SETI Cheating scandal--Stats meaningless?

[networkman]

If a ton of results are coming from a single IP it would have to be either a badarse mainframe or a cheater.

Not so! I have cable internet access thru ATTBI and a single IP at the router/firewall. At one point, I was cranking out 70+ WUs/day just from home. No cheating there! Just a guy who's admittedly a little nuts and has the space/power/time to feed his stats addiction.

 

[Orange Kid]

Originally posted by: Baldy18
An interesting point has been brought to my attention. SETIQUEUE OWNERS BEWARE! If you operate a team SetiQueue and rsubmit dupes you may be under suspicion. The best thing for SetiQueue owners to do would be only resubmit dupes once and if they don't go through then delete them or not try to send them at all.

Sorry to disappoint you all, but I no longer resubmit dupes. After taking a good look at them, they are just that--duplicates--submitted and gone through once to Berkeley.
Mine all just get deleted every so often..............same goes for any results that were dropped.
Like it or not, this seems to be the best way to do it and by far the easiest for me, as I don't seem to have the time I once had to play with the Q

 

[conjur]

The only time I've had my SetiQueue show results as dupes is when I was on a PC and it was sending results but for some reason the result didn't upload to the server properly (I was having network cabling issues and from my laptop at work I sometimes lost dial-up connections right as I was uploading...that happened twice). I'm afraid to upload the six dupes I have out there now. I'll go ahead and trash them as six WUs aren't going to amount to a hill of beans for me.

 

[Freewolf]

It seems that some people have forgotten the reason we do seti in the first place! I don't do seti just for numbers althought that is a cool part of it. I do seti because I believe in the project's goals and want to help reach those goals and anyone who cheats aren't helping the cause, they are hurting the project and must not believe in it and should be kicked out of seti forever!

 

[Eponymous]

Seti@Home should be able to identify these people without voiding all the statistics.

The real question is how much CPU time are they willing to throw at it.

Fact: Each WU submitted has the name of the person who supposedly crunched it.

Process One: Full audit. (lots of time)

Sort the WU's by the most number of times submitted.
A WU submitted More than 3 or 4 times and this is a "duplicated" WU.

Cut off all the ones submitted less then this. Those are legit WUs.

Then sort this list by who is submitting them. And there is your list of losers/cheaters.

(However your two person "twin" ring might fall through) Worst offenders will however be caught.

Process Two: Double check history

A quick double of the amount of history they go back in time to check for duplicates
will pop up these "twin"ers really quick. However this is no long term fix.

Possible fix: *sigh* new rev...

Each wu would be assigned a unique id number (in addition to the one they already have)
so each handed out WU is uniquely identified.

When asking SETI@HOME for the WU you would provide the ID number of the person who
is going to crunch it.

Therefore SETI@HOME has a list of who was given what WU and can compare against this
list when it is turned back in.

Doing this will break lots of peoples stuff.

EDIT: Ooops we don't really care who check it out. Just that only one person checked it in.
Doesn't break as much stuff that way.

And for a matter of record. I have the WU id numbers of the last 46,000 WU's I have
submitted if you wish to audit them for duplicates. (That is when I started using SetiQueue).

 

[Crazee]

I think at this point in time we need to all calm a little bit. We are all accepting this as if it is fact. Do the methods of cheating in that thread sound plausible? YES Has anyone from here verified that they work? NO

I don't know Fragile from ARS (and I am not disparing the guys integrity), but unless someone from here can vouch for either the guy or the methods then I will wait for Dr. Anderson to tell me if it is happening or not. I am not opposed to sending them an email requesting that they look into it and let us know if it is going on, but I am not going to take the word of someone I don't know, who doesn't work at Berkeley until I have it confirmed either by a TeAm member who knows him or someone who has verified that what he is talking about works.

As for Tim, he is my TeAmmate and I will stand by him and believe him until someone shows me solid proof otherwise. As for him having to prove anything he doesn't have to do it for me. They have to prove he is guilty. That is the way the system of justice I believe in works.

 

[Robor]

Originally posted by: networkman
If a ton of results are coming from a single IP it would have to be either a badarse mainframe or a cheater.

Not so! I have cable internet access thru ATTBI and a single IP at the router/firewall. At one point, I was cranking out 70+ WUs/day just from home. No cheating there! Just a guy who's admittedly a little nuts and has the space/power/time to feed his stats addiction.

When I say from a single IP I mean the *machine* that analyzed the WU, not the machine that sent it. Thinking about it more that wouldn't matter though 'cause many of use have private IP's behind our firewalls. Maybe we could get Microsoft to use their snazzy XP license encryption to uniquely identify each PC and...

 

[Robor]

Originally posted by: Crazee
I think at this point in time we need to all calm a little bit. We are all accepting this as if it is fact. Do the methods of cheating in that thread sound plausible? YES Has anyone from here verified that they work? NO

I skimmed the thread over at Ars but I think the fact that some of the Ars members were talking about setting up a "Cheater Protest Team" or something is a pretty good indication they've proven the alleged cheats work.

Originally posted by: Crazee
I don't know Fragile from ARS (and I am not disparing the guys integrity), but unless someone from here can vouch for either the guy or the methods then I will wait for Dr. Anderson to tell me if it is happening or not. I am not opposed to sending them an email requesting that they look into it and let us know if it is going on, but I am not going to take the word of someone I don't know, who doesn't work at Berkeley until I have it confirmed either by a TeAm member who knows him or someone who has verified that what he is talking about works.

I have a feeling Berkeley is aware. After all, Ars is the #1 SETI team. We'll just have to wait for a response.

Originally posted by: Crazee
As for Tim, he is my TeAmmate and I will stand by him and believe him until someone shows me solid proof otherwise. As for him having to prove anything he doesn't have to do it for me. They have to prove he is guilty. That is the way the system of justice I believe in works.

Couldn't agree more. Tim could've crawled away and hidden after he pulled his account and was banned from this forum. Well, he didn't and he's even rejoined the TeAm. I think it's time some of us go over to the offsite forum and give Tim a little support. If a thread hasn't been started I will start one.

 

[Postman]

I've been reading Ars for quite some time and Frag seems like a stand up guy to me. Check his posting history to see all the stuff he's done for TLC.

 

[Freewolf]

I don't believe we have anybody on our team cheating. But it is clear reading the Ars thread they believe there is some truth to it.

 

[lane42]

If someone has 1 PC at say IP address 195.195.195.65 or computer name home-bob and we can see that they are pumping out more than 15 wu's a day then we know we have an issue. Really a high powered computer is only going to pump out 7 a day if your lucky.

What happens if you sneakernet. I have 8 machines at work and upload all 8 on my laptop at home from my zip drive.

 

[Assimilator1]

This a sad day ,I hope the cheating doesn't go too far ,I am not aware of anyone on this team doing it.

I also believe that Cobra is innocent ,in fact I thought that Berkley had already checked his account out throughly & found it to be legite!
I'm gonna tell the Ars guys that too!

Robor
Don't worry ,there are still some oldies left who remember you starting a longtime ago with far fewer WU's

 

[MoFunk]

Originally posted by: lane42

If someone has 1 PC at say IP address 195.195.195.65 or computer name home-bob and we can see that they are pumping out more than 15 wu's a day then we know we have an issue. Really a high powered computer is only going to pump out 7 a day if your lucky.

What happens if you sneakernet. I have 8 machines at work and upload all 8 on my laptop at home from my zip drive.

I do this as well but it is from home to work. I was just curious if the actual WU kept an IP log. That way if you have 8 computers with different IP addresses or names and upload and download from one computer it would show up that way.

 

[Smoke]

4) You look for big hitters that never post

You guys don't need to worry one bit! LOL

 

[gogeeta13]

Man, that is crappy, good luck TA SETI, i hope nobody in TA has taken part in this.

 

[Rattledagger]

Hmm, from same thread:

by fragile:
Assimilator - sorry d00d

His (COBRA-ICE) account was deleted by S@H for "an amazing amount of duplicates" following a lot of checking.

This is not common knowledge, but was confirmed to me by two people that should know the truth.

 

[Poof]

As some of you all know, I post at Ars quite a bit. I have known and interacted with fragile (his old forum name was "notfragile") for some time, especially since he is a *nix guy.

Assim1 - he just responded to your post regarding Cobra-Ice...

I don't know what more we can do except to have teams self-monitor. Some of the suggestions on the other thread will be difficult. Data from Arecibo is scarce compared to the number of active users/machines crunching, so current users will inevitably have to crunch the same WU again at some point. Perhaps the interval for submission of such could be moved out by several months, but even then, the overhead of managing this server-side, seems enormous.

Cheaters will always be around but if discussions like this appear on a regular basis, we can at least discourage the casual cheater from even thinking about doing it under the banner of TA.

 

[Rattledagger]

Skimming the threads...
Ah, so cheat 2 is the reason to have multiple accounts. Well, I've managed to download exactly the same wu for more than 1 user, so I've got some multi-account-doublets, but stupid me crunched all from start to finnish.
Hmm, I flushed the duplicates-folder for some months ago also.

3) You look for twins or worse, in sync A and B producing very similar numbers and shaddowing each other.
Well, I've got zig-zag-production, but the total production is the same regardless of how many accounts I'm using... If I'm not re-routing to other projects.

 

[Smoke]

Been doing more reading on the Ars DC Forum. Found this comment by:

Hat Monster
Ars Centurion
et Subscriptor

The Knowledge
No forum or resource anywhere has the sort of knowledge that we do in the field of DC. Nobody comes close. Free-DC are still oxymorons, Anandtech don't really give a damn either way and DPC speak foreignese. The rest are small-timers who don't accumulate the know-how like we do. Have the flashest stats and fastest CPUs in the world, but that doesn't tell you how the client works.

Nice to know we are held in such low regards. :disgust:
________________

And concerning "fragile's" reply to Assimmilator1

This is not common knowledge, but was confirmed to me by two people that should know the truth.

OH, now that sounds really convincing ... doesn't it? IMO I think fragile needs to put up or shut up. :frown:

 

[ys]

What's fragile's background? Is he some kind of senior SETI authority?

 

[JWMiddleton]

Originally posted by: Baldy18
An interesting point has been brought to my attention. SETIQUEUE OWNERS BEWARE! If you operate a team SetiQueue and resubmit dupes you may be under suspicion. The best thing for SetiQueue owners to do would be only resubmit dupes once and if they don't go through then delete them or not try to send them at all.

I was just thinking the same thing. I had planned to check my dups on Sunday before heading out of town, but forgot to do so. Guess I'll delete them.

This is not the first cheating scandal. The first one involved ARS. I never saw any resolution to it, nor did I see their stats drawn down. I sure hope they (S@H) address this problem publicly so that the project does not lose any more credibility! I agree with Robor that it is a real bummer to bust your a$$ and have other flying by you in the standings then find out that they might have been cheating.

 

[CADsortaGUY]

Originally posted by: Smokeball
And concerning "fragile's" reply to Assimmilator1

This is not common knowledge, but was confirmed to me by two people that should know the truth.

OH, now that sounds really convincing ... doesn't it? IMO I think fragile needs to put up or shut up. :frown:

Yeah - that's what I was thinking - If he has "proof" then so be it - but anyone posting that Cobra cheated without CONCRETE PROOF should be put in their place. I don't mean this to flame fragile but it is put up or shut-up time. I wasn't around for the "cobraICE" issue but why would berkely knowingly let him create new accts if he was guilty of cheating? Hmmm.

As far as that attempt of hat monster to pat himself on the back...

Who do these people think they are? Jeez

CADkindaGUY

PS - I know my post adds nothing to the current cheating subject but I also don't think we(or Tim) need to be blasted by Ars or anyone else. PUT UP OR SHUT UP! <-- preferably SHUT UP!

 

[Def]

*sigh*

Dishonest people are everywhere. Even in a VOLUNTEER project to help further science and mankind as a whole.




People never cease to surprise me with how utterly sleazy they can be...




Defster

 

[OhioDude]

There are a couple of new cheats in what I understand to be common usage. I have spoken to Berkeley about them, proven that they work on dummy accounts, and have received information from a couple of sources that lead me to believe that these cheats are in common use not only here, but also amongst TLC's nearest competitors.

This is not common knowledge, but was confirmed to me by two people that should know the truth.

Well, I know that at least one member of <insert name of team here> beats his wife and has forced his children into child prostitution. This is not common knowledge but was confirmed to me by two people who should know the truth. I can't tell you how I know nor can I divulge the name of the person in question but I know that it's true and I know that it happens on other teams as well. Besides, hundreds of thousands of clinical psychologists, social workers, and law enforcement personel can provide documented proof that this type of behaviour is possible.

This is how witch hunts are started. :disgust:

I agree with CAD and Smoke. Put up or shut up. :|

 

[MoFunk]

Originally posted by: Smokeball
Been doing more reading on the Ars DC Forum. Found this comment by:

No forum or resource anywhere has the sort of knowledge that we do in the field of DC. Nobody comes close. Free-DC are still oxymorons, Anandtech don't really give a damn either way and DPC speak foreignese. The rest are small-timers who don't accumulate the know-how like we do. Have the flashest stats and fastest CPUs in the world, but that doesn't tell you how the client works.

Wow. Man that is kind of a low blow. What the hell do we have to do with this? Sounds a little cocky to me. Bet this guy hasn't had a date ever! Unlike us super sexy AT guys and gals!