Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
World Community Grid Forums
Category: Support Forum: Suggestions / Feedback Thread: Why do we get badges based on run time? |
No member browsing this thread |
Thread Status: Active Total posts in this thread: 22
|
Author |
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
You asked about the base of the badges...stick to that Great the selfproclamed Internet police/forum moderator strikes. Ok I leave my own thread then, an alternative had been you stoped read it but that is not an option i guess? |
||
|
KLiK
Master Cruncher Croatia Joined: Nov 13, 2006 Post Count: 3108 Status: Offline Project Badges: |
It also eliminates the benefit of point cheating that has occurred in other Boinc projects. +1 |
||
|
SekeRob
Master Cruncher Joined: Jan 7, 2013 Post Count: 2741 Status: Offline |
You asked about the base of the badges...stick to that Great the selfproclamed [sic] Internet police/forum moderator strikes. Ok I leave my own thread then, an alternative had been you stoped [sic] read it but that is not an option i guess? If you want to change the topic, start a new thread. You see mistakes, were it not that long thought went into various elements of the crediting system and when to credit. I.e. when you are credited for a valid results [Quote]Then must they be credited on wrong day after they got valid.[/Quote], it's on the day it was validated, not the day you returned the result unless you want statistics to be retroactively revised? Just think of the consequences such an action would have. |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
for me it is unlogic, it is not runtime that change the world it is the number of results that returns... Why dont credit the thing what mather? I prefer it. If you were here at the beginning the computer you used will be much weaker than even a cheap one is now, but your contribution is still valuable. so they reward with run time. |
||
|
adriverhoef
Master Cruncher The Netherlands Joined: Apr 3, 2009 Post Count: 2070 Status: Offline Project Badges: |
There are already two 'systems' that get you rewards, one is based on runtime (with badges) and the other is based on (claimed/granted) 'credit' (that comes with 'points').
Rewarding your work (on your computerdevice) with badges is fair, the system doesn't judge how 'strong' or how fast your device is, it just measures the time spent doing calculations. Just that. Nothing fancy. It is what it is. The other system (with 'points' that you earn for your spent time, already awarded with badges) seems fair, too, because it makes decisions based on the speed of your device. If you spend 1 hour to earn 50 points for 1 result, you get 50 points, or if you spend 1 day to get 50 points for 1 result, you still get 50 points. If there were a system that was rewarding you with badges for the number of results that you were returning, and it was everybody's job to get as many badges as possible, everybody would turn to the science application that has the shortest WUs (to calculate as many results as possible in the shortest amount of time). How fair is that? |
||
|
Eric_Kaiser
Veteran Cruncher Germany (Hessen) Joined: May 7, 2013 Post Count: 1047 Status: Offline Project Badges: |
I give you an other example:
----------------------------------------My points per day dropped to less than a half on crunching HST1 only compared to ZIKA only. A HST1 wu (most of them) need more than 20 hrs to complete. A ZIKA wu (most of them) need 2 hrs to complete. Guess which project I would choose if badges are based on results or points. |
||
|
KLiK
Master Cruncher Croatia Joined: Nov 13, 2006 Post Count: 3108 Status: Offline Project Badges: |
I give you an other example: My points per day dropped to less than a half on crunching HST1 only compared to ZIKA only. A HST1 wu (most of them) need more than 20 hrs to complete. A ZIKA wu (most of them) need 2 hrs to complete. Guess which project I would choose if badges are based on results or points. the one with most points per h... as our projects do have different points per h...some might get neglected by that criteria! while all projects have same time to complete the badge...day in one projects = day in another! fair system... |
||
|
ThreadRipper
Veteran Cruncher Sweden Joined: Apr 26, 2007 Post Count: 1317 Status: Offline Project Badges: |
But when a task has been finished, isn't it logged in the result sent back as to what computational effort was needed for that task - in GFLOPs?
----------------------------------------Badges based on that would eliminiate reliance on points, computation time or results returned. Join The International Team: https://www.worldcommunitygrid.org/team/viewTeamInfo.do?teamId=CK9RP1BKX1 AMD TR2990WX @ PBO, 64GB Quad 3200MHz 14-17-17-17-1T, RX6900XT @ Stock AMD 3800X @ PBO AMD 2700X @ 4GHz [Edit 1 times, last edit by flodisar at Aug 10, 2016 10:41:04 PM] |
||
|
SekeRob
Master Cruncher Joined: Jan 7, 2013 Post Count: 2741 Status: Offline |
The Gflops is a function of benchmark and runtime. Everyone knows, except you maybe, it's extremely easy to manipulate the benchmark values. The fun is then, when you alter the benchmark, for instance double it, any work assigned will show in the queue as needing half the time, until it runs for real and it comes out the benchmark is a big lie. Short for, flop counting is of a long time past.
|
||
|
ThreadRipper
Veteran Cruncher Sweden Joined: Apr 26, 2007 Post Count: 1317 Status: Offline Project Badges: |
Ok I see...well, runtime it is then :)
----------------------------------------Join The International Team: https://www.worldcommunitygrid.org/team/viewTeamInfo.do?teamId=CK9RP1BKX1 AMD TR2990WX @ PBO, 64GB Quad 3200MHz 14-17-17-17-1T, RX6900XT @ Stock AMD 3800X @ PBO AMD 2700X @ 4GHz |
||
|
|