Why is the 'new' credit system still erratic.

Message boards : Number crunching : Why is the 'new' credit system still erratic.
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · Next

AuthorMessage
W-K 666 Project Donor
Volunteer tester

Send message
Joined: 18 May 99
Posts: 19012
Credit: 40,757,560
RAC: 67
United Kingdom
Message 1059579 - Posted: 25 Dec 2010, 10:46:08 UTC

Consistency is what I am looking for. Just reported three AP tasks. Two with 0% blanked and the other with 2.5% it took about 10 mins extra to complete. The credits granted are 1,029.45, 769.20 and 727.54.
Now without peeking say which task had the blanking.
ID: 1059579 · Report as offensive
Profile Frizz
Volunteer tester
Avatar

Send message
Joined: 17 May 99
Posts: 271
Credit: 5,852,934
RAC: 0
New Zealand
Message 1059596 - Posted: 25 Dec 2010, 13:42:11 UTC - in response to Message 1059579.  
Last modified: 25 Dec 2010, 13:49:42 UTC

Consistency is what I am looking for.


Yes me too!

Just look at my Astropulse host: Aldebaran

E.g. one task with 78% blanking got 713 points, while one with 2% blanking got 1.801 points.

For my most recent work units, I got between 149 and 1.801 points per unit.

Most of my wingman run stock application with more than 10 validated results.

Still the points given are somehow unpredictable.

I would go so far to say: The current credit system is a FAIL.
ID: 1059596 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1059598 - Posted: 25 Dec 2010, 14:05:14 UTC - in response to Message 1059596.  
Last modified: 25 Dec 2010, 14:07:10 UTC

Consistency is what I am looking for.


Yes me too!

Just look at my Astropulse host: Aldebaran

E.g. one task with 78% blanking got 713 points, while one with 2% blanking got 1.801 points.

For my most recent work units, I got between 149 and 1.801 points per unit.

Most of my wingman run stock application with more than 10 validated results.

Still the points given are somehow unpredictable.

I would go so far to say: The current credit system is a FAIL.


You can't expect to a full load of Credits if the Wu finishes early with 'Found 30 single pulses and 30 repeating pulses, exiting.'
Which some/most of the Sub 600 Credits are,

Claggy
ID: 1059598 · Report as offensive
Profile Frizz
Volunteer tester
Avatar

Send message
Joined: 17 May 99
Posts: 271
Credit: 5,852,934
RAC: 0
New Zealand
Message 1059599 - Posted: 25 Dec 2010, 14:17:45 UTC - in response to Message 1059598.  
Last modified: 25 Dec 2010, 14:19:34 UTC

You can't expect to a full load of Credits if the Wu finishes early with 'Found 30 single pulses and 30 repeating pulses, exiting.'


Even when excluding those tasks, the range for "normal" tasks is from 648 to 1801 points per WU. Independent of blanking and other factors I can think of.

As in the example I mentioned: One task with 78% blanking got 713 points, while one with 2% blanking got 1.801 points.

300%. That's just a too big variance for my taste.
ID: 1059599 · Report as offensive
Cosmic_Ocean
Avatar

Send message
Joined: 23 Dec 00
Posts: 3027
Credit: 13,516,867
RAC: 13
United States
Message 1059606 - Posted: 25 Dec 2010, 15:07:01 UTC

Both of my hosts are AP-only hosts. And I've been keeping a spreadsheet of every WU I've done and I try really hard to catch the granted credit for each one (less than six hours before a validated WU disappears is really annoying). There is a huge variance, and from what I can see, has absolutely nothing to do with run time OR blanking. Seems like a random number generator set for the ranges of 200-5000, it picks a number 5 times and then averages the 5 picks together and that's what you get. I know that sounds drastic, but I have yet to see any correlation between granted credit and runtime, blanking, or the number of pulses found.
Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)
ID: 1059606 · Report as offensive
Profile perryjay
Volunteer tester
Avatar

Send message
Joined: 20 Aug 02
Posts: 3377
Credit: 20,676,751
RAC: 0
United States
Message 1059611 - Posted: 25 Dec 2010, 15:29:05 UTC - in response to Message 1059568.  

lunatics op apps is pretty well recognized by the SETI staff, they seem comfortable with it, and my understanding the source is readily available to them. If the lab was uncomfortable with it, I would uninstall it in a heartbeat.


Morning S^S, Merry Christmas. Just as a side note, a large portion of what is now the stock apps started life as Lunatic's work.


PROUD MEMBER OF Team Starfire World BOINC
ID: 1059611 · Report as offensive
W-K 666 Project Donor
Volunteer tester

Send message
Joined: 18 May 99
Posts: 19012
Credit: 40,757,560
RAC: 67
United Kingdom
Message 1059614 - Posted: 25 Dec 2010, 15:47:12 UTC - in response to Message 1059611.  

lunatics op apps is pretty well recognized by the SETI staff, they seem comfortable with it, and my understanding the source is readily available to them. If the lab was uncomfortable with it, I would uninstall it in a heartbeat.


Morning S^S, Merry Christmas. Just as a side note, a large portion of what is now the stock apps started life as Lunatic's work.


The stock AP app would probably be 3* slower if it wasn't for the Lunatics crew. Think they also got rid of a "few" bugs as well.

To add to my last now reported AP with 22.33% blanking and the granted has gone down again to 611.32cr.

As Cosmic_Ocean suggested they would be better off with a random number generator.
ID: 1059614 · Report as offensive
1mp0£173
Volunteer tester

Send message
Joined: 3 Apr 99
Posts: 8423
Credit: 356,897
RAC: 0
United States
Message 1059687 - Posted: 25 Dec 2010, 21:19:55 UTC

Credit is the ultimate Tar Baby.

If they don't change it, somebody's going to be mad.

If they try to fix it, more people are going to be mad.

Seems like there's a choice: you can touch the Tar-Baby, like Br'er Rabbit, or you can leave him be.

That said, everyone here is pretty much following Br'er Rabbit.
ID: 1059687 · Report as offensive
Josef W. Segur
Volunteer developer
Volunteer tester

Send message
Joined: 30 Oct 99
Posts: 4504
Credit: 1,414,761
RAC: 0
United States
Message 1059745 - Posted: 26 Dec 2010, 5:41:46 UTC - in response to Message 1059564.  

-BeNt- wrote:
...
The way I read his post he is saying that using an optimized app and running 2 WU's at a time is somehow cheating the credit system by making it take longer for your WU to complete hence causing more credit to be granted.
...

Sometimes I'm too wordy, this time obviously I erred on the other side. The example was merely to illustrate that changing elapsed time directly affects the effective claimed credit in the short term. If there were a significant granted credit advantage to that kind of change, a cheater would simply arrange to falsify the reported elapsed time. That's why high claims are given less weight (or ignored when all results are from anonymous platform hosts).
                                                                 Joe
ID: 1059745 · Report as offensive
Profile BilBg
Volunteer tester
Avatar

Send message
Joined: 27 May 07
Posts: 3720
Credit: 9,385,827
RAC: 0
Bulgaria
Message 1059751 - Posted: 26 Dec 2010, 6:47:26 UTC - in response to Message 1059535.  
Last modified: 26 Dec 2010, 6:53:58 UTC


Yes, it's slightly unfair to those who use anonymous platform ethically. But as was thoroughly proven by S@H Classic there's a minority who get too competitive for credits and will cheat if possible. So BOINC always mistrusts anything a user can change.

Even with the best of intentions, there are things users may do which make the credit calculations deviate quite a bit. One example is shifting from doing one WU at a time on a Fermi GPU to doing more. Shifting from 1 to 3, for instance, might extend the elapsed times by about 2.5 times which would increase the claimed credit by the same factor until the server-side averages adapted.
                                                                Joe


Wait so doing more work is 'cheating'? ...
(... and more misunderstanding of Josef W. Segur's words)


Josef W. Segur only said deviate (NO word "cheating" in last paragraph)

"So BOINC always mistrusts anything a user can change" means:
using Anonymous platform every programmer can get the open source of the app,
change it so it claims (very) big credits, compile it, create app_info.xml
and use the app for cheating the credit system
if the credit is calculated by just average of claimed credits.

E.g. "normal" (fair) user of Anonymous platform (optimized app) claims 100
and the cheating user of Anonymous platform claims 100 000 000 credits
the average will be 50 000 050

So the decision was to use just the minimum claim (100 in this case)

The stock app as delivered to BOINC in binary (executable) form can't be changed
(patched, cracked) by the user (or programmer) as it have signature (checksum) protection
so it is more trusted.


Edit:
Obviously I have to write faster and not deviate ;)
- Josef W. Segur made his position clear meanwhile


 


- ALF - "Find out what you don't do well ..... then don't do it!" :)
 
ID: 1059751 · Report as offensive
-BeNt-
Avatar

Send message
Joined: 17 Oct 99
Posts: 1234
Credit: 10,116,112
RAC: 0
United States
Message 1059756 - Posted: 26 Dec 2010, 7:52:17 UTC - in response to Message 1059751.  
Last modified: 26 Dec 2010, 8:03:20 UTC

Thanks for clearing it up Josef. In my haste I may have taken it wrongly which was implied before hand, guess some people understand others don't.
Traveling through space at ~67,000mph!
ID: 1059756 · Report as offensive
Profile WHOSIT
Avatar

Send message
Joined: 12 Nov 10
Posts: 29
Credit: 162,958
RAC: 0
United States
Message 1060061 - Posted: 27 Dec 2010, 1:02:53 UTC

Greetings all,

Richard Haselgrove,

Re: Stat counter malfunction

Sir,

As I have not received a reply from the alternative stat site I chose to use, not seeing any change to my SETI numbers in mine nor any other's for over twenty four hours, and in seeing here within the topic thread so many people seeming to be interested in the gaining of statistical numbers, rather than in the scientific work being performed itself, I have decided to eliminate any stat counter all together and replace it with a plain graphic.

Besides, IMHO, It is a needlessly redundant practice to include my oun stat numbers on my oun postings, where as I can just glance off to the left side and see perfectly matched numbers to those displayed on my programs' screen savers underneath my avatar.
For those who are of the numerical statistic interested inclined, I have listed the programs in my signature I partake of. Feel free to seek and look my ten to the infinite stat numbers up. ;-) :-)

To those who may be offended by this post, with my mention of "number collectors" included within, I wish to state, that there is nothing what so ever wrong with this practice! By all means, I wish each of you good luck, with more and more success in your numerical gain endeavors.
This was / is just not my personal point of interest in deciding to participate in the certain "@home" projects I have chosen.

I also found it repulsive to read and discover here, that some persons have chosen to cheat, attempted to, or still attempt to do so, in order to ONLY increase their stat(s)' numbers. WHY!? To what end!? Did I miss mention of some sort of grand, ultimate prize to be given to the holder of the highest stat number(s) at some unknown point in time? :-)
To take something good such as this, to that level and degree of anti-social action, is an indication (to myself anyway), of a person with some type of mental problem and / or disorder. Hopefully, these people seek out counseling and treatment, while also being banned from ANY "@home" program usage.

Wishing you continued science "fun" and at least, always valid results,
Laters,
Rick "WHOSIT" W.
Participating in: Einstein/MilkyWay/Rosetta/SETI
ID: 1060061 · Report as offensive
-BeNt-
Avatar

Send message
Joined: 17 Oct 99
Posts: 1234
Credit: 10,116,112
RAC: 0
United States
Message 1060160 - Posted: 27 Dec 2010, 9:51:09 UTC

Ric, the Seti stats are only generated for download for the other sites once a day. There have been issues going on in Berkley, so they may not have gotten an updated set. So I don't get it, is it really that big of a deal?
Traveling through space at ~67,000mph!
ID: 1060160 · Report as offensive
Profile Bill Walker
Avatar

Send message
Joined: 4 Sep 99
Posts: 3868
Credit: 2,697,267
RAC: 0
Canada
Message 1060209 - Posted: 27 Dec 2010, 14:40:15 UTC

I'm with Sten-Arne on this one. If they don't improve the obviously flawed and unfair credit system within 5 to 10 years, I will probably quit.

ID: 1060209 · Report as offensive
Profile WHOSIT
Avatar

Send message
Joined: 12 Nov 10
Posts: 29
Credit: 162,958
RAC: 0
United States
Message 1060280 - Posted: 27 Dec 2010, 19:06:12 UTC - in response to Message 1060160.  
Last modified: 27 Dec 2010, 19:09:26 UTC

Greetings all,

Ric, the Seti stats are only generated for download for the other sites once a day. There have been issues going on in Berkley, so they may not have gotten an updated set. So I don't get it, is it really that big of a deal?



BeNt,

Thanks for your response, but I have been aware of the 24 hour delay in reporting, which is not the cause.

It's obvious you didn't and don't "get it" at all.

It has always been of no importance to me to display my stats, be they up to date or not, in my signature. Rather, I was looking for something to "dress up" the plain printed signature area and chose a stat display I liked to do so.

A part of that signature graphic was not operating correctly (btw it still isn't on ANY site's stat graphic as of this current posting), therefore it was replaced with something else. If something I have isn't working right, can't, will not be, or isn't worth being fixed (this latter being the case), I dispose of it. I don't keep broken anythings hanging around.

". . ., is it really that big of a deal?"
In more than one sense, no it isn't.
However, It is the cause of the odd single program stat malfunction, which has peaked MY interest, so in turn, that WOULD make it a very minor "big deal" to me and perhaps no one else.

Also, I am very aware of the MANY issues affecting SETI and other "@home" project end users emanating from both Berkley and the BOINC system.
Wishing you continued science "fun" and at least, always valid results,
Laters,
Rick "WHOSIT" W.
Participating in: Einstein/MilkyWay/Rosetta/SETI
ID: 1060280 · Report as offensive
1mp0£173
Volunteer tester

Send message
Joined: 3 Apr 99
Posts: 8423
Credit: 356,897
RAC: 0
United States
Message 1060302 - Posted: 27 Dec 2010, 21:20:51 UTC - in response to Message 1060280.  

It's obvious you didn't and don't "get it" at all.

It has always been of no importance to me to display my stats, be they up to date or not, in my signature.

If it's not important, then why waste so much time explaining how it's not important -- or how beNt doesn't get it?

IMO, the biggest issue that SETI has is the fact that we as a society demand perfection: 100% availability and absolute agreement on all stats sites at all times.

BOINC doesn't work that way. BOINC isn't designed for 100% availability, it's designed so that parts can fail and the overall system still works.

... and that includes stats.

I think beNt does get it, in that he doesn't expect the project to do more than they've promised (a "best effort" to periodically deliver work -- done with a small staff and a tiny budget), and he doesn't expect the stats sites to always have the latest.

We live in an imperfect world, and simulating perfection in an imperfect world is expensive. You do it with redundant connections into diverse locations on different power grids, different infrastructures -- and even then you're hoping and praying that there isn't some common element that one idiot can break.

BOINC doesn't try to simulate perfection. It recognizes that things break, and does a good job of working anyway.
ID: 1060302 · Report as offensive
-BeNt-
Avatar

Send message
Joined: 17 Oct 99
Posts: 1234
Credit: 10,116,112
RAC: 0
United States
Message 1060314 - Posted: 27 Dec 2010, 22:16:51 UTC - in response to Message 1060302.  


If it's not important, then why waste so much time explaining how it's not important -- or how beNt doesn't get it?

IMO, the biggest issue that SETI has is the fact that we as a society demand perfection: 100% availability and absolute agreement on all stats sites at all times.

BOINC doesn't work that way. BOINC isn't designed for 100% availability, it's designed so that parts can fail and the overall system still works.

... and that includes stats.

I think beNt does get it, in that he doesn't expect the project to do more than they've promised (a "best effort" to periodically deliver work -- done with a small staff and a tiny budget), and he doesn't expect the stats sites to always have the latest.

We live in an imperfect world, and simulating perfection in an imperfect world is expensive. You do it with redundant connections into diverse locations on different power grids, different infrastructures -- and even then you're hoping and praying that there isn't some common element that one idiot can break.

BOINC doesn't try to simulate perfection. It recognizes that things break, and does a good job of working anyway.


Exactly, thank you.

Traveling through space at ~67,000mph!
ID: 1060314 · Report as offensive
Profile WHOSIT
Avatar

Send message
Joined: 12 Nov 10
Posts: 29
Credit: 162,958
RAC: 0
United States
Message 1060404 - Posted: 28 Dec 2010, 3:53:13 UTC - in response to Message 1060314.  


If it's not important, then why waste so much time explaining how it's not important -- or how beNt doesn't get it?

IMO, the biggest issue that SETI has is the fact that we as a society demand perfection: 100% availability and absolute agreement on all stats sites at all times.

BOINC doesn't work that way. BOINC isn't designed for 100% availability, it's designed so that parts can fail and the overall system still works.

... and that includes stats.

I think beNt does get it, in that he doesn't expect the project to do more than they've promised (a "best effort" to periodically deliver work -- done with a small staff and a tiny budget), and he doesn't expect the stats sites to always have the latest.

We live in an imperfect world, and simulating perfection in an imperfect world is expensive. You do it with redundant connections into diverse locations on different power grids, different infrastructures -- and even then you're hoping and praying that there isn't some common element that one idiot can break.

BOINC doesn't try to simulate perfection. It recognizes that things break, and does a good job of working anyway.


Exactly, thank you.


Greetings all,

1mpO*173 and BeNt,

I greatly dislike repeating myself and thought I had made my communications quite very clear to anyone, IF they had bothered to take the time to actually read and understand my postings on the matter, which is apparently something both of you are incapable performing, as you both have strayed far off the point and initial intent of the original posted question.

Thanks so much for both of your invaluable assistances rendered, but they are no longer desired, requested, nor required, as the answer I've sought has been finally received from the third party stat graphic provider earlier today, explaining the fault in their generation of only my SETI statistics.

Richard,

Your suggested course of action concerning the site I chose has worked. :-) However, I have chosen to no longer display a stat graphic in my signature in any case, so the whole matter is mote.

Thanks very much for your help.

Wishing you continued science "fun" and at least, always valid results,
Laters,
Rick "WHOSIT" W.
Participating in: Einstein/MilkyWay/Rosetta/SETI
ID: 1060404 · Report as offensive
-BeNt-
Avatar

Send message
Joined: 17 Oct 99
Posts: 1234
Credit: 10,116,112
RAC: 0
United States
Message 1060429 - Posted: 28 Dec 2010, 5:42:02 UTC - in response to Message 1060404.  
Last modified: 28 Dec 2010, 5:43:59 UTC


Greetings all,

1mpO*173 and BeNt,

I greatly dislike repeating myself and thought I had made my communications quite very clear to anyone, IF they had bothered to take the time to actually read and understand my postings on the matter, which is apparently something both of you are incapable performing, as you both have strayed far off the point and initial intent of the original posted question.

Thanks so much for both of your invaluable assistances rendered, but they are no longer desired, requested, nor required, as the answer I've sought has been finally received from the third party stat graphic provider earlier today, explaining the fault in their generation of only my SETI statistics.

Richard,

Your suggested course of action concerning the site I chose has worked. :-) However, I have chosen to no longer display a stat graphic in my signature in any case, so the whole matter is mote.

Thanks very much for your help.


It's moot not mote, mote is a small particle of dust. However you have contributed so much to these forums and project I beg you to stick around. You're so important keep up the good work! Oh and btw how exactly is talking about your personal decision not to display your stats due to the updates being pushed to the 3rd party sites or their failure to update their site have anything to do with the credit system here again? Oh yeah we are all off topic, please stay on it nobody cares about your personal choices in signature pictures.
Traveling through space at ~67,000mph!
ID: 1060429 · Report as offensive
1mp0£173
Volunteer tester

Send message
Joined: 3 Apr 99
Posts: 8423
Credit: 356,897
RAC: 0
United States
Message 1060739 - Posted: 29 Dec 2010, 7:03:56 UTC - in response to Message 1060404.  


1mpO*173 and BeNt,

I greatly dislike repeating myself and thought I had made my communications quite very clear to anyone, IF they had bothered to take the time to actually read and understand my postings on the matter, which is apparently something both of you are incapable performing, as you both have strayed far off the point and initial intent of the original posted question.

... and I greatly dislike people who take a technical question, and choose instead to attack the people behind the answers they don't like.

Personal attacks show that the attacker is more interested in winning than in seeking the truth.

Please give my warmest regards to everyone on your planet.
ID: 1060739 · Report as offensive
Previous · 1 · 2 · 3 · Next

Message boards : Number crunching : Why is the 'new' credit system still erratic.


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.