• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 35,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

Changes to the Weekly Competition website

Mike Hughey

Super Moderator
Staff member
Joined
Jun 7, 2007
Messages
9,694
Likes
1,659
Location
Indianapolis
WCA
2007HUGH01
YouTube
MikeHughey1
Thread starter #282
It would be cool if events that are not part of the weekly comp anymore are displayed in a seperate field on the PB-sheet^^
Yes, I saw this suggestion the first time. It's a good one; I'll try to get to it someday. I apologize; I haven't had as much time lately for further improvements, and most of my recent improvements have been behind the scenes to help with administration of the competition, but I do hope to get back to them again in a month or so.
 
Joined
Dec 24, 2015
Messages
1,459
Likes
912
Feature request #1: make the session timeout (much) longer. Maybe two hours or more?

I think it's currently half an hour or so, which means that I have to log in again after each FMC attempt. It's not a big deal, but it is a little bit annoying to have to do this all the time.

Feature request #2: after submitting a solution on the manual entry page, it shouldn't redirect to the "main" page, but stay on the manual entry page. (Subjective; I don't know would prefer this over the current behaviour.)

Same thing; I use manual entry for FMC and I have to go to the manual entry page for each attempt, unless I enter all three attempts at once.
 
Joined
Jan 2, 2018
Messages
5
Likes
2
WCA
2016IHLE01
Feature request #1: make the session timeout (much) longer. Maybe two hours or more?

I think it's currently half an hour or so, which means that I have to log in again after each FMC attempt. It's not a big deal, but it is a little bit annoying to have to do this all the time.

Feature request #2: after submitting a solution on the manual entry page, it shouldn't redirect to the "main" page, but stay on the manual entry page. (Subjective; I don't know would prefer this over the current behaviour.)

Same thing; I use manual entry for FMC and I have to go to the manual entry page for each attempt, unless I enter all three attempts at once.
Never really thought about it but, yes please I need those changes!
 
Joined
Aug 28, 2016
Messages
58
Likes
34
Location
Durban, South Africa
WCA
2017LAWR04
Looks like a bug in MBLD ordering - quicker 1/2 DNFs are being ranked higher than a slower 0 points success, e.g. from this week:
Week 2019-17 3x3x3 multiple blindfolded said:
13 Caleb Arnette (Caleb Arnette) | 1 | 2/3 |14:38
14 Abhijeet Ghodgaonkar (abunickabhi) | DNF | 1/2 | 2:06
15 Harry Owen (Alpha cuber) | DNF | 1/2 | 13:16
16 Mitchell Hum (MCuber) | DNF | 1/2 | 13:44
17 Andreas Lambropoulos (2017LAMB06) | 0 | 3/6 | 43:26
18 Balder Henke (fun at the joy) | DNF | 0/2
 

Mike Hughey

Super Moderator
Staff member
Joined
Jun 7, 2007
Messages
9,694
Likes
1,659
Location
Indianapolis
WCA
2007HUGH01
YouTube
MikeHughey1
Thread starter #287
Looks like a bug in MBLD ordering - quicker 1/2 DNFs are being ranked higher than a slower 0 points success, e.g. from this week:
Sorry, this is a known bug; I'm short on extra time right now to work on fixes like this. I haven't worried too much about it because I consider it a somewhat low-priority bug - the ordering corrects itself after the week is over - the ordering bug only appears when calculating the current week.
 

Mike Hughey

Super Moderator
Staff member
Joined
Jun 7, 2007
Messages
9,694
Likes
1,659
Location
Indianapolis
WCA
2007HUGH01
YouTube
MikeHughey1
Thread starter #290
I just noticed that even tho me and Malte have the same AVG in skewb, we seem to have a different kinch value for that event, which seems a bit odd
The difference is due to the fact that I currently use the unrounded averages for this calculation. Your average is actually 4.6533..., while Malte's is actually 4.6466... That makes the difference.

I admit that I'm not sure if that is the same way the Kinch is calculated "officially" (admittedly, there's no such thing as an "official" Kinch score, but I don't remember what Daniel set up as his intended way of calculating this). If there is a general consensus that the rounded average is what should be used for calculating Kinch, I can try to change it. But my gut reaction is to think that using the unrounded average is somewhat more "fair", for what that's worth.
 
Joined
Oct 19, 2017
Messages
475
Likes
256
WCA
2015KUCA01
YouTube
channel/UCKY9uCII2wIfPmLidA_KXMA
Here's why unrounded averages are bad (and why the WCA doesn't use them):

Taking the average of numbers that only go to the hundreths shouldn't give you thousanths. If you had the thousanths in there originally, and averaged them out, then that might give a different result.

Example:
.96 .96 .97 = .963
.96 .96 .96 = .960

.961 .961 .971 = .964
.967 .967 .967 = .967

So the one with the actually faster average would get a worse result
 

Mike Hughey

Super Moderator
Staff member
Joined
Jun 7, 2007
Messages
9,694
Likes
1,659
Location
Indianapolis
WCA
2007HUGH01
YouTube
MikeHughey1
Thread starter #293
It doesn't always help. Since we truncate the thousandth digit, your problem can happen even if we don't keep the thousandths digit on an average.

Example:
.96 .96 .96 = .960
.96 .95 .95 = .953

.961 .961 .961 = .961
.969 .959 .959 = .962

The one with the faster average (.961 vs. .962) still gets a worse result (.96 vs. .95).

To me it seems that choosing where we do and don't round is somewhat arbitrary. If there's consensus on the preferred way, I'll happily change it, but I want to be sure there's actually consensus. I guess I'm curious what the pseudo-official Kinch ranks do (wcadb, I guess?).
 
Joined
Oct 19, 2017
Messages
475
Likes
256
WCA
2015KUCA01
YouTube
channel/UCKY9uCII2wIfPmLidA_KXMA
I guess I'm curious what the pseudo-official Kinch ranks do (wcadb, I guess?).
I just went to wca.cuber.pro and looked at Blake's vs Danny's 3x3 averages (6.753 and 6.747 respectively) and both have an 84.30 kinch for 3x3. I don't think it really matters too much though, as the difference is very marginal, so I don't think people will care too much either way.
 
Joined
Dec 24, 2015
Messages
1,459
Likes
912
Here's why unrounded averages are bad (and why the WCA doesn't use them):

Taking the average of numbers that only go to the hundreths shouldn't give you thousanths. If you had the thousanths in there originally, and averaged them out, then that might give a different result.
This seems like a better argument for including at least one more decimal place in the averages than the singles, actually. (Why should singles and averages be written to the same precision anyway?)

If I could have my way with things, I'd change the WCA regulations to start recording milliseconds and use four decimal places for averages, I guess. And Kinch ranks aren't ever calculated by hand in the first place; there's no need to "simplify" the calculation process by rounding some of the intermediate values to three or however many decimal places.
 
Joined
Oct 19, 2017
Messages
475
Likes
256
WCA
2015KUCA01
YouTube
channel/UCKY9uCII2wIfPmLidA_KXMA
This seems like a better argument for including at least one more decimal place in the averages than the singles, actually. (Why should singles and averages be written to the same precision anyway?)

If I could have my way with things, I'd change the WCA regulations to start recording milliseconds and use four decimal places for averages, I guess. And Kinch ranks aren't ever calculated by hand in the first place; there's no need to "simplify" the calculation process by rounding some of the intermediate values to three or however many decimal places.
Singles and averages should be written to the same precision because of sig figs, which I thought I explained why but clearly I was wrong.

As for milliseconds, the timers aren't completely accurate so having more precision isn't going to help, and plus it would look so much messier. We don't have an event at this point in time where milliseconds are going to matter.
 
Top