" and convinced us <img src="style_emoticons/<#EMO_DIR#>/smile-fix.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile-fix.gif" />

    Now, we also know we have many parameters to play with, but I've yet to see a suggestion has convinced me that it'll solve the situations without creating new ones. So consider this an open challenge to all of you -- device a model for how to dynamically scale the parameters in the game depending on teamsizes.
  • PriestlyPriestly Join Date: 2006-10-30 Member: 58098Members
    edited March 2007
    <!--quoteo(post=1612907:date=Mar 9 2007, 09:25 AM:name=Dirm)--><div class='quotetop'>QUOTE(Dirm @ Mar 9 2007, 09:25 AM) [snapback]1612907[/snapback]</div><div class='quotemain'><!--quotec-->
    My thoughts:
    Balancing large pubs should not be done at the expense of non-large pubs, as large pubs do not constitute a vast majority of pubs.
    <!--QuoteEnd--></div><!--QuoteEEnd-->

    Let's take a look at Thief's Snapshot, taken at 1pm est, and go through the 3.2 servers to see what we have:

    On the larger servers we have 119 players in game.
    On the smaller servers we have 103 players in game.

    Of coarse that isn't counting bots that are playing instead of people, smaller servers normally have these on to try and populate their server when they're half empty to try and illicit people to drop by and have something to do other than idle in readyroom.

    So dirm, while you are correct in this snapshot that Large pubs do not constitute a vast majority of pubs, in the snapshot we have 6 large servers and 8 small servers, only close to half or 43%.

    Large Servers constitute the largest contribution of people who are actually playing NS, and this is important, around the clock. I think people tend to lose sight of that.

    A possible rebuttal I can think of from you would be that this was not a good time as it was early in the afternoon on the west coast, so let's do a survey around the clock starting now at 6:20 am cst on a Friday.

    Results from the early morning usa crews and afternoon european crews?

    Large Servers: 70 Players
    Small Servers: 18 Players

    <a href="http://img136.imageshack.us/my.php?image=clipmg9.jpg" target="_blank">Image Shack Hosting</a>

    I'll take another survery later on this afternoon and then this evening when some of the smaller servers might actually have a population. Also, funny enough, I dont see any 12 or 14 person servers listed there that are 3.2, the smallest seem to start at 16 players.
  • the_x5the_x5 the Xzianthian Join Date: 2004-03-02 Member: 27041Members, Constellation
    <!--quoteo(post=1612915:date=Mar 9 2007, 05:24 AM:name=tankefugl)--><div class='quotetop'>QUOTE(tankefugl @ Mar 9 2007, 05:24 AM) [snapback]1612915[/snapback]</div><div class='quotemain'><!--quotec-->
    Now, we also know we have many parameters to play with, but I've yet to see a suggestion has convinced me that it'll solve the situations without creating new ones. So consider this an open challenge to all of you -- device a model for how to dynamically scale the parameters in the game depending on teamsizes.
    <!--QuoteEnd--></div><!--QuoteEEnd-->

    How about this newbie's idea. Sometimes it's good to get a new perspective:

    <a href="http://www.unknownworlds.com/forums/index.php?s=5015360473011884032&showtopic=100728" target="_blank">http://www.unknownworlds.com/forums/index....howtopic=100728</a>
  • TOmekkiTOmekki Join Date: 2003-11-25 Member: 23524Members
    edited March 2007
    try taking the numbers from every server once an hour for a week and you might get some plausible statistics. however it is hardly that essential to the point of this thread if the "small server / large server" player numbers are 45-55 or 55-45%.

    also a philosophical question springs from this kind of number worshipping. having played ns on pubs for countless hours i can objectively say that small server games are much more rewarding and fun than large server games. are numbers the only thing that really matter to you


    ---------------------


    the number one problem with large servers is that marines can be everywhere at once. in fact there are no other major problems. to fix this, give pg's a cooldown: after 5 or 6 marines have went through the pg stops working for x seconds. or make them have an adrenaline bar or something similar to the mg42 in dod.

    also make turret factories cost more and reduce their hp a lot
  • PriestlyPriestly Join Date: 2006-10-30 Member: 58098Members
    edited March 2007
    <!--quoteo(post=1612941:date=Mar 9 2007, 01:27 PM:name=TOmekki)--><div class='quotetop'>QUOTE(TOmekki @ Mar 9 2007, 01:27 PM) [snapback]1612941[/snapback]</div><div class='quotemain'><!--quotec-->
    also a philosophical question springs from this kind of number worshipping. having played ns on pubs for countless hours i can objectively say that small server games are much more rewarding and fun than large server games. are numbers the only thing that really matter to you
    <!--QuoteEnd--></div><!--QuoteEEnd-->

    Dirm made a statement saying "Balancing large pubs should not be done at the expense of non-large pubs, as large pubs do not constitute a vast majority of pubs."

    So I decided to look at the screen shot thief posted, my thoughts were confirmed there but I thought maybe it was just a bad time so I took another standing and the same held true, I think you're also missing the largest part of my post which is the amount of people playing natural selection around the clock and the size of the servers they play on in relation to what should be balanced for. I'll be taking multiple standings but I think every hour on the hour is a bit much, I do have a bit of a life and college classes.

    As to the philosphical question and number worship, how else are you supposed to look into the point Dirm brought up, would you rather if I just said I dont feel that there are as many small server players around the clock as large server players and therefore they're missing a large if not the largest segment of people who still play ns? Numbers help to back up claims like that.

    I'll agree with you competitive play and pugging is extremely fun, it's just you don't see enough of it and Savants suggestion was balance for the larger servers where a greater portion of players spend their time as well as a tournament mode for those sunday evenings for three hours when 6v6 action takes place sounds like a fine idea. The best of both worlds.
  • the_x5the_x5 the Xzianthian Join Date: 2004-03-02 Member: 27041Members, Constellation
    That's way too much TOmekki. I think fixing marine movement as described in the <a href="http://www.unknownworlds.com/forums/index.php?s=5015360473011884032&showtopic=100668" target="_blank">what's wrong about skulks thread</a> would do so much for all levels and sizes of the game. And then you could make SLIGHT dynamic scalability based on server size as was <a href="http://www.unknownworlds.com/forums/index.php?s=5015360473011884032&showtopic=100728" target="_blank">this newbie's idea</a>.



    <img src="style_emoticons/<#EMO_DIR#>/tsa.gif" style="vertical-align:middle" emoid="::tsa::" border="0" alt="tsa.gif" /> That's my action plan. <img src="style_emoticons/<#EMO_DIR#>/smile-fix.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile-fix.gif" />
  • tankefugltankefugl One Script To Rule Them All... Trondheim, Norway Join Date: 2002-11-14 Member: 8641Members, Retired Developer, NS1 Playtester, Constellation, NS2 Playtester, Squad Five Blue
    edited March 2007
    Well, yeah. One of the basic ideas has always been dynamic scaling. It's not new; the vital question is, though -- which parameters should be scaled and in what way? And will, for example, the scaling of resource gains compencate for the marines' ability to be virtually everywhere on large servers? (BF2 does this well by scaling the map, but we cannot do that, at least not without heavy rework.) What problems can arise from the scaling?

    (As for gathering statistical data; I doubt this is the time to gather any data for finding trends, as the gameplay is still changing and adapting as players learn the new patch. It could give insight to how specific servers are experiencing the game at the moment, but the data would be very weak in a larger timespan.)
  • TG-CainTG-Cain Join Date: 2006-10-31 Member: 58105Members
    <!--quoteo(post=1612932:date=Mar 9 2007, 01:45 PM:name=Priestly)--><div class='quotetop'>QUOTE(Priestly @ Mar 9 2007, 01:45 PM) [snapback]1612932[/snapback]</div><div class='quotemain'><!--quotec-->
    Let's take a look at Thief's Snapshot, taken at 1pm est, and go through the 3.2 servers to see what we have:

    On the larger servers we have 119 players in game.
    On the smaller servers we have 103 players in game.

    <snip>

    Results from the early morning usa crews and afternoon european crews?

    Large Servers: 70 Players
    Small Servers: 18 Players
    <!--QuoteEnd--></div><!--QuoteEEnd-->
    One could argue several different ways based on these screenshots. First of all, where do we draw the line between large and small servers? Some have stated that up to 24 players is balanced, others claim 20 players to be the upper limit or possibly 18. That alone would shift the statistics significantly.

    Then there's the issue of bots. This could swing both ways. On Tactical Gamer, bots are used on co-maps only and can fill the server up to 32 slots, but as you point out, this will possibly have greatest impact on small servers while seeding. Filtering out co-maps is therefore possibly a good thing, removing a bias towards a game mode that inherently isn't balanced, but it could also have a negative impact on small servers waiting to fill up or, indeed, popular servers of a game mode that is after all officially supported.

    Seeding is connected to time zone bias. A server should really only be judged by its stats during prime time, which of course could be 24/7 in some rare cases. Anything else is probably not representative of the active community in that particular time zone.

    Filtering out passworded servers is another difficulty. Not all servers with a password means a clan match. And even if it did, one can hardly argue that they are not part of the community as a whole.

    The final question is something I touched upon in my original post, but perhaps is beyond the scope of a server study: actual player behaviour. How much time does a player spend on different servers? How many players visit a particular server during the week and how long do they stay? What is the total number of players active in the community? Making statistics based on 100-200 players is only relevant if they are representative of the community at large. Now that 3.2 has been released it should be easy to get stats on the number of downloads, provided the devs will release such data.

    Without taking sides, one really shouldn't make general statements about the community before answering at least some of the above, regardless of personal preference.
  • the_x5the_x5 the Xzianthian Join Date: 2004-03-02 Member: 27041Members, Constellation
    edited March 2007
    <!--quoteo(post=1612953:date=Mar 9 2007, 09:10 AM:name=TG-Cain)--><div class='quotetop'>QUOTE(TG-Cain @ Mar 9 2007, 09:10 AM) [snapback]1612953[/snapback]</div><div class='quotemain'><!--quotec-->
    provided the devs will release such data.<!--QuoteEnd--></div><!--QuoteEEnd-->

    And...?

    <!--quoteo--><div class='quotetop'>QUOTE</div><div class='quotemain'><!--quotec-->Without taking sides, one really shouldn't make general statements about the community before answering at least some of the above, regardless of personal preference. <!--QuoteEnd--></div><!--QuoteEEnd-->

    I would <i>love</i> to have more data on this stuff.

    Balancing NS for all server sizes while imploying important RTS and teamplay elements seems to be a holy grail of development on this mod since it's inception.
  • kamikazegoatkamikazegoat Join Date: 2007-03-08 Member: 60292Members
    The ever-elusive solution is likely going to involve a few thousand <img src="style_emoticons/<#EMO_DIR#>/pudgy.gif" style="vertical-align:middle" emoid="::gorge::" border="0" alt="pudgy.gif" /> 's, a webcam, some ###### <img src="style_emoticons/<#EMO_DIR#>/marine.gif" style="vertical-align:middle" emoid="::marine::" border="0" alt="marine.gif" /> 's, and a URL like ... www.gorge.org/y
  • tankefugltankefugl One Script To Rule Them All... Trondheim, Norway Join Date: 2002-11-14 Member: 8641Members, Retired Developer, NS1 Playtester, Constellation, NS2 Playtester, Squad Five Blue
    <!--quoteo--><div class='quotetop'>QUOTE</div><div class='quotemain'><!--quotec-->Now that 3.2 has been released it should be easy to get stats on the number of downloads, provided the devs will release such data.<!--QuoteEnd--></div><!--QuoteEEnd-->

    Considering the various mirrors for the patch, I'd say it's unlikely that anyone has that data.
  • PriestlyPriestly Join Date: 2006-10-30 Member: 58098Members
    edited March 2007
    <!--quoteo(post=1612951:date=Mar 9 2007, 02:07 PM:name=tankefugl)--><div class='quotetop'>QUOTE(tankefugl @ Mar 9 2007, 02:07 PM) [snapback]1612951[/snapback]</div><div class='quotemain'><!--quotec-->
    (As for gathering statistical data; I doubt this is the time to gather any data for finding trends, as the gameplay is still changing and adapting as players learn the new patch. It could give insight to how specific servers are experiencing the game at the moment, but the data would be very weak in a larger timespan.)
    <!--QuoteEnd--></div><!--QuoteEEnd-->

    While I can agree there are players adapting and learning to the new 3.2 patch that doesn't change the servers where the players do the learning and adapting. Most of the NS community has been playing the game for months or years, especially the segment that posts on these forums. A new patch wont change where they have been playing for any appreciable length of time. Even with all the ###### and moaning that happens after a patch, people still play where they know the people on their server, have a good time, and have fun. The only variable I can think of that 3.2 would of changed is perhaps a small number of people who haven't played ns before or had it uninstalled previously, and decided to install to give 3.2 a whirl.
  • PriestlyPriestly Join Date: 2006-10-30 Member: 58098Members
    edited March 2007
    And now back with the mid-afternoon report. This is taken as of 2:17 pm cts.

    <a href="http://img252.imageshack.us/my.php?image=survery2xv4.jpg" target="_blank">Afternoon Survey of NS Servers</a>

    Large Server Populations = 238 Players
    Small Server Populations = 115 Players

    So over double the amount of players are showing up as playing on larger servers. I took the suggestion of including the passworded servers in this one as well. I'm sure the evening survey will be more even as that is prime time and all of the smaller servers should be filled to max capacity and not half filled with bots playing combat.
  • SavantSavant Join Date: 2002-11-30 Member: 10289Members, NS1 Playtester, Contributor
    First off, on the 'server size' definition, I think it is reasonable to consider a 'small' server to be any server that holds 50% or less than the maximum number of players supported on the HL engine. Since the maximum is 32 players, I consider a 'small server' to be one that is set to hold no more than 16 players. 'Large' servers are those that hold 17+ players, IOW, anything 9v9 or above is what I consider a 'large' server. Overall, the balance in NS seems to be ideal for a small 6v6 to 8v8 match, but once you pass the 10v10 point you are left with a game that is not balanced for that form of play. So with that clarification in mind, let me continue.

    <!--quoteo(post=1612915:date=Mar 9 2007, 05:24 AM:name=tankefugl)--><div class='quotetop'>QUOTE(tankefugl @ Mar 9 2007, 05:24 AM) [snapback]1612915[/snapback]</div><div class='quotemain'><!--quotec-->
    Hehe, Savant -- you're arguing that there is a resistance for changes that helps larger servers, I'm saying there isn't. It's a bit pointless to go on arguing that there is while attributing the statements of previous development members to the current development team. So let's end that debate, shall we? Consider it like you've "won" and convinced us <img src="style_emoticons/<#EMO_DIR#>/smile-fix.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile-fix.gif" /> <!--QuoteEnd--></div><!--QuoteEEnd-->I appreciate the sentiment, and I hope you don't see me as a cynic when I ask if you mean that you actually genuinely believe/agree that there are balance issues on larger servers, especially up towards the top end 16v16.

    <!--quoteo--><div class='quotetop'>QUOTE</div><div class='quotemain'><!--quotec-->Now, we also know we have many parameters to play with, but I've yet to see a suggestion has convinced me that it'll solve the situations without creating new ones. So consider this an open challenge to all of you -- device a model for how to dynamically scale the parameters in the game depending on teamsizes.<!--QuoteEnd--></div><!--QuoteEEnd-->I initially refrained from providing suggestions since I wanted the concept to be acknowledged that there is an inherent balance flaw on larger servers.

    First off, we need to know if you intend to detach competitive play balance from public play balance. The main reason that every other balance initiative for large servers has failed is because it was deemed 'too disruptive' to the balance of smaller servers used for competitive play. If we can't separate the balance of large servers from small servers than we are faced with the same problem, any suggestion for balance on large servers will be deemed to create "new problems" for small competitive servers and be rejected. The framework for this concept really needs to be developed before the balance can change.

    So we first need to recognize that there are unique balance issues on larger servers that are independent of balance on smaller servers. We also need to recognize that competitive play balance is not possible if you try and balance for large servers. As such, NS first needs a separate 'competitive play mode' to remove any chance that competitive play will be injured by the subsequent balance changes.

    Next we need to define the low end of balance for public servers. At present, I would suggest that the smallest server NS be balanced for is 8v8. If server operators want to set limits lower than that, they can play in competitive mode which would be balanced for 6v6. (Seeing as there are zero 12 man servers online right now, I don't see this as an issue) So with a low end of 8v8 and a high end of 16v16, we can assume that the current level of balance is sufficient for 8v8 play and move to finding an ideal balance for a 16v16 game, ignoring other server sizes <i>for the time being</i>. Once balance is achieved for 16v16, we can then look at a linear scale to adapt balance to server sizes between 8v8 and 16v16.

    Will a linear scale for balance numbers work? Honestly, I don't know. No one has ever attempted to balance for large servers, so I really don't know how far we need to go. However, I think there is enough flexibility that we could devise a scheme that would allow us to apply balance to the vast majority of server sizes, instead of picking an arbitrary server size, balancing the game for that and then saying that if the game is unbalanced for other server sizes that there is 'nothing we can do about that'.

    IOW, something is better than nothing.

    As for suggestions on variables to tweak, there are a number of key variables that need to change for larger servers. The following is just off the top of my head, given that I'm still not sure if this concept is actually going to go anywhere...

    -Alien spawn rate: In a 16v16 game, sitting in the respawn queue is a game ender for aliens since the marines then get the run of the map. Spawn times should be reevaluated to make sure aliens never wait longer than X seconds to respawn, even if most of their team is down. While marines can scale their respawn points by spending res, the aliens can only do so by adding hives. As such, early game on a large server can mean a long respawn wait, which is death as it gives up territory to the marines.

    -Skulks: Honestly, I really think we need to examine and/or reevaluate how many skulks a marine should be able to kill with one weapon clip. In small competitive games it may be important to be able to kill a skulk in 10 bullets, but in large games it sets the bar too high when you consider the large groups of marines that romp around on big servers. Since the LMG has shoots 10 rounds a second, that's 1 second for a kill from one marine, less if a group is firing. Alternately, the skulks bite takes much longer, and to kill a stock starting marine it takes 3 bites. That's 2.4 seconds that you have to 'connect' with a marine that is jumping around like a noob on crack to exploit the flawed bite model of the skulk. Which leads me to...

    -Skulk flawed bite model: It's been said that marines are masters of ranged combat, and <b>I agree completely</b>. If a skulk rushes a marine head on and/or from range, they are dead, no question about it. However, the problem happens when a skulk is able to ambush a marine and starts biting. The bite model is such that your accuracy needs to be perfect or your bites will miss. All this while trying to stay on top of a jumping marine and being shot at. The big problem with the bite model is that (from my understanding) it's like a shotgun of sorts. While the spread is wide, that spread is AWAY from the skulk's body. So it means you end up needing to be slightly away from the marine to benefit from the 'spread'. I've watched people attack marines time and time again, they will be right on top of the marine and the marine lives, even though the skulk has gotten 6+ bites in. IMO, the skulk bite model should be inverted, so that the further away you are from the marine, the more accurate you have to be. Alternately, once a skulk is 'on top of' a marine, <b>every bite should connect</b>, so long as the skulk is remotely facing the marine. In short, if a skulk is biting you, then you are pretty well dead - in the same was as a skulk is dead if they charge down the hall at a marine. I would also suggest increasing the bite ROF to compensate for the imbalance to kill times, and perhaps we could balance that by decreasing a skulk's damage to structures so the net result won't impact skulks vs structures. A stock skulk should be able to kill a stock marine in the same amount of time it takes a stock marine to kill a stock skulk. I don't think that is unreasonable.

    Resource flow: In larger games this becomes a big issue for aliens, since they each have their own resource pool. The model for resource flow to aliens should give them the same relative resource flow regardless of the number of players on the team. Whether it be 8v8 or 16v16, they should still be able to accumulate resources at the same rate given the same circumstances. This creates a potential problem though, which is that aliens will end up with more overall resources with which to build structures. While this could cause issues, I don't think that would be the case. Remember, more aliens mean more marines, which means aliens need more 'higher level' evolutions. If aliens 'lame up the map' with all their extra res, they will have no res to go fade etc, which will impact their survivability. If increasing alien res flow causes more problems than it solves, the perhaps we could scale the cost of upper evolutions downwards as the server increases in size. These 'cheaper' evolutions would allow aliens the power needed to defend their territory, while not giving aliens a blank cheque to drop chambers everywhere. Something needs to be done though, since it's painful in large games.


    I'm going to stop here since I'm not sure where this thread is going, but I'll wait and see how serious the developers are on tackling this problem. If we could get some actual movement on this I would be really happy, since it has been a thorn in the side of NS from the beginning. I'd love to see NS balance viable on larger servers, and I think that would bring back a lot of fun to the game. (IE, the 'old days')

    Regards,

    Savant
  • Heavy_DHeavy_D Join Date: 2002-12-14 Member: 10816Members
    <!--quoteo(post=1613092:date=Mar 9 2007, 04:42 PM:name=Savant)--><div class='quotetop'>QUOTE(Savant @ Mar 9 2007, 04:42 PM) [snapback]1613092[/snapback]</div><div class='quotemain'><!--quotec-->
    -Alien spawn rate: In a 16v16 game, sitting in the respawn queue is a game ender for aliens since the marines then get the run of the map. Spawn times should be reevaluated to make sure aliens never wait longer than X seconds to respawn, even if most of their team is down. While marines can scale their respawn points by spending res, the aliens can only do so by adding hives. As such, early game on a large server can mean a long respawn wait, which is death as it gives up territory to the marines.
    <!--QuoteEnd--></div><!--QuoteEEnd-->
    I think there are two separate issues here which end up lumped together. One is that the spawn time is long, which is an aggravation for the player - sitting around, doing nothing for a greater period of time. The spawn queue is meant to be a "punishment" for dying but in larger servers the punishment doesn't fit the crime because it's too long.

    The other issue is that long spawn times don't affect the two sides symmetrically, as marines can compensate by adding more spawn points. This does affect the early game balance as the marines will take advantage as you described. However, my feeling is that the best way to fix the balance issue here would be to slow the marine spawn down to the same level rather than bringing aliens back faster. This of course does nothing to help with the first problem, in fact making it worse because it'll now affect marines as well. Perhaps if this were done both sides would be less aggressive in the early stages of the game, but I think it's unlikely people would change their playing style on a large server just because of the queue.
  • SavantSavant Join Date: 2002-11-30 Member: 10289Members, NS1 Playtester, Contributor
    In looking back at the thread, I think one of the biggest problems here is what I like to call 'developer arrogance'. Now before I go further I want to qualify my remark and say that it is in no way meant as an insult to developers. Furthermore, this 'developer arrogance' is something that is present in nearly <b>every game </b> that I have beta tested. It's funny since I have beta tested games in which the developer was so different from the developer of the last game I tested, and yet the views are very much the same. BTW, my comments in this post relate to developers in general, and not the NS developers specifically.

    In my opinion, it basically boils down to the fact that developers see their creations as 'their baby', and much like a parent of a child, they don't want advice on how to raise their kids. To them suggestions on how to raise their kids are presumptuous and insulting, even if the suggestions have merit. Hence the 'arrogance', in that even if the suggestions of others have merit, they don't want to hear them.

    What developers want (and usually crave) is 'positive feedback' on their creations. If a person doesn't like something the developers really don't care, since they see it as a difference in opinion. If people have suggestions, the developers don't want to hear it since taking suggestions from the public would be akin to the public developing the game. They feel that doing as such would cause them to lose developmental control of the game.

    In the end I totally respect the will of developers to do what they want with their creations. Hey, they created it, they programmed it, and they put their blood, sweat and tears into it. If they want to put out a game that isn't balanced in a way that the public approve of, then that is their right and I totally support them on it. I'd rather see developers happy with their creations then unhappy just to appease public opinion.

    However, the flipside of this is that I don't have a lot of sympathy for developers that put out games that don't appeal to the public. Ive seen (and played) more than enough 'flops' in my life, and so while it may have been painful for that developer to see his creation rejected by the public at large, that is part of the learning experience.

    Does this mean all developers ignore public opinion? No. I've beta tested a few titles where beta tester input was actually quite influential in making the game a success.

    People shouldn't harbour any ill will towards the NS developers. If this is the game the developers want, that's their call. Like I said in the beginning of the post, this particular balance issue has been around since NS existed, and I never really expected the developers to fix it now. Based on the decline of NS servers, I expect NS will see the bulk of the remaining playerbase evaporate throughout the year, and once Christmas hits things will likely have dropped to the point where they will stay. Balance has been (and will always be) a gamekiller, and so this really shouldn't be a surprize to anyone. While it's natural for players to get angry when something that is fun for them become 'unfun', it's best not to let things get carried away.

    If anything, I think we should all thank Charlie for the great ride, since I know there are many of us who were around for the fun times.

    Regards,

    Savant
  • enigmaenigma Join Date: 2004-09-11 Member: 31623Members
    <!--quoteo(post=1613093:date=Mar 9 2007, 04:50 PM:name=Finawin)--><div class='quotetop'>QUOTE(Finawin @ Mar 9 2007, 04:50 PM) [snapback]1613093[/snapback]</div><div class='quotemain'><!--quotec-->
    Remove knockback on skulk bites and it will level it a bit more.
    <!--QuoteEnd--></div><!--QuoteEEnd-->
    clueless
  • tankefugltankefugl One Script To Rule Them All... Trondheim, Norway Join Date: 2002-11-14 Member: 8641Members, Retired Developer, NS1 Playtester, Constellation, NS2 Playtester, Squad Five Blue
    <!--quoteo--><div class='quotetop'>QUOTE</div><div class='quotemain'><!--quotec-->I appreciate the sentiment, and I hope you don't see me as a cynic when I ask if you mean that you actually genuinely believe/agree that there are balance issues on larger servers, especially up towards the top end 16v16.<!--QuoteEnd--></div><!--QuoteEEnd-->

    You are being cynic <img src="style_emoticons/<#EMO_DIR#>/wink-fix.gif" style="vertical-align:middle" emoid=";)" border="0" alt="wink-fix.gif" />

    But I'll answer with my exact thoughts on the situation:
    - I've seen that NS have had large-game imbalances in 3.1 and earlier.
    - I've not played 3.2 on a so-called large server yet, but I do believe many of the same problems are carried over from 3.1.
    - I am currently not a developer, so I cannot speak on behalf of the dev team or Charlie.
    - We have been reviewing mechanics to shift balance according to teamsizes (both much larger than 6v6 and smaller than 6v6), but so far I've seen no overwhelmingly good solution. It was not done (while I was developing) because we thought many other changes to the NS balance would possibly skew balance back and forth and produce unpredictable results.
    - Re-designing and separating game modes further will probably not happen on NS.

    You got to understand, it's less a lack of will, and more a lack of time and resources that is the reason why things haven't changed or been implemented. I understand what you say about developer arrogance, and possibly everyone who has created something has a bit of that as well, but that is not the main obstacle here.

    We've been maintaining this on our spare time, which means we've had to pick features and changes that we know we can overcome while still keeping this as a hobby. More bang for less coin, basically. I hope this clears things up, and I hope you believe me this time.
  • UnderwhelmedUnderwhelmed DemoDetective #?&#33; Join Date: 2006-09-19 Member: 58026Members, Constellation
    Can you spec a few scrims to see how balanced the game is between players of roughly equal ability?
  • DuoGodOfDeathDuoGodOfDeath Join Date: 2002-08-01 Member: 1044Members
    That is impossible Sir! NS was clearly designed with the intention of 15v15 in mind since the beginning of NS.
  • the_x5the_x5 the Xzianthian Join Date: 2004-03-02 Member: 27041Members, Constellation
    <!--quoteo(post=1613093:date=Mar 9 2007, 05:50 PM:name=Finawin)--><div class='quotetop'>QUOTE(Finawin @ Mar 9 2007, 05:50 PM) [snapback]1613093[/snapback]</div><div class='quotemain'><!--quotec--> Remove knockback on skulk bites and it will level it a bit more. <!--QuoteEnd--></div><!--QuoteEEnd-->

    While I dislike the degree of knockback, completely removing it would require MAJOR rebalancing. So in that sense it's impratical.



    <!--quoteo(post=1614093:date=Mar 13 2007, 05:19 PM:name=Savant)--><div class='quotetop'>QUOTE(Savant @ Mar 13 2007, 05:19 PM) [snapback]1614093[/snapback]</div><div class='quotemain'><!--quotec--> In looking back at the thread, I think one of the biggest problems here is what I like to call 'developer arrogance'. Now before I go further I want to qualify my remark and say that it is in no way meant as an insult to developers. Furthermore, this 'developer arrogance' is something that is present in nearly <b>every game </b>that I have beta tested. It's funny since I have beta tested games in which the developer was so different from the developer of the last game I tested, and yet the views are very much the same. BTW, my comments in this post relate to developers in general, and not the NS developers specifically.

    In my opinion, it basically boils down to the fact that developers see their creations as 'their baby', and much like a parent of a child, they don't want advice on how to raise their kids. To them suggestions on how to raise their kids are presumptuous and insulting, even if the suggestions have merit. Hence the 'arrogance', in that even if the suggestions of others have merit, they don't want to hear them.

    [...]
    People shouldn't harbour any ill will towards the NS developers. If this is the game the developers want, that's their call. Like I said in the beginning of the post, this particular balance issue has been around since NS existed, and I never really expected the developers to fix it now. Based on the decline of NS servers, I expect NS will see the bulk of the remaining playerbase evaporate throughout the year, and once Christmas hits things will likely have dropped to the point where they will stay. Balance has been (and will always be) a gamekiller, and so this really shouldn't be a surprize to anyone. While it's natural for players to get angry when something that is fun for them become 'unfun', it's best not to let things get carried away.
    <!--QuoteEnd--></div><!--QuoteEEnd--> You are very correct with pointing out what you call developer arrogance. Although I disagree with the ride being over. If you want change, work as a team. Even go check out Zunni's mod: <a href="http://idmod.net/" target="_blank">http://idmod.net/</a>



    <!--quoteo(post=1614144:date=Mar 13 2007, 11:04 PM:name=gumhat)--><div class='quotetop'>QUOTE(gumhat @ Mar 13 2007, 11:04 PM) [snapback]1614144[/snapback]</div><div class='quotemain'><!--quotec-->
    can you play a few rounds on the largest server (g4b2s) to see what we're putting up with <!--QuoteEnd--></div><!--QuoteEEnd--> <b><!--coloro:#33cc00--><span style="color:#33cc00"><!--/coloro-->YES!!!<!--colorc--></span><!--/colorc--></b> We need developers and playtesters to get involved. It's stupid to do developement work on a project you never test the end results in real-world situations. How can you hope to understand your community?
  • F4tManMGS2F4tManMGS2 Join Date: 2004-04-10 Member: 27842Members
    <!--quoteo(post=1614147:date=Mar 13 2007, 10:17 PM:name=DuoGodOfDeath)--><div class='quotetop'>QUOTE(DuoGodOfDeath @ Mar 13 2007, 10:17 PM) [snapback]1614147[/snapback]</div><div class='quotemain'><!--quotec-->
    That is impossible Sir! NS was clearly designed with the intention of 15v15 in mind since the beginning of NS.
    <!--QuoteEnd--></div><!--QuoteEEnd-->

    Haha CAL died but the big pubs didn't.

    hahahahahahahahahahahahahaha
  • DuoGodOfDeathDuoGodOfDeath Join Date: 2002-08-01 Member: 1044Members
    Still have ANSL and ENSL. Not like the pubbing community is doing awesome either thar buddy. I might prefer my 6v6 but I don't mind annihilating your 15v15 servers :/
  • F4tManMGS2F4tManMGS2 Join Date: 2004-04-10 Member: 27842Members
    you're good at NS! congratulatoriations, you can bring the pwn-house to the 15v15 servers!

    doesn't stop the fact that CAL dropped NS. Wonder what the people over there think of competetive NS?
  • DuoGodOfDeathDuoGodOfDeath Join Date: 2002-08-01 Member: 1044Members
    Well it really is no big deal. Fact is whining about 15v15 is imbalanced is retarded. The game is pretty well balanced and maybe a nerf on the HMG would be nice and then the world will be a sphere. Marines arent over powered and the aliens are okay as it is. Only reason marines seem overpowered is because pubbers straight line down super long corridors hoping by some hand of hades they can get to the marine before they die.
  • FirewaterFirewater Balance Expert Join Date: 2002-12-12 Member: 10690Members, Constellation
    edited March 2007
    I posted this a while ago on the intelligent design forums. Its a framework for design considering the Quasi RTS/FPS genre that NS has created.

    --------------------------------------
    The purpose of this post is to provide a framework for the mod intelligent design.

    The fundamental premise of this game is ranged vs. melee. This concept has been tried before (Nautral Selection, aka NS) and it is very difficult to achieve balance within the mod. To understand this concept, one must define balance. I define balance as having an equal amount of viable features available to both teams so that it is the player that determines success or failure, not the game.

    Ranged vs. Melee is very difficult to balance simply because the ranged team will almost always have initiative (aka first shot) against the melee team. Having this primary advantage allows for players to potentially kill an enemy before the enemy can strike them. This can cause a momentum shift in the game as early as the first minute or so in NS.

    This momentum can allow for the ranged team to acquire resources faster, than the melee team, and allow them to protect them easier. Because of this resource gap, the ranged team would be able to tech faster as a team, as they can not only secure more resources for the team, but at the same time destroy enemy resources. So the ranged team speeds up and the melee team slows down.

    Since the development team has it locked in stone that they want to do a quasi RTS/FPS with primarily ranged with primarily melee, I will not attempt to sway them away from that model, rather than attempt to provide a theoretical backbone for an attempt at balance for this model.

    A potential solution to this problem is to give the melee team a chance to "close the gap" between it, and the ranged player, but at the same time, still give the ranged player some sort of advantage for having a ranged weapon. It would be wise at the beginning of the team that the melee team could have a way to reduce the effect of a ranged weapon, but not eliminate it.

    --------------------------------------------------------------------------------

    Class structure in NS, what can be learned from it?

    Skill/Firepower/Time Theory.

    In NS marines need 3 things to allow them to achieve a kill. They need enough skill, that is brains/dodging ability/aim to make a kill. One must be able to shoot at the target basically. One also needs enough firepower. Firepower is one of the larger variants as you can go from needing very little firepower (i.e. skulks) to needing very heavy fire (i.e. Onos). Firepower is defined as the amount of damage a player can deal to the enemy. The final aspect that a player needs is time. One must have enough time to fire enough bullets/shells to achieve a kill.

    As previously discussed, the skulk (1 hive), requires a decent amount of skill (which based entirely on the skulk) very low amount of firepower, and not very much time at all. The gorge on the other hand requires less skill than the skulk to kill, BUT requires more firepower and time to kill. The lerk is a difficult one to assess, because it doesn't require much time to kill, but one needs a good amount of skill, and decent amount of firepower.

    Fades on the other hand, require the greatest amount of from all 3 aspects of achieving an alien kill. The fade requires the highest amount of skill, firepower and time to be killed.

    The Onos on the otherhand, requires very little aim, but the highest amounts of firepower and time to kill. Since the onos requires such little skill to kill based on its size and speed, I determine it a skilless class.

    The onos messes the balance up for both sides. It is the most expensive alien unit and can either be dominant or useless SOLELY BASED ON THE MARINE'S TECH.

    When designing the game, please make sure that the most expensive melee unit is the game is skill based, and the most dominant melee unit. What the developers should avoid is having "powerless" situations (i.e. Vanilla marine vs an Onos, 1 hive skulk vs. Heavy with HMG) by making all the classes in the game skill based.

    --------------------------------------------------------------------------------

    Time/Resource/Location (TRL) theory.

    Currently, marines need 2 aspects to aquire an advantage over the other team. They need time, and resources. Enough resources to purchase the upgrade, and enough time to wait for it.

    Aliens are restricted with 3 aspects to aquire an advantage over the other team. They are restricted with Time, Resources, and 1 of 2 specific locations in the map that they can expand to. It is very possible (and quite common in pub play) that the aliens can have both time and resources, but not the location because of the two hive lockdowns. Lockdowns takes away from the skill of the game because a player fighting the lockdown will usually fail because of game controlled entities (i.e. electricity and turrets) rather than actually players. All the marines need to do is contain them, tech dominating tech, and walk all over them.

    The aliens will ALWAYS be at a disadvantage because they do not follow the same rules as the marines with regards to teching. In order to balance the game (ESPECIALLY RANGED VS MELEE) one needs to allow both teams upgrade under the same rules.

    Furthermore NS's hive system gives TOO much of an advantage to the aliens when they get the second hive. (i.e. Faster spawn times, unlocked weapons, enhanced armor protection, greater map control, unlocked chamber upgrade). 1st hive is not enough, 2nd hive over compensates, 3rd hive is overkill.

    This also leads to a lot of obselete weapons (i.e. primal scream, acid rocket, webs) because they are rarely, if at all, used in a close battle.

    The alien expansion method hampers strategy because you will not see a viable situation where hive 3 weapons are used tactically. Commanders use a set of commonly used strategies to achieve victory. Aliens always try to get that second hive, because with out it their chances of winning are significantly hampered.

    The solution to this is make both the melee team and ranged team be only restricted by TIme and resources. Have the dual restriction for both sides will unlock unlimited strategies with both teams, commanders will have to think on their toes and there will be more tactics involved as well. The strategies and tactics that can evolve, the more the community will converse either postively, or negatively, BOTH of which are better than not talking at all about the game at all. Having an active community can only sure success of the mod.

    Take this as you will.

    FireWater
    ------------------------------------------------

    The game will never be balanced on the simple fact that both teams do not tech similarly, that is one of the major points I made. Another point I made is that the onos is a skilless class, it either rapes or gets slaughtered, there rarely is an in between.

    Most of these points revolve around an overhaul of the gameplay as we know, which is why these neccessary functions could not be executed in the current NS. Hopefully ID will take this framework and use it for their theoretical basis of their game design.
  • the_x5the_x5 the Xzianthian Join Date: 2004-03-02 Member: 27041Members, Constellation
    edited March 2007
    Well said firewater.

    I really don't know what else to add to that except that it would be nice to have a mod with greater attention to RTs aspects like that.
  • DoggDogg Join Date: 2003-03-31 Member: 15063Members
    Yup, firewater's looking at the big picture, not just "but last game I won as alien so it must be balanced."

    It's interesting to see the same issues come up year after year.

    Last time this resulted in untying lifeforms from the hives... now the aliens aren't as dependent on predicable map locations for their power. People thought this would ruin the game. But having more options proved to be better.

    Next time will they unchain the chambers? If so then the aliens could get stronger even if they don't secure a second hive (just like the marines can). The counter argument is that unchained chambers would be too much like combat mode and make the game all about RTs instead of hives. Personally I'd rather have multiple skirmishes over various resource nodes than one game breaking skirmish over the second hive (as it is now). Especially since in some matches, that hive skirmish never happens (i.e. early two hive lockdown or mid-hive lockdown).

    Very frustrating for aliens, who either get the 2nd hive or lose. For marines its get the hive, or tech and attempt a comeback (very possible).
  • Sephiroth2kSephiroth2k Join Date: 2002-04-21 Member: 481Members, Constellation
    edited March 2007
    <!--quoteo(post=1612465:date=Mar 8 2007, 02:34 AM:name=Sarisel)--><div class='quotetop'>QUOTE(Sarisel @ Mar 8 2007, 02:34 AM) [snapback]1612465[/snapback]</div><div class='quotemain'><!--quotec-->
    Perhaps a solution would be to buff RFK for aliens. Have the base +2 or +3 res per kill for an individual. Then add a balancing factor that will basically give each alien a little bit of res for a team-mate's kill. Ideally this balancing factor would take into account the number of players on the team.

    In essence, the balancing factor would be about +1 res for each player (except the person that got the +3 res for the kill). If you test how it plays out, then you could make the balancing factor to account for increasing server sizes by adjusting it to +(6n)^0.5/6, which would give 1 res per person at 6 players, about 1.4 res per person at 12 players, and adjust it as necessary.
    <!--QuoteEnd--></div><!--QuoteEEnd-->

    I think they could definitly make this work effectively if they borrow on the CO experience gaining, as long as it's not too much. Just make it so you get similar exp to say lvl3 alien with one hive, lvl 6 with 2 hives and lvl10 for 3 hives. When you get a level, you get 5 resource, and the level resets itself, and you dont really gain a 'level' you just get 5 res. No matte how many hives, you get 5 res. Kills will come in more with more hives, therefore res will come in more, but that is why the level is like it is @ lvl 10 combat... eh? eh?
  • NiebelungNiebelung Join Date: 2006-11-12 Member: 58565Members, Constellation
    <!--quoteo--><div class='quotetop'>QUOTE</div><div class='quotemain'><!--quotec-->
    Perhaps a solution would be to buff RFK for aliens. Have the base +2 or +3 res per kill for an individual. Then add a balancing factor that will basically give each alien a little bit of res for a team-mate's kill. Ideally this balancing factor would take into account the number of players on the team.

    In essence, the balancing factor would be about +1 res for each player (except the person that got the +3 res for the kill). If you test how it plays out, then you could make the balancing factor to account for increasing server sizes by adjusting it to +(6n)^0.5/6, which would give 1 res per person at 6 players, about 1.4 res per person at 12 players, and adjust it as necessary.<!--QuoteEnd--></div><!--QuoteEEnd--> something like <a href="http://www.unknownworlds.com/forums/index.php?s=2424529450534609408&showtopic=100840&view=getnewpost" target="_blank">this?</a>
Sign In or Register to comment.