• 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

Joined
Feb 14, 2014
Messages
1,703
Likes
1,262
Location
Over there
WCA
2013GOOD01
YouTube
DGCubes
#41
Hey, Mike. Just a question (and a suggestion):
I just finished doing this week's FMC attempt, and I started my solution by doing EO. Only 2 moves were required (F2 U'), however these were (accidentally) the inverse of the last two moves of the scramble. The U' ended up canceling out with insertions, so it wasn't in my final solution, but it was originally used nonetheless. Would this result be DNFed or is it okay? I know the WCA has been harsh in the past with things like this, but I'm not sure if 2 of the same moves would be enough to DNF.

Also, if it's not too hard, I think it would be a good idea to append R' U' F to the beginning/end of the scrambles like the WCA does.

Thanks for all you do for the weekly competitions! :)
 

Mike Hughey

Super Moderator
Staff member
Joined
Jun 7, 2007
Messages
9,879
Likes
1,902
Location
Indianapolis
WCA
2007HUGH01
YouTube
MikeHughey1
Thread starter #42
I think we should count your solve as a valid solve still; prior to the change in scrambles, the WCA regulations were really quite unclear on what invalidated a solve, and we like to be reasonably generous on gray areas here. But I agree that this is a real problem (I actually ran into something similar with using a premove that matched the beginning of the scramble this week, so I was already thinking about it); we will try to add R' U' F to the beginning/end of the scrambles as soon as possible - hopefully next week.

Thank you for bringing this up!
 

Mike Hughey

Super Moderator
Staff member
Joined
Jun 7, 2007
Messages
9,879
Likes
1,902
Location
Indianapolis
WCA
2007HUGH01
YouTube
MikeHughey1
Thread starter #43
So I tried making the change for fewest moves to add R' U' F to the front and back of the scramble, but didn't realize that I needed to look for canceling moves (TNoodle does this). So we have a canceling move in this week's fewest moves scramble. I'm inclined to just leave it that way for this week; it's better than not having those moves added. But it could be confusing - please watch out for that canceling set of moves in the scramble. I can try to come up with a better solution for next week.
 

Mike Hughey

Super Moderator
Staff member
Joined
Jun 7, 2007
Messages
9,879
Likes
1,902
Location
Indianapolis
WCA
2007HUGH01
YouTube
MikeHughey1
Thread starter #45
When viewing someone's personal records, all the ordinal numbers default to having a th, even those that should not, such as 22th and 23th.
Thanks. I would fix that, but I'm working on redoing that page anyway to make it more like the WCA page, so I probably won't.
 

Mike Hughey

Super Moderator
Staff member
Joined
Jun 7, 2007
Messages
9,879
Likes
1,902
Location
Indianapolis
WCA
2007HUGH01
YouTube
MikeHughey1
Thread starter #47
I'm just getting started with it, but if you go to the Statistics page now, you can see an overall Kinch ranking for the weekly competition, for any historical week.

The Kinch ranking here is calculated based on the all-time best results on the weekly competition website, not the current world record. (That was necessary because a lot of our events aren't official! And besides, this lets you know how you're doing relative to people who have competed here.) Keep in mind this will still only support those who enter their results on the website rather than in the forum.

There's much left to do with this. You can't really see yet how the scores are calculated (although they work like proper Kinch scoring, except using our records instead of WCA world records), but eventually you will be able to. The multiBLD Kinch scores aren't calculated correctly yet - they simply use the WCA representation to calculate percentages instead of doing it the "proper" way. I will fix that someday. And eventually we will have personal results pages for everyone. But at least now you have another way to compare your results with everyone else each week!

For those who don't know, "proper" Kinch rankings are done like this, so this is what my rankings are based on:
http://www.kinch2002.com/kinchranks/
 

Mike Hughey

Super Moderator
Staff member
Joined
Jun 7, 2007
Messages
9,879
Likes
1,902
Location
Indianapolis
WCA
2007HUGH01
YouTube
MikeHughey1
Thread starter #48
Okay, I've finally started releasing some of the new statistics pages. Please let me know of any bugs you find in them.

Note that our current database is a little messed up in that the 23456 Relay event is conflated with the old Magic event, and the 234567 Relay with the old Master Magic event. So you'll see that records are quite messed up there; the Magic / MM results completely blow away the relay results. I intend to separate these two events in the database sometime soon, so hopefully that will clear up the situation.

They are a work in progress, so sorry for the messy construction and the current lack of some functionality. The ultimate goal is to get our statistics pages to be a near mirror-image of the WCA statistics pages; we have a long way to go. It is still true that the view of a user's results for the current week is in the old format, so you'll notice that is different. Also, there is temporarily no way to see overall records; I will add those soon, hopefully this week.

Hopefully you will enjoy some of the new features; an underline appears under an average (or single if there is no average) to allow a tooltip showing any comment you might have had for that week. The Persons page shows everyone who has ever competed on the website, and is sortable by # of competitions participated in and by # of completed solves (non-DNF, non-DNS). And it's nice to have some of the extra statistics that the WCA pages had that we didn't. Still a bunch left to go, but it's a start.
 
Joined
Jan 26, 2016
Messages
239
Likes
258
Location
South-Germany
WCA
2013HUBH01
#50
Wow, just "fell" over the new statistics - really great!

Just three things:

in 2016-32 I had a single 8.00 in 7x7x7. Can't remember what went wrong there :D, but as I'm ranked in place 52 overall, there have to be even better results. Is this a general bug or have so many competitors entered wrong results? For me it would be ok if it is replaced by DNF.

and: means-of-3 in 2,3,4,5bld could be added (I just checked, I have them all :rolleyes:)

and: Thank you!!!
 

Mike Hughey

Super Moderator
Staff member
Joined
Jun 7, 2007
Messages
9,879
Likes
1,902
Location
Indianapolis
WCA
2007HUGH01
YouTube
MikeHughey1
Thread starter #51
Thanks for the input!

in 2016-32 I had a single 8.00 in 7x7x7. Can't remember what went wrong there :D, but as I'm ranked in place 52 overall, there have to be even better results. Is this a general bug or have so many competitors entered wrong results? For me it would be ok if it is replaced by DNF.
I suspect it is just wrong results. I found a few of my own when I first enabled this and was looking at my own page; I've corrected mine to DNFs when I found them. I can do the same for you if you like. As for the ranking, I suspect this means I'm not taking DNSs into account for the rankings; I'll try to fix that tomorrow. Sorry about that!

and: means-of-3 in 2,3,4,5bld could be added (I just checked, I have them all :rolleyes:)
That's disappointing - they were working yesterday! (I was rather enjoying looking at mine, so I know they worked yesterday.) I'll try to fix that tomorrow too.
 
Joined
Sep 22, 2013
Messages
406
Likes
380
Location
UK
WCA
2013COPP01
#52
Okay, I've finally started releasing some of the new statistics pages. Please let me know of any bugs you find in them.
Love it!

A low priority feature request: it'd be nice to see your PB single / average on the Compete page for an event.

Edit: Also, should we report data errors here? I just noticed that my one-handed results for OH for 2016-45 are clearly wrong and should be DNF'ed. I must have done 2-handed solves, inadvertently...
 

Mike Hughey

Super Moderator
Staff member
Joined
Jun 7, 2007
Messages
9,879
Likes
1,902
Location
Indianapolis
WCA
2007HUGH01
YouTube
MikeHughey1
Thread starter #53
A suggestion: just so this thread doesn't become too cluttered, please just PM me with data corrections that you'd like done. I'll try to keep up - hopefully they won't be too overwhelming. But give me a few days just in case I get backed up by the requests. It is fairly clear that the database has a lot of untidy data in it.

Oh, and @bubbagrub, I do like that suggestion!
 

Mike Hughey

Super Moderator
Staff member
Joined
Jun 7, 2007
Messages
9,879
Likes
1,902
Location
Indianapolis
WCA
2007HUGH01
YouTube
MikeHughey1
Thread starter #54
Update for @Jacck:
The bug that counted the zeros for best singles rank has been fixed. There were 6 entries for 7x7x7 that were sub-10; one of those I had noticed and already fixed (because it was mine :) ), and two of the others belong to the same person, so you are currently ranked 4th in all of Speedsolving.com. (SR == "Speedsolving.com Record", in case you were wondering.)

I can go ahead and DNF that solve if you want. Or perhaps the best approach would be to assume there was some sort of bug in the website that caused those to happen, in which they are probably just the minute portion of your solve time. (I think that's very likely. I think I remember it being true that there was a way to misformat your entry such that it threw away everything after what you intended to be the minute. Another thing I should add if I have time - validation for manual time entries to prevent mistakes like that from happening.) If that's a good assumption, I could go ahead and change it to 8:59.99, which is probably a safe overestimate of your actual solve. I'm inclined to go do that to the others. Shall I do that for you too?

As for the BLD means, I found my bug. You now have BLD means showing in your records.

Thanks again for pointing those bugs out to me so I could fix them!
 
Joined
Jan 26, 2016
Messages
239
Likes
258
Location
South-Germany
WCA
2013HUBH01
#55
You should DNF the 7x7x7 attempt, could be that I tried to enter a DNF manually. But in general it could be a good idea to correct a 5 to 5:59.99.
By the way: I checked my other results, they should be ok (hmmm, there is a surprising 3.75 in 2x2x2, but it was a easy scramble, so it is possible).

But are you sure about the bld-mean-rankings: my mean-rankings are worse than my single-rankings. Could be at 2bld, but I wonder if 31 have made a 5bd-Mo3 better than 19:51 and only 27 a single better than 16:36. Or does it count all means that were ever made?
But just seeing the means is great!!!
 

Mike Hughey

Super Moderator
Staff member
Joined
Jun 7, 2007
Messages
9,879
Likes
1,902
Location
Indianapolis
WCA
2007HUGH01
YouTube
MikeHughey1
Thread starter #56
Oops - you're right! I did a quick database query and discovered that there are only 6 of us who have ever even gotten 5x5x5 BLD means. I realize my mistake now; I had assumed that I could simply compare overall results, but had forgotten that for the BLD events, the overall result is the best solve, not the mean. I'll fix it soon.

And that was one reason I released it before it was all done - I figure others can help me find the bugs, and you are. Thank you!

Many of these bugs will become quite obvious once I add the next page that I am working on - the overall rankings by event. Then we can compare that list against the numeric rankings and be sure they're calculated correctly. Until that's done, I'd suggest taking the numeric rankings with a grain of salt.
 
Top