PDA

View Full Version : Proper scoring for troll postings, per seanadams



pfarrell
2010-03-17, 20:48
We seem to be attracting a lot of, er, unfounded and uninformed postings about the audio quality of various SD/Logitech devices, the brains or lack thereof of Logitech management, etc.

The Touch seems to bring out nearly as many as the Audiophile forum does.

I'm a fan of reputation based systems, where the community provides information on the value one should place on a posting's author.

A while back, a well regarded forum member wrote into one of the Audiophile threads:

"2/10
I think we should start scoring the trolls. This post lacks originality - you were doing better earlier with the bath tub business."

The author seemed to know a lot about the legacy Slim Devices products, he uses a handle of "seanadams".

What's not clear to me is how the SeanTrollScore system works. is it like a signal to noise ration, where smaller numbers means worse (more noise, less signal) or does it mean "this posting is only a 2 out of a 1 (no troll) to (10 100% troll food) so that a highly troll score is 10 and a good posting is 1 or 2.

Seems to me that lower is more trollish, like a bad signal to noise ratio. But I'm not sure, and need input from the user community.

Thanks
Pat

Mark Miksis
2010-03-17, 23:25
FWIW, these forums have both thread rating and user reputations enabled. They could possibly benefit from some tweaking and publicity, but they are there. Almost no one ever uses them.

erland
2010-03-17, 23:46
FWIW, these forums have both thread rating and user reputations enabled. They could possibly benefit from some tweaking and publicity, but they are there. Almost no one ever uses them.

Where is the user reputations seen ?
To make it useful it feels like it at least needs to be indicated somehow in the heading of each forum post.

Still, I suspect in most cases the post count or registration date of the user will give you a better indication unless we get people to actually start adding points to the reputations of different users. The trolls often have a low post count or have just registered. Another problem is that many senior members uses the mailing list and not the forum.

If we like to get rid of pointless posts, it might be a better way to do more moderation. Although, this is a tricky road as it can easily backfire and give a lot of negative critique creating even more pointless posts.

Mark Miksis
2010-03-18, 00:53
Where is the user reputations seen ?
To make it useful it feels like it at least needs to be indicated somehow in the heading of each forum post.

Heh, this is less than obvious at best...

In the upper right hand corner of each post is a block of info with a small colored square at the bottom. If you hover your mouse over that square, it will display an alt-text such as "erland is on a distinguished path". To make matters worse, that particular phrase means you are somewhere between +10 and +30, but there's no way for users to know that.

AFAIK, the text descriptions can be changed, but we are stuck with the tiny "traffic light" icons instead of just text.

Mark Miksis
2010-03-18, 00:54
I'm also not sure how to give another user negative reputation points.

Siduhe
2010-03-18, 02:56
Ok, I've been searching for the last ten minutes - can someone point me to the Sean Adams response'o'matic multiple choice auto response post? Sounds like we could update and resurrect it...

bobkoure
2010-03-18, 04:33
Heh, this is less than obvious at best...
OK - this is the first time I've ever noticed those.
Can the forum software change that graphic based on the score?

That said, I don't think we have a problem. Spam posts are very much under control and "trolling"... well one readers "troll" may be some writer's honest rant. I haven't seen any that I would regard as "trolling" (post written to specifically be so "off the wall" that it gets at least some forum members riled up, and generates a long stream of responses). As at least some of us remember, this was a real problem with usenet discussion groups. There was even trolling in the .lang groups, and I hear that it got downright malicious in some of the .pets groups.

And, frankly, that "honest rant", although it's sometimes a bit of a pain to deal with, might serve as some sort of "early warning system" for any Logitech folks reading the forum. OK, well, knowing that some folks are getting antsy about the Touch delay probably isn't news... On the other hand, knowing that there's a strong constituency for funding 3rd party developer support and a stable API, might be good. There's likely a bit of a fight over where development dollars go, and those posts might give the folks on "our" side a bit of ammunition.

Sorry to be so long-winded...

fragfutter
2010-03-18, 05:14
and in usegroups i had my very own scorefile, could ignore users, or threads, could highlight other users, or subjects, or ...

usenet was superior to any forum i've seen.

gregklanderman
2010-03-18, 07:59
>>>>> On March 18, 2010 fragfutter <fragfutter.480r1b1268914441 (AT) no-mx (DOT) forums.slimdevices.com> wrote:

> and in usegroups i had my very own scorefile, could ignore users, or
> threads, could highlight other users, or subjects, or ...

> usenet was superior to any forum i've seen.

You can subscribe to the mailing list and use a newsreader
to read the mail. That's what I do, works great!

greg

fragfutter
2010-03-18, 08:40
getting off-topic, but i need to ask...

how would i read a mailinglist with slrn?

snarlydwarf
2010-03-18, 08:44
been a while since i played on lusenet, but nntps://snews.gmane.org/ should
work as a news server.

iPhone
2010-03-18, 09:14
Don't know if the green block is any help or even a so so indicator. I have yet to see any color but green. Even a new Forum member with only 5 total posts "is on a distinguished path".

When I look at my User information, I can see when people have taken the time to use the scales icon to rate/leave feedback but I don't believe others can see that only the member (please correct me if this is wrong).

I think a good point was made that most trolls or trolling is done on purpose by somebody that just signed up. So again I will also voice that Forum Join Date and total number of posts has some degree of bearing on this subject. The few that become trolls over time tend to get to aggressive and end up getting banded from the Forum.

I would also have to agree with Pat that some of the recent comments and frustration over the Touch release delay have been unwarranted and over the top at times. I responded to several of them out of attempting to correct errors and rumors but that only seemed to create more posting. I would guess the one that caused Pat to start this tread is the one that I started to answer 3 separate times and ended up deleting my reply each time before posting as typing it let me get my frustration out and better judgement enabled me to realize that if the poster actually felt that way then my post wouldn't help and would probably just cause more posting along the same lines.

All in all this is an excellent Forum with level and equal moderation (great job administers). The members by and large follow the Forum Rules and I believe more often then not attempt to help other members. Lastly, this is one of the few Forums where members can speak their minds about the Company and Products without fear that negative posts will be deleted just because they are not glowing posts about company and its products.

It is also my opinion that Squeezebox would not be where it is today without this open environment of a pretty much free to speak what is on ones mind Forum, open source software, pluggins, and the company listening to customers wants, dreams, and wishes as well as actively reading and replying to the Forum.

seanadams
2010-03-18, 09:16
Points are awarded for the effectiveness of the troll, with consideration for style, originality, and stealthiness. Judges especially favor a convincing appeal to newly discovered pseudoscience and techno-jargon.

JJZolx
2010-03-18, 09:22
Points are awarded for the effectiveness of the troll, with consideration for style, originality, and stealthiness. Judges especially favor a convincing appeal to newly discovered pseudoscience and techno-jargon.

Transporter: The greatest trick the devil ever pulled was convincing the world he did not exist.

iPhone
2010-03-18, 09:24
Points are awarded for the effectiveness of the troll, with consideration for style, originality, and stealthiness. Judges especially favor a convincing appeal to newly discovered pseudoscience and techno-jargon.

Is it a seven judge panel where the high and low scores are discarded then the remaining scores averaged? Are the judges appointed for life or on a rotating panel?

And as far as what the judges favor, does one get bonus points for using newly discovered pseudoscience and techno-jargon in the Audiophile Forum then say General Discussion?

Ending on a serious note, great to see you still checking in on the Forum, Sean.

seanadams
2010-03-18, 09:35
Is it a seven judge panel where the high and low scores are discarded then the remaining scores averaged? Are the judges appointed for life or on a rotating panel?

Yes, yes, and yes. However, note that we refer to "judges" in the pluralis majestatis sense.

pfarrell
2010-03-18, 09:39
seanadams wrote:
> Points are awarded for the effectiveness of the troll, with
> consideration for style, originality, and stealthiness. Judges
> especially favor a convincing appeal to newly discovered pseudoscience
> and techno-jargon.


OK, so if a post talks about newly invented pseudoscience, it gets more
from the judges. If it uses well established nonsense such as jitter, it
gets some, etc.

The best score is 10/10, for perfect execution, creativity, and freshness.

Note, a 1/10 is still troll food, just not as perfect.

--
Pat Farrell
http://www.pfarrell.com/

seanadams
2010-03-18, 09:46
if it uses well established nonsense such as jitter

3/10.

Phil Leigh
2010-03-18, 09:47
3/10.

:-) - nice one.

erland
2010-03-18, 10:53
Just to get a reference point, I'm really interested if there are any known 10/10 trolls ?
Or is this like perfect audio quality, an unreachable level which no one is ever going to reach ?

pfarrell
2010-03-18, 10:56
erland wrote:
> Just to get a reference point, I'm really interested if there are any
> known 10/10 trolls ?
> Or is this like perfect audio quality, an unreachable level which no
> one is ever going to reach ?

I've never seen one in six years. I think its the perfection,
unreachable. Most common trolls seem to get 2/10 or 3/10, some rise to
5/10. The judges are fair, but not generous.

--
Pat Farrell
http://www.pfarrell.com/

Phil Leigh
2010-03-18, 11:08
I reckon anyone who mentions Firmware 15 must get 6...

seanadams
2010-03-18, 11:56
Just to get a reference point, I'm really interested if there are any known 10/10 trolls ?

The perfect troll, by definition, is perfectly stealthy. Thus it could never be detected so as to receive a rating.

This may seem paradoxical but it is elegantly explained by a sort of warping of the "thread fabric", which occurs as the limit is approached.

iPhone
2010-03-18, 12:08
The perfect troll, by definition, is perfectly stealthy. Thus it could never be detected so as to receive a rating.

This may seem paradoxical but it is elegantly explained by a sort of warping of the "thread fabric", which occurs as the limit is approached.

So if I follow this, as the elusive perfect Troll approaches perfection basically nearing something similar to an Event Horizon, as the Troll reaches 9.99/10 we still see him but he has actually disappeared and its really only a past image of him that we still see hence we never actually see the perfect Troll?

Are we still on topic or did the Tread cross over the Event Horizon while I was typing?

gregklanderman
2010-03-18, 15:13
> how would i read a mailinglist with slrn?

I'm not sure every newsreader can necessarily read mail, but some do.
I use Gnus, an emacs based newsreader, which also handles email.

If you're motivated, you could probably run a local news server, pipe
the slim mailing list mail into it, and use any newsreader you want.

greg

Mnyb
2010-03-18, 15:28
Did anyone remember those fellas that claimed that the SB sounded different if you paused the music and then started it again ?
They got the tread going for 5-6 pages and claimed that they would be back with measurements, but then it just stopped ?

Id just now realized what it was 2 years after the fact, that was good trolling ! Nobody called them out when it was an active tread .

It was in the Audiophile forum, they must have laughed their socks off.

What would that score ?

Siduhe
2010-03-18, 15:42
Aha! Found it - who wants to have a go at a non-audiophile forum version?:

You claim that an
( ) audible
( ) measurable
( ) hypothetical

improvement in sound quality can be attained by:
( ) upsampling
( ) non-oversampling
( ) increasing word size
( ) vibration dampening
( ) bi-wiring
( ) replacing the external power supply
( ) using a different lossless format
( ) decompressing on the server
( ) removing bits of metal from skull
( ) using ethernet instead of wireless
( ) inverting phase
( ) reversing “polarity” of resistors
( ) ultra fast recovery rectifiers
( ) installing bigger connectors
( ) installing Black Gate caps
( ) installing ByBee filters
( ) installing hospital-grade AC jacks
( ) defragmenting the hard disk
( ) running older firmware
( ) using exotic materials in cabinet
( ) bronze heatsinks
( ) violin lacquer
( ) $500 power cords
( ) a universally applicable omnidirectional tweeter

Your idea will not work. Specifically, it fails to account for:
( ) the placebo effect
( ) your ears honestly aren't that good
( ) your idea has already been thoroughly disproved
( ) modern DACs upsample anyway
( ) those products are pure snake oil
( ) lossless formats, by definition, are lossless
( ) those measurements are bogus
( ) sound travels much slower than you think
( ) electric signals travel much faster than you think
( ) that's not how binary arithmetic works
( ) that's not how TCP/IP works
( ) the Nyquist theorem
( ) the can't polish a turd theorem
( ) bits are bits

You will try to defend you idea by:
( ) claiming that your ears are “trained”
( ) claiming immunity to psychological/physiological factors that affect everyone else
( ) name-calling
( ) criticizing spelling/grammar

Your subsequent arguments will probably appeal in desperation to such esoterica as:
( ) jitter
( ) EMI
( ) thermal noise
( ) quantum mechanical effects
( ) resonance
( ) existentialism
( ) nihilism
( ) communism
( ) cosmic rays

And you will then change the subject to:
( ) theories are not the same as facts
( ) measurements don't tell everything
( ) not everyone is subject to the placebo effect
( ) blind testing is dumb
( ) you can't prove what I can't hear
( ) science isn't everything

Rather than engage in this tired discussion, I suggest exploring the following factors which are more likely to improve sound quality in your situation:
( ) room acoustics
( ) source material
( ) type of speakers
( ) speaker placement
( ) crossover points
( ) equalization
( ) Q-tips
( ) psychoanalysis
( ) trepanation

(c) Sean Adams 2007

Peter314
2010-03-18, 16:33
I reckon anyone who mentions Firmware 15 must get 6...

*sigh* 6/10


You set 'em up, I'll knock 'em down... :)

iPhone
2010-03-18, 17:35
Rather than engage in this tired discussion, I suggest exploring the following factors which are more likely to improve sound quality in your situation:
( ) room acoustics
( ) source material
( ) type of speakers
( ) speaker placement
( ) crossover points
( ) equalization
( ) Q-tips
( ) psychoanalysis
( ) trepanation

(c) Sean Adams 2007

Exploring Sean's Q-tips tip has been found to be the biggest bang for the buck tweak as far as achieving better sound at the human receiving source!

I am surprised that Sean left the almighty cryogenics treatments off his very through list.

Mnyb
2010-03-18, 21:37
A really cool thing is if someone could write a nice bot software that auto responded to to posts with Seans form prefilled.

Searching for key phrases like quantum purifiers, "wav sounds better than flac"

cunobelinus
2010-03-19, 00:40
.....and it would change its behaviour when observed. Can also be active in several places at once and affect directly at enormous distances the behaviour of others unfortunate enough to be entangled with it. Quantum troll.

On 18 Mar 2010, at 18:56, seanadams wrote:

>
> erland;526148 Wrote:
>> Just to get a reference point, I'm really interested if there are any
>> known 10/10 trolls ?
>
> The perfect troll, by definition, is perfectly stealthy. Thus it could
> never be detected so as to receive a rating.
>
> This may seem paradoxical but it is elegantly explained by a sort of
> warping of the "thread fabric", which occurs as the limit is approached.
>
>
> --
> seanadams
> ------------------------------------------------------------------------
> seanadams's Profile: http://forums.slimdevices.com/member.php?userid=3
> View this thread: http://forums.slimdevices.com/showthread.php?t=76315
>
>

esbrewer
2010-03-19, 08:33
The trained educators among us know that a simple X/10 scale as Sean has proposed does nothing to illuminate:

1. the criteria/attributes that best describe the ideal troll
2. explicit descriptors of levels of performance for such criteria
2. a particular troll's performance for such criteria for a given thread

I'm sure there are some valid troll rubrics out there. They might just need slight modifications to fit this particular forum.

How can we expect our trolls to achieve excellence in their work without first establishing a proper standard?

seanadams
2010-03-19, 08:52
1. the criteria/attributes that best describe the ideal troll
2. explicit descriptors of levels of performance for such criteria
2. a particular troll's performance for such criteria for a given thread


As in audio testing, the details of our methodology are proprietary. Furthermore, we have finely honed senses and decades of experience that no layman could hope to match. There is no point disclosing the criteria since you have no means to evaluate them.

Siduhe
2010-03-19, 08:59
as in audio testing, the details of our methodology are proprietary. Furthermore, we have finely honed senses and decades of experience that no layman could hope to match. There is no point disclosing the criteria since you have no means to evaluate them.

4/10

;-)