Making maps competitive

2»

Comments

  • ScardyBobScardyBob ScardyBob Join Date: 2009-11-25 Member: 69528Forum Admins, Forum Moderators, NS2 Playtester, Squad Five Blue, Reinforced - Shadow, WC 2013 - Shadow
    3del! wrote: »
    That being said, I still believe Aliens are in a better position, when they have close spawns, since they can more easily capture and secure their expansion Harvesters and Hive.

    tl;dr: The problem is not the frequency of engagements, but the number of engagement possibilities before Marines can attack Alien infrastructure.

    PS: A good Alien team will not get egg locked by engaging too frequently. When you realize that there are less than 4 eggs during a wavespawn, usually you avoid dying for a minute and all is good. But if Marines have easy access to Alien Spawn, they can take advantage of that.
    I think we're largely talking about the same thing here, as both will lead to egglock problems. My analysis above doesn't include marines killing eggs directly, but assumes aliens will intercept the marines to prevent egg-killing. Its technically different but will lead to much the same result (e.g. alien dying while preventing egg-killing is equivalent to an egg getting killed with regards to the total number of eggs available).
  • blindblind Join Date: 2010-04-17 Member: 71437Members, Squad Five Gold
    Scardy, as much as I adore your work, I feel egglocking isn't the real issue on close spawns. What Scrajm wrote is more about the metagame in NS2 why far spawns should be established in a good competitive map. I see that there are many opinions about this and I think the strongest argument vs. far spawns is the lack of variety which I can see. Having not to position your marines to prevent spawn camps and reacting as aliens to it, does in fact play very different from cross spawn games (if you look at summit gameplay for example). But different doesn't mean better.

    Close spawns include a much higher gamble than cross spawns where you can focus on your metagame. Thus it might be interesting to some people, but in a competitive environment this is not necessarily the best solution. True, poker for example is very popular and includes a high randomness (although it is common belief that skill has the upper hand). But if we look to closer examples, say StarCraft 2, we find that over the time the focus has shifted towards macros games (small maps vs. big maps). Having some randomness in it is still possible and viable (see the 4 player maps), but they all have one in common: they are still macro-oriented, the spawns are still very far. This is also what excelled good NS1 maps - you had a randomness where aliens spawn, but the spawn distance was always far (look at veil for example). Then even the feared pipeline hive wasn't that of a big issue anymore. (The issues of pipeline hive in ns2_veil has more to do with cyst and res mechanics.)

    Also looking at the balancing of close spawns, you will find that the second hive is considerably easier to get for aliens than on cross spawns. Imagine that, in terms of SC2, that your natural expansion is set random - this would be unplayable for at least 1 race (Zerg) and would force all-in moves pretty much just the second you notice your starting location. This is IMO a way bigger concern than egglocking. In practical terms of NS2, a close spawn is mostly a comparable garuanteed second hive and removes the marine's strategic option to prevent that hive from growing up. This leads to more predictable and thus boring gameplay by both teams.

    Egglocking in itself is an important valid option for marine gameplay, but the moment you balance it versus close spawns you probably will remove, or at least heavily change, a strategical option for marines within the normal macro game.

    Also I'd like to remember that focusing maps on macro oriented gameplay does, by no means, remove the possibility for rushes or "cheesy" strats. It just makes them more special and tends to reward the actual better team more. Even on cross spawns you can execute a base rush by bypassing the marines, catching a situation where the opponent is out of position etc. Cross spwans do not remove base rushes.

    In terms of randomness, you can have both randomness AND scrajm's commandment of far spawns in it. This is pretty much exactly what competitive SC2 maps fulfill today. And yeah, I know this isn't SC2, but I think in terms of metagame due to layout of the maps, especially in the spawning location issue and the consequences on the strategical metagame, this is a fairly good analogy.

  • ScardyBobScardyBob ScardyBob Join Date: 2009-11-25 Member: 69528Forum Admins, Forum Moderators, NS2 Playtester, Squad Five Blue, Reinforced - Shadow, WC 2013 - Shadow
    edited March 2013
    Yeah, other things besides egg-locking can be a problem between close vs cross spawns, but I don't think its a coincidence that the aversion towards close spawns happened at roughly the same time as the alien egg spawning system got overhauled (e.g. the wave-spawning). I fear it may have led to some inadvertent conditioning, where comp NS2 players started associating close spawns with the terrible games that resulted from the early versions of wave-spawning.

    I also understand where you're coming from with the SC2 analogies (as I'm a big fan of RTSs), but I just don't think they translate well into NS2, because NS2 violates too many of the underlying mechanics of RTS gameplay. I'd love for NS2 to take a much more macro focus, but it'd probably take some serious redesign to make it workable.

    For example, the SC2-equivalent of the core 'extractor/harvester on res node' mechanic would be like having max mineral/vespene gas flow upon completion of the command station/hive/nexus. On even the smallest SC2 maps (which are huge in comparison to even the largest NS2 maps; e.g. here is AkilonFlats, a 1v1 map which has the equivalent of 12 tech/res nodes), the optimal early game strategy would devolve into massive res-grabbing expansion because the reward (supercharging your early macro) greatly outweighs the risks (there are more than enough res locations for both sides to drown in res and the distances make it impractical to contest your opponents macro in any meaningful way). If you were Blizzard and stuck with the 'full res flow from one structure' policy, you'd do exactly what UWE did with NS2; severally limit the number of res locations and shrink the map size to bring the expansion risk vs reward back into balance. However, once you do that, micro starts to become the largest factor determining the outcome of a match.

    This is exactly the problem I see with NS2; maps need to be small to ensure a proper balance between risk and reward for expansion, but that leads to games determined primarily by micro (e.g. shooting skill, positioning, where to attack, etc). As such, removing close spawns eliminates a whole selection of potential micro for a mostly inconsequential improvement in macro.

    Rushing and other 'cheese' tactics can certainly still occur on larger maps/cross spawns, but part of the reason increased spawn distance benefits macro is because it reduces the likelihood of these tactics. However, I largely find the success or failure of rushing to be highly dependent on micro, such that something as simple as increasing the distance has a large impact.
  • matsomatso Master of Patches Join Date: 2002-11-05 Member: 7000Members, Forum Moderators, NS2 Developer, Constellation, NS2 Playtester, Squad Five Blue, Squad Five Silver, Squad Five Gold, Reinforced - Shadow, NS2 Community Developer
    The alien egg spawning system in NS2 suffers from both being underpowered AND overpowered. Without shift hive, aliens are extremly prone to egglocking, especially in large a games - average skulk lifetime to avoid egglocking is 7 seconds * number of players, so in 12v12, you need to live on the average twice as long as in a 6v6 to avoid egglock - which is of course ridiculous; I would argue that skulk lifetime goes DOWN with the server size, not up.

    The basic flaw is in how often eggs are generated - if you are out of eggs, eggs are generated at a fixed rate no matter the teamsize.

    WITH shift though, the alien commander can keep spamming eggs (at cost) which guarantees that all aliens always respawn every 15 seconds, allowing them to use "skulk wave attacks" if they so choose. The cost of those eggs are high though.

    Compared to the NS1 system, where each hive basically worked as an infantry portal, spawning one alien every 6 seconds, first come, first serve, and adjusting down those 6 seconds with teamsize.

    NS1 had the advantage that if you died alone, you were pretty certain to spawn back inside a few seconds, instead of the 15 seconds you have to wait now (this, bw, is why Xenocide worked in NS1, and does not work in NS2). Also, there were no way for marines to egg-lock aliens; you knew that you would have one alien coming at you every few seconds, so trying to camp a hive was difficult and required a much larger commitment (especially with hive teleport).
  • 3del!3del! Join Date: 2009-05-11 Member: 67386Members
    Scardy, let's refrain from linking to wikipedia articles that are clearly off-topic, and on top of that use them to support ad hominem arguments from the textbook. Thanks.

    You cannot compare the scale of a Starcraft 2 map to the scale of a Natural Selection 2 map. It can't work for obvious reasons. In any case, if i was to compare map sizes, i'd say SC2 maps are smaller, since units can travel the maps faster. But that's beside the point. What you can compare though, is the setup of the maps. So let's translate the close and cross spawn setup on Summit to a SC2 setup:

    Close Spawn: Bases are only separated by one natural expansion. On the other side of these bases, there are two expansions, which are farther from the opponent, than the main base is.
    Cross Spawn: On each side of the main bases, there is one expansion that is closer to the enemy than the main base, but still in the players side of the map.

    This view leaves tech points out of the equation, since SC2 doesn't have an equivalent to this. Just remember, that one of the expansions on close spawn is a tech point.


    Back to my point: What i was trying to say, is that egg locking happens when marines enter the main hive. A team cannot really egg lock themselves, as i said. The spawns may take a little longer, but unless there are marines in the hive, every spawn wave will spawn at least 2 aliens. So egg locking is not a passive thing that happens circumstantial.
    Therefore you cannot prevent egg locking with higher egg-spawn-rates on close spawns. This would only draw out the time it takes to destroy the hive, once the egg lock is started.
    Apart from that, increasing the spawn rates dependent on spawn would be a really hard to grasp concept. And, as i said, aliens feel more powerful in close spawns anyway, so buffing them wouldn't be a good idea.

    One more anecdote: Before the ensl forced cross spawn, i would always hope to get cross spawns as marines and close spawns as aliens. There was one tactic, that i know some teams played, mine included, where you would go phasegate first, if you had close spawns and arms lab first, if it was cross.

    There are a lot of sound arguments in this thread against close spawns. Apart from "variety" i have yet to see an argument for it.
    BUT, i have an argument against variety, and that is arbitrariness. In case of Summit, there are 12 possible spawns per side, if spawns are totally random. That makes 24 possible scenarios for each team in a match. It's not possible for teams to prepare "special taktiks" for that amount of spawn possibilities. Which in turn makes the early game very generic.

    And a short reminder: we have random spawns in the first place, because of the lack of actually playable maps at the time. I think we're having a good variety of maps coming up, and hope to see more in the future.
  • OnosFactoryOnosFactory New Zealand Join Date: 2008-07-16 Member: 64637Members
    "But if map instead is decided by 100 encounters it will result in a higher likelihood of the better team winning a majority of the engagements"

    ^^ This is a really interersting point.
  • ScrajmScrajm Join Date: 2011-10-17 Member: 127859Members
    About descent.

    1.1. Introduction
    All right, time for descent. Once again I will use the commandments as a basis for the analysis. At a first glance I thought to myself that this map is ready for comp play pretty soon. But the vents are an obvious problem. And yet again, I hope that the responsible map maker could give a reply to what you think about these points I am making.

    If anyone feels that any of my points are wrong in relation to the commandment, feel free to point it out so it can be discussed and hopefully we can reach a consensus.

    1.2. The map
    Descent.png

    2.1. Both sides at all spawn points should always have somewhat easy access to exactly two additional res nodes.
    Ok.
    This map gets ”ok” on this. I think fab and mono might be rough. There are 8 cyst distance from fab to energy, which means holding energy might be tricky. This however gets somewhat mitigated from the fact that once you go save energy, you are actually very “centered” on the map, which actually might be an interesting trade of. Further to defend, but gaining position instead. Compare that to drone bay for instance, where the rts are closer, but worse position wise, which means that when the field players try to save an rt they will have further travel distance until they can be “active” again. I am unsure what I think about this at this point, but it should be a pretty neat trade of, but I’m unsure however of how it plays.

    From mono there are 8 cysts to obs, and 7 to energy. That is a lot for both sides. Cus it’s not just about the fact that it costs cysts early game, it’s also about that it will take longer time to reach them from when you spawn. Also, from mono, you don’t gain that much position when saving obs for instance. Maybe its ok having 7 cyst distance from mono to energy since you gain position, but saving obs will always be a hazzle.

    2.2. If the map is big, the RTs should always be forward on to the map from spawn
    Ok
    This map is HUGE. But thank god, defending RTs from almost all spawn points gives you the possibility to be somewhat active somewhat fast again. The only concern I have here is still observation, which feels like “backwards” from both drone and mono.


    2.3. Both sides need a 2nd hive spot to expand into that cannot be too easy taken down, but not too easy to defend.
    Good.
    At this point it feels like both sides have viable expansion.


    2.4. All tech points should be equally viable for both sides to expand to, given the proximity
    Ok.
    The fact that hydra is so super mega easily seiged basically makes it not viable for aliens as 2nd hive. Ever. However, since aliens can choose not to expand into that room, and instead go for something else as 2nd hive, the map is still playable on this point, making sure this map still gets an “ok” here.


    2.5. Where marines spawn, there mustn´t be more than two entrances, and no vents unless marines can defend the vents.
    Bad.
    This map has horrible layout of the vents. This is actually such a concern that it might basically make the map unplayable. Both launch and mono are like “hey aliens, baserush me plz". The vents from water into mono, and from south of receiving into launch needs to be fixed.

    2.6. The chokepoints cannot be too choky.
    Ok.
    I said earlier that it might be an interesting trade of with spawning in fab, since you sacrifice distance for position to save energy. But another issue with this is that if aliens spawn in fab, and rines in say drone, I feel that marines will have it a little bit too easy attacking energy and plaza. Lets say I have two rines in receiving, and I make a scan in energy. See 3 skulks ready to ambush. Then I can just send them to plaza, and there is nothing the aliens can do about that. Plaza will go down, even though aliens had both good scouting and positioning.

    2.7. No hive should be able to get sieged from an absurd location.
    Ok.
    Hydra is very easily seiged from south of receiving as I mentioned earlier. The reason why it still gets “ok” here is that aliens can expand into something else. So this might not hurt balance per se, but it will reduce options for alien com, making the map slightly more predictable and less varied.

    2.8. There cannot be too many or too few res nodes
    Good.
    Not that much to say about it

    2.9. Spawns should always be far
    Bad.
    This map has random spawns. I have played about 4-5 matches on this map, and early egg lock is so EZ mode it’s boring. If I just invest some res on medpacks with three marines moving into hive, it feels guaranteed as a win. The hives just feel easier to egg lock than other hives on other maps

    2.10. No spot should be bile bombable where marines cannot kill the gorge
    Good.
    At this point I couldn´t find any spot.

    3.1. Comments
    This map got 3 good, 5 ok and 2 bad. If you agree with the underlying commandment, and if you also agree to my take on that commandment in relation to this map, you should also agree to that the bad ones obviously need to be fixed in order for the map to be playable. And these are the spawn and the vents. The spawns can be fixed with the mendasp mod, however, the vents should be fixed asaply.
    I remind the reader once again, that I am in no way saying this is the only way to solve the maps issues. If you feel that these changes I’m suggesting furthers the discrepancy between the map and the commandments, feel free to argue for it.


    3.2. Suggestions
    For starters, the vents leading into mono and launch should be removed.

    My take is that we also need a vent from energy to plaza in order to mitigate the issues aliens will have defending both energy and plaza. This might solve the issue with commandment #6.

    I also believe that maybe observation needs an overhaul in some way. I am unsure how, but the idea is that it shouldn’t feel so “backwards”. Perhaps obs can be moved a little bit “north west”, closer to in between mono and drone?

    We also need a vent from that “new” obs to water treatment after that, so the aliens can flank the marines attacking that RT.

    In general, I have to say that I am unsure of the best placement for the vents. Feel free to suggest more possible vent locations. But my idea of how vents should be placed is that they should be countered by good marine positioning to some extent. There is no way for me as a com to counter a base rush vs mono with good marine positioning, and that is silly.

    My last point is that maybe south of receiving should be reworked so you cannot arc hydra from there. I think it would be cool if aliens could choose to hold hydra as a 2nd if they wanted to, without the fear of silly arc position. I believe that it would add more depth to the map.
    That is my thoughts anyways. I hope you enjoyed reading!
  • swalkswalk Say hello to my little friend. Join Date: 2011-01-20 Member: 78384Members, Squad Five Blue
    ScardyBob wrote: »
    Yeah, other things besides egg-locking can be a problem between close vs cross spawns, but I don't think its a coincidence that the aversion towards close spawns happened at roughly the same time as the alien egg spawning system got overhauled (e.g. the wave-spawning). I fear it may have led to some inadvertent conditioning, where comp NS2 players started associating close spawns with the terrible games that resulted from the early versions of wave-spawning.
    I've been against the close spawns since the day random spawns got implemented and I expressed it on these forums. For the reasons blind said, though he is better at putting it into words than I am.
  • JonacrabJonacrab Join Date: 2003-08-02 Member: 18705Members, NS2 Playtester, NS2 Map Tester
    Both sides at all spawn points should always have somewhat easy access to exactly two additional res nodes.

    while you are correct on the distances, I have modified the starting cysts to shorten the distance to energy flow RT, and to observation RT if you spawn in mono or fabrication. Currently it should be 5 cysts to expand to energy flow from mono or to obs from mono, also from fabrication to energy flow. It's not the best solution but the easiest. This has been in the map from day 1, so im not sure how you are testing cysts.
    This map is HUGE. But thank god, defending RTs from almost all spawn points gives you the possibility to be somewhat active somewhat fast again. The only concern I have here is still observation, which feels like “backwards” from both drone and mono.

    observation is different from the way energy flow connects, Ive had my concerns about keeping it this way, but in the end i think it makes for more interesting/asymetric differences in the map depending on where you spawn, and if i can avoid changing it, i will. Observation has been a concern as well, but im more concerned about the lack of a connection between hydro and obs, its a long walk and if there are suggestions on changes im open, but part of the problem is how i designed the location.... its hard to cut into that satellite relay area without forcing massive changes.

    Hydroanalysis IS an easy siege and that is on purpose. Holding that position on this map is very advantageous and i dont want one team to win just because they can take it. I have made it very assaultable for both teams, with easy access vents for the aliens, and an easy siege for marines. Ive never been comfortable with 5 tech point layouts, they dont really allow you to equalize the distance between ALL hives, and while the "wagon wheel" seems to be a good compromise, 4 tech point maps should be the norm IMO, the issue here is that the game is currently more balanced for 5.
    This map has horrible layout of the vents.

    I dont necessarily disagree, but I dont agree either. The vents on this map have been one of the biggest challenges, because i always wanted the alien team to have a "flank vent" that they could access from the hive room. I never liked forcing aliens to attack a marine assault directly from the front, they need options. This is why the vents are set up like they are now. Ideally i wanted some form of vent welding to be in the game when i released the map... this never happened, however, there is hope in the near future that something may be added in that regard, and if it is added, i would probably make any vent that connected into a tech point location weldable.

    In specific about the vent leading into monorail, it might be good for baserushes, until you consider that the most ideal observatory locations for that room will show any baserush coming from a mile away, and i recently modified the vent coming out of mono, and it should now be a mineable vent exit.

    Other vents for the time being I wont change unless there is more of a consensus that it needs to be changed. If there is a consensus, some suggestions would help, I still want to have a vent coming out of each hive for the sake of flanking, and its not something i want to just remove, but I will if necessary.
  • xtcmenxtcmen Join Date: 2004-04-20 Member: 28040Members, Squad Five Blue
    Interesting criticism. I like it.
Sign In or Register to comment.