Yeah, those are valid points. Just like if you give points for killing a structure, what if someone comes in at the last minute and shoots the hive when you have been shooting it down from 100%. But like I said, there is no perfect system. You can't take into account all of the complexities of the game and determine exactly how much points you should get, because every situation is different. It's much harder to kill a skulk when there are 4 of them attacking you than when one is just sitting still in the middle of a corridor. But it's impossible to have a perfect system so you just have a simple system and accept that it's imperfect.
Perhaps they could add in some form of assit system, idk. Generally I find the better players tend to get the most kills anyway, because they are the ones doing the most damage and are therefore the most likely to get the kill. Kill steals happen sometimes but I don't know if it's such a big deal that it needs an assist system in the rankings. It will work itself out over time anyway, when the noob that got the kill steal dies 5 times over to the better player.
Whatever they decide to do I think they should keep it simple as possible.
DghelneshiAims to surpass Fana in post edits.Join Date: 2011-11-01Member: 130634Members, Squad Five Blue, Reinforced - Shadow
edited August 2012
I would actually be interested in stats for spawn combinations instead of just stats for each spawn point. It's a lot more data, but more meaningful.
Accuracy stats could also be cool, but also potentially bring down server performance. It could possibly be mitigated by trying to use local variables as a cache.
<!--quoteo--><div class='quotetop'>QUOTE </div><div class='quotemain'><!--quotec-->I don't think you should get points for wins/losses as it encourages stacking and it's too hard to work out exactly what that win or loss is worth. I feel like it would punish players for being on a team with noobs when there is nothing they can do to avoid losing points.<!--QuoteEnd--></div><!--QuoteEEnd-->
win/loss is absolutely needed imo. it is the reward for contributing for the greater cause, covering things other stats can not e.g. sacrificing yourself to save a phasegate. i agree stacking could be a problem. if they use ELO to actually balance out the teams this should be fine right? as long as there is some motivation to stick with your team instead or switching to the winners (which is rarely possible anyway).
<!--quoteo--><div class='quotetop'>QUOTE </div><div class='quotemain'><!--quotec-->You can't take into account all of the complexities of the game and determine exactly how much points you should get, because every situation is different.<!--QuoteEnd--></div><!--QuoteEEnd--> this will always be true. but if you create a final ELO-score, you should at least try hard imo. the more stats you combine, the more situations (possible contributions) you can cover. and the more simple it is, the easier it is for people to find obvious exploitation methods.
two more things i'd like to bring up: what happens if the versions change? it is probably not a big deal i guess, just asking. what happens if the servers have other mods activated? this is currently not an issue, but certain mods might drastically alter certain stats like the score.
Are there any elo ratings for games without any proper rules or systems behind it? I dont think so, because its kinda worthless...
For proper elo raitings you need rules, 12 players, 2 commanders 10fieldplayers, no afking, no leaving, no winning team joining, no stacking(if its not something like "party ranked") best of 3 or whatever. Set timers to start, pause rounds etc(so you dont end up in the ready room for 30min and a playtime of 10min), also you should have a reconnect feature in case of crashes, or timeouts etc. server needs to have relative fair pings and proper performance, you basically need a freaking big system behind all this so you cant cheat it and ppl can get punished by ###### up other players experience, and to be as fair as possible.
supsuJoin Date: 2012-04-24Member: 151023Members, Squad Five Blue
edited August 2012
<!--quoteo(post=1963902:date=Aug 16 2012, 02:21 PM:name=Laosh'Ra)--><div class='quotetop'>QUOTE (Laosh'Ra @ Aug 16 2012, 02:21 PM) <a href="index.php?act=findpost&pid=1963902"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->two more things i'd like to bring up: what happens if the versions change? it is probably not a big deal i guess, just asking. what happens if the servers have other mods activated? this is currently not an issue, but certain mods might drastically alter certain stats like the score.<!--QuoteEnd--></div><!--QuoteEEnd--> By versions I think you mean ns2 patches and they shouldn't be that big of a problem: We only have some function calls in some of the original ns2-lua files so it won't take us more than 15 minutes to make a new version suitable for the new patch and server admins just need to update their stats mod. We'll probably try to make this even more easier in the future but it isn't our first priority at the moment.
And there shouldn't be any conflicts with another mods unless they use the same ns2-lua files that we do and in case there's a conflict you can just contact me and I'll make a version where your mod and our mod will merge which shouldn't take more than couple of minutes to create.
<!--quoteo(post=1963909:date=Aug 16 2012, 02:52 PM:name=Koruyo)--><div class='quotetop'>QUOTE (Koruyo @ Aug 16 2012, 02:52 PM) <a href="index.php?act=findpost&pid=1963909"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Are there any elo ratings for games without any proper rules or systems behind it? I dont think so, because its kinda worthless...
For proper elo raitings you need rules, 12 players, 2 commanders 10fieldplayers, no afking, no leaving, best of 3 or whatever. Set timers to start, pause rounds etc(so you dont end up in the ready room for 30min and a playtime of 10min), also you should have a reconnect feature in case of crashes, or timeouts etc. server needs to have relative fair pings and proper performance, you basically need a freaking big system behind all this so you cant cheat it and you can get punished by ###### up other players games.<!--QuoteEnd--></div><!--QuoteEEnd--> We'll make separate elo-ranking for competitive matches that are solely based on PCWs and gathers, or purely PCWs if the community grows, and even for clans in the future but there'll always be public elo-ranking too even if it won't be anywhere as accurate as competitive elo-ranking but it'll give nice touch to player stats and it certainly shouldn't ever be taken too seriously, unless you take public seriously.
<!--quoteo(post=1963720:date=Aug 16 2012, 03:03 AM:name=Dghelneshi)--><div class='quotetop'>QUOTE (Dghelneshi @ Aug 16 2012, 03:03 AM) <a href="index.php?act=findpost&pid=1963720"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->I would actually be interested in stats for spawn combinations instead of just stats for each spawn point. It's a lot more data, but more meaningful.
Accuracy stats could also be cool, but also potentially bring down server performance. It could possibly be mitigated by trying to use local variables as a cache.<!--QuoteEnd--></div><!--QuoteEEnd--> And yes you are right, we can make that spawn combination stat pretty easily and probably will be added soon. And accuracy will probably try to be done in the future but we have to look out not to decrease server performance; we'll probably even make it so that you can see your accuracy with rifle or with shotgun or with bite etc.
<!--quoteo--><div class='quotetop'>QUOTE </div><div class='quotemain'><!--quotec-->And there shouldn't be any conflicts with another mods unless they use the same ns2-lua files that we do and in case there's a conflict you can just contact me and I'll make a version where your mod and our mod will merge which shouldn't take more than couple of minutes to create.<!--QuoteEnd--></div><!--QuoteEEnd--> thats good to hear, but i actually wanted to point out a different issue. to give an exaggerated scenario: "play on our server, we have a mod to double your score!" i doubt anyone will do anything like this intentionally, but it could be a side effect. e.g. you will not get any building stats when playing the already working combat mod.
The database probably was crashed when you tried to access the site; I just tested it, the website works fine for me. You should try again to open it now and it will probably work for you, too.
supsuJoin Date: 2012-04-24Member: 151023Members, Squad Five Blue
edited August 2012
Yeah, database was down for couple of hours today, it should be back up now.
<!--quoteo(post=1963972:date=Aug 16 2012, 06:39 PM:name=Laosh'Ra)--><div class='quotetop'>QUOTE (Laosh'Ra @ Aug 16 2012, 06:39 PM) <a href="index.php?act=findpost&pid=1963972"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->thats good to hear, but i actually wanted to point out a different issue. to give an exaggerated scenario: "play on our server, we have a mod to double your score!" i doubt anyone will do anything like this intentionally, but it could be a side effect. e.g. you will not get any building stats when playing the already working combat mod.<!--QuoteEnd--></div><!--QuoteEEnd--> Score/kill cheating is possible but if we notice something like that we can ip-ban that server and ignore the stats that server has sent.
<!--quoteo(post=1963986:date=Aug 16 2012, 07:33 PM:name=DualSight)--><div class='quotetop'>QUOTE (DualSight @ Aug 16 2012, 07:33 PM) <a href="index.php?act=findpost&pid=1963986"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->For the player section can you put a search function thanks.<!--QuoteEnd--></div><!--QuoteEEnd--> There's already search but it's hard to notice; it's on top of the playerlist in players-page. We'll make search feature better in soon; the site is no-where finished and almost 90% will be different looking in 4-5months - site has been under construction for a little over a week now, so patience. :>
<!--quoteo(post=1963983:date=Aug 16 2012, 09:23 AM:name=Angelo85)--><div class='quotetop'>QUOTE (Angelo85 @ Aug 16 2012, 09:23 AM) <a href="index.php?act=findpost&pid=1963983"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->The database probably was crashed when you tried to access the site; I just tested it, the website works fine for me. You should try again to open it now and it will probably work for you, too.<!--QuoteEnd--></div><!--QuoteEEnd-->
Now detailed map stats sound awesome. Great tactical tool. But just imagine recording entire gameplay (in a simplified way ofc), and then.. playing it back in the browser on a map overview, with a time slider, see how the game progressed with positions of players/buildings, meds dropped, given orders etc.
Now detailed map stats sound awesome. Great tactical tool. But just imagine recording entire gameplay (in a simplified way ofc), and then.. playing it back in the browser on a map overview, with a time slider, see how the game progressed with positions of players/buildings, meds dropped, given orders etc.
mind = blown :D<!--QuoteEnd--></div><!--QuoteEEnd--> Data for this is already being recorded and all this has already been planned and will definitely be on our site at some point, will take a lot work so won't happen for couple of months but it'll happen. :>
Coool! I would really die to see that, especially that all competitive games take place on the other side of the planet so can't really participate in first person due to ping 400 haha. So following the strategical overview while watching the streaming would be killer
Perhaps I might be of some assistance on the front-end side after all.. ;)
I installed it on my server inside the DAK folder as it's the mod I'm using and while it appeared to work as it displayed the welcome info of the stats mod on join, it crashed the server after a few minutes of the server starting the match, had to take it off.
<!--quoteo(post=1963902:date=Aug 16 2012, 12:21 PM:name=Laosh'Ra)--><div class='quotetop'>QUOTE (Laosh'Ra @ Aug 16 2012, 12:21 PM) <a href="index.php?act=findpost&pid=1963902"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->win/loss is absolutely needed imo. it is the reward for contributing for the greater cause, covering things other stats can not e.g. sacrificing yourself to save a phasegate. i agree stacking could be a problem. if they use ELO to actually balance out the teams this should be fine right? as long as there is some motivation to stick with your team instead or switching to the winners (which is rarely possible anyway).<!--QuoteEnd--></div><!--QuoteEEnd-->
IMO you just need to accept that imperfections in the system. Wins/loss will just lead to stacking in public games. If you sacrifice yourself for a phase gate then tough ######. Again, there will always be situations where you lose points where you "shouldn't" have. How often do those situations really come up? It would lead to a small difference in points, but it's not really a big deal. Most people want to win the game and aren't going to care about a tiny amount of points being lost from an individual encounter. If you lose the phase gate it will probably lead to you dying more often later in the game and then losing more points in the end anyway. I doubt most servers will use the rankings to balance public games, what happens when people join in the middle of a match etc.
<!--quoteo--><div class='quotetop'>QUOTE </div><div class='quotemain'><!--quotec-->this will always be true. but if you create a final ELO-score, you should at least try hard imo. the more stats you combine, the more situations (possible contributions) you can cover. and the more simple it is, the easier it is for people to find obvious exploitation methods.<!--QuoteEnd--></div><!--QuoteEEnd-->
I completely disagree. This is for public games. If you keep it simple with kills/deaths and buildings then you really can't exploit it. Playing the game and getting more kills will be the way to get points, which is what people do anyway. Trying to create this near perfect system for public games is just futile. The public ranking is just a fun thing it's not a big deal and shouldn't be obsessed over to try and make it perfect.
how can it lead to excessive stacking if the teams are autobalanced by ELO? and why would they not be stacked in comparison, if the goal is kdr and score (which is also heavily influenced by having a good team)?
i am concerned with very low morale (and i've seen in other threads that several people share this concern) as in people giving up risks in favor of turtling the rest of the match in order to gain some kills and score... the other team possibly trying to prolong the game as long as possible (spawnkill etc.) will further encourage this behavior.
in fact i have seen this issue today on one of the stat-servers. our second marine base was heavily attacked (power down). i tried to motivate some teammates to come with me to take it back, 2 guys refused (they even had jetpacks) as in "it's already lost". we managed to take it back on the first try though, even with low numbers. another time were 3-4 people standing around in the base, not going through the phase gate as the other side was attacked. they were either not aware of this for some reason (although it was pointed out via voice) or nobody wanted to go first into this death-trap. of course it will never be flawless. but there is no point in adding extra flaws imo. if stacking will be a problem then i'd really prefer elo-based teams, or simply randomly assigned teams. people might be annoyed as in "damn i wanted to play on the other side this time" but it's better than stacking or people putting their greed above the team effort.
found the major thread where people heavily discussed about this (~10k views, 14 pages): <a href="http://www.unknownworlds.com/ns2/forums/index.php?showtopic=117168" target="_blank">http://www.unknownworlds.com/ns2/forums/in...howtopic=117168</a> it also holds some points for your side, i just want to point out several people apparently have very strong oppinions about this.
This is why ranking based on stats in pubs is a bad idea. I like the idea of being able to go and check out the stats based on lifetime, 30 days, 1 day, etc. Any attempt at ranking these beyond time-function "Most Kills in xx time" is going to end up poorly.
I saw this in global agenda with ASM-whoring. ASM was a weighted mechanic that took into account about a hundred different factors such as damage, kills, deaths, win, loss, building damage, objective points. It also weighed the teams. IE if you had 1000 ASM and were on a team with nine other 10 ASM players (noobs), but the other team was 10 1000 ASM players, then you would not lose very much ASM when your team inevitably lost. Eventually, ASM could actually predict the winner of the match in a "slightly-favored, steamroll, or even" style with decent accuracy. The creator of ASM tried to rank people via their points. This lead to people making characters, playing the minimum 30 games to get ranked, and playing super-meter-###### for those 30 games, then quitting for the month. I don't think there is any way to create a ranking system without this happening.
I think ELO should be kept to matches and be team based, if anything. I don't think the competitive community is to the point where we need a strict ranking for anything.
stats in general are a bad idea honestly i have no idea how you guys can be surprised by the things that result from stats like this when they have been in so many games previously and always cause the same issues....
Either you accept the problems the stats bring or you dont log them.
Having an ELO is honestly completely pointless in NS2.. its pretty obvious who the best players are, we dont need some complicated numerical system to find out the exact same information. Maybe if the playercount increases by about 1000x, but honestly its more fun to play the game not knowing exactly how good your opponents might play.
because it will not fix team stacking in any way, you cannot fix team stacking in a game like NS2... Are you going to prevent two people that are higher ELO from playing on the same team? Sounds to me like I would be finding a new server to play on...
ELO/truskill rankings and ratings IMO are something that should be exclusive to competitive play only. And even then the nature of the game makes it pretty hard, if not impossible, to perceive skill by looking at stats alone.
Don't think I've ever played a game that was more enjoyable with public stats, only less enjoyable because of people trying to farm or protect their score.
ScardyBobScardyBobJoin Date: 2009-11-25Member: 69528Forum Admins, Forum Moderators, NS2 Playtester, Squad Five Blue, Reinforced - Shadow, WC 2013 - Shadow
<!--quoteo(post=1963660:date=Aug 15 2012, 02:22 PM:name=Sint)--><div class='quotetop'>QUOTE (Sint @ Aug 15 2012, 02:22 PM) <a href="index.php?act=findpost&pid=1963660"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Yes ELO rating is pretty tricky to do in NS2, it could be made simply win/loss, but that has several problems. One of the worst is that it encourages stacking. You would get better ELO rating if you stack on pubs always. So to get good rating you would need to know who is good and stack with them or play with friends. Another problem is that when you play with team which has 1500 average ELO rating vs opposite team that has same average rating, then when your team loses you would get lot of negative rating if your own rating would be for example 1800, even if you had 3 times the score of 2nd best player in your team.
So we have some ideas to make system which uses ELO rating but points what you get will depend also on your ranking inside your team. This ranking could be based on score for example, but maybe also something else. So for example first person in scoreboard would get 15 points, second 14.5 points etc. This score would then be reduced or increased by ELO ratings when we compare marine team average ELO to alien team average ELO. Also if player has not been playing the whole round(joined middle) ELO rating change would be reduced. (<a href="https://docs.google.com/drawings/d/16QisluLHMofn39UgPOpZp-tcp8JjEyVufu6PgIvZzYo/edit" target="_blank">here</a> is some draft about this). This would mean that even if you lose you would not lose that much points if you are high in scoreboard, but if not then you would.
<u>Anyway nothing is decided yet, we are still thinking what we are going to do about it.</u>
But we are planning on making sv_randomByEloRating command or similar, which would randomize teams using each players ELO rating, so that both teams would have about the same average ELO rating. Rating would be fetched from website before teams are arranged. People could also vote for it in the game, or there could be setting to use it always. This needs more thinking also.<!--QuoteEnd--></div><!--QuoteEEnd--> So, if I understand this correctly, the total Elo rating points added or subtracted is based on the win/loss, but the distribution of the points added or subtracted within the team is based on their K/D? I like that, but I'd prefer it based on points rather than K/D.
I'm glad there has been some discussion about the downside to the KDR and score metrics. Tonight while playing I chomped 3 extractors down from 100% only to have a jonny-come-lately swoop in and score the killing blow. The other classic example is a lerk and a fade go in to a bar... the lerk lays down spores which makes it easier for a fade to score a killing blow. If all glory goes to the KDR then we can say the lerk made two mistakes: the first is evolving to lerk, a lifeform less able to score killing blows than a fade. The second was artificially raising someone else's stats.
What concerns me is that some people might change what they do in the game in order to better their stats, and it’s especially concerning if they do that to the exclusion of the team goals. Let's not forget that one team wins while the other loses. Individual stats are interesting, but it is a team game.
Another concern is that people's attitudes might change for the worse. In the extractor example above, should I feel resentful that another player 'stole' my points or should I feel grateful that my team worked together to achieve a common goal?
Also of concern is where the outcome of a game is readily apparent long before the actual end. Ironically, the better the losing team plays, the longer the game takes and during that time scoring kills or points is easier for the winning team (think shotgun/jetpack marines vs skulks without leap; onos vs jetpack-less marines). The points system doesn't scale for those scenarios. So should the losing team f4, ragequit or simply go hiding in order to preserve their stats? Don't the winning team deserve to play out their victory?
The way to mitigate these problems is to better measure what is actually going on. The best way to measure what's going on is to measure more things. So for me the question is what things can be measured that reflect how effectively a player is contributing to the team's objective.
There are two pieces of information I'd be interested in tracking in addition to the currently available stats. I'm not sure what's possible in terms of technical limitations, so I'm just going to throw this out there...
The first is damage dealt. This helps to show what actually happened for the extractor example above. It would show that I dealt 2000 damage while the other skulk did 100. That information could be broken up in to damage to players, buildings and maybe even extractors/harvesters (since that's a big part of the game). In another example, a pair of skulks chew an extractor down to 10% and are then killed by a grenade. In that example, the skulks have been more effective than simply dying without earning a kill or any points. This system helps to measure that kind of scenario. It's not a holy grail, but it is measurable and has some relevance, just like KDR and points. If particuarl weapons (like AoE) throw these stats out then they could be scaled or categorised.
The next piece of info I'd be interested in tracking is something that measures the relative worth of my kills. A basic formula I came up with is myKills * enemyTeamKiills / myTeamKills. The more kills I get, the higher my score is. The more kills the enemy got, the more worth each of my kills had. The more kills my team got, the less worth each of my kills had. If the winning team killfarms the losing team it lowers their score while boosting the enemy's score.
Comments
Perhaps they could add in some form of assit system, idk. Generally I find the better players tend to get the most kills anyway, because they are the ones doing the most damage and are therefore the most likely to get the kill. Kill steals happen sometimes but I don't know if it's such a big deal that it needs an assist system in the rankings. It will work itself out over time anyway, when the noob that got the kill steal dies 5 times over to the better player.
Whatever they decide to do I think they should keep it simple as possible.
Accuracy stats could also be cool, but also potentially bring down server performance. It could possibly be mitigated by trying to use local variables as a cache.
Really, this is really awesome work!
Love the site.
win/loss is absolutely needed imo. it is the reward for contributing for the greater cause, covering things other stats can not e.g. sacrificing yourself to save a phasegate.
i agree stacking could be a problem. if they use ELO to actually balance out the teams this should be fine right? as long as there is some motivation to stick with your team instead or switching to the winners (which is rarely possible anyway).
<!--quoteo--><div class='quotetop'>QUOTE </div><div class='quotemain'><!--quotec-->You can't take into account all of the complexities of the game and determine exactly how much points you should get, because every situation is different.<!--QuoteEnd--></div><!--QuoteEEnd-->
this will always be true. but if you create a final ELO-score, you should at least try hard imo. the more stats you combine, the more situations (possible contributions) you can cover. and the more simple it is, the easier it is for people to find obvious exploitation methods.
two more things i'd like to bring up:
what happens if the versions change? it is probably not a big deal i guess, just asking.
what happens if the servers have other mods activated? this is currently not an issue, but certain mods might drastically alter certain stats like the score.
I dont think so, because its kinda worthless...
For proper elo raitings you need rules, 12 players, 2 commanders 10fieldplayers, no afking, no leaving, no winning team joining, no stacking(if its not something like "party ranked") best of 3 or whatever. Set timers to start, pause rounds etc(so you dont end up in the ready room for 30min and a playtime of 10min), also you should have a reconnect feature in case of crashes, or timeouts etc. server needs to have relative fair pings and proper performance, you basically need a freaking big system behind all this so you cant cheat it and ppl can get punished by ###### up other players experience, and to be as fair as possible.
what happens if the versions change? it is probably not a big deal i guess, just asking.
what happens if the servers have other mods activated? this is currently not an issue, but certain mods might drastically alter certain stats like the score.<!--QuoteEnd--></div><!--QuoteEEnd-->
By versions I think you mean ns2 patches and they shouldn't be that big of a problem: We only have some function calls in some of the original ns2-lua files so it won't take us more than 15 minutes to make a new version suitable for the new patch and server admins just need to update their stats mod. We'll probably try to make this even more easier in the future but it isn't our first priority at the moment.
And there shouldn't be any conflicts with another mods unless they use the same ns2-lua files that we do and in case there's a conflict you can just contact me and I'll make a version where your mod and our mod will merge which shouldn't take more than couple of minutes to create.
<!--quoteo(post=1963909:date=Aug 16 2012, 02:52 PM:name=Koruyo)--><div class='quotetop'>QUOTE (Koruyo @ Aug 16 2012, 02:52 PM) <a href="index.php?act=findpost&pid=1963909"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Are there any elo ratings for games without any proper rules or systems behind it?
I dont think so, because its kinda worthless...
For proper elo raitings you need rules, 12 players, 2 commanders 10fieldplayers, no afking, no leaving, best of 3 or whatever. Set timers to start, pause rounds etc(so you dont end up in the ready room for 30min and a playtime of 10min), also you should have a reconnect feature in case of crashes, or timeouts etc. server needs to have relative fair pings and proper performance, you basically need a freaking big system behind all this so you cant cheat it and you can get punished by ###### up other players games.<!--QuoteEnd--></div><!--QuoteEEnd-->
We'll make separate elo-ranking for competitive matches that are solely based on PCWs and gathers, or purely PCWs if the community grows, and even for clans in the future but there'll always be public elo-ranking too even if it won't be anywhere as accurate as competitive elo-ranking but it'll give nice touch to player stats and it certainly shouldn't ever be taken too seriously, unless you take public seriously.
<!--quoteo(post=1963720:date=Aug 16 2012, 03:03 AM:name=Dghelneshi)--><div class='quotetop'>QUOTE (Dghelneshi @ Aug 16 2012, 03:03 AM) <a href="index.php?act=findpost&pid=1963720"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->I would actually be interested in stats for spawn combinations instead of just stats for each spawn point. It's a lot more data, but more meaningful.
Accuracy stats could also be cool, but also potentially bring down server performance. It could possibly be mitigated by trying to use local variables as a cache.<!--QuoteEnd--></div><!--QuoteEEnd-->
And yes you are right, we can make that spawn combination stat pretty easily and probably will be added soon. And accuracy will probably try to be done in the future but we have to look out not to decrease server performance; we'll probably even make it so that you can see your accuracy with rifle or with shotgun or with bite etc.
thats good to hear, but i actually wanted to point out a different issue. to give an exaggerated scenario: "play on our server, we have a mod to double your score!"
i doubt anyone will do anything like this intentionally, but it could be a side effect. e.g. you will not get any building stats when playing the already working combat mod.
Error 500
CDbConnection failed to open the DB connection.
thx
<!--quoteo(post=1963972:date=Aug 16 2012, 06:39 PM:name=Laosh'Ra)--><div class='quotetop'>QUOTE (Laosh'Ra @ Aug 16 2012, 06:39 PM) <a href="index.php?act=findpost&pid=1963972"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->thats good to hear, but i actually wanted to point out a different issue. to give an exaggerated scenario: "play on our server, we have a mod to double your score!"
i doubt anyone will do anything like this intentionally, but it could be a side effect. e.g. you will not get any building stats when playing the already working combat mod.<!--QuoteEnd--></div><!--QuoteEEnd-->
Score/kill cheating is possible but if we notice something like that we can ip-ban that server and ignore the stats that server has sent.
<!--quoteo(post=1963986:date=Aug 16 2012, 07:33 PM:name=DualSight)--><div class='quotetop'>QUOTE (DualSight @ Aug 16 2012, 07:33 PM) <a href="index.php?act=findpost&pid=1963986"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->For the player section can you put a search function thanks.<!--QuoteEnd--></div><!--QuoteEEnd-->
There's already search but it's hard to notice; it's on top of the playerlist in players-page. We'll make search feature better in soon; the site is no-where finished and almost 90% will be different looking in 4-5months - site has been under construction for a little over a week now, so patience. :>
Working now. Thanks
Now detailed map stats sound awesome. Great tactical tool. But just imagine recording entire gameplay (in a simplified way ofc), and then.. playing it back in the browser on a map overview, with a time slider, see how the game progressed with positions of players/buildings, meds dropped, given orders etc.
mind = blown :D
Now detailed map stats sound awesome. Great tactical tool. But just imagine recording entire gameplay (in a simplified way ofc), and then.. playing it back in the browser on a map overview, with a time slider, see how the game progressed with positions of players/buildings, meds dropped, given orders etc.
mind = blown :D<!--QuoteEnd--></div><!--QuoteEEnd-->
Data for this is already being recorded and all this has already been planned and will definitely be on our site at some point, will take a lot work so won't happen for couple of months but it'll happen. :>
Perhaps I might be of some assistance on the front-end side after all.. ;)
i agree stacking could be a problem. if they use ELO to actually balance out the teams this should be fine right? as long as there is some motivation to stick with your team instead or switching to the winners (which is rarely possible anyway).<!--QuoteEnd--></div><!--QuoteEEnd-->
IMO you just need to accept that imperfections in the system. Wins/loss will just lead to stacking in public games. If you sacrifice yourself for a phase gate then tough ######. Again, there will always be situations where you lose points where you "shouldn't" have. How often do those situations really come up? It would lead to a small difference in points, but it's not really a big deal. Most people want to win the game and aren't going to care about a tiny amount of points being lost from an individual encounter. If you lose the phase gate it will probably lead to you dying more often later in the game and then losing more points in the end anyway. I doubt most servers will use the rankings to balance public games, what happens when people join in the middle of a match etc.
<!--quoteo--><div class='quotetop'>QUOTE </div><div class='quotemain'><!--quotec-->this will always be true. but if you create a final ELO-score, you should at least try hard imo. the more stats you combine, the more situations (possible contributions) you can cover. and the more simple it is, the easier it is for people to find obvious exploitation methods.<!--QuoteEnd--></div><!--QuoteEEnd-->
I completely disagree. This is for public games. If you keep it simple with kills/deaths and buildings then you really can't exploit it. Playing the game and getting more kills will be the way to get points, which is what people do anyway. Trying to create this near perfect system for public games is just futile. The public ranking is just a fun thing it's not a big deal and shouldn't be obsessed over to try and make it perfect.
and why would they not be stacked in comparison, if the goal is kdr and score (which is also heavily influenced by having a good team)?
i am concerned with very low morale (and i've seen in other threads that several people share this concern) as in people giving up risks in favor of turtling the rest of the match in order to gain some kills and score...
the other team possibly trying to prolong the game as long as possible (spawnkill etc.) will further encourage this behavior.
in fact i have seen this issue today on one of the stat-servers. our second marine base was heavily attacked (power down). i tried to motivate some teammates to come with me to take it back, 2 guys refused (they even had jetpacks) as in "it's already lost". we managed to take it back on the first try though, even with low numbers.
another time were 3-4 people standing around in the base, not going through the phase gate as the other side was attacked. they were either not aware of this for some reason (although it was pointed out via voice) or nobody wanted to go first into this death-trap.
of course it will never be flawless. but there is no point in adding extra flaws imo. if stacking will be a problem then i'd really prefer elo-based teams, or simply randomly assigned teams. people might be annoyed as in "damn i wanted to play on the other side this time" but it's better than stacking or people putting their greed above the team effort.
also, whats the advantage of keeping it simple?
<a href="http://www.unknownworlds.com/ns2/forums/index.php?showtopic=117168" target="_blank">http://www.unknownworlds.com/ns2/forums/in...howtopic=117168</a>
it also holds some points for your side, i just want to point out several people apparently have very strong oppinions about this.
I saw this in global agenda with ASM-whoring. ASM was a weighted mechanic that took into account about a hundred different factors such as damage, kills, deaths, win, loss, building damage, objective points. It also weighed the teams. IE if you had 1000 ASM and were on a team with nine other 10 ASM players (noobs), but the other team was 10 1000 ASM players, then you would not lose very much ASM when your team inevitably lost. Eventually, ASM could actually predict the winner of the match in a "slightly-favored, steamroll, or even" style with decent accuracy. The creator of ASM tried to rank people via their points. This lead to people making characters, playing the minimum 30 games to get ranked, and playing super-meter-###### for those 30 games, then quitting for the month. I don't think there is any way to create a ranking system without this happening.
I think ELO should be kept to matches and be team based, if anything. I don't think the competitive community is to the point where we need a strict ranking for anything.
Either you accept the problems the stats bring or you dont log them.
Having an ELO is honestly completely pointless in NS2.. its pretty obvious who the best players are, we dont need some complicated numerical system to find out the exact same information. Maybe if the playercount increases by about 1000x, but honestly its more fun to play the game not knowing exactly how good your opponents might play.
i don't like the idea of public elo that much either. but since they will add it either way (i guess) why not use it to prevent team stacking?
So we have some ideas to make system which uses ELO rating but points what you get will depend also on your ranking inside your team. This ranking could be based on score for example, but maybe also something else. So for example first person in scoreboard would get 15 points, second 14.5 points etc. This score would then be reduced or increased by ELO ratings when we compare marine team average ELO to alien team average ELO. Also if player has not been playing the whole round(joined middle) ELO rating change would be reduced. (<a href="https://docs.google.com/drawings/d/16QisluLHMofn39UgPOpZp-tcp8JjEyVufu6PgIvZzYo/edit" target="_blank">here</a> is some draft about this). This would mean that even if you lose you would not lose that much points if you are high in scoreboard, but if not then you would.
<u>Anyway nothing is decided yet, we are still thinking what we are going to do about it.</u>
But we are planning on making sv_randomByEloRating command or similar, which would randomize teams using each players ELO rating, so that both teams would have about the same average ELO rating. Rating would be fetched from website before teams are arranged. People could also vote for it in the game, or there could be setting to use it always. This needs more thinking also.<!--QuoteEnd--></div><!--QuoteEEnd-->
So, if I understand this correctly, the total Elo rating points added or subtracted is based on the win/loss, but the distribution of the points added or subtracted within the team is based on their K/D? I like that, but I'd prefer it based on points rather than K/D.
What concerns me is that some people might change what they do in the game in order to better their stats, and it’s especially concerning if they do that to the exclusion of the team goals. Let's not forget that one team wins while the other loses. Individual stats are interesting, but it is a team game.
Another concern is that people's attitudes might change for the worse. In the extractor example above, should I feel resentful that another player 'stole' my points or should I feel grateful that my team worked together to achieve a common goal?
Also of concern is where the outcome of a game is readily apparent long before the actual end. Ironically, the better the losing team plays, the longer the game takes and during that time scoring kills or points is easier for the winning team (think shotgun/jetpack marines vs skulks without leap; onos vs jetpack-less marines). The points system doesn't scale for those scenarios. So should the losing team f4, ragequit or simply go hiding in order to preserve their stats? Don't the winning team deserve to play out their victory?
The way to mitigate these problems is to better measure what is actually going on. The best way to measure what's going on is to measure more things. So for me the question is what things can be measured that reflect how effectively a player is contributing to the team's objective.
There are two pieces of information I'd be interested in tracking in addition to the currently available stats. I'm not sure what's possible in terms of technical limitations, so I'm just going to throw this out there...
The first is damage dealt. This helps to show what actually happened for the extractor example above. It would show that I dealt 2000 damage while the other skulk did 100. That information could be broken up in to damage to players, buildings and maybe even extractors/harvesters (since that's a big part of the game). In another example, a pair of skulks chew an extractor down to 10% and are then killed by a grenade. In that example, the skulks have been more effective than simply dying without earning a kill or any points. This system helps to measure that kind of scenario. It's not a holy grail, but it is measurable and has some relevance, just like KDR and points. If particuarl weapons (like AoE) throw these stats out then they could be scaled or categorised.
The next piece of info I'd be interested in tracking is something that measures the relative worth of my kills. A basic formula I came up with is myKills * enemyTeamKiills / myTeamKills. The more kills I get, the higher my score is. The more kills the enemy got, the more worth each of my kills had. The more kills my team got, the less worth each of my kills had. If the winning team killfarms the losing team it lowers their score while boosting the enemy's score.