Moving old multi-BLD results to the new list?

qqwref (2009-02-24 00:07:12 +0000)
As far as I understand the only change in the multiBLD rules is that any attempts which take over an hour are DNF. Thus I suggest that ALL old multiBLD attempts which were under an hour should be moved to the new list, whether they were under the first-generation or second-generation rules. Basically the only differences in the multiBLD rules that have ever happened were changes in how attempts were 'scored' - it was not harder or easier to get a 4/6 in 50 minutes in 2005 than it is now, it was just given a different rank compared to other attempts. The main reason I suggest this is that I looked under the rankings and Bernett Orlando has the "world record" with a 4/6 in 60 minutes. This is very clearly not the best sub-hour multiBLD performance in competition that has ever been done. In fact Dennis Strehlau did a 10/10 in 44:09. I can see no reason why the old sub-hour performances should not be counted in the new list.
BryanLogan (2009-02-24 04:26:10 +0000)
Agreed. I would say that changing Square-1 to an average from a mean is much much more of a radical change than putting a time limit on multi. But all the Square-1 and Clock time weren't touched.
Clement Gallet (2009-02-25 22:41:02 +0000)
[quote="qqwref":1lwdhes2]As far as I understand the only change in the multiBLD rules is that any attempts which take over an hour are DNF[/quote:1lwdhes2] Wrong : [code:1lwdhes2]H1b1) When the total time is reached, the attempt is stopped and the number of solved and not solved puzzles is counted.[/code:1lwdhes2] Also, only keeping results that have a total time below one hour is unfair, because for exemple Istvan did a 7/7 in 1:06:24, so he probably solved some cubes before one hour, then regarding the new rules, he should have had a result, and not DNF. Why are you using quotation marks for "World Record" ? Bernett has the World Record, that's all.
StefanPochmann (2009-02-26 10:28:33 +0000)
The idea I believe is to not throw away achieved results [b:telnnlym]unnecessarily[/b:telnnlym]. If your old result was under one hour, it's also perfectly valid under the new rules, so throwing it away is completely unnecessary. If your old result was over one hour, we simply have no good way to tell what the result was at the one hour mark, so what are we supposed to do with it?
jbcm627 (2009-02-27 00:13:12 +0000)
What about normalizing results that were over 1 hour? Eg, Tim Habermaas 24/24 2:15:57 multiply the 24/24 by 60.00min/135.95min -> 10.59/10.59 cubes in 1:00:00
blade740 (2009-02-27 20:14:58 +0000)
That would be complicated and would leave a lot of non-integer results in the database that would be confusing to anyone in the future who doesn't know about this rule change.
jbcm627 (2009-02-28 01:09:47 +0000)
[quote="blade740":2fv2wgjf]That would be complicated and would leave a lot of non-integer results in the database that would be confusing to anyone in the future who doesn't know about this rule change.[/quote:2fv2wgjf] It could be confusing to people in the future, but still at least provides a way to rank those times. The result could also still be listed as 24/24 in 2:15:57, but ranked as 10.59/10.59 in 1:00:00.
timhabermaas (2009-03-29 19:55:57 +0000)
[quote="BryanLogan":2hyxdd3q]Agreed. I would say that changing Square-1 to an average from a mean is much much more of a radical change than putting a time limit on multi. But all the Square-1 and Clock time weren't touched.[/quote:2hyxdd3q] Why? Square-1 just became sligthly easier. But it's so much harder to break your own multi bld record with a 1h time limit. [quote="jbcm627":2hyxdd3q]What about normalizing results that were over 1 hour? Eg, Tim Habermaas 24/24 2:15:57 multiply the 24/24 by 60.00min/135.95min -> 10.59/10.59 cubes in 1:00:00[/quote:2hyxdd3q] Uhm, please not. Ok, here's my suggestion for the new records: Keep all old results which are under one hour and throw all other results away.
BryanLogan (2009-04-01 17:26:39 +0000)
[quote="timhabermaas":1w8jw2u2][quote="BryanLogan":1w8jw2u2]Agreed. I would say that changing Square-1 to an average from a mean is much much more of a radical change than putting a time limit on multi. But all the Square-1 and Clock time weren't touched.[/quote:1w8jw2u2] Why? Square-1 just became sligthly easier. [/quote:1w8jw2u2] Well, when moving to an average of 5 from a mean of 3, you end up eliminating "lucky" solves from the average now. So let's say someone can solve Square-1 in 30 always. However, if they get a very lucky solve, it could be 15. So if they get a 30,15,30, their average is 25. Now, moving to the average of 5, if they got a lucky solve and the rest were normal (30,15,30,30,30), their average is 30. This is just a simplified version. You can argue about the probability of a PLL skip and the fact that "unlucky" solves are now discarded, but we don't keep track of worst, we keep track of the best. But the fact is, we make the average more reflective of true skill (which I agree with doing), I'm just pointing out that records may be more out of reach if they were set under the old rules.
Ron (2010-01-03 12:01:51 +0000)
Dear members of our community, Today we promoted the old Multiblind results to new Multiblind results, if they are compliant with the new regulations. We used the following design principles: 1) we want all old results that are valid according to the new regulations to be listed in the new category 2) we do not want to change the results of past competitions 3) we do not want results to appear double in past competitions 4) we only want old results to appear on the Competition page, not on the other pages 5) we want to keep old and new regional records From a technical point of view: a) We copied all valid old results (eventId=333mbo) to new results (eventId=333mbf). So these valid results are now in the database twice. b) We changed the software accordingly to the design principles. Thanks to Clément Gallet for his help in creating this solution. If you find any problems, then please let me know. Have fun, Ron