Should BIS Implement A Player Stats Database?

[infiltrator_2k 29]

I'm probably going to get ripped a new one for asking this as I'm sure this has been mentioned before. But, at the weekend I was pleasantly surprised to see a few players join my public server who worked cleverly together on taking on I&A's missions with tough AiSkill settings. It was evident from observing them they were all on coms as they worked as a closely knit highly organised team.

Anyone who admins a public server will know there's a lack of this kind of team play and there continues to be an abundance of idiots who do the usual stupid sh*t. Behaviour that AFAIK can only really be controlled by ban lists and private servers. But, what if BIS created and kept a player database of a player's kill record? I can totally understand many not wishing to have this information made public, but I'm sure a basic ratio of the friendly vs enemy kill stats could be created and kept with this information however detailed only available to the player. But, the basic kill ratio information available for the server lobby to filter and servers to reference before allowing players to join.

Just as the BE option is available, it woild be good for players and admins to be able to host and play servers that have an option that enabled a feature that only allowed players to join who had a good or bad kill ratio that was dependant on what the server admin configured. It could even be implemented on a simple percentage scale.

Personally I can't see any down sides to such an idea. It's optional like BE, the information can be private or public, the information can be as detailed or as basic as BIS liked. And it would literally eliminate the intention TK overnight on servers who had such a feature enabled and configured.

It force all the idiots to either change their ways and work on improving the TK kill ratio to have access to more severs, or to play on servers that hasn't got such a feature enabled.

Can anyone think of why such a system wouldn't work?

[MrCopyright 107]

A system like this would never be implemented by BI. They have given us plenty of ways to resolve the issues you have outlined. You can easily create a team damage detection script which punishes those who team damage too often. And a flaw in your system is that a group of 'trolls' could easily set up their own server, legitimately kill each other to improve their team kill ratio, and then they'd be on your server. 

[spidypiet 17]

after reading your post i'm thinking its comming down on KD ratio's again 

if some one want's to play for there K/D ratio's there is COD or BF series

K/D ratio dosnt mean you'r a good team player or will follow orders correctly

many times i have watched players with a high KD ratio abandon there team 

go lonewolf even using there team mates as lures to get a kill

playing arma should be more then just a good KD ratio

medics who save you'r lives heal/revive lots of people but also die allot in the process see there KD ratio plummed

and will be abandoned from your server

[Greenfist 1863]

My t/tk ratio is about 12,000 after that one nuke mission. I should have no trouble getting into servers.
 

What's wrong with scripts that kick after the first TK, and ban after the second?

[ineptaphid 6413]

My t/tk ratio is about 12,000 after that one nuke mission. I should have no trouble getting into servers.

 

What's wrong with scripts that kick after the first TK, and ban after the second?

i was just about to say the same-There are plenty of ways to get a high K/D ratio. Like in real life, I am sure there are a lot of B-2 or Apache pilots wondering what all the fuss is about when people talk about sniper records of 100 kills etc :) And there are plenty of great players witha low ratio too-the medics, or transport chopper pilots, reconnaissance guys...etc etc

[fn_Quiksilver 1636]

My t/tk ratio is about 12,000 after that one nuke mission. I should have no trouble getting into servers.

 

What's wrong with scripts that kick after the first TK, and ban after the second?

 

most TK is accident, ban for 2 TK is a good way to run a 0/100 server

 

 

 

 

stats database for personal player achievements (k/d, kills, score,etc) is a good way to discourage teamwork

[da12thMonkey 1943]

There are leaderboards for the time trials and firing ranges. And as you can see, people just cheat to get to the top of them.

Other stats would be no different in how people manipulate them.
 
 

Indeed. Let's just say that it's quite hard to make tamper-proof given the openness of the Arma 3 platform. We do plan to tighten the system for Leaderboards. The Leaderboards are mostly nice for comparing records with your Steam friends, rather than with the global top (whose times are indeed not all feasible without exploits). Just hope that your friends are above board ;)

 
https://forums.bistudio.com/topic/143930-general-discussion-dev-branch/?p=3061405
 
So yeah, BIS know it'll be exploited and isn't really a worthwhile as a global ranking system.

[froggyluv 2136]

 What special kinda loser cheats to reach leaderboard status in a video game? Never understood that..

 

 *scampers off to cash $42,000 tax return check

[infiltrator_2k 29]

Guys just to be clear, I didn't mean any kind of scoring records as such. I'm simply talking about a simple calculation to decide if a player's team kill percentage exceeds a threshold set by a server in order to be the determining factor whether or not to allow them to join. Not the amount of kills, but if a player has for example over 50% TK to an accumulative score regardless of what it is. I wouldn't want a scoreboard as that's totally arcade and it would indeed encourage people to play lone wolf.

Of course, it would be unfair for players to be penalised for accidentally crashing a chopper full of players. But given the game's engine can detect how a player is killed those kind of events could be whitelisted. Your player profile could simply display a pie chart oppose to numbers to give a graphical representation of your TK percentages so you could monitor your performance. If the engine knows how players are killed and by who or what an algorithm could iron out any flaws in such a system.

Scripts are good and I've used BTC's punishment script. But I'm talking about something that's an integral part of the game and of course optional like BE. If it could be designed to be fair and to gave admins control it can only be a good thing.

[Tajin 349]

This is not gonna be very reliable information no matter what.

Sometimes people just like to mess around with their friends on a server and teamkilling might occur. In some missions "teamkilling" is actually a normal part of the gameplay. (just think about the independant faction in wasteland)

Not to mention that there are several nasty ways how this could be exploited to harm specific players.

Only similar thing I could imagine to work would be some sort of voting/endorsement system for players but even that tends to be somewhat unreliable because it often leads to statpadding.

You're better off using some sort of whitelisting system on your server. There are plenty of ways to do that.

[infiltrator_2k 29]

This is not gonna be very reliable information no matter what.

Sometimes people just like to mess around with their friends on a server and teamkilling might occur. In some missions "teamkilling" is actually a normal part of the gameplay. (just think about the independant faction in wasteland)

Not to mention that there are several nasty ways how this could be exploited to harm specific players.

Only similar thing I could imagine to work would be some sort of voting/endorsement system for players but even that tends to be somewhat unreliable because it often leads to statpadding.

You're better off using some sort of whitelisting system on your server. There are plenty of ways to do that.

 

 

This isn't about me and my needs, it's just an overall idea. I'm happy enough to use Dart and BTC's Punishment script.  But if such a system could be devised and implemented it could potentially solve the problem of prolific intentional team killing and there wouldn't be any more need for ban lists and scripts. It could work flawlessly with enough thought and testing.  More importantly, players and admins would have the 'choice' whether to use it or not. It was mentioned about troll servers exploiting the system.  But if the ratio was only valid on enabled servers it would be prove fruitless..

I think by nature Arma players are subconsciously always continuously trying to exploit the game to gain the upper hand, and the fact they fail is proof such a system would eventually become reliable and trustworthy.  You've only got to look at how crude ban lists work to realize how sophisticated and much more reliable such a system would be in comparison.  I'm wondering what Dwarden has to say about it.

[R0adki11 3949]

Personally i don't see the need for such a system, we aren't playing COD are we? I've always thought BIS games (OFP to Arma3 etc) were about team work and not about getting as many kills as possible. 

[jarrad96 1940]

I'm backing up Roadki11 here- A global stat tracking kind of system would be cool, and has actually already been done by ALiVE War Room, but it should not be used in any official way or added by BI. Examples have already been given in regards to using K/D ratios as a way to measure 'Good' players, but even something more tailored such as a 'teamwork rating' for example, using heals/ repairing/ completing objectives could still be exploited in some way, without even considering mods- what is considered 'medical help' in ACE, and how is that recorded? What is considered an objective, is it just a editor designed missions, or do Zeus placed objectives factor in?

Arma's own sandbox plays against having any sort of universal rating system, when nothing can be paralleled between a hardcore Milsim vehicle engineer (example) , a Breaking Point ghille Survivor and a Battle Royal or KOTH player. 

[Tajin 349]

It could work flawlessly with enough thought and testing.

I doubt that.

 

There is just no way it could take all the vastly different modifications, scripts and gamemodes into account and provide viable data that is equally true for all of these variations (and safe from tampering).

ArmA is simply too open and flexible (we love it for beeing so) for such a thing.

[andersson 285]

I join a server and TK 10 persons. My stats are 100% TK

 

I join my own server where I have 100 unarmed enemy AI standing still and shoot them down with my machinegun. My stats are now 10% TK.

I replay my mission 10 times. My stats are now 1% TK.

 

Or I make my own mod, a weapon so powerful it kills the complete island full of enemy AI in 1 sec. Now my TK is so small it's next to none.

 

Then I join a server again and TK some.

 

 

Stats doesn't say much in an open and free environment as this game is.

[infiltrator_2k 29]

I join a server and TK 10 persons. My stats are 100% TK

 

I join my own server where I have 100 unarmed enemy AI standing still and shoot them down with my machinegun. My stats are now 10% TK.

I replay my mission 10 times. My stats are now 1% TK.

 

Or I make my own mod, a weapon so powerful it kills the complete island full of enemy AI in 1 sec. Now my TK is so small it's next to none.

 

Then I join a server again and TK some.

 

 

Stats doesn't say much in an open and free environment as this game is.

Are you saying Arma's engine couldn't detect the Ai were unarmed, who killed them, where, when and how? An algorithm would prevent that kind of exploitation. But of course, nothing's 100% But it begs the question of would it be practical to use a more sophisticated method to circumvent a player's kill ratio if it were time consuming and a tedious process? Not to mention a ban - albeit temporary - could be imposed on player's who are detected attempting to exploit such system to prevent them joining kill ratio enabled servers That I'm itself would act as a deterrent.

I think to assume that it wouldn't work is only undermining what Arma's currently capable of. But as I've previously mentioned, what's important is people have a choice whether or not to play on servers with this kind of feature enabled, and server admins also have the choice whether or not to enable it. Giving players the choice whether or not to penalise players for TK like BTC's punishment script would give players a lot more control too.

I think many are sceptical, but not because it couldn't work, but because it's not something they'd like to see implemented. But, it does have the potential to eradicate the excessive malicious team killing. I believe in time such a system would be welcomed.

[andersson 285]

For sure I looked at it from an unconstructive way, maybe it is possible for a certain gametype and servers? But overall in A3 I think it is nearly impossible.

[spidypiet 17]

a global TK punishment isnt the way to go like many said before it can easely be altered or avoided

on the other hand being strickt on your own server is a good thing and there are many way's to implement that already

[Vigil Vindex 64]

I think this could be a good idea if it is limited to BIS authored and unmodded competitive missions and perhaps servers. We already have the tools to make this ourselves. But without some constraints I think it would lose meaning as it would be too easy to manipulate the stats by for example spawning lots of unarmed AI and killing them.

[Tajin 349]

Are you saying Arma's engine couldn't detect the Ai were unarmed, who killed them, where, when and how? An algorithm would prevent that kind of exploitation.

There is more to it than that.

 

Trust me, I've been scripting in this game series since ofp in the very beginning. I know quite well how wide and complex the possibilities are. (for example, I could convince your own bullets to come back and fly circles around your head while singing the bananaphone song...)

 

 

 

The thing is that even without the intention of exploiting, these stats could easily be messed up by modified gamemodes.

 

Now if someone actually has the intentioin of messing around with it, then there are so many ways he could do it, that it would be near impossible to prevent.

 

It might work by having to register your mission/server for the system, including a checksum, restricted config settings and whatnot. But who would take the time to check and approve those missions in detail? I really don't see it happening.

 

Besides, the benefit is way too little compared to the effort it would require and we already have all the necessary tools to take care of it ourselves.

[infiltrator_2k 29]

There is more to it than that.

 

Trust me, I've been scripting in this game series since ofp in the very beginning. I know quite well how wide and complex the possibilities are. (for example, I could convince your own bullets to come back and fly circles around your head while singing the bananaphone song...)

 

 

 

The thing is that even without the intention of exploiting, these stats could easily be messed up by modified gamemodes.

 

Now if someone actually has the intentioin of messing around with it, then there are so many ways he could do it, that it would be near impossible to prevent.

 

It might work by having to register your mission/server for the system, including a checksum, restricted config settings and whatnot. But who would take the time to check and approve those missions in detail? I really don't see it happening.

 

Besides, the benefit is way too little compared to the effort it would require and we already have all the necessary tools to take care of it ourselves.

It is possible to implement. It's just finding the right balance of control and transparency. It's a no-brainer people who intentionally TK are going to cry foul if they are kicked or banned from playing on enabled servers. So there would have to be some kind of event log that lead up a ban. Then there's the question of how long a ban lasts? It could be an incremental system which ban period increases for repeated offenders. 10mins, 1 hour, 1 day, 1 week etc.

An algorithm could suspend penalising players until definitive intentioal TK has been clearly established. Then penalise the player retrospectively for the number of team kills made in a session. Of course, everyone is going to want to know exactly how it works, and players will test it through curiosity. I think it would take a lot of initial testing to iron out flaws and exploits. The community would have to be heavily involved. But it could work well.

I think it could be one of the best thing to happen to the game, purely because by nature people crave discipline and rules. Just like a child will test the boundaries, it's because the child by nature needs them. The virtual world is no different to the real one. Without rules, law and order there's anarchy.

I just believe that such an official automated system is the way forward to bring a single standard to the series that brings a fair and balanced method of maintaining order on servers who's admins and players opt to host and play on. Computers are afterall smarter than humans when given a formula to work by and they don't discriminate either which invalidates a penalised offender's accusation.

[clawhammer 10]

I think such features dont make sense in arma. I can create a mission where i put 100 unarmed ai's, save the mission and kill them all in mp just for the k/d. Makes no sense because it is too easy to manipulate that.

[andersson 285]

What about death match missions? That is by computer logic TK.

[Tajin 349]

@clawhammer

Don't bring up the example of killing AI again. That is clearly beside the point and I agree with infiltrator that at least THIS can easily be filtered out.

 

However, I don't agree with his "magical" solutions:

"an algorithm"

"the right balance of control and transparency"

"automated system"

"smarter than humans"

 

That is all very vague and our computers are only as smart as the people programming them, they don't posess any magical powers.

 

 

 

Here, let me give you a detailed example, since you don't seem to know where I'm coming from...

 

Lets say I'm running a scripted gamemode on my server that pretty much puts all players on the same team and allows players to dynamically create groups to work together. Essentially that means the only teamkilling that can happen is when players inside the same group decide to kill each other.

 

There would however be no way for your "magical" algorithm to know that. It would either assume that everyone is making teamkills all the time since they're all on the same team or it may be smart enough to recognize it as a FFA gamemode and ignore teamkills completely.

 

 

Both outcomes are wrong.

[Greenfist 1863]

@clawhammer

Don't bring up the example of killing AI again. That is clearly beside the point and I agree with infiltrator that at least THIS can easily be filtered out.

Would this only apply to killing players then? That would rule out all coop players.

You're right, detecting unarmed AI would be easy. But what if they're armed but no ammo? Or their skills are set to zero? They simply couldn't hit the player. Or the player is equipped with nukes? Or maybe just allowdamage false, or impenetrably armored? The algorithm would check if the used weapon and armor are realistic and within boundaries? It would somehow know if the fight was fair to approve the stats? 

 

No mass killings in one go? Ok, I'll just leave my character standing on a server and go AFK. The script with then spawn AI units in front of me and fire my weapon automatically every 5 seconds for the rest of the day.

 

Or should the enabled servers be allowed to run only approved whitelisted mods and missions?

 

Sorry, but I really can't even imagine how this could ever work with a game like Arma. A nice idea, but very much a pipe dream unfortunately.