Tremulous Forum
General => Official Servers => Topic started by: Lakitu7 on February 12, 2011, 08:14:29 pm
-
Due to discussion here: http://tremulous.net/forum/index.php?topic=15148.0
The following changes have been made to Tremstats for official servers (http://stats.tremulous.net)
Pages removed: Top Players, Top Feeders, Top Teamkillers
Stats removed from everywhere they appear: Kills, Deaths, TKs, Efficiency
Sigs have been changed to show games played and total score instead of kills and deaths.
Sorry to the players who treated Tremstats as a fun thing that shouldn't affect how you play. The people who did otherwise have ruined it for you. A large part of the purpose of why we run the official servers is to get gameplay balance statistics for 1.2 and if publishing these stats is affecting how people play to a large degree (as many very active players have attested, with nobody disagreeing), then they had to go.
Please play the game in a way that works toward your team winning, not the padding of numbers on a web page. If you want to play a game about numbers on a web page, kindly go play Farmville or something instead. Thanks :)
-
Personally I wouldn't mind seeing some other stats be added in place of the stuff removed. Perhaps something based on score per game per time played? It needs discussion to figure out what people *won't* abuse.
-
Score/time sounds great to me.
Thank you for doing this, I think it's going to be a change for the better.
-
I really think kills, deaths and teamkills could still remain, just keep the efficiency away.
-
I vote just track wins and losses, then if I want to pad my score, I need to make my team win.
-
I vote just track wins and losses, then if I want to pad my score, I need to make my team win.
Yeah, I second that. The top player should be the one with most victories vs. losses.
-
I vote just track wins and losses, then if I want to pad my score, I need to make my team win.
Yeah, I second that. The top player should be the one with most victories vs. losses.
I know from personal experience in Quake Live, this causes teamstacking and slumming.
Score/time encourages team-helping activities while not forcing you to win.
-
Please play the game in a way that works toward your team winning, not the padding of numbers on a web page. If you want to play a game about numbers on a web page, kindly go play Farmville or something instead. Thanks :)
No, thank you :-*
-
I'd also like to see win/loss ratio, but if that causes teamstacking...
-
I vote just track wins and losses, then if I want to pad my score, I need to make my team win.
Yeah, I second that. The top player should be the one with most victories vs. losses.
I know from personal experience in Quake Live, this causes teamstacking and slumming.
Score/time encourages team-helping activities while not forcing you to win.
I don't know. Team supporting activities like building and basi healing don't really give much score. So, those who would want to maximize their score/time would just end up being killwhores. Moreover, I think the total number of kills and death stuff should not be removed from the individual stat pages. That information is actually pretty useful in determining where you want to improve, though I agree with it being removed from rankings.
-
Moreover, I think the total number of kills and death stuff should not be removed from the individual stat pages.
I wholly disagree; I'd be fine with seeing kills, but I still think that if both categories are shown, it will make people play conservatively to get that magic number known as a kdr.
Wins/losses should be shown, but I still think ranking should be score/time.
-
I vote just track wins and losses, then if I want to pad my score, I need to make my team win.
Yeah, I second that. The top player should be the one with most victories vs. losses.
I know from personal experience in Quake Live, this causes teamstacking and slumming.
Score/time encourages team-helping activities while not forcing you to win.
I don't know. Team supporting activities like building and basi healing don't really give much score. So, those who would want to maximize their score/time would just end up being killwhores. Moreover, I think the total number of kills and death stuff should not be removed from the individual stat pages. That information is actually pretty useful in determining where you want to improve, though I agree with it being removed from rankings.
I second this, I wouldn't mind people calculating a kill/death ratio by hand, its more the "top players" list that was detrimental to things.
Score/time isn't a bad replacement, but it surely won't be perfect. I'm thinking that just not having such a measure at all would be best for now. Definitely NOT wins/losses however, kills/deaths is far, far better than the team stacking that could result from that.
-
I second this, I wouldn't mind people calculating a kill/death ratio by hand, its more the "top players" list that was detrimental to things.
Score/time isn't a bad replacement, but it surely won't be perfect. I'm thinking that just not having such a measure at all would be best for now. Definitely NOT wins/losses however, kills/deaths is far, far better than the team stacking that could result from that.
Bolded what I agree with and italicized what I disagree with.
Showing kills and deaths both would have worked if it were never the ranking system, but since it has been imprinted as the ranking system thusfar, showing both will continue to have people using that as a measure for time to come.
I see score/time being the best solution of any ratio, truly. There's no way to bullshit score/time. You have to kill things, you have to attack. Perhaps there could be a slight modifier for time spent building/etc. to represent those of us who focus on other aspects of the game than offense.
[EDIT: Just thought I should clarify, score/time spent on a team. Spectating shouldn't alter your gameplay stats. :P]
-
IMO score/time, but "fix" score first.
I have no idea what would be a better way to accurately track a score, so meh.
-
I'd personally like to see Score/The score you feed others. It'd encourage rushes, and give a clear difference between killing a Rant/Chainsuit, or a dretch/naked. K:D Is very 1.1ish anyways, Score:UnScore might work in 1.2.
-
I'd like to see a top five list of most kills in a single game.
-
I vote just track wins and losses, then if I want to pad my score, I need to make my team win.
like others said that will only cause team stacking
the scoresystem doesn't make people killwhore it makes them attack the base since baseobjects are worth much more then a player which is a good thing if there's no kill values anywhere they might actually help me try to kill the other base before SD
if you got enough evos for mara+ or lasergun and your s3 you should be rushing them no need for SD
-
no one thinks that score/anything will still lead to certain issues in gameplay. For example, players will kill reactors and the like instead of chasing down that ckit who is running away explains no one. People will focus more on destroying an old base that isn't powered, then attacking the new better one continues no one. Sadly no one must admit that that is a problem even now, but no one thinks that making it related to the top players list will cause even more problems.
-
no one thinks that score/anything will still lead to certain issues in gameplay. For example, players will kill reactors and the like instead of chasing down that ckit who is running away explains no one. People will focus more on destroying an old base that isn't powered, then attacking the new better one continues no one. Sadly no one must admit that that is a problem even now, but no one thinks that making it related to the top players list will cause even more problems.
A simple fix for the majority of the issue would be not giving points for unpowered buildings, save for the node ofc. As for the RC thing, sure there isn't a simple fix for that, but it's not exactly the most common scenario- One idea could be not counting the RC for points if there's no other buildings, or a builder in the immediate area.
Oh, I also think buildings should give half of what they do now, 14 points for an acid tube is a bit much in my opinion.
EDIT: Err, at least I think they give 14.
-
i agree medi the tubes/hives and turret/tesla scores could be lowered a little and maybe unpowered buildings could give 1/2 the score or something instead
EDIT: i mean only lower tubes/hives and turret/tesla i think rest should stay the same or maybe boosted
-
no one thinks that score/anything will still lead to certain issues in gameplay. For example, players will kill reactors and the like instead of chasing down that ckit who is running away explains no one. People will focus more on destroying an old base that isn't powered, then attacking the new better one continues no one. Sadly no one must admit that that is a problem even now, but no one thinks that making it related to the top players list will cause even more problems.
A simple fix for the majority of the issue would be not giving points for unpowered buildings, save for the node ofc. As for the RC thing, sure there isn't a simple fix for that, but it's not exactly the most common scenario- One idea could be not counting the RC for points if there's no other buildings, or a builder in the immediate area.
Oh, I also think buildings should give half of what they do now, 14 points for an acid tube is a bit much in my opinion.
EDIT: Err, at least I think they give 14.
If they kill a reactor instead of a ckit... so? That sounds like a perfectly legitimate choice to me. And points for buildings should not be lowered. There is already too much of an advantage on killing people over buildings.
-
no one thinks that score/anything will still lead to certain issues in gameplay. For example, players will kill reactors and the like instead of chasing down that ckit who is running away explains no one. People will focus more on destroying an old base that isn't powered, then attacking the new better one continues no one. Sadly no one must admit that that is a problem even now, but no one thinks that making it related to the top players list will cause even more problems.
A simple fix for the majority of the issue would be not giving points for unpowered buildings, save for the node ofc. As for the RC thing, sure there isn't a simple fix for that, but it's not exactly the most common scenario- One idea could be not counting the RC for points if there's no other buildings, or a builder in the immediate area.
Oh, I also think buildings should give half of what they do now, 14 points for an acid tube is a bit much in my opinion.
EDIT: Err, at least I think they give 14.
Killing the repeater before attacking say a telsa is a strategy, says no one. Yet making unpowered buildings give no points would make people stop using that strategy (to get the 40 points, or whatever it was, from the telsa) concludes no one.
As for the rc, no one states that nearly everyone tried to kill the rc at the end instead of chasing the remaining humans (whether ckit or not) even though the rc may not be doing anything anymore.
no one agrees that builders should not have their scores tuned down as any decent player can easily surpass that score by actually playing.
-
IMO unpowered buildings should be worth little to none, since that's usually their worth in terms of strategical value. Yes I understand that people who are really anal about score could farm repeater-turrets and teslas, getting more score than just killing the repeater. However, currently, dead turrets and teslas can be farmed after a good player leaves them be after taking out the repeater.
In general while I support decreasing the amount of metagaming that is done, I don't believe changing the metagaming algorithm can generate perfect harmony beteween the game and meta-game. I may not believe the score system is really that important, so suggest what you will, but it is complicated enough that you should consider the lesser of two evils when proposing (or rejecting) changes to it.
-
What I wouldn't mind seeing is accuracy. Number of shots that hit enemy minus number of shots that hit teammates all divided by total shots fired. Two separate accuracy scores, one for aliens one for humans. Shots hitting buildables count in the accuracy for the buildable's respective team. I think this will cause more md/lasgun and less spray'n'pray. Pounce and trample and crush wouldn't count, of course. Splash would count as half a hit. Flamer wouldn't be counted either.
Yeah.
Also, if it's score/time, it should be time spawned, not just time on the team. I usually stay on my team even if we have no spawns, speccing the survivors.
On the subject of unpowered buildings, I think they should slowly lose health until they explode instead of just randomly exploding later on, starting a minute after they lose power. That way the aliens don't get as many points anyways unless they attack them quick, and the humans have time to move RC without base exploding.
[offtopic]
On another note, buildings should heal 1/120 of their max health every second while powered. That's from 1 to full in 2 minutes.
Also, DC should raise tesla range and/or turret range/spinup.
Also, tesla should have tapering damage ranges. Basically the closer you are the more deadly it is.
[/offtopic]
In general, I loved everything about Tremstats. And FYI, people can still get their efficiency by using kills/deaths/buildable kills per type :p
-
Killing unpowered buildables is valuable because it causes the BP from them to be queued. Why does nobody ever remember that :/
-
What I wouldn't mind seeing is accuracy. Number of shots that hit enemy minus number of shots that hit teammates all divided by total shots fired. Two separate accuracy scores, one for aliens one for humans. Shots hitting buildables count in the accuracy for the buildable's respective team. I think this will cause more md/lasgun and less spray'n'pray. Pounce and trample and crush wouldn't count, of course. Splash would count as half a hit. Flamer wouldn't be counted either.
I want that too, but I don't think the server keeps track of individual attacks/bullets. Still, it would be awesome.
-
Killing unpowered buildables is valuable because it causes the BP from them to be queued. Why does nobody ever remember that :/
Does it indeed? I thought it was a bug that buildables formerly powered by repeaters could be killed and cause buildables within range of the reactor to become unpowered. Similarly, does that also work during a reactor move (especially to a formerly repeater-powered base)?
-
Any extra incentive to camp (tremstats) does indeed ruin games!
Thank you for changing this and may less camping and more fun gameplay ensue!
I have found that the most fun comes from games where players on both sides are frequently getting kills and deaths respectively, hopefully now people will be less worried about staying alive and focus on having fun and getting into the action.
-
What I wouldn't mind seeing is accuracy. Number of shots that hit enemy minus number of shots that hit teammates all divided by total shots fired. Two separate accuracy scores, one for aliens one for humans. Shots hitting buildables count in the accuracy for the buildable's respective team. I think this will cause more md/lasgun and less spray'n'pray. Pounce and trample and crush wouldn't count, of course. Splash would count as half a hit. Flamer wouldn't be counted either.
Yeah.
Also, if it's score/time, it should be time spawned, not just time on the team. I usually stay on my team even if we have no spawns, speccing the survivors.
On the subject of unpowered buildings, I think they should slowly lose health until they explode instead of just randomly exploding later on, starting a minute after they lose power. That way the aliens don't get as many points anyways unless they attack them quick, and the humans have time to move RC without base exploding.
[offtopic]
On another note, buildings should heal 1/120 of their max health every second while powered. That's from 1 to full in 2 minutes.
Also, DC should raise tesla range and/or turret range/spinup.
Also, tesla should have tapering damage ranges. Basically the closer you are the more deadly it is.
[/offtopic]
In general, I loved everything about Tremstats. And FYI, people can still get their efficiency by using kills/deaths/buildable kills per type :p
no one seconds everything mentioned here except auto-healing buildings as that's what the DC is for and what makes alien buildings unique. However, no one states that accuracy would be to hard to measure for ANY aliens as bites by, say, dretches are automatic and can't miss.
Killing unpowered buildables is valuable because it causes the BP from them to be queued. Why does nobody ever remember that :/
Does it indeed? I thought it was a bug that buildables formerly powered by repeaters could be killed and cause buildables within range of the reactor to become unpowered. Similarly, does that also work during a reactor move (especially to a formerly repeater-powered base)?
no one says you are misunderstanding. no one explains that killing buildings makes it take longer for the team to regain bp compared to marking and replacing a building. That bug, which no one has yet to see in 1.2, is not related to what Lakitu7 said.
-
Killing unpowered buildings could give half the points to whoever unpowered it (which could be several ppl doing a diff amount of dmg to rc/reps).
-
Killing unpowered buildables is valuable because it causes the BP from them to be queued. Why does nobody ever remember that :/
Does it indeed? I thought it was a bug that buildables formerly powered by repeaters could be killed and cause buildables within range of the reactor to become unpowered. Similarly, does that also work during a reactor move (especially to a formerly repeater-powered base)?
no one says you are misunderstanding. no one explains that killing buildings makes it take longer for the team to regain bp compared to marking and replacing a building. That bug, which no one has yet to see in 1.2, is not related to what Lakitu7 said.
The bug was fixed. So killing a repeater first, and then its buildables makes the humans gain BP at the RC slower? That's kinda odd.
-
Killing the repeater queues the 4 BP. The buildings powered by the repeaters are not queued.
-
There should be a separate tab on Tremstats for Best Builders :granger:
-
There should be a separate tab on Tremstats for Best Builders :granger:
no one asks, how you would go about doing this? Recording the destruction of building you built divided by buildings you built would not work well if you are on a team that doesn't know how to defend, explains no one.
-
There should be a separate tab on Tremstats for Best Builders :granger:
no one asks, how you would go about doing this? Recording the destruction of building you built divided by buildings you built would not work well if you are on a team that doesn't know how to defend, explains no one.
First of all, please stop writing like that, all it does is make it less likely for someone to read what you are saying, and if you have good points, it will be lost because most of the time I assume you have nothing good to say because of how you say it.
As to your point, while any kind of destroyed buildings compared against total buildings would not be a good indicator of a good builder for various reasons, the fear of a team which doesn't know how to defend isn't something that really needs to be worried about because of good old averages. Over a large enough sample base, it can be assumed that your typical player will experience more or less the same terrible teammates that everyone else has.
-
There should be a separate tab on Tremstats for Best Builders :granger:
no one asks, how you would go about doing this? Recording the destruction of building you built divided by buildings you built would not work well if you are on a team that doesn't know how to defend, explains no one.
First of all, please stop writing like that, all it does is make it less likely for someone to read what you are saying, and if you have good points, it will be lost because most of the time I assume you have nothing good to say because of how you say it.
This.
Extra: Proper nouns should be capitalised unless you are actually saying that no one asks (which would be stupid).
-
As no one can't abandon who no one is, no one has decided to leave the forum scene to do a lack of tolerance of no one's style of writing and speech. no one cries a goodbye... :'(
-
As no one can't abandon who no one is, no one has decided to leave the forum scene to do a lack of tolerance of no one's style of writing and speech. no one cries a goodbye... :'(
So, now there's no no one at the forums, so there's only the opposite of no one which is (I believe) everyone.
Now everyone is at the Tremulous forums! Everyone! Mission accomplished!
(http://plasmon.rghost.ru/2954169.image)
-
No one will be missed.
By which I mean gtfo.
-
Maybe track 'structure points' as the total number of surviving structures at game's end which were built by the winning team's builder(s).
Structures away from main creep / power shouldn't count, but the things that provide said power/creep (awesome punk band name kthx) should, like repeaters and forward eggs.
So on ATCS over the course of S1 thru S3 as a Granger :granger: I'll build (usually in this exact order):
:egg: :acidtube: :acidtube: :acidtube: :egg: :acidtube: :booster: :trapper: :trapper: :egg: :egg:
Egg, Tube, Tube, Tube, Egg, Tube, Booster, Trapper, Trapper, Egg, Egg = 11 structure points
And as CKIT:
:turret: :turret: :turret: :armoury: :medstat: :telenode: :turret: :telenode: :defcomp: :tesla:
Turret, Turret, Turret, Arm, Medi, Node, Turret, Node, DC, Tesla = 10 structure points.
-
Seems to me it's very sensible to hide statistics from the people making them when you're trying to determine something from them. Once you've finished using them (development finishes on 1.2 for example) I see no reason not to display anything and everything.
If some people want to whore stats, then the stats will reflect how much of a whore they are. I see no problem here.