PDA

View Full Version : Sieve Stats Questions



Stromkarl
02-18-2005, 05:56 PM
MikeH et al,

I have been submitting my Fact, Factexcl, and Factrange (less than 20M) files per the discussion in this thread: excluded/duplicate factors - what's the difference? (www.free-dc.org/forum/showthread.php?s=&threadid=8026)

The score page had counted all these factors as I submitted them.
I understand that factors found in the latter two files usually score 0 points.
Out of curiosity, I have also been checking out the gap report pages.

What would cause the FactD and FactE factors on the score page to be removed?

Why would the entry for my range 678000T to 678500T be removed from the gap pages? Is is because there are no gaps due to the way I have been submitting the Factexcl and Factrange files?

BTW, MikeH, thanks for the good work with the sieve stats pages, even though I don't completely understand them fully yet. :notworthy

Stromkarl

maddog1
02-18-2005, 06:25 PM
Ah, a fellow TeamRetro siever :)
Guess I should try to give some answers here! (of course, Mike or anybody else can correct me if I get something wrong)

Non-scoring factors -practically almost all D and E factors- are not displayed it the 2005 stats page, but they are displayed in the all time (http://www.aooq73.dsl.pipex.com/scores.htm) scores. According to this, you have submitted 2 uniques, 3 duplicate and 115 excluded factors as of 21:00 18 Feb. You should be able to check yourself if this is correct.
I assume it was Mike's choice to not include non-scoring factors in the 2005 scores to avoid unnecessary clutter. I can't give a 100% certain explanation as to why a few D actually score some low points, while others listed as D score zero. Mike?
As for the gaps analysis, it counts all factors submitted in. This means with 120 factors submitted in total for a 500G range, your average yield is one factor every 4,16 G (and the largest gap between two factors in a row can't be THAT much higher than this) which is far lower than the gaps shown in the relevant stats page, so it disappers completely from this page.

Hope this helps and I haven't made any critical errors here :blush:

MikeH
02-19-2005, 08:33 AM
Non-scoring factors -practically almost all D and E factors- are not displayed it the 2005 stats page, The few duplicates that we see are ones where both the unique and duplicate were found in 2005. It shouldn't be possible for 'excluded's to appear.


As for the gaps analysis, it counts all factors submitted in. This means with 120 factors submitted in total for a 500G range, your average yield is one factor every 4,16 G (and the largest gap between two factors in a row can't be THAT much higher than this) The gap analysis has knowledge of whether a factor is unique or not. The information that is displayed relating to largest or smallest accepted gaps is actually referring to the gaps between unique factors. The non-unique factors (duplicates and excluded) are used as an extra aid to determine where work has been done, and thus 'acceptance' of the gap.

Thanks for the interest. Good to see folks are still using this stuff. :cheers:

Joe O
02-19-2005, 10:00 AM
If you would like to see your factrange and factexcl files put to use, please send them to factrange@yahoo.com. VJS and I process these files looking for factors > 20M or < 1M.
Thanks in advance.

Stromkarl
02-23-2005, 08:20 PM
Originally posted by Joe O
If you would like to see your factrange and factexcl files put to use, please send them to factrange@yahoo.com. VJS and I process these files looking for factors > 20M or < 1M.
Thanks in advance.

I will send them in when I am done with this range. It looks like another 6 days or so.

Stromkarl