Self-proclaimed Hall Monitor OG 667 here. Is there any proper procedure for getting AFK players actually reviewed? I just played against another idiot tonight who had the following K/D spreads on their last 16 Firefight games (all today):
meaning that this player played two games (badly) out of 16… and it may be interesting to see how many of those 79 and 69 kills were in the bonus round. Looking through their game history from only a couple of days ago shows similar patterns.
This is the type of player who ruins FF for most of the community. Yes, yes, some people like the AFKers because they get more kills but some of us like to finish a game every now and then. Someone mentioned PMing an admin before but I’m hesitant to do that unless they do not mind getting random PMs.
Wouldn’t waste your time. I have a list of over 200 gamertags who have been AFKing for months.
Everytime I set a thread up about this it gets locked. I love playing firefight but have to play doubles on my own or wait for a friend to come on as almost everygame I come across an AFKer.
Nothing will be done because 343 don’t have the balls to do a reset or permaban on these players.
We should at least be able to post their gamertags so when someone from 343 see how many players AFK FF they might do something about it, but no we can’t do that as its against forum rules.
All I do now when I come across AFKers is to kill them and make sure I don’t kill many enemies so that they don’t get any free perfections.
The banhammer is fully automated, but unfortunately, it’s only able to guess if a player is AFKing (by checking if they have zero kills). It’d take a pretty extensive (and technologically-risky!) TU to fix that.
if it’s any consolation, people have devised tons of methods for dislodging AFKers from exploit spots. Search YouTube; you’re sure to find a few that suit your needs. :\
People can easily get around the 0 kills problem the AFK’ers face. All they do is wait till the last wave or bonus round and then get a few kills. It really annoys me going into a game with four people and having at least one person AFK. The most so far I’ve had was two people AFK. Something needs doing about this really.
> People can easily get around the 0 kills problem the AFK’ers face. All they do is wait till the last wave or bonus round and then get a few kills. It really annoys me going into a game with four people and having at least one person AFK. The most so far I’ve had was two people AFK. Something needs doing about this really.
Something does indeed need to be done, but unfortunately, nothing will be.
Adding real AFK detection would probably require releasing a TU that tampers with the code that handles button presses. Delicate stuff. Not something you wanna do to a game developed by an entirely different studio, when most of your programmers and testers are fully occupied by a separate (and arguably more important) project. :\
> Wouldn’t waste your time. I have a list of over 200 gamertags who have been AFKing for months.
I’m glad I’m not the only one doing this :-D. There are some social forums where you can post names but yep, I don’t understand the rationale behind not publicly naming them here.
> If it’s on holdout I grab a concussion rifle and bounce them off the map. Take that, your K/D!
Corvette - death by flying boxes
Outpost - death by gravity
Courtyard - death by being pushed into the wrong place
Waterfront - death by inevitability
…
> Adding real AFK detection would probably require releasing a TU that tampers with the code that handles button presses. Delicate stuff. Not something you wanna do to a game developed by an entirely different studio, when most of your programmers and testers are fully occupied by a separate (and arguably more important) project. :\
If detection required a TU, I would fully agree with you. I have worked with developers both during crunch and just before a code commit freeze and there can be completely different attitudes in those phases regarding new code. After release, it is even more delicate.
However, a TU should not be required. The stats are presumably held in a database or most of this website would be infeasible. Assuming the database is structured properly (I am sure it would be), it is a simple task to query it and get all the information you need. As an example, something like this:
SELECT COUNT(numberofkills) FROM GameStats WHERE PlayerID=‘OG 667’ AND numberofkills < 30 AND GameDate <= somestartdate and GameDate >= someenddate;
would return the number of potentially dodgy games I played in a range of dates. If they keep stats on which rounds the kills occurred in then you could subtract the kills in the bonus round.
Detection should all be possible server-side with no TU needed.
Perhaps if there was a way to record them somehow and categorize the film under a specific name (343AFK) so they’re easily searchable by the community. Then perhaps the folks who want to help curtail this kind of behavior could view and upvote films of legitimate idling behavior. Every week one person from 343 could take a few minutes to review the top films and manually drop the banhammer on idlers. Right now players are idling with absolute impunity because they can easily bypass the automatic detection system by getting 1 kill. If idlers get wind that their gameplay can be viewed and voted on by their peers then passed on to 343 I think this will seriously discourage AFK’ers from ruining this playlist any further.
^
This is a good idea. The cynical part of me wonders whether bans won’t happen until Halo 4 is released though; there may be worry on the business side that mass bans will affect sales.
> SELECT COUNT(numberofkills) FROM GameStats WHERE PlayerID=‘OG 667’ AND numberofkills < 30 AND GameDate <= somestartdate and GameDate >= someenddate;
>
> would return the number of potentially dodgy games I played in a range of dates. If they keep stats on which rounds the kills occurred in then you could subtract the kills in the bonus round.
>
> Detection should all be possible server-side with no TU needed.
Except that’s not detection. That’s just more guesswork, and it’s just as dodgy and easily-evaded as the current “0 kills” guess.
^
What are we trying to detect? Whether people are playing the game or not. If someone consistently has a low number of kills, they are not playing. If you can break it down by round or just discount the bonus round where dying doesn’t affect commendations, then you have an even better indicator.
Many proven scientific methods are based on measurement and extrapolation. Short of watching individual movies of FF games - which just isn’t going to happen - automated methods and basic, common sense are the way to solve this problem. It’s not guesswork, it’s heuristic detection. There may be a small number of false positives but those are likely 4-year-olds, people with broken thumbs, or complete muppets. The “I had to take a crap” defense does not hold 16 games in a row.
In terms of being easily-evaded, the only way to avoid getting a low K/D spread would be to… um… play the game? That’s kind of the point, here.
