I don't see why people care so much about this...it's not like the prize was $1,000. Mike tried to do something to give people a little extra prize. I really don't think people would start dropping 4+ events just to get a 3x3...
I'm happy that you won. I was actually figuring that the cubes would probably last until 30th place due to no-shows, but almost everyone in the top 30 stayed through awards, which was really nice. Sorry for the people just below you, though.
Oh, don't get me wrong, I am thrilled that I won! I just thought I had done better, that's all. Also I truly do think this new ranking system is a great idea since there are many people consist in getting top ten in a lot of events but not quite getting 3rd or higher. It really gives them a chance to earn something.
I don't have any specific suggestions, but it seems to me that people are penalized for trying to compete in more events (unless they're good at every event), which to me goes against the spirit of WCA competitions. For example, had I only competed in Clock, I would have had first choice at the puzzles. I don't personally care myself, but I don't like the idea that people are persuaded to compete in fewer events under this percentile system. But maybe you like this as an organizer.
Not knowing the system beforehand, I think it did a good job of recognizing cubers for a good overall performance, but if this stays the same for 2014, I feel like some people may only sign up for a handful of events with the intent of winning that award. I'm probably making a bigger deal out of this than necessary though, it is just a few puzzles, after all.
I totally understand your perspective here, and I really sympathize with it, since I like competing in lots of events I'm terrible at. But the problem I had in coming up with a decent alternative is that I couldn't find an alternative metric that didn't either penalize people trying to compete in more events (like this), or else actually REWARD them for competing in more events (like, say, the scheme used for the weekly competition points on Speedsolving.com). And to me it felt equally unfair to reward people for competing in more events. Should Austin Moore have finished way down the list simply because he didn't compete in other events, because he doesn't like to compete in them? I find that at least as hard to justify as what I did.
So if you have an alternative system that would walk the line between these two extremes, I'd love to hear it. You're the statistician, after all - I'd love to see what you can come up with!
I totally understand your perspective here, and I really sympathize with it, since I like competing in lots of events I'm terrible at. But the problem I had in coming up with a decent alternative is that I couldn't find an alternative metric that didn't either penalize people trying to compete in more events (like this), or else actually REWARD them for competing in more events (like, say, the scheme used for the weekly competition points on Speedsolving.com). And to me it felt equally unfair to reward people for competing in more events. Should Austin Moore have finished way down the list simply because he didn't compete in other events, because he doesn't like to compete in them? I find that at least as hard to justify as what I did.
So if you have an alternative system that would walk the line between these two extremes, I'd love to hear it. You're the statistician, after all - I'd love to see what you can come up with!
Since this is "overall ranking", I think competing in an event should only benefit you.
For a given person, take the average of the ranks in each event; if you did not compete in an event, then you should be ranked last in that event.
This encourages people to compete in more events even if they are not so great in them.
This of course would be a bad idea in Indiana competitions since there would be so many more people doing so many more events, and from an organizational viewpoint this would be hard to manage.
I totally understand your perspective here, and I really sympathize with it, since I like competing in lots of events I'm terrible at. But the problem I had in coming up with a decent alternative is that I couldn't find an alternative metric that didn't either penalize people trying to compete in more events (like this), or else actually REWARD them for competing in more events (like, say, the scheme used for the weekly competition points on Speedsolving.com). And to me it felt equally unfair to reward people for competing in more events. Should Austin Moore have finished way down the list simply because he didn't compete in other events, because he doesn't like to compete in them? I find that at least as hard to justify as what I did.
So if you have an alternative system that would walk the line between these two extremes, I'd love to hear it. You're the statistician, after all - I'd love to see what you can come up with!
What I would do is determine a sort of "base level" for each event. As an example, let's make the base level for Megaminx the cutoff of 2:30.00. Then, if someone doesn't compete in this event, calculate whatever their percentile ranking would have been had their result been the base level, and give them that.
For determining the base level, I'd make it a time that is achievable with a little practice. This is rather subjective, but I think it tries to not penalize competitors for trying an event, while also not being too harsh on those who don't care to compete in an event. These baselines become tricky with BLD events, I'd probably just make the ranking one higher than DNFs (say 10:00 for 3BLD) so that there is some penalty for not completing an attempt.
A number of people have asked about the ranking system I used for the prizes for overall results. I believe Chris Krueger was the one who quickly observed it is essentially an average percentile - an average over your percentile ranks in the events that you competed in. The formula we used was actually this:
(sum over events competed in of ((rank of the person for a given event) / (number of competitors for that event))) / (number of events competed in)
If there were multiple rounds, the final round was used for the ranking of those people who made the final, the first round placement used for the rest, but for all of those, the number of competitors for that event was the number of competitors in the first round.
The actual ranks and scores for everyone are given below, so you can see how you did:
Austin Moore 0.064887
Lucas Etter 0.149258
Andrew Kang 0.169761
John Brechon 0.174564
David Ludwig 0.214158
Chris Tran 0.218794
Brennan Arnold 0.233656
Rami Sbahi 0.235560
Nathan Dwyer 0.254216
Christopher Olson 0.258504
James Hildreth 0.259857
Corey Sakowski 0.284200
James LaChance 0.290712
Sam Nave 0.291917
Jared Benson 0.302356
Walker Welch 0.315360
Anthony Brooks 0.327824
Chris Krueger 0.349615
Trevor Mahoney 0.356695
Michael DeLaRosa 0.359300
Tolik Borisov 0.370937
Mike Hughey 0.396025
CJ Sheng 0.411299
Daniel Garcia 0.413761
Jim Mertens 0.415640
Hongfei Tian 0.420499
Elliott Ross 0.423901
Sydney Weaver 0.430269
Dylan Laudenschlager 0.436170
Ryan DeLine 0.442085
Chris Braiedy 0.460364
Daniel King 0.471337
Ilkyoo Choi 0.473331
Nakai Velasquez 0.483892
Brandon Mikel 0.501651
Ayman Hussein 0.512122
Trevor Spitzley 0.531926
Carson Penticuff 0.558439
Evan Brown 0.562590
Kit Clement 0.573339
Michael Czerpak 0.578554
Kalib Walling 0.589961
Cooper Bossert 0.592047
Jack Engledow 0.593539
Kevin Voor 0.598583
Cameron Dowless 0.618578
Andy Evans 0.619616
Joshua Sampson 0.630320
Benjamin Skoning 0.630721
Praval Telagi 0.638428
Jackson Carroll 0.639206
Richard Tran 0.649325
Jack Haviland 0.649984
Christopher Resendiz 0.664375
Brandon Haynes 0.670735
Yasser Alsilwadi 0.688226
Jacob Naranjo 0.688466
Nathan Garner 0.688895
Kevin Tatooles 0.709950
Daniel Kottenstette 0.710177
Justin Rausch 0.733766
Abraham Levine 0.751501
Michael Conard 0.753133
Scott Guenard 0.755572
Jacob Galloway 0.760040
Eli DeLaRosa 0.770887
Bobby Krupa 0.776655
Megan Le 0.777288
Marie Hughey 0.780099
Sam Wyckoff 0.790088
Young Kim 0.797052
Michael Womack 0.800516
Quentin Mounts 0.808511
Collin Guenard 0.815906
Carson Brown 0.829657
Brandon Kesner 0.829787
Ryan Crum 0.834036
Hannah Phipps 0.836203
Dae San Kim 0.850372
Ben Crum 0.861702
Ian Phipps 0.864624
Kaleb Vicary-Rzab 0.878179
Caden Hines 0.886611
Tyler Gross 0.893617
Rebecca Hughey 0.895823
Kelsey Vicary 0.928516
Ross Vardaman 0.929943
Keegan Fischer 0.935713
Adam Earnst 0.936170
Dana Conard 0.946809
Danielle Harrison 0.956668
Gloria Hughey 0.968085
Humberto Velasquez 0.978723
Rachel Hughey 0.989362
Dave Haviland 1.000000
I'm very happy with how this came out - it gave some pretty good solvers who are never quite able to make a podium to be recognized for their achievements. It seems like this formula isn't a bad way to do an overall ranking, but if you have suggestions for an alternative, it might be interesting to hear them.
This is the formula we use at Vancouver competitions. It sums percentile ranks as well as a ratio between their average and the average of all averages in the round, that way there's a benefit for competing and doing okay across many events as well as completely dominating one or two of them. We've used it at our last 3 competitions and has been working pretty well. I think Pedro is planning to use it at his competitions too, now.
This is the formula we use at Vancouver competitions. It sums percentile ranks as well as a ratio between their average and the average of all averages in the round, that way there's a benefit for competing and doing okay across many events as well as completely dominating one or two of them. We've used it at our last 3 competitions and has been working pretty well. I think Pedro is planning to use it at his competitions too, now.
This is nice - thanks, I appreciate the input. But it seems to me like it might still give quite an advantage to someone doing lots of events okay over someone who just does one or two of them well. One interesting thing about it, though, is that adding scores for multiple rounds would help the one-or-two-event person a bit, if you did lots of rounds of each event. Unfortunately, I usually don't have lots of rounds.
All things told, though, I'm really liking this formula as a starting point.
Then shouldn't baseline for all other events be the worst you can do in the allotted time? Such as 10:16.00 (assuming gaining all +2's and that you can still get +2's on a 10:00.00 solve) for any event without hard limits?
Thank you for hosting a great competition Mike!
Thank you Mike Hughey's family for showing us that a family can accomplish amazing things by working together.
Thank you to everyone who said hi to me, it was great to meet you.
Thank you to Chris Krueger and Chris Tran for tagging along and making this trip possible.
And finally thanks to everyone for showing me again how amazing cubing is.
Cheers, hope to see everyone again soon.
If anyone has some "generic" pics from Indiana, it would be very awesome if you could link them for me. I need some for a project I'm working on. The pics don't need to be super high resolution... as long as they are at least "web quality" and are in focus, they will be fine.
Then shouldn't baseline for all other events be the worst you can do in the allotted time? Such as 10:16.00 (assuming gaining all +2's and that you can still get +2's on a 10:00.00 solve) for any event without hard limits?
Yeah, I never said this was a good idea for BLD events.
What I think would make a good baseline time would be to design a query that pulls all of the first average each competitor gets in each event, and take the average of these averages. Again, this is a poor way of determining this for "Best of X" events, and I'm not sure what the best method for determining a baseline is in this case. For average events though, it would give an idea as to what the average competitor does on their first time with an event, which I feel is the best way to accomplish what we want - a system that won't too harshly penalize those who want to compete in all events, but still make it possible for those who don't compete in said events to win. If this is still seemingly harsh on those who compete in a handful of events, then it may be a good idea to give stronger weights to those events which the competitor actually competed in.
If anyone has some "generic" pics from Indiana, it would be very awesome if you could link them for me. I need some for a project I'm working on. The pics don't need to be super high resolution... as long as they are at least "web quality" and are in focus, they will be fine.
If anyone has some "generic" pics from Indiana, it would be very awesome if you could link them for me. I need some for a project I'm working on. The pics don't need to be super high resolution... as long as they are at least "web quality" and are in focus, they will be fine.
