Possible idea to help resolve "pubstomping"

simbasimba Join Date: 2012-05-06 Member: 151628Members
Many suggestions I have seen are very detrimental and harsh to either a new player or a veteran player. I am a strong believer that any solution to this should not involve banning players from servers because they're too good or too bad. The game should not have any setbacks on launching the game and just joining a server to play. Many people assume veterans get on a pub server and maliciously kill everyone of lesser skill than they for the "lulz" and purposefully try to empty a server. 

My suggestion requires both work from UWE and work from server managers. The first part is UWE needs to come up with a "ranking" system that does some algorithmic math to determine as accurately as possible, a players "skill" relative to other players. If UWE wants these ranking to be publicly visible, that's fine. It can also remain hidden from players to not be a distraction. Every server would need an update to display in some way, the average of the player skill on a given server in the server browser. If we're trying to hide the "ranking" from the players, merely an indicator of "this server has players that are ranked much higher than you" or "this server has players that are on near your ranking" or "this server has players that are slightly lower ranked than you" would suffice. 

This gives players the ability to see on the server browser what they're getting into when joining a server. So if they're new, they'll know there's a good chance they'll get dominated or that the server has many other newbies like them. This would also allow server owners to optionally limit players that are too low or too high in relative skill from joining. The hardest part of this is coming up with a good "ranking" algorithm. I did a quick google search on the topic and there comes a few examples and ideas on how to create a relative ranking system and I don't imagine it's out of UWE's reach to implement. 

This solution in my opinion lets the new or veteran player join servers that are more enjoyable for them. If server admins want to limit players, it's always relative, so if you're not allowed to join a server one night, the next day the relative skill might have changed and you'll be allowed. It will still allow new players to join veteran servers (if they want a challenge and that server allows it). It also allows veterans to pub stomp if they really want to (if the server allows it). It allows players to have an idea ahead of time what they're getting into. A lot of the time a veteran player doesn't join a server knowing they're going to be 100x better than everyone else and pub stomp and it happens on accident. Same applies to a new player. This just gives more information to a player to make a better informed decision. It also gives server operators a tool to better tailor the server environment that they wish. 

Comments

  • BestProfileNameBestProfileName Join Date: 2013-01-03 Member: 177320Members
    The problem with ranking is you need the player base to do it, we simply do not.
    Don't understand the logic behind that. Might as well not have a competitive scene with that logic.


    I agree with the idea of a ranking system. It won't be perfect, so any comments saying, "well kd isn't necessarily indicative of skill"...Yes we know that.

    But perhaps an algorithm that takes into account, as a working definition, the following:
    1. Number of hours played
    2. Registered in the ENSL
    3. Kill/death ratio
    4. Average score per minute

    It's just a start. I'm not saying it is easy. We can split atoms, though, and we already have the NS2stats thing, which is a start. Players registered to ENSL by steamid or something must be really easy. Number of hours is really easy. Now all you need is to make sure it is built into the client that k/d and avg score/min is logged.

    Doable? I think so. Favourable for the community? Yes. I can't tell you how many people in that free weekend were turned away from the game for the reasons the OP has suggested. Equally, a lot of people do not persist because a substantial number of rounds are very lopsided and thus of little fun.
  • Ghosthree3Ghosthree3 Join Date: 2010-02-13 Member: 70557Members, Reinforced - Supporter
    edited May 2013

    The problem with ranking is you need the player base to do it, we simply do not.
    Don't understand the logic behind that. Might as well not have a competitive scene with that logic.

    With more numbers the skill would be easier to seperate, with the low players that we have, the difference in skill shown in the servers in YOUR area (100 ping or less) probably won't vary much.
    It could be added sure, and might occasionally work out, but I'm not sure a system this sophisticated would be worth the time and resources to probably not help the community out any more than the issue is being helped already by people joining the servers every night and knowing the community that plays on it. Also it's another thing that players WILL be able to whine about in some way. Eg. "Your team only won because you had that guy with a really low ranking! Go join another server!" Or a kick.

    EDIT: There's also the fact that some high skilled players will intentionally (or could) join low skilled servers to rape face.
  • simbasimba Join Date: 2012-05-06 Member: 151628Members
    Yes, at the beginning, with the current population, the score wont help much because in your area, there might be only 10 servers operating in your area. But with this change, it will help new players choose better which server they want to be and possibly increase the population. Once population increases, then this tool will be more valuable as there's most options for servers. 

    It is problematic to assume the community alone can solve this problem because there are only so many tools we can do. Sure someone can make a mod that does all of this, however we can't force every server to use it, therefore it wont be effective because not every server WILL use it. If it is in vanilla, it will be used. 

    For calculating a players "ranking", there are many many ways. This is why I mentioned a quick google search. You can look at some documents on Microsoft's ranking system for Call of Duty. You can look at how other games do it. It's not trivial and generally a lot of mathematics is involved. If it's not crucial for that amount of detail, yes it can be simpler. Factors that should be taken into account for determining a player's rank:

    -number of hours played
    -K/D of last 50 games (or 100 or w/e) the idea is players get better so old k/d should decay off the ranking decider
    -Number of game wins (because many people play mostly commander)
    -score per minute
    -activity (hours played in a given period of time)

    I don't see the ENSL website being much help to determine a ranking. Many are inactive and it would require a lot of extra work from ENSL to maintain some accurate representation. 
  • ijustpwnedu96ijustpwnedu96 Join Date: 2013-01-27 Member: 182320Members
    Good idea. However, the factors you listed to determine whether or not a player is a veteran or not can be inaccurate.

    For example, hours played can be manipulated by idling in a server without auto-kick. I have played for 335 hours without idling, but my K/D is not that good. Also, I have lost many games not because I was a rookie or a veteran, but because we lacked teamwork or other factors that made us lose the game. For scores, a rookie can get lots of points by killing structures such as extractors easily. 
  • simbasimba Join Date: 2012-05-06 Member: 151628Members
    Good idea. However, the factors you listed to determine whether or not a player is a veteran or not can be inaccurate.

    For example, hours played can be manipulated by idling in a server without auto-kick. I have played for 335 hours without idling, but my K/D is not that good. Also, I have lost many games not because I was a rookie or a veteran, but because we lacked teamwork or other factors that made us lose the game. For scores, a rookie can get lots of points by killing structures such as extractors easily. 
    There is not much else that is tangible from a programming perspective to estimate a players "rank". Sure, all of those criteria can be inaccurate, it is going to be generally equally distributed inaccuracy across all players. Good players play on bad teams all of the time. The total wins wont factor in to that much of the score. Also, if a new player is racking up points by "just killing harvesters", that, to me, is being good at the game. You know what you should be doing and how to win the game. Why is that an inaccurate reflection of player ranking? 
  • piratedavepiratedave Join Date: 2012-03-10 Member: 148561Members
    some players on NS2stats pad their score by pubstomping on rookie servers with NS2stats installed (silly server admins)

    you know, if new players could

    A) learn the game in relative peace
    B) be in a low pressure environment so they could appreciate the atmospherics
    C) experience at least 1 very fun good game

    i bet you more noobs would stick around. Everyone i have gotten to purchase this game, have quit, even though they thought it was awesome at 1st (they said ...performance, bad balance(team stacking), hackers (pubstompers), too complex, couldnt get into it ... were some of the reasons they came up with)

    im getting dejavu, its like NS1 all over again, a slowly declining player base


  • simbasimba Join Date: 2012-05-06 Member: 151628Members
    edited May 2013
    some players on NS2stats pad their score by pubstomping on rookie servers with NS2stats installed (silly server admins)

    you know, if new players could

    A) learn the game in relative peace
    B) be in a low pressure environment so they could appreciate the atmospherics
    C) experience at least 1 very fun good game

    i bet you more noobs would stick around. Everyone i have gotten to purchase this game, have quit, even though they thought it was awesome at 1st (they said ...performance, bad balance(team stacking), hackers (pubstompers), too complex, couldnt get into it ... were some of the reasons they came up with)

    im getting dejavu, its like NS1 all over again, a slowly declining player base


    We're not talking about ns2 stats. We're also not talking about "evil pubstompers who log in only to improve their ns2 stats standing" which is itself a poor generalization and misleading assumption. I doubt many, if any, people care about their ns2 stats standing in this game. We're not identifying the problem new players have with the game. 

    The problem is identified, we're offering criticism to the idea originally posted by myself (OP) in this thread that is a first step attempt to help rectify the problem. We're discussing a ranking system type implementation into vanilla that offers players extra tools they can use to make better server decisions that might increase their overall experience in NS2. 
  • 0ni0ni Join Date: 2012-08-30 Member: 156991Members
    We've got a similar discussion going on our own forums.

    The idea being that if we track each individual users stats on each team and apply a custom nerf to that player based on their k/d ratio. When they go over a 2:1 ratio they earn a 1 star badge and bullets do 1 less damage and bites do 10% less. If they can maintain that 2+:1 for a while they get a 2 star badge and so on and so on. The exact numbers and balance would be whatever they need to, I'm no expert in this. The reason I say k/d ratio instead of other variables is because the real problem with pub stomping is that these players just slaughter less skilled players solo, they don't win by helping build structures or repairing things. A pubstomper is a guy with a jetpack and a shotgun, not a gorge. :p

    Anyway, I think this would allow comp players to play on well populated/low ping pub servers on the team of their choosing without ruining the game for the people already playing there. I understand a lot of comp players prefer Marines and as a person who greatly prefers alien I understand how it feels to be forced onto the team you'd rather not play. At the same time it would provide a much greater challenge for these players of exceptional skill.

    Another fun side effect would be what it would do to aimbotters. Can you imagine the rage when an aimbotter empties a whole clip into a skulk and doesn't get a kill?

    This mostly came about because we don't want to ban players for being skilled but at the same time they ruin games and they tend to stack marines when there is more then one.

    So this is just an idea. I just think it's less invasive then any team balance style mod because it allows people to play on the team of their choice.

    Thoughts?
  • TinkerTinker Join Date: 2003-03-11 Member: 14395Members
    I would never play on that server.
    You're doing well, here's a penalty.

    Tactical gamer uses a balance mod that tracks players wins on the server and puts teams together that way (When triggered by an admin). It's not perfect but it's several times better than random in my experience. Between that, an active a reasonable admin system, and the server rules requiring players be able to communicate with their team there aren't too many straight stomps on the server.
  • ezekelezekel Join Date: 2012-11-29 Member: 173589Members, NS2 Map Tester
    edited May 2013
    Hours don't determine skill
    Being registered in ENSL does not determine skill
    K/D barely determines skill
    Total matches won, can be used to determine skill but still barely
    Accuracy can be determined to skill
    Completing objectives can be determined to skill

    You don't want anything that requires individuality to determine skill, hence why ns2stats is garbage; I literally avoid servers with that, sure it's useful info, but stats in general make people care about themselves and not winning. So anything that requires teamwork to determine skill is probably the best bet, because the goal is to win the match, not to kill 50 players and die once

    All in all a system can be put in place, when someone keeps winning rank them higher, then they will face other people who are also constantly winning. And eventually you'd get...? That's right, two teams with competition! Just logging into this game everyday and seeing no servers for me to play on.. so what do I do, I just exit the game really

    I have no interest in anything above 16 slots (minus balance test, i love playing there) the game simply goes out the window with so many players, dying becomes a non-issue, there's so much more room for mistakes, there's an extra five people around the corner, game performance worsens and blah blah I just hate it
  • BacillusBacillus Join Date: 2006-11-02 Member: 58241Members
    I think having the possibility to easily force random teams and making sure uneven games are finished quick enough are the best approach right now.
  • turtsmcgurtturtsmcgurt Join Date: 2012-11-01 Member: 165456Members, Reinforced - Supporter
    i'd rather the developers not waste their time working on this and put it toward something useful, such as performance fixes.
  • ezekelezekel Join Date: 2012-11-29 Member: 173589Members, NS2 Map Tester
    i'd rather the developers not waste their time working on this and put it toward something useful, such as performance fixes.

    I'm 110% sure that's already being worked on, and I believe sewlek is doing most of this so technically I don't think any time is being wasted?
  • 0ni0ni Join Date: 2012-08-30 Member: 156991Members
    ezekel wrote: »
    Hours don't determine skill
    Being registered in ENSL does not determine skill
    K/D barely determines skill
    Total matches won, can be used to determine skill but still barely
    Accuracy can be determined to skill
    Completing objectives can be determined to skill

    You don't want anything that requires individuality to determine skill, hence why ns2stats is garbage; I literally avoid servers with that, sure it's useful info, but stats in general make people care about themselves and not winning. So anything that requires teamwork to determine skill is probably the best bet, because the goal is to win the match, not to kill 50 players and die once

    All in all a system can be put in place, when someone keeps winning rank them higher, then they will face other people who are also constantly winning. And eventually you'd get...? That's right, two teams with competition! Just logging into this game everyday and seeing no servers for me to play on.. so what do I do, I just exit the game really

    I have no interest in anything above 16 slots (minus balance test, i love playing there) the game simply goes out the window with so many players, dying becomes a non-issue, there's so much more room for mistakes, there's an extra five people around the corner, game performance worsens and blah blah I just hate it

    I think the point of this thread isn't to determine who's more skilled or how to quantify it. It's more about how do we stop pub stomping from being an issue.

    Pub stomping for the most part is just one or more really good players who just slaughter less skilled players. There's some really skilled players with a good sense of strategy and some with none, so it's really hard for anyone to measure that. Hence in formulating my thoughts I figured the best way to make sure it would only apply to comp level players was to only affect a players damage out put. By keeping it to one variable you can avoid unnecessary complications. I mean I could be totally wrong, I'm just trying to explain my train of thought.

    The other side of it was forcing people onto teams they don't want to be on blows. I'd imagine most comp players given the choice between a handicap and playing on their team of choice would usually play on their team of choice. Again, I don't really know that for sure. I myself usually hold just positive of a 1:1 so I'm hardly qualified to make the call on what comp players prefer. That said, I don't enjoy playing on Marines. Partially because I'm not a great shot but mostly because running on the roof and swooping in for a quick bite is wicked fun. That's how I enjoy this game, others are different. Some people don't care what team they're on and that's great.

    I just would like to see a system where everyone can play the way they want and have FUN on pub servers. I personally don't enjoy being forced onto marines and I don't enjoy being crushed every round by comp level marines. I like when the game is decided by well timed strategies and team work rather then a couple of clan team mates who want to play on the same team together but easily out class everyone else. Why shouldn't they play on the same team though? They're team mates and maybe even friends, it's part of their fun too, but in doing so they can ruin a server. There has to be a way to make these things work. Maybe not MY way but surely we can think of something better then what we have. /random totally sucks skulk balls. :p
  • simbasimba Join Date: 2012-05-06 Member: 151628Members
    That is exactly the point, Oni. You can most definitely use hours played as a value in this "ranking" system. If you've put in 1000 hours in ns2, and you still suck, It doesn't matter to this ranking system. It will still match you up against higher ranked players because you've played the game that long. You know how the game is supposed to be played, and you're relatively good at the game. It's not measuring skill. It's measuring your "capability to play". It's to let new players learn with other relatively new players and let those who know how to play, play with others who know how to play.

    How "skilled" you are is another matter.
  • WakeWake Join Date: 2003-03-05 Member: 14351Members, Constellation
    The deal would be to introduce negative handicap or positive boost based on indicators / metrics from the current game.

    These metrics would have to put in evidence a strong balance default in order to trigger the effect.
    It could obviously be pertinent only in the early stages of game.

    You would have also to make sure that a team is not able to lure the system in the early stage so they earn an advantage based on a faked inferiority.
    In another hand, if a superior / stacked team restrain itslef from stomping at the very start, goal is somewhat achieved.

    The boost or handicap should be limited in its effects in weight and/or duration.
    - What are the metrics available ? whole teams K/D, RT's builded, collected res ratio ...
    - You would need a minimum time in order to get enough metrics, 3(?) minutes.
    - If the compensation is triggered, then it should last no more than 3(?) minutes.
    - It could affect spawn / eggs hatch rates, trigger a fake team auto-balance calculation (on 8v8, lure the system it's 6v8)
  • maD_maX_maD_maX_ Join Date: 2013-04-07 Member: 184678Members
    This idea has been tossed around forever, and is nothing new, as everyone has said we dont have a player base....

    Solution, create the ranking system keep it hidden, and tie it into RANDOM... such that RANDOM does not stack.... Of coarse there would need to be a randomizer function as well or the teams would repeat match after match so a tolerance would be needed where teams are random but total value of player skill cannot differ by more then 5-10%
  • ezekelezekel Join Date: 2012-11-29 Member: 173589Members, NS2 Map Tester
    In my BT suggestion thread I suggested removing the rookie filter, extending player rookie time to 15 hours, if a server wants to be rookie friendly they can include it in the server title

    And adding a competitive filter, which filters out those specific servers only and they require a 100-150+ hour playtime to join, surely the matches may not be even but at least the player will by then have some type of game sense even if they're still no good
  • mushookeesmushookees Join Date: 2008-03-26 Member: 63967Members
    0ni wrote: »
    We've got a similar discussion going on our own forums.

    The idea being that if we track each individual users stats on each team and apply a custom nerf to that player based on their k/d ratio. When they go over a 2:1 ratio they earn a 1 star badge and bullets do 1 less damage and bites do 10% less. If they can maintain that 2+:1 for a while they get a 2 star badge and so on and so on. The exact numbers and balance would be whatever they need to, I'm no expert in this. The reason I say k/d ratio instead of other variables is because the real problem with pub stomping is that these players just slaughter less skilled players solo, they don't win by helping build structures or repairing things. A pubstomper is a guy with a jetpack and a shotgun, not a gorge. :p

    Anyway, I think this would allow comp players to play on well populated/low ping pub servers on the team of their choosing without ruining the game for the people already playing there. I understand a lot of comp players prefer Marines and as a person who greatly prefers alien I understand how it feels to be forced onto the team you'd rather not play. At the same time it would provide a much greater challenge for these players of exceptional skill.

    Another fun side effect would be what it would do to aimbotters. Can you imagine the rage when an aimbotter empties a whole clip into a skulk and doesn't get a kill?

    This mostly came about because we don't want to ban players for being skilled but at the same time they ruin games and they tend to stack marines when there is more then one.

    So this is just an idea. I just think it's less invasive then any team balance style mod because it allows people to play on the team of their choice.

    Thoughts?

    I think its genius. Rookie servers or noob friendly servers desperately need something like this, I think anybody that would have a problem with this is probably one of those people who enjoy pub stomping on rookie servers :P

    My suggestion is that you make it clear to people that they are being handicapped and why they are being handicapped, eg. Have a message show up saying " Your k/d is above 2:1, your damage will be reduced by xx%, this is a rookie friendly server using xxxxxx mod"
  • ScardyBobScardyBob ScardyBob Join Date: 2009-11-25 Member: 69528Forum Admins, Forum Moderators, NS2 Playtester, Squad Five Blue, Reinforced - Shadow, WC 2013 - Shadow
    KISS

    Elo-rating based solely on win/loss data (as that's the primary goal of playing NS2). Keep the ratings secret, but sort players into groups/leagues. Then add a column to the server browser that shows the number of players in the same or near group/league in that server.
  • ezekelezekel Join Date: 2012-11-29 Member: 173589Members, NS2 Map Tester
    ScardyBob wrote: »
    KISS

    Elo-rating based solely on win/loss data (as that's the primary goal of playing NS2). Keep the ratings secret, but sort players into groups/leagues. Then add a column to the server browser that shows the number of players in the same or near group/league in that server.

    Win + loss / objectives completed / accuracy / assists can all be combined as well

    A team player with good accuracy and a high win ratio can prove the players skill
  • simbasimba Join Date: 2012-05-06 Member: 151628Members
    ezekel wrote: »
    ScardyBob wrote: »
    KISS

    Elo-rating based solely on win/loss data (as that's the primary goal of playing NS2). Keep the ratings secret, but sort players into groups/leagues. Then add a column to the server browser that shows the number of players in the same or near group/league in that server.

    Win + loss / objectives completed / accuracy / assists can all be combined as well

    A team player with good accuracy and a high win ratio can prove the players skill

    His is very similar to my original idea. I agree that sole ranking on k/d may not be best. Especially considering some players might be strong comm, and don't get a K/D. Either way, the ranking system wont force people onto servers, it just will give the server browser another visual queue for players to look at when joining a server and make a more educated decision about which servers to play on. Now veterans can't join and say they didn't know the server would be full of rooks, and rooks don't accidentally join a server that stomps their face in and isn't fun.
Sign In or Register to comment.