Simon Crooke has GONE w/23,178 Seti WUs

[Smoke]

Yes, he has chosen to leave just as he arrived with nary a word.

I asked him to share with me the source of his crunching power. I guess that was his response.

His production in the last 30 days was most suspicious. His lack of communication with the TeAm and the pattern of his output was a perfect match for someone doing the Duplicate WU scheme.

I will monitor his account to see if it is deleted by S@H and report back to you all if that happens. If his was true honest production I believe he would have responded to me as others have done and not just quit.

I refer you all to a chart of his production in the last 30 days.

Chart of Simon Crooke

This chart will disappear within the next 24 hours so I have saved a copy that I will put up when and if necessary.

Here is a history of his account with TeAm AnandTech. He joined in February, 2002.

History of Simon Crooke with TeAm AnandTech

At this time whether Simon Crooke was doing anything wrong is pure speculation. Only Seti@Home can really determine that fact.

I'd like everybody to look at this link to a special sort of CkG's Q:

WU History of CkG?s Q

Notice the number of duplicate WUs. Scroll from page to page. :Q

I wrote Simon Crooke on three separate occasions in the last month or two and he never responded to any of them. I must admit I was a little strong on my last message that I sent to him just this morning. I hope I was not too severe. You all be the judge:

Hello Simon,

I have written you previously but I have not gotten a response.

I am writing you once again about your fantastic output of SETI Work Units. This output has increased significantly in the last 30 days and has really boosted the total output for the TeAm as well. For this I am most appreciative. However, there is a matter that I need to bring to your attention.

There has recently been a lot of publicity about how some individuals have taken advantage of different loopholes in the SetiAtHome Statistical Accounting to enhance their Seti WU production. SetiAtHome and TeAm AnandTech have taken a firm stand that this type of "cheating" will not be allowed to continue.

The "cheating" has taken several different forms.

There is the "Duplicate WU Cheat". This is where previously crunched WUs are SAVED in another location prior to being submitted to SetiAtHome-Berkeley. At a later time these OLD COMPLETED WUs are resubmitted to Berkeley as Newly Crunched WUs. Any New WUs that are still being crunched honestly are added to the SAVED batch and the resubmitted batch of SAVED WUs continues to grow exponentially.

Then there is the "Fast WU Cheat". This is when an overclocked computer starts returning garbled information. There is a built-in component in the Seti Client that upon receiving this garbled data aborts any further analysis, submits the results to Berkeley where the user gets credit, and then downloads and starts analyzing another WU. This all can take only a few minutes or even a few seconds. Now when this happens the total WUs submitted by the USER will start growing at a larger rate than previously experienced. This can also happen unknowingly to an honest USER if he does not monitor his SETI production. It is the job of each Team Founder to monitor each member's production, and when a sudden increased output surfaces, to verify if that output is genuine or not. If it is discovered that the newly increased output is due to an errant computer, that computer needs to be taken out of service, as far as analyzing SETI is concerned, immediately. If the offending computer is allowed to continue, then not only will the USER be considered cheating but the TEAM FOUNDER and officials will be construed as harboring this dishonest production

Finally, there is the "Hacked Client Cheat". This will have all the signs of the Fast WU Cheat and all comments I made above will apply to it.

Now, I have no way of knowing if your recent production increase is due to a recent assimilation of computer power or is due to one of the above. However, that can be determined by SetiAtHome-Berkeley on a special case by case basis. The staff of the SetiAtHome project is small and their primary job is the analysis of the data we, the Seti At Home Users, provide them for signs of extraterrestrial intelligence. In the past they have not put forth much effort in monitoring for dishonest WU production. They just have discarded the garbled WUs and gone on with the SCIENCE. That situation has changed. In the last few weeks SetiAtHome officials have taken action against USERS who have used dishonest methods to enhance their "stats".

TeAm AnandTech SETI has taken the position that we will not allow cheating in any form.

So, I am asking you once again to please explain to me the source of computer power you have engaged to produce your WUs. Only 30 days ago you were rocking along around 40 WUs @ day and over the course of the last month your production has been steadily increasing to where you now are putting out hundreds and hundreds of WUs @ day. If this is all due to the assimilation of additional computer power then I will want to lead the TeAm in cheering you on.

In light of the recent cheating in the SetiAtHome project all efforts must be made to insure the honesty of all submitted WUs. If you are legally producing WUs, and I certainly hope that is the case, please provide me with some detail.

Thank you,

Greg Lawrence
Aka Smokeball
TeAm AnandTech SETI
Membership Administrator

 

[Rattledagger]

"Simon Crooke" is living up to his name.

Based on federation-queues logs he has passed 4010 results here, and a quick sum gives 3996 dublicates.

37 times of wu 20my99aa.24221.32561.573572.144
36 times of wu 12fe02aa.27961.3312.572140.50
35 times of wu 09se02ab.18048.28722.11064.102
35 times of 11au01aa.7088.27665.29822.226
... and so on and so on.

 

[deerslayer]

If he is a cheater, then good riddance, we don't need his kind around here :disgust::|

 

[SoulAssassin]

get the fvck out and don't come back. No love for cheaters. We're all driven by the stats but's it's not the reason. Is this the first semi-confirmed cheater on AT?

Anyone on the TeAm who is cheating, leave now. Do not pass Go. Do not collect $200. Do not say goodbye. Just leave. We're better without you and your company is not welcome.

 

[Tarca]

Good work Smoky!

We don't need any cheaters here! And those numbers are suspicious!

Major Tom Tarca

 

[zodder]

Nice try, loser.

Nice catch, Greg. The sheriff appoints you as an honorary deputy.

 

[Wiz]

Thanks Greg, I appreciate your work forn the TeAm!

There is no room for cheaters here, plain and clear.

 

[JWMiddleton]

Incredible!! Great catchGreg!!

oh yeah, don't let Kinguni see this thread or we won't live it down!

 

[Baldy18]

The integrity of this team is one of the things that makes it great. Good riddance! We will do this thing and conquer it fairly.

 

[Assimilator1]

Damn

Letter looks ok to me SB ,especially seeing as it was your 3rd.
Btw that link to a SETIQ you've posted ,it doesn't seem to do any sort of special sort for me

Wierd how he downloaded a load of WU's & then stopped..I guess that's further confirmation?

Rattledagger
How do you get that information?

 

[CADsortaGUY]

Here might be a better link to my Queue's stats - not that the other one didn't show it loud and clear enough
http://mysterymachine.no-ip.com:443/history.htm?user=821885&historysortby=1d&historyline=0

CADkindaGUY

 

[networkman]

Greg - YHPM.

Yeah, this sucks to have had a cheater on AT, but at least he's been found out and is gone.

 

[Rattledagger]

Assim, it's easy to get the logs...
From my link, click on Queue Log, and it's just to change the dates (increase size-limit), cut-n-save.

Afterwards, I sorted all, removed all lines not starting with
time: Result wu-name completed by ...Simon Crooke...
and sorted it.

From 28.10.2002 to 18.11.2002 at 4:43:34pm queue-time
there was 4010 Result..., but only 14 unique wu-name.

There's 533 doublet-messages, but I'm not sure if these are repeaded or not... Looks like around 3800 was passed successfully to Berkeley.

 

[wirelessenabled]

Good job Greg!

 

[Smoke]

OK, this is a step by step for Ass1.

Go to CkG's Q

Click on HISTORY
Then Click on Ra, Dec

Now just scroll down the page and go to the next page and the next and the next .... and look for DUPLICATE WUs!

 

[CADsortaGUY]

Ooops - forgot the link in my last post

http://mysterymachine.no-ip.com:443/history.htm?user=821885&historysortby=1d&historyline=0

CADkindaGUY

 

[Freewolf]

At least you took action Smokeball instead of looking the other way. We are better off without him.

 

[Rattledagger]

Yes, even with the limited history-data there's showing up probable triplets...
Now, going to the logs, there's no doubt of 37-tuplets... or thatever it's called...

BTW, I just got an idea...
For the queue-manager, it's just to look on Result.csv for a given user.### This isn't timed-out with the other logs.
The 1st part here is the wu-name.
Even users that's left a queue can be found under Client, but it's much more difficult to sort...

 

[Robor]

Sucks that we'll lose 23K+ in production but it's better than having a cheater on the team. I think you did the right thing Greg.

 

[Crazee]

If he was a cheater then we never lost the production because it wasn't real.

Good catch Greg and CKG

 

[micron]

 

[Wolfsraider]

i think that you were not hard on him smokeball but rather straight forward in giving him the benifit of the doubt.he chose not to respond three times and finally up and left,anybody following the cheating on seti would understand the implications and provide proof of their work history to try to work out the problem,unless there was nothing to work out


on a related note:

i have my main rig overclocked although not as major an overclock as most would like to see

how can i monitor my production to ensure that i am not sending bad wu's? i have dsl so upload times are minimal.

should i also check for identical wu's?

there is no sense in doing this unless its fun,hurting the team by cheating takes away the fun factor imho

great catch smokeball and its nice to see there is more than one way to check for cheats

mike

edit spelling

 

[Smoke]

Here one nice easy way

If you know your rank on the team, go to TeAm AnandTech Stats on the S@N Website ... and no you don't need to know Dutch

Click on the bracket that would include your rank, for instance today since you are ranked 1276 on the TeAm you would select the 1201 - 1300 bracket and then go down until you find your name. Place a check mark next to it and then up at the top click on GRAPH.

You would then see the graphical representation as I gave you in the first link.

If you saw your production increase dramatically and you didn't add any more computers to account for it, then you would know that one of your computers is putting out FAST WUs. This is just one of the ways to monitor your production but I think it is rather easy and gives you a 30 day snapshot.

The stats available on the S@N website are about the nicest you will find anywhere ... all brought to you by a fine young man by the name of Pascal.

 

[Kinguni]

Originally posted by: JWMiddleton
oh yeah, don't let Kinguni see this thread or we won't live it down!

Too late! Good catch TeAm.

 

[Wolfsraider]

thanks smokeball much appreciated boss

(man thats a great way to check production stats)