Language In Comments

When did GS started with this pending for language use in comments. So much for freedom of speech.

F-U-C-K... this seems to work and everyone gets the message. Seriously if you get offended by bad language you should stay indoors with ear plugs and never visit any websites...unless they are for christies and their jebus lol

The filter is dumb, we know. However, it's either this or those people that are being intentionally offensive in their use of such language get all their comments go through as well - the filter has actually been in place since long before the site revamp as well, it's just been changed from auto-hide (no feedback to the user) to auto-pending (feedback to the user, and easier for mods to correct false positives too), making it clearer to the posting user when it's happened and at the same time allowing us to allow false positives.

What you you rather have?

1: No filter - all those offensive users get their comments going through and we have to rely on users manually flagging several hundreds of comments a day just to keep it under control because we can't monitor every stream 24/7. This would also likely mean that such offensive comments aren't removed in nearly as timely a manner as most of them are atm, potentially turning the comments into a cesspool that would in turn likely result in stricter actions being taken than now, as well as the filter being reapplied. We had some streams turning into cesspools under the old filter due to a lack of robustness as is, so is obviously not a very practical option.

2: Old filter - everyone's comments containing any words that the filter twigs onto get auto-blocked without them receiving any feedback that the comment has been removed by the automated system beyond people simply not responding to them, and variations and other commonly used to insult words get added to the filter over time, false positives slip through the net very frequently and in many cases never get allowed as a result.

3: Current filter - everyone's comments containing any words that the filter twigs onto get auto-pending, informing the user that their comment has to wait for mod or admin approval before it will go through, in precisely the same manner that posts containing media such as images/videos are auto-pending. Due to the very large amount of comments to go through, false positives still aren't reliably dealt with in an ideal time frame, but the vast majority are approved so that they actually appear quite fast, as opposed to being hidden without the user knowing.

Livefyre's filtering functionality is limited and a tremendous amount of comments pass through it on a daily basis - while I'm sure we'd all love a smart filtering system that auto-blocked everything inappropriate but also had no false positives, that's simply not realistic and at best is a target to strive towards.

On that particular word: it's not outright disallowed, several comments using said word are allowed through manually, but several inappropriate uses will obviously not be allowed as well. Being able to bypass the filter just means it's not robust enough though, something that will hopefully be corrected in the near future. We can more easily correct false positives than we can find inappropriate comments when they're not flagged in any way, so this largely comes down to practicality. I'll agree that there's far too many false positives atm however.

It's worth pointing out that freedom of speech does not apply to privately owned websites such as GS and that the site tries to keep the forum and comment areas as ones that allow people to feel safe when they're using them - as such harassment or being intentionally offensive in general is disallowed.

Just drop the filter. Entirely. What you need to do is actually start moderating comments. As in - temp bans for people who post truly objectionable material. Right now, all you do is delete the posts, and there's no indication or justification that it was done. This means no feedback for offenders to know what is or isn't actually out of line, and no real punishment. This is utterly terrible and not a way to run a comments section.

Either let everything fly, or actually do something about it. This current system is the worst of both worlds. You say that dropping the filter would require mods to have to deal with tons of flagging - well right now they have to deal with what is likely an order of magntitude more "pending" post approvals. Sounds horrible.

I would MUCH prefer no filter at all. I believe people should learn to just ignore "offensive" comments, as the people who write them simply feed off of negative replies anyway. Ignore them long enough and they'll go away. All the filter does is restrict word choice, which leads people to either use different words to post their offensive replies or to use punctuation hacks to post the exact same words anyway.

Meanwhile, I take the time to write a lengthy response, as I'm prone to doing, to try to actually discuss something at length and I'm banging my head against the wall because the stupid filter is stopping me even though I'm not using any "offensive" words. The word "pending" may as well mean "rejected" because I've never seen one of those posts show up after any period of time, and it seems like comment moderation takes days anyway. By that point, the conversation I was attempting to take part in is essentially dead. At least with the old auto-blocking, it's immediately clear that your post is rejected.

I agree with Thanatos2k, this current system is the worst of both worlds. I can probably count the number of truly offensive comments that I've read on Gamespot on one hand. There really is no need for such high levels of moderation. I thought that Gamespot had gotten lax on the language, since they started using words like "shit" in videos (by the way, you can bleep it out, but when we know what you're saying, the effect is exactly the same as if it hadn't been bleeped out) and I've been seeing comments all over the place with that word and other "obscenities" in them, but now all of the sudden I can't even post a comment without any profanity in it.

I understand that freedom of speech doesn't apply to Gamespot. But freedom of speech teaches us to be tolerant of the speech of others, and that's a lesson that I think is valuable to a community like Gamespot. Treat people like children and they will act like children. Try treating them like adults. I know not everyone here is an adult, but that doesn't mean they can't act like one.

Just now, another comment "pending," because I said "only a moron would suggest that the climate isn't changing." Obviously I'm not attempting to offend, I'm simply pointing out an absurdity.

Comment filter sucks.

Well i just wrote a comment with no foul language and linked gameplay video from Dailymotion just to show quality is better then youtube and its also pending comment. Im not saying just foul language attacks on other comments but in general, i often use "fucking hell" in speech or if i have opinion on something lol. Nothing bad about that.

Filter needs to be dropped all together. As i said if you get offended by language on the internet, dont use the internet.

Before I continue I want to be clear: I'm not in any position to decide whether or not it stays or goes, that decision is entirely down to the admins, this is largely just to get any idea of how people feel about the current set up, seen as it's popped up a few times now. As such, I can't promise anything will change. My opinions here also do not necessarily reflect the opinions of the rest of the moderator team nor GS on the whole. Filter changes will continue to be made in the future, but whether or not further changes will be clearly visible to the community on the whole is a different matter. Obviously filter bypassing to insult people is an issue, as are false positives. Unfortunately dealing with the former also causes more of the latter in general.

Ignoring and flagging would be best, but a lot of the time that simply doesn't happen, often people flag and reply but at the same time they often reply without flagging - similarly some people mass post offensive comments for whatever reason and the filter often helps to massively reduce the damage they can cause when a moderator isn't actively watching a stream or livefyre on the whole. Let's not forget that a fair few false positives come through users flagging comments as well - especially via the 'disagree' flag. The old filter was also designed to make users think they were being ignored until they realised they were actually being blocked as the comment still displayed while that particular account was logged in. Punctuation and the like being used to bypass the filter is obviously an issue however...one that has made the more transparent setup of using pending instead of auto-hiding a problem due to the amount of people intentionally bypassing it now.

A significant number of pending posts are allowed almost immediately, especially in new articles, including some with inoffensive use of profanities - due to the sheer volume of comments some do admittedly end up getting left for days however, which ideally is only going to be a short-term issue but I can't make any promises on that one I'm afraid. That was only the case with auto-blocking if you viewed it from a different account or while logged out (which meant a separate page and finding it in the stream) - the old system only hid the comments for other users, not the user that posted it. We had very few complaints as a result, but over time thousands of false positives were deleted by the filter with no one the wiser in most cases.

At a glance, there's consistently more than 100 comments removed each day, sometimes 500+, not including any comments stuck in pending. Some days this can account for 10% or more of the total comments posted. These vary in extremity from porn&death threats to people just throwing insults at each other repeatedly and random spam rather than posting anything worthwhile. If not for filters&moderators I'd be willing to bet it'd be far worse than system wars, to say the least. I'd say we're reasonably lenient on what's allowed and what isn't tbh - the problem is that the filter isn't particularly smart and the few ruin it for the many when it comes to several words.

Surely you can see how the term 'moron' can be used offensively and hence why it is filtered? Yes there are false positives (although not many for that particular word), the filter isn't really able to read context, only words and phrases, and as such there are almost always both upsides and downsides to adding terms to it. Personally I'd actually like the filters tweaked/split - one list of words/phrases that are outright disallowed, and then one for terms that are merely often used to insult others, which would allow us to clear up false positives more readily and potentially even eventually get the filter smart enough to block most inappropriate comments with a minimal amount of false positives. That is of course an ideal, and not necessarily going to prove practical, especially in the short term - livefyre in general currently only supports a single customisable 'profanity' filter and a 'spam' filter that to my knowledge cannot be modified directly. It's worth mentioning that the filter was made stricter at one point already due to many users being frequently particularly offensive towards certain editors as well as other users in a manner that resulted in a significant amount of manual removals being necessary.

We have quite a few mods taking part in the comment sections frequently, but it's not really practical to moderate every article all the time, especially with the comments often moving far faster overall than any other specific area of the site. We also have handed out bans (perm ones at that in some cases) on multiple occasions for people being rude (or worse) in the comments. Increasing strictness is certainly something we can consider, but that's not directly related to the whether or not the filter stays. Almost certainly not a removal of the filter to coincide with being stricter at that. The livefyre system is currently still poorly integrated and it's quite impractical to send a PM manually for every single comment that has to be removed - Strike PMs are still sent however when the acting moderator feels it is warranted, this is the same as how the forum currently works, merely that more comments are removed without a PM explaining precisely why due to how impractical it is to do so in the current system. There's a difference between flagged things getting overwhelming and pending ones: flagged ones may be crucial to be removed immediately, especially in extreme cases, whereas pending prevents such things from being visible to begin with, so it doesn't matter so much if we're a little slow to get to it..the biggest problem with pending atm is the fact that there's so many false positives to begin with, but that has been the case for a long time already - the only difference now is that users can see that their comments haven't gotten through the automated filter more easily by the pending marker, which is obviously causing some unrest.

All media is automatically sent to pending because there have been cases of images/videos/links being used to post porn in the comment sections. That filter has been in place for a long time now. I'm fairly certain the pending will be staying for that one too as it's quite practical for us to check them all quickly...at least when they're not hidden amongst many others, anyway. An account on the site only 'requires' an age of 13, which includes access to (almost) all the forums/comment streams, so we obviously want to be restricting adult content - having it visible even briefly is detrimental to the user experience for many users regardless of age as well, they don't come here for that sort of thing. Similarly users obviously don't want to be harassed or attacked in general - and expletives, among other terms, are commonly used for such things, hence why the filter blocks them.

The livefyre system is currently still poorly integrated and it's quite impractical to send a PM manually for every single comment that has to be removed - Strike PMs are still sent however when the acting moderator feels it is warranted, this is the same as how the forum currently works, merely that more comments are removed without a PM explaining precisely why due to how impractical it is to do so in the current system.

Well yes, that is the biggest reason why the current commenting system does not work. Either fix this issue or ditch Livefyre. I'd recommend the latter.

I believe GS are already in the process of working with the livefyre team to make sure the system is better integrated, and ditching it if they fail to meet targets, we can only be patient until that's sorted out in the end as it's well above anyone in the mod team as well. In the meantime we just have to make do with what we have, no matter how much we may dislike the livefyre system as it stands.

I understand that it's not up to you whether to change it or not, but I do appreciate the fact that you're at least giving it some attention and actually listening to us.

For instances of mass posts (if I'm understanding you correctly, that's what I'd call spam), the filter should be used to detect and eliminate them, not necessarily for being offensive but for being mass posts. Similarly, I would hope the filter would be able to detect a jibberish ad post and deal with it (I haven't seen any of them on Gamespot lately, so maybe it's already doing this?).

I agree that flagging isn't the only answer. Some people like to flag posts simply because they don't like them. I would think that flagging a post should require at least 10 flags from different users or something like that. I mean, if it's truly an offensive post, then many people will flag it. If only one or two people flag it, then it's probably not really all that offensive.

Yeah, I understand why moron is blocked, but I still feel it's a mistake. I mean, I can be quite offensive without using any kind of derogatory words at all, if I really want to. And even when used as an insult, is the word "moron" REALLY all that offensive? I don't think so. "Dude, you're a fucking moron." That's your opinion, fine. I've moved on with my life already.

I really don't care how many posts are removed. I'm fine with leaving them all there. I think what's even more annoying than the intentional troll posts are the random drive-by Facebook comments. It's far more annoying to see several dozen single-shot Facebook comments on a story than a couple completely off-topic bitchfests.

I actually don't dislike Livefyre, I just don't see the value of the profanity filter, especially as it works currently, and especially if profanity will be occasionally allowed through anyway. Spam filter, sure, and let's flag the offensive posts, but this profanity filter is bullshit.

The spam filter is catching most ad posts, though not all - we do have to manually remove some every now and then still at this point in time, that's one of the few areas that the forums are worse for, because we have to do the whole lot manually on the forums. Most of the ad comments get auto-blocked just because they tend to post the same/similar text such as URLs in all their comments, and the spam filter then twigs on and gets rid of the whole lot. It has a few false positives in itself, but only because of users repeatedly posting the same things, normally in an attempt to force a post past a filter, ironically getting them caught by an additional filter that makes it harder for moderators to correct false positives.

I don't believe there's ever been a comment that has been flagged (correctly) more than 6 times - and more than 3 is extremely rare too (roughly 0.1% of flagged comments are flagged 4 or more times). Quite a large portion of offensive comments only get flagged once or twice before being removed as is, some before being flagged (not even by the filters) at that too. More people flagging inappropriate comments is helpful though, especially for really bad stuff - it allows us to prioritise the really bad stuff better. 10 wouldn't be practical in the slightest as many users seem to never touch the flag feature.

I've got pretty thick skin when it comes to such things tbh, a lot of things I don't personally find offensive - it's mainly about whether or not a significant portion of the community may find it offensive, and of course the person or group targeted themselves in particular. We obviously can't block every instance of offensive content via just the filter, but that's where flagging comes in. Interestingly, I can't actually find moron (or idiot either, for that matter) on the filter list, despite it clearly being blocked by livefyre. Either it just isn't included on the current copy I can view or it's on a separate list somewhere. That said, even that is clearly insulting someone's intelligence - even if they're not offended by it it's not particularly appropriate use of comments under the existing ToU. Depending on what else is in the comment such things may well be removed, or may be allowed to slide based on context if it doesn't seem like it's actually being used to attack someone - potentially with the offending part only being removed too. However, it is very rare a moderator will edit a comment to remove just the offending part due to practicality reasons so anyone including offensive comments in their posts risks removal regardless of what else is written in them. That's of course the user's responsibility, whereas ours is simply to keep the peace and make sure the comments remain reasonably civil....and to get rid of spam and other unwanted junk posts.

From what I can tell, there's not been a single comment from facebook in the GS livefyre streams since February. As such, outside of old streams you shouldn't be seeing any facebook comments at all any more. I'm not sure if they simply disabled that crossover or livefyre is being buggy again, but I'm not really bothered either way on that one.

I'll agree that the profanity filter is somewhat flawed in it's current state, but I feel that can be corrected and is better than outright removing it. The spam filter and media requiring moderator approval parts are working fine as is on the other hand, though the spam filter itself does have a few minor flaws. If profanity was disallowed outright, the profanity filter would have far less 'false positives' as well - it's just that that isn't the case atm (and as far as I'm aware, isn't changing), which complicates matters when it comes to using an automated filter.

I don't think you can ever correct a filter. For example:

http://www.gamespot.com/articles/teenager-robs-bank-is-caught-hours-later-after-buying-new-gaming-system/1100-6419760/

A comment of "What a moron" would be completely applicable to this article and not insulting another user. But a filter will never figure that out. So if you can't fix that then don't bother with an automatic filter.

@Thanatos2k: That may well be the case for such a comment if the term moron remains blocked, but that's not going to be a common case, and is a false positive that is easily enough corrected manually - my main concern would be simply to minimise false positives to the point that all false positives can be easily found and manually corrected quickly without delaying the process of handling flagged content, which is not currently the case, as many have noticed. Of course, if necessary that may also mean removing some terms from the filter entirely, or better replacing them with a specific set of phrases in general such that false positives are vastly reduced while still catching most inappropriate content. What I'm thinking of isn't entirely impractical - merely a manner of giving at least some sense of context to the filter, without going so far as to require a complete rewrite of the system.

Being able to remove the filter entirely and yet maintain a safe environment for users in the comments is even less practical atm though due to the sheer volume of comments and the frequency of accounts made exclusively to cause problems - the moderator team is a team of volunteers and as such there are going to be stretches of time every now and then when no one is monitoring livefyre after all. Perhaps better integration and limits on newer accounts, as well as stamping down on problem accounts more (and users actually using the flag feature on livefyre more/properly) would help enough to make it work, but that'd have to wait on technical back-end changes.

That's part of the problem with the pending thing. Just reject the comment outright if I'm using language that's not allowed so that I can edit the comment and try again. I know that you have to wait for a pending comment to be approved, but I'd rather just edit until it goes through since I'll never know exactly why the filter rejected it in the first place. As it is, I don't even know whether the pending comment ever gets approved without actually going back to the conversation and checking to see if it shows up. After a couple days of doing that and it's still not there, I give up. It's a waste of time waiting around for days checking to see if a comment ever appears.

I don't believe there's ever been a comment that has been flagged (correctly) more than 6 times - and more than 3 is extremely rare too (roughly 0.1% of flagged comments are flagged 4 or more times). Quite a large portion of offensive comments only get flagged once or twice before being removed as is, some before being flagged (not even by the filters) at that too. More people flagging inappropriate comments is helpful though, especially for really bad stuff - it allows us to prioritise the really bad stuff better. 10 wouldn't be practical in the slightest as many users seem to never touch the flag feature.

Well yeah, if they're being removed after 2 flags, then they're not going to get flagged more than 6 times. The point is that in order to weed out spurious flags, you need more than just a couple from unique users. The reason why most people, like myself, don't touch the flag feature is because they probably don't find the comments worthy of flagging. Part of that may be due to the fact that the ones that got flagged appropriately are already gone. Maybe I'd actually use the flag feature if I saw some of the really bad ones, but as a rational adult I just don't care enough about what people write on the internet to try to have them removed. But my point is that if these posts weren't being so aggressively removed, then I still think 10 unique flags would be feasible, and that if a post can't even generate that much, then it's probably not offensive enough to worry about.

A filter is only good for filtering words, period. A filter cannot detect offensiveness. A post can only be considered offensive in context, and a filter cannot detect context. If you want to filter out bad words, then fine, but I see no point in filtering out words like "moron." Note that our entire conversation in this thread would have been filtered out by the dumb filter even though we're having an intellectual conversation here. No one is being offensive, and we're using our words like adults. Now go read some of the childish drivel that is rampant in the comments on any given article that doesn't get filtered out because it doesn't have any of the offending words in it.

If Gamespot wants to use a word filter, fine. Tell me what words I can and can't use. Stop pending my comments. Reject them outright if the filter doesn't like it. Nothing pisses me off more than writing several paragraphs of discourse only to have the comment pending while a bunch of asinine fanboys litter the comments with personal insults. I've had an account on this site since 2005. The biggest BS about all of this comment filtering is that it's not making the comments section any better, it's actually doing the opposite.

How about instead of rejecting posts or pending their approval, you just make the filter replace the offending words with asterisks? That's what a lot of other sites have done in the past. Then at least the comment goes through without the words showing up and without people having to resubmit their comments or moderators approve them. In fact, I've never understood why the filter didn't work like this in the first place.

@rarson: I can certainly understand your frustration, most of my comments tend to be quite long as well - and on the rare occasion they contain words that the filter would flag them for at that. Replacing the words with asterisks doesn't really help the fact that a post is being offensive most the time though - and more importantly I don't believe livefyre actually supports such a feature. Rejecting them outright was pretty much what the filter did previously of course, but I always considered that as hiding the bulk of the problem rather than fixing it.

As for a list - ignoring the fact that changes are made every so often - releasing a full list of filtered words would make it far too easy for users to bypass the filter, and we have enough issues with livefyre and the filter (including filter-bypassing inappropriately at times) as is.

This is a reasonable suggestion, I'll add it to the feedback queue to see if this is something we can implement.

So today I discovered that "Xbone" is an auto filter flagger. What in the &#@* is that? PS4 is not a filtered word. Talking about consoles is moddable now?

So my post gets the famous pending despite having no objectionable content, and hours later it's just gone.

Your claims that the false positives are low and will get handled by mods is completely untrue. The filter has to go.

@Thanatos2k:

I guess people tend to use the term "Xbone" in a derogatory manner, like MS is "boning" them, but it's still easier to write Xbone than Xbox One (and thanks to Microsoft's **** naming convention, it can't be something simple like XB3).

Oops, I said "****." That's offensive. Quick, someone delete my post.

From what I can tell, every mention of false positives here has been in agreement that there are too many of them, to the point that they are difficult to handle. I did however also point out that previously these false positives auto-hid comments in a way that made it seem as though they went through, and no correction was being done at all.

Xbone is one term that is certainly questionable (I believe I even brought this up very briefly a while back) and could arguably be removed as well - rarson's reasoning makes sense as to why it was added to begin with however - tbh, there's a couple like that, it's not exactly a small job to sort through every term in the filter and determine whether or not it's worth keeping though with the amount of words on the list. Only staff members (or site admins as livefyre calls them) being able to actually update the filter also means E3 is detrimental to the speed at which changes are made...not to mention that the moderator's themselves don't exactly have a limitless amount of time to spend on the site.

It took me half an hour to post a comment because the word "b-l-o-w" was preventing the comment from going through. My patience on this website is running thin as it is.

I would like to point out here that this is the system which was used for the comment system before the LiveFyre one. While it lets posts through and masks phrases, it won't mask the intent of the posts.

Implementation of this is simply a revert to times past. It's not a novel solution. Furthermore, I recall when it was around, it did nothing to nip crude and crass hostility at the bud.

With that said, I personally prefer a tweak to the "old filter" which RobotOpBuddy mentioned. To be specific, the tweak should include a message to the poster on what has been blocked and why it was blocked.

As much as this is a sledgehammer, this "do-it-this-way-or-no-way" system would encourage people to be more cunning in expressing their hostility, or simply don't.

Whether it's new or not has no bearing on whether it is more effective than the current way of doing it, and the current method isn't eliminating crudeness or hostility, either (interesting that you only care about the "crass" hostility).

By telling people why their posts are being rejected, you're effectively telling them how to circumvent the filter. A person who is just trying to write a legitimate comment will edit out their language but a person who is just trying to be rude will replace letters with symbols or choose different words for the same effect. As I said 9 posts ago, a filter cannot filter out offensiveness, all it can filter out is words. That's why it's not working. It's not capable of doing what you guys want it to do.

This is all but admitting an automatic filter does absolutely nothing to prevent your true purpose which is to stop "hostility."

First of all this is an impossible task and one that is questionable to begin with, but a filter does not help you reach this goal. If automatically replacing "banned" words with asterisks doesn't solve the problem, how does allowing people to do the exact same thing manually and letting the posts through help?

As with any website in the world, you will never win a profanity arms race with a filter. People simply adjust their vocabulary to avoid your filter. Filter "retard" and people will start using "mongoloid." Filter that and they'll use something else. Trying to control people's "intent" never works, has never worked, and will never work.

You cannot stop people from posting what they want to post, so your only choice is reactionary punishments to those who break the rules, namely ACTUALLY doing something in a transparent way to people who post objectionable material. Tell people what they did wrong. Temp ban them from posting. If they keep doing it after the ban wears off, ban for longer, and then permanently. If Livefyre doesn't let you do this, get rid of it immediately.

The word "moron" is autofiltered, but "douchebag" is not. This tells me all I need to know about how ineffective trying to constantly update a filter is. Stop trying to do it, and drop it entirely. Make people responsible for the words they post, and let them post them.

We are not going to remove the filters that we have in place to prevent abuse. Allowing the abuse to happen only to take care of it after the fact is not an option given the current state of volunteers that we have for the moderation team. If you have any suggestions on how to improve the filter I'm happy to hear them. But the filter is not going to go away, at least not anytime soon.

@LiquidGold According to Livefyre, the comment was "Submitted as pending by LiquidGold at 2:27 PM on Jun 5, 2014" then "Edited by LiquidGold at 2:28 PM on Jun 5, 2014." Not quite sure what happened here to be honest. I'll submit a report with the service.

@LiquidGold:

Either that's a glitch, or LiveFyre's filter actually considers entire sentences.

I have a slight suspicion that LiveFyre may have some zealous scripts which filter out what they suspect to be suggestions of domestic violence, e.g. your mention of violent old men with wrenches. I don't believe that LiveFyre's automated scripts are that complex, but then, I don't work at LiveFyre.

The filter could be easily improved by simply using it to filter out profanity, as it was designed to do. By profanity, I mean profane words, not every single word that might be construed as at all offensive to any person. In addition, automatically editing those profane words out, using asterisks or whatever, instead of pending the comments and having moderators reject them later would be simpler and more effective in achieving that goal. A profanity filter is fine with me, as long as it doesn't turn posting into the chore that it already is.

I know you guys have your terms of use, and it's your site and all that, but the only reason the site exists is that people want to use it. If you keep pushing away the users who are asking to converse intelligently, then all that will be left is the trash that you're trying to eliminate. You get a paycheck because people like me use your site.

Your filter is broken. You don't need to remove it, you need to fix it.

From this I'm pretty sure it's related to the word beating tbh, livefyre is a bit odd at times with the filters, sometimes even allowing things through despite being on the filter list (though it still flags them, at least) or on the flipside, blocking words just because they're a minor variation of one on the filter (things like tense differences). so something like this occurring isn't all that surprising. An additional oddity I noticed with this comment is that it's marked as pending in some places, and as bozo'd in others..which is obviously a bug. Easy to manually correct, but shouldn't be happening to begin with. In theory it should've been marked as pending everywhere still seen as the comment was changed back to the original one, which was marked as pending initially.

It is crystal clear the filter is not working properly. The only solution is removal. It's not preventing abuse - it is CAUSING abuse, and far more problems than it is fixing.

Stop being stubborn, admit when you messed up, and correct the problem.

If you still refuse to accept this, at the very least mitigate its awfulness and change the filter to replace "banned" words with asterisks and remove the "Pending" nonsense. Additionally, remove about 2/3 of the words on your filter list like "Xbone" These words are completely valid words that can be used within a normal comment. Filtering actual profanity is fine, but all other words must be allowed.

Also, let people post images, videos, links, and gifs again. It's mind boggling you took this feature away, which was the source of so much joy in the comments.

That feature is still there, we just have to allow them 1st..which is perfectly reasonable, especially seen as users were previously using it to post completely inappropriate content such as pornographic content and other blatantly inappropriate things. We're not having any issues keeping on top of them in general either - at this point in time there's not even 10 comments containing media in the pending list and that can easily enough be cleared out too. None of them are yours either for that matter.

The filters most certainly are preventing abuse, even if not all of it, they've even completely prevented multiple spambot accounts from getting any messages through at all. Of course you won't be able to see that, as what it does prevent never becomes visible to the vast majority of users to begin with. The only reason the filter is potentially causing abuse is the fact that people can now actually tell when the filter catches their comment, whether incorrectly or correctly - previously they could not and there were no complaints about the filter as a result. That could be corrected simply by reverting it back to auto-delete (the type where the commenter can't even tell it's deleted when they're logged in).

Personally however, I see that as a step backwards because it's then infinitely more difficult for moderators to correct false positives and ultimately only hides the root problem rather than fixing it. Yes, the filters have issues - complete removal is not a suitable solution at all though. Removing some words such as Xbone (I'm not even entirely sure why that one's on there tbh) is perfectly reasonable, though such words don't constitute anywhere remotely near 2/3rds of the profanity filter list...more like 1/20th, at the most.

@robotopbuddy said:

@Thanatos2k said:

Also, let people post images, videos, links, and gifs again. It's mind boggling you took this feature away, which was the source of so much joy in the comments.

That feature is still there, we just have to allow them 1st..which is perfectly reasonable, especially seen as users were previously using it to post completely inappropriate content such as pornographic content and other blatantly inappropriate things. We're not having any issues keeping on top of them in general either - at this point in time there's not even 10 comments containing media in the pending list and that can easily enough be cleared out too. None of them are yours either for that matter.

Well sure, like everone else, I've long since stopped attempting to post any links or images because they vanish into the pending void. I don't think you understand that when a conversation is happening, slapping a multiple hour waiting period between content entry and other people being able to see it destroys any kind of enjoyment from the experience. That's why the new pending thing doesn't work and needs to be dropped.

Also, images and gifs, and youtube embeds don't expand like they did before. Having to click to view is patently worse.

Before, the filter was not deleting posts because people said things like "Xbone" or "beating." That is why the new filter is terrible - it's hopelessly overaggressive. At the most it should be filtering the scant few profanity words that are disallowed on television (you know which ones), but that's about it.

I had somewhat of a long comment that kept showing up as pending even though I used no foul language or offensive words. After waiting for at least 2 hours and not seeing any resolution, I started doing a trial and error by immediately deleting the comment to see what would go through. I broke the comment down into alternating groups of paragraphs and found out that a certain number of "?" and "!" were triggering it. I removed a few (some had two "!!") and that worked. However, when I would paste it along with the whole comment, it would get "pending" again. So I had to resort to breaking it down into 3 parts and it all went through.

Don't know the criteria here exactly but it was a frustrating experience and it's probably one reason why you see no meaningful discussions taking place. Maybe that's the point, to keep the comments short and carry on the discussions to the forums but then they loose the immediate relevance a passing reader would have by having it right under the article itself. Maybe have a direct link to its respective forum thread for longer discussion purposes at the end of an article?

Honestly, that sounds a lot like a bug to me, mass symbols are probably getting marked by the spam filter or something, but just a couple in a row shouldn't be..if 3 parts go through individually then it should be letting it through when they're combined as well. That said, while I can find the 1st 2 parts of your 3-part message as approved, the 3rd part either isn't marked as clearly or didn't get through the filter, suggesting that something in the 3rd part of it was an issue to the filter for some reason.

I am sorry, but your claims are unsubstantiated which you used as the foundation of your argument. To my knowledge, there has not been a single moderator who has stated the process of moderating comments is as you say.

If you had proof, your comment would likely be considered as more credible.

Then again, moderation policy is not supposed to be discussed in the forums, anyway.

I know for a fact that I have had a post deleted before, and I received no indication or feedback that it happened or why it happened.

I have had posts go into pending with no swear words and never come out of pending. So clearly the mods are either not getting to the stack or they actually denied the post, AGAIN, with no indication to the user that the post was denied or any feedback about why it was denied.

What exactly is incorrect in what I said? If you're going to suggest I'm "wrong" (I'm not) then one would hope you'd actually elaborate on what you're talking about.

The point is that the moderation policy is so bad and poorly explained that it appears as though your commenting system is bugged.

I understand your point and if you would like to discuss moderation policy more thoroughly, you are welcome to private message the moderation team.

The pending notification has already been explained by RobotOpBuddy; it is part of the LiveFyre auto-filter.

So you're not going to tell me what "unsubstantiated claims" you were talking about? Or the "proof" I lacked that made my claims not "credible"?

Yes, the filter was explained, and that explanation reveals why such a filter is a bad idea, and unsustainable idea, and a currently but predictably broken idea with many innocuous words inexplicably being flagged and posts vanishing into the void never to return. You can say "It's not a bug, it's a feature!" but that feature is still a problem that is disrupting the enjoyment of your site and must be fixed, like any bug. Sometimes it's ok to admit you made a bad design decision and correct it.

Sure I will, in a private message granted you would like to know. I did not say anything to the contrary.

I can indeed say that as can you respond with constructive criticism as well as any other member of GameSpot. Also, it may come to some as a surprise, but LiveFyre is actually not made by GameSpot. It is its own entity and GameSpot utilizes the software of that company. In addition, DigitalDame told you and others in another thread over a week ago that the continued use of LiveFyre is being considered. Thirdly, as good of a time I am sure being staff of GameSpot is, I am a volunteer.

http://www.gamespot.com/forums/bug-reporting-feedback-1000006/please-ditch-livefyre-comments-asap-31261382/

All that is requested of you and anyone with concerns or criticisms is that they are directed at the correct people and within reason.

I *know* Liverfyre's terribleness is to blame. Have you noticed I already posted in that thread you linked over a month ago? Clearly nothing has been done, and it's been a long time since the redesign that added this wretched commenting system and greatly decreased the enjoyment of using this site. Instead of saying "We're considering it" actually make a decision and drop it. This thread and the one you linked are over a month old.

How long does it take to come to a decision? Who are the "correct people" to talk to? Who *does* have the power to make this decision and where are they? Why aren't they responding to the consequences of their decision to plague us with a bad implementation of Livefyre and ruin their own site?

I wrote:

In addition, DigitalDame told you and others in another thread over a week ago that the continued use of LiveFyre is being considered.

I do not mean to be forward. I believed that my awareness of your participation of that thread was blatantly apparent considering the above statement.

That thread was created over one month ago, but her response is more recent as are several in this thread. Your wishes are understood and are being taken into consideration. Please have some patience. For now, I would like to know what you propose is used instead. After all, there is no sense in trading one thing disliked out for another that will be. Before you answer, I will reiterate to you what has already been said: The comments section will not be left without some kind of filter system. That would be worse, believe it or not.

How long does it take to come to a decision? Who are the "correct people" to talk to? Who *does* have the power to make this decision and where are they? Why aren't they responding to the consequences of their decision to plague us with a bad implementation of Livefyre and ruin their own site?

That depends on the subject, staff for both of the next questions (some of whom are available to converse), and that is up for debate.

Answer the question, please. Who is the decision maker on this subject? Who will decide whether to keep Livefyre or not? Who has the power to decide that the site will switch?

The alternative I propose is Disqus, a commenting system widely adopted by numerous other gaming sites and lacking a language filter. (Or if it has one, I've never seen it used). It basically does the exact same thing the current commenting system does but better in every way. For example, it nests comments in reply chains better than the current system which after you get two deep just displays everything on the same horizantal plane making it impossible to see what someone is replying to. THAT may be an actual bug, but I don't know for sure. It allows the posting of pictures and gifs. It allows embedding video content.

You say that a filter would be "worse" but we know that isn't true because Gamespot didn't used to have one for years, and was far, far more enjoyable than it is now. (We could post pictures and gifs too. Now, we cannot. Even links may triggered the dreaded pending) We are not evaluating a completely new system, we are comparing it to the old system, and the current situation is demonstrably worse.

@Thanatos2k: The depth of the nesting has 3 settings on Livefyre to my knowledge: 2 deep, 3 deep, and 4 deep; the default for the site atm is 3, not 2, but that does have the potential to be increased slightly even in the current system (not saying it will be though). Livefyre also allows pictures/gifs/videos/etc, it's simply that comments containing media have to be approved due to having a pre-moderate setting. Note that Disqus also contains a pre-moderate links option that would likely be enabled if implemented here, and media display has to be enabled on Disqus for it to even allow images, etc - on Livefyre they're simply optionally pre-moderated rather than outright disabled. It also has other pre-moderation including a word filter much like Livefyre's 'profanity' filter and a spam filter as well. Also note that it would have similar implementation issues to Livefyre in terms of how well it meshes with GS behind the scenes.

By old system are you referring to the old in-house GS system? I believe the filter has always been in place since livefyre was implemented, even if the comments it was removing previously were hidden in a way that mislead the person that made them as if it was an attempt to make them simply think people were ignoring them. The in-house system I believe used the same filter as the forums at the time, which has since been removed due to the site taking a slightly more lenient approach on the use of profanity when not used offensively.

That's simply not true - I know for a fact that we could swear as necessary with no filtering involved before the site redesign. (And use the word "troll") So it sounds like in addition to implementing this new system in place there were some pretty bad policy changes coming from the top. Yes, I know Disqus allows you to implement/enable some of the same undesirable fun-destroying features that have made it into the site currently, but I would hope that during a switch you'd just....not do that.

What Disqus also provides is reply notifications, something that used to be on Gamespot (through emails) and has since vanished. The notifications for forum replies seems to work just fine though....

Another thing that's flawed in the current system is the performance. On any article with a large amount of comments, every time you click "Show More" it starts taking exponentially longer to load each new batch of ~50 comments. Once you get to about the 500 range it's taking upwards of 5 to 10 seconds each time you load more. From what I've seen Disqus doesn't have this performance problem.

By nesting I mean that when you have multiple replies to replies, they look like this:

1. Original post

One tab -> 2. reply to OP

Two tabs -> 3. reply to 2

Two tabs -> 4. reply to 3

Two tabs-> 3.1 another reply to 2nd

Two tabs-> 5. reply to 4

Two tabs-> 3.2. Another reply to 2

This gets extremely confusing, especially because it starts prepending "@Whoever @Someone, ..." in front which quickly makes it confusing about who's actually being responded to, and it's even worse because people can delete those callbacks from their post. (In Disqus it always displays who you're replying to) There's got to be some kind of solution for this. I recommend placing the most recent reply in the chain at the top instead of the bottom of the row of replies, which makes it FAR more readable.

@Thanatos2k: I answered you to the best of my current knowledge. That is the GameSpot administration who oversees those decisions.

@Thanatos2k: The Livefyre system for article comments didn't change with the new site itself - I imagine when it was 1st implemented there were a lot less words in the filter list (it may even have been using the default one rather than a custom one), the word troll in particular wasn't on there previously and the list would've been quite short at 1st, but that is just the word list being updated over time (both before and after the site revamp), even if not always with the best word choices (and Livefyre seemingly taking it upon itself to block similar words every now and then not helping). As I mentioned before, previously the comments were hidden in a way that would make comments appear visible to the commenter but remove it from everyone else's vision. It's not that there wasn't a filter, merely that it was working in a way that it was never clear if it had done anything or not for a normal GS user that's logged in all the time when viewing article comments. If you're referring to before GS even used Livefyre, then there was a filter initially but it was removed before Livefyre was adopted so it wasn't always there.

Can you be more specific on reply notifications? There's emails for every comment posted in a comment stream you were following, which obviously caused inbox flooding when combined with the auto-follow feature whenever you commented on any particular article comment stream. Other than that there's emails for 'liked' comments and for comments replied directly to you (incl. the @username part) I believe. I'm not sure what sort of separations they have option-wise for the notification types but hopefully it's possible to get replies for the latter 2 only.

I'll agree with you on the performance issues at the least - it's one of many reasons why I was glad that GS used the GS/GB chat system for E3 - Livefyre simply doesn't handle comments well when they're being posted very fast, and the performance issues become incredibly obvious at times. That said, I've never seen a Disqus chat system that's had to put up with the amount of comment traffic that Livefyre has to on GS either, so it could well have similar issues.

The nesting/reply stuff you mentioned is fair enough - whether or not it's implemented is down to the Livefyre team (assuming we even keep Livefyre) and the GS staff though. As mentioned the nesting level itself could also be increased if deemed beneficial.

Who? What are their names? How does one contact them? Why do you keep avoiding this question?

Before the site redesign, you would receive emails whenever someone responded to a comment of yours, with the message of the reply and a link to the article where that comment was present. You'd also get emails whenever someone liked a comment, but often these would get clusered together into some kind of digest mail indicating how many likes your comment got. The like emails were pretty worthless and spammy and I'm glad they're gone, but the reply ones were much more useful.

After the site redesign (and implementation of Livefyre) these notifications have ceased completely from comments posted under articles and reviews. Ironically, you replying to me in this topic on the forums causes an email notification, but comment reply notifications are completely broken, or they have just been removed as part of the new system.

Disqus on the other hand has a notification menu in its interface that tracks every single reply to one of your comments (across all sites that use Disqus, no less) and has links that lead you directly back to those comments and replies. No needing to click "Show more" until your comment is in view. This feature alone should be enough motivation to move to Disqus because it makes conversation tracking actually possible. And of course every time someone clicks to view their comment it'll open the article page up which means more page views for Gamespot. Win win for everyone I'd say.

I'm sure many sites don't deal with the total comment throughput over the entire site that Gamespot does (although Destructoid uses Disqus and they seem to get a lot of comments too), but I've seen plenty of sites which have articles with the same magnitude of comments on an article that Gamespot gets on some (you know, the 1k+ comments usually found on those awful clickbait console war articles), and I've yet to see the performance issues creep in there. There has to be some kind of inefficiency in the code of how Livefyre does the loading, because it should not take more and more time to load the same number of new comments unless it's somehow loading stuff related to the existing loaded comments as well, and I'm not entirely sure that's something you guys can fix.

@Thanatos2k: Sir, you just accused me of avoiding answering you as as a direct response to a concise answer I gave you.

For the sake of everyone interested in knowing the step by step process I recommend:

1. Go to the Bug & Feedback Forum
2. Select "Start a New Thread"
3. When making a title, direct it at the GameSpot staff by including the words "GameSpot Staff" at some point.
4. If you have any criticisms, be assertive and polite! The staff have made this thread for constructive criticism (Bug Reporting & Feedback), but being a jerk while doing so makes any feedback self-defeating. Trust me!
5. If there is anything on GameSpot you can explain, you might be able to also take a screencapture (ctrl+prtSc, open Paint or Irfanview+paste) or record a video of (with proper hardware and software) take a screenshot of. This is an immense help to developers and makes a big difference in responses you will see.

I wish I could tell you all more, but that is all I currently know. I like GameSpot overall, so users will see me commending the site developers, though there are bugs and things I would like to see back on GameSpot or on this site for the first time.

May your feedback be listened to.