It seems that you're using an outdated browser. Some things may not work as they should (or don't work at all).
We suggest you upgrade newer and better browser like: Chrome, Firefox, Internet Explorer or Opera

×
Did someone say 'bigger fish to fry'? Check it out. A German fisherman caught this yesterday, it's a new world record.
Attachments:
fish.jpg (47 Kb)
avatar
tinyE: Did someone say 'bigger fish to fry'? Check it out. A German fisherman caught this yesterday, it's a new world record.
Holy crap, what type of fish is that? Also, what the hell was he fishing for that meant he had strong enough line to haul that thing in?
avatar
tinyE: Did someone say 'bigger fish to fry'? Check it out. A German fisherman caught this yesterday, it's a new world record.
avatar
wpegg: Holy crap, what type of fish is that? Also, what the hell was he fishing for that meant he had strong enough line to haul that thing in?
It is a Halibut, great eating if you've ever had one. The article it came from was a bit vague but evidently it wouldn't fit in the boat so he had to tow it in. Nothing about what kind of lure or line he used. Also, it pointed out where the fisherman was from and where Halibut are commonly found but it didn't say anything about WHERE he caught the damn thing.
avatar
wpegg: Holy crap, what type of fish is that? Also, what the hell was he fishing for that meant he had strong enough line to haul that thing in?
avatar
tinyE: It is a Halibut, great eating if you've ever had one. The article it came from was a bit vague but evidently it wouldn't fit in the boat so he had to tow it in. Nothing about what kind of lure or line he used. Also, it pointed out where the fisherman was from and where Halibut are commonly found but it didn't say anything about WHERE he caught the damn thing.
Norway it would seem, though nothing more than that. Still no mention of the line. I'm guessing that either the thing was close to death, or the stats I've heard about a fish being able to pull 1/3rd of its weight underwater do not linearly scale.
avatar
wpegg: It would need to be confident of not resulting in false positives. Once again, here I'm interested to know what your hypothesis test would be. I understand that stats are never perfect and a z test will only deliver a degree of confidence in a result, but you could suggest the criteria (in a x > 5 etc. way) so I understand how you're planning on assessing this.
Of course, false positives would be avoided but how much effort needs to be put into avoiding them depends on the action you take on them. An inconvience like losing the abiltiy to rate posts would need only.. 90% confidence maybe? Forum bans would be more like 99%, obviously. If we're talking about the criteria for being reporting to GOG by myself, with my limited dataset would mean it would have to pretty low or it'd never be triggered.. 80% or so?

avatar
wpegg: This was more a statement of fact that GOG won't do it, as a qualification of the later argument. I wasn't suggesting you were expecting them to go on your word, in fact I was implying that we both accepted they wouldn't.
Ah okay. Sorry that I mistook this part!

avatar
wpegg: This hasn't really addressed the point that I was making. The information you would provide them with is inconsequential to any action they might take. Either they decide to act or they don't, telling them a specific person is most likely doing this does not make a difference to that, or prompt them into any action. It's clear that it is happening, so that's all the information that might push them to do something.
They aren't doing anything. If they are not doing anything knowing that this is going on, those bad people have more say here than I do because they are abusing the system where I won't. I'm not going to participate on a forum where this is a condoned situation.

avatar
wpegg: If GOG have an alt matching system, then sure they could use it, problem solved. If they don't it would be expensive to write. I suspect they don't have one given the furore with Anasto. However I'm not suggesting that this form of bullying isn't nasty, it is a bitter thing for people to be doing and I agree it should be stopped. I'm just questioning whether this course of action can actually have any effect.
Why do you think it would be expensive to write one? One could be made in an hour or so.

And they definitely don't have one. They don't even have a test suite for the site :P

avatar
wpegg: I hope you won't leave over this, as I think the response (or lack of) from GOG is increasingly suggesting that they have bigger fish to fry.
It would take one minute to post in this topic that they are aware of the problem and are going to do something about it. Pray tell, what could possibly be going on where they have time to banter in [url= http://www.gog.com/forum/general/gogcom_will_be_recruiting_at_gdc_europe?staff=yes]forums but not to address the community about a problem they are having?

I find your conclusion that silence indicates something other than apathy to be a strangeone :P
Post edited August 17, 2013 by xyem
avatar
xyem: <snip>
Ok, I just can't handle all the quoting overhead, so I'll bunch this up about in order to what you said.

The hypothesis test, I was hoping you'd provide criteria for passing or failing the test, rather than assessment of what levels of confidence would mean. A bit like our stats courses (I only went up to A-Level further maths, so if you've done degree work I'll probably have missed something) where we would say "I think this person is doing this, I will prove it by providing these metrics over these statistics, demonstrating that with a 99% level of confidence that my hypothesis is correct". I was hoping for the metrics you would apply to demonstrate complicity.

Alt testing, if you can write one, I'll happily write corresponding tests to highlight any faults (black box, no need for the code). I'm happy to set up a test site, and the work would probably be worth something beyond GOG - your copyright. You could then offer that as a bargaining chip.

As for GOG's response. You're not the only one that has written tools to overcome user problems with the site (not that you suggested you were). Barefoot essentials being an obvious mention. I myself also encountered a situation where someone was abusing the site, and wrote something that would enable a secure GOG game trading platform. It turns out people didn't want to participate, but I was also annoyed by the cold shoulder that GOG gave me in seeking permission, or resolving the issue themselves. However in that case they did eventually respond and sort the situation out. So maybe the wheels are turning slowly, however I think I agree with you here, it's apathy.
"I'm not going to participate on a forum where this is a condoned situation."

" It would take one minute to post in this topic that they are aware of the problem and are going to do something about it."

Well it is clear to me that they are not explicitly condoning it but in their silence, they are implicitly saying something that effectively translates to "We condone this behavior" or something akin to that. Inaction and silence are messages too.
avatar
Momo1991: Inaction and silence are messages too.
However you only know they are guilty of the latter.
avatar
Momo1991: Inaction and silence are messages too.
avatar
wpegg: However you only know they are guilty of the latter.
True ;-)
avatar
xyem: "and" gets much higher precedence than "rather than". In fact, I can't see any way, even with added punctuation, to make the "track" be on both sides of the "rather than". It would require being explicted added as far as I can tell..
Nice try, but the question was what to do AFTER tracking them down. And besides, you are very wrong anyway, hitting the - button once in a while is obviously much less bother than creating and sending reports to customer service.
LESS UGLY THAN THIS THREAD = http://www.gog.com/forum/general/as_n_when_ya_numpties
avatar
wpegg: The hypothesis test, I was hoping you'd provide criteria for passing or failing the test, rather than assessment of what levels of confidence would mean. [snip] I was hoping for the metrics you would apply to demonstrate complicity.
I can't provide the metrics because I don't know the available data set. Are ratings available? If so, one metric would be "only interactions with the site are ratings". The metrics would increase/decrease the confidence level and the user of the system would set the confidence level for output.

So running this:
findalt -c 90
would output a list of all accounts that had 90+% confidence of being alts (along with possible "parent" accounts) and a verbose mode could output a breakdown of why that level of confidence.

avatar
wpegg: Alt testing, if you can write one, I'll happily write corresponding tests to highlight any faults (black box, no need for the code). I'm happy to set up a test site, and the work would probably be worth something beyond GOG - your copyright. You could then offer that as a bargaining chip.
Thanks for the offer, but I doubt I'll have reason to call up on it (unless it extends to my other projects :D).
I'm up I'm down I'm up I'm down
What the hell am I!? The rep system itself must have ADD. :P
avatar
jamotide: Nice try, but the question was what to do AFTER tracking them down. And besides, you are very wrong anyway, hitting the - button once in a while is obviously much less bother than creating and sending reports to customer service.
I refuse to speak with you again.

If anyone wants to know why, I'll happily answer. If not, feel free to take my refusal as you will.

EDIT: Added quote so people know who I even said that too. Good job missing that, Xyem :P

EDIT #2 (maybe?): By the way, I find it hilarious that this post got low-rated. I didn't know my sarcasm was that scathing :P
Post edited August 17, 2013 by xyem
avatar
xyem: <snip>look above</snip>
I don't mean to patronise you, I suspect you already know this, but I'm trying to extract the information in the form I work best with.

Could you provide the specification of the z-test that you would apply on the data that you will gather? In a standard formal definition language (any, I'll learn it).