How to access to this http://forum.grangerhub.org/t/player-specific-gameplay-statistics/33/2?
I agree, and there is a potential problem with coming up with a ranking formula and then fine-tuning it: it may piss off people to see their ranking mysteriously changing around. I would make player ranking not based on some unknown formula, but directly with karma. And karma should not be only automatic, but also people-driven, like a review system. Intermissions would the perfect place to distribute a few karma points to your best teammates. This would avoid any type of abuse of an automated system because it would be people-driven, and people are smart. 
And reviews are a time-tested way to provide feedback.
In addition to the karma/review system, all stats should be available - including a complex total formula for player competitiveness. However, this should not be used to give advantages in game so that we don’t end up like in League as @DarkMicrobe explained. Instead, the stats should be available to filter and rank players freely and can be used by clans, etc.
1 Like
How you avoid the camp with this? By doing for everything x*time_outside_base ?
To avoid camp, we need the karma/review system, not the stats. Players would know when a teammate has been camping the whole game. The stats should be simple. Complex/mysterious calculations can hurt as much as they help. They should be left as optional statistics that could be used if someone is looking into competitiveness.
1 Like
Floor trappers are still better than no trappers. What if all players online are new/have low Karma, nobody would be able to build and games would be shit.
The Karma system should attempt to make them think “oo I would like that feature, I should raise my karma or stop lowering it by misbehaving” Not “Fuck,this is annoying. Fuck this game/server”.
What if the Karma was compared by other builders present?
If you have 3 builders, the one with the highest karma could be designated/lead builder. The one with the second highest Karma could have normal building rights, the one with the lowest karma has restricted building (can only make or decon rets) If the 2nd highest Karma builder changes weapon or dc’s the 3rd highest karma builder would now be able to build normally. This would prevent games with no builders at all.
The issues with this is, we are setting a precedent that GOOD KARMA = GOOD BUILDER which is very BAD. Perhaps the building restrictions should be scrapped from the karma system and be built into some sort of “Builder review” system.
Building is just too vital to the game to restrict, especially automatically. Also just because someone misbehaves doesn’t mean they are a bad builder. (maebong)
That is almost never true.
The first purpose of the building restrictions is to prevent immediate griefing upon evading (as an example a player who is banned, then evades and attempts to maliciously decon the om/rc as a “revenge action”), the secondary purpose is to protect completely new players from themselves. But for what I have in mind is that full building abilities could be obtained after a week or two of regular activity. In tremulous, if you are new to the game, that is generally not long enough to even become a good builder, but you should at that point be at least somewhat familiar with the basics, and could start to become an ok builder.
Very good idea. In general various abilities should take into consideration the availability of other active (not afk ofc) players on the server and their karma levels in relation to a specific player and his/her karma level. This would help a new server with an unestablished playerbase grow.
Almost never is better than nothing.
I understand this, but it’s worse than NSA monitoring video games to catch terrorists. Building restrictions could piss off way more people than the actual deconning.
1.1 building is actually more simple than the building in GPP. Whilst they may not know everything there is to know about building in 1-2 weeks most competent individuals can grasp how to build a standard and proficient base in mere minutes by simply reading (or watching) a guide.
[trailing posts]
enough of the discussion of rating systems. they have been discussed before, and this attempt yields nothing new.
what significant behavioral property do u expect to derive from karma ?
- eg., the player has passed the initial (restricted) startup stage, and on top of that, he/she has a karma value granting some non-gamelogic-related capabilities (eg., kick vote must have a ≥85% acceptance ratio) — on a particular server group, that is.
so fucking what ? ok, it shows that the player isn’t an ultimate newbie (unless the player avoided the system by becoming „friends” with a manager of the server group, or broke the server group’s security systems).- also, eg.; on all other servers, the player only disrupted games, for many years; but „u don’t have to know anything about that.”
- suppose that, in addition to the karma value, a CRIMINAL™ history is shown for a player, and suppose that it shows a few malicious decons.
so fucking what ? it happens every now and then that a game is so gay due to others, that someone perceives it as insignificant.
eg.; /dev/humancontroller, in his first year of Tremulous, was a „full-time” game disruptor, uniquely known for things like chat flooding (DOSing)1 and the development of an aimbot and laghax. later, he became a prominent contributor to Tremulous, ioQuake 3, and open source in general.2
1 anti-flood systems in ioQuake 3 (and derivatives) were developed in response to his chatfloodkick-based server-takeover activities.
2 however, he has recently been losing faith in and support for GrangerHub, due to sum 2f4gg0tz.
all-in-all, the karma value will be a very weak metric.
first of all, the system proposed here is all about automation. second of all, an also-ppl-driven system will be arguably as hard to use and predict as a fully automatic one, but the point will be that the latter will continuously need work from all users.
@DarkMicrobe explained not only why whore metrics shouldn’t give in-game advantages, but also why whore metrics shouldn’t be shown at all.
and here u’ve just suggested a metric that directly changes the way ppl play the game (not the meta behavior (such as whether they grief), but what they regard as winning). this is exactly the opposite of what the karma system aspires to influence.
WRONG. in general, they take away buildpoints from properly composed base layouts.
headshot ! this is an example of what could have been fucked up.
also, there’s a fuckup: if the 2 most „karmish” players don’t want to build, or if they can’t build fast enough, then the team could remain in a lack of other buildables, eg. only 2 telenodes for 20 ppl, no armory, etc.
I’m going to have to agree with DevHC here. The Karma system explained thus far seems vague and even if I did understand it, would require multiple iterations to get to a point where its considered useful.
I believe if such a system is going to be used, it should be opt-out once you are registered to the server. A proper tutorial system implemented after the release of 1.3 would be a better alternative than spending resources on a Karma system before the release of 1.3.
Addressing the problem of preventing most occurrences of a very new player from making a very big mistake that disrupts game play would only be a secondary effect of the Karma system. The karma system is not intended to do the job of dedicated tutorial systems.
The primary purpose of the Karma system is to provide a downside from being banned and evading, provide encouragement to behave well, and protect against new accounts maliciously griefing (which many times are evaders).
Yes it is true that in-game behavior of a player known to disrupt game-play could improve, as well as in-game behavior of a player known to not disrupt game-play can deteriorate. However, those are not counter examples against the karma system, as the karma system is meant to be dynamic, and any player could have a net gain/lose in karma based on change in behavior. That is, unless you have zero karma, in which case it should not be possible to obtain negative karma, as that defies one of the primary purposes of providing a downside to evasion.
on top of both being primary, what is the purpose weight between (1) and (2) ?
WRONG, and u’re missing the point. the point shows that even with the hypothetic additional information, one can’t seem to derive a really valuable behavioral property for the purposes of considering to accept the player in a clan.
eg., one who plays 30 hours per day can permit himself/herself to decon every now-and-then.
No mine was just a question to @cengique
I would like to see a game with a level of competitiveness comparable to games like battlefield for pc.
If we continue to give against all possible ideas, not finding at least a decent solution, I’ll settle for seeing the same number of players who currently populate urban terror.
In any case, camping is a part of all games. You can fight it, not abolish it.
A good clan member should have a good mix of decent skill/decent behavior.
There are ways to addressing camping, however, the Karma system is not meant to address camping nor other game-play behaviour, but rather like DevHC mentioned, Karma is meant to address poor meta game behavior (such as griefing and other violations of game server rules).
The problem of camping in Tremulous is discussed in this topic: Is camping in Tremulous a problem?
They are two parts of the same primary purpose and carry equal weight in the design.
The maximum amount of karma you could earn each day would be capped.
i believe proper game design can nail it.
which is exactly what u can’t derive.
then u instead advance 30 accounts per day.
1 Like
ppl ! let’s not forget that invincibility to griefers should be established through technology to reverse the effects of griefing, as opposed deterrence of griefers. and with that said, the design of the karma system is to be scrapped,1 as the remaining purposes r:
- to deter ppl from bad behavior in general — this has no justification to be implemented by punishing by revoking potentially useful capabilities (see below);
- to protect n00bs from „their alter ego” — and this has already been argued to be only detrimental in general, if forced.
elaboration on revoking „potentially useful capabilities”:
- what if use of weapons stronger than X were forbidden ? then, the player would feed more, and thus be more detrimental to the team.
- what if the power to kick was revoked ? then, the admin would be unable to kick griefers, and thus be less useful to the server.
- one could now say: „but the power to kick includes the power to abuse !” yes, but in what circumstances is revocation warranted ? when the admin starts teamkilling ? that is rather irrelevant. instead, when the admin starts abusing the power to kick ? this, however, is something the karma system can’t detect.
though, by contrast:
- some cautious applications of values like „value of time played minus value of teamkills, etc” is acceptable, such as X in „need ≥X% support to be vote-kicked” (however, this example metric is very weak).
- there is nothing bad in affecting immaterial values, such as one’s position on the scoreboard.
1 well, there is almost nothing to scrap yet.
LoL’s review system is used for a very large community, and as such is directly dependant on the quality of said community (which is too often shit to be good). The same system is usually better at smaller scales when admins review the reports. The process is exactly the same except that the admins are behind the decisions, the results can sometimes be achieved faster than with LoL’s tribunal and on top of that, you are certified the result is not bullshit or retarded.
You usually don’t need to involve players in handling reports on these kind of games where each server has their own smaller community and assigned admin team. That is unless you have an overwhelming amount of reports and not enough people to deal with all of them so they keep stacking up. (In which case, in Tremulous you would seriously need to recruit admins.)
This is what we are doing already on GHub at the moment though. Our only problem is the lack of a report system in-game, (which is, soon to be fixed, until then @DevHC can keep fucking my mom for creating so many bugs) and serverside demos.
2 Likes