Jump to content

Featured Replies

The first result is a complete mystery for me ... there's no reason why the score should turn up there ... I'll keep looking :-/

 

The second is an issue we've noticed recently. The change in code had an effect on the way score are ranked, in particular how nda-hardware scores affect the boints. The issue is not that easy to solve, so we're looking for work-around alternatives at the moment.

  • Author

As for the 32M rankings, I'm loosing ~30 points for me (and my team) and Stummie looses another ~50 points or so.... :(

Yeah. There's a fix coming in REV3.5, together with the WR points, but it's maybe too long. I'll see if we can get a temporary fix.

and one more problem

 

if some result gives me 0.1 point i could just disable participation in ranking and top places are loosing points

i mean its not fair play but it would be legal as far as i know

0.1 point rule ist useless

sombody who thinks a little bit would just disable ranking

a dumb one thinks 0.1 points are still something but some other concurent is getting for this 0.1 maybe even 2-3 points extra

 

i think this system is wrong

my opinion just

and one more problem

 

if some result gives me 0.1 point i could just disable participation in ranking and top places are loosing points

i mean its not fair play but it would be legal as far as i know

0.1 point rule ist useless

sombody who thinks a little bit would just disable ranking

a dumb one thinks 0.1 points are still something but some other concurent is getting for this 0.1 maybe even 2-3 points extra

 

i think this system is wrong

my opinion just

 

A dumb one refuses to accept boints for a result. EOD.

The i5 750 is in fact a 4xCPU ... but the ranking isn't really right. I think I have fixed it, so should be correct in the next ranking update.

I bench for my team. Every point, even 0.1, is worth to bench for.

 

yes but if your 0.1 point gives the concurent side 2 points ?even then?

you dont win 0.1

in this case you lose 1.9points :)

Hm. The score wasn't detected as "best submission". I forced the engine to rescan and now it's receiving points again.

 

Ranking will update in a few minutes, I think.

 

Score is insane LOL

link to submission in question?

 

http://hwbot.org/community/submission/935049_kingpin

 

HEY!.....I "lost" my Super-Pi 1M score with the 8600!..... :o

What happent to it?..... :o

 

http://hwbot.org/community/submission/923068_hipro5_superpi_core_2_e8600_3.33ghz_6sec_734ms

 

http://hwbot.org/rankings/benchmark/superpi/rankings

 

 

.

 

Same issue again :-/

 

Fixed it for you, points should be back in no time.

  • Author

Can't these issues be fixed automaticaly?..... :D

Each one HAS to check his submissions once a time every day or so?..... :o

 

EDIT: when the Super-Pi 32M list will be fixed?.....C'mon - c'mon, we are paying you a lot of money to be ALWAYS here...... :P:D

Edited by hipro5

Well, normally this happens automatically. There's an issue that prevents the 'automatic' to go correctly :P

  • Author

:D

 

Question:

 

IF there are TWO IDENTICAL scores, HOW does the engine put them into rankings?

By the "submission date" FIRST (the OLDER FIRST - the NEWER FIRST)?

OR by the CPU FREQUENCY (the LOWER FIRST)?

 

:)

I don't know how it was before, but we hardcoded in rev3 that submissions are ranking by: score,date.

 

So, first the score counts, then whoever submits first.

 

We were thinking about ranking on frequency instead of date ... but it's way to easy to cheat with that.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...