2009 : BLD average

Edouard Chambon (2009-01-08 20:23:13 +0000)
I would like that we take an average also in BLD event to decide who is the winner. Everybody solves 3 times. Then, we do the mean between the 2 best solves out of the 3.
Lucas (2009-01-08 22:06:49 +0000)
I am against this, only because it introduces a new format. It's not a bad format (it'd get my vote for 6x6x6 and 7x7x7), but I don't think it's good to introduce. And why not median of 3? If this is introduced, I would prefer if it were an optional round format. Perhaps it would be adopted for finals at larger comps? Once several people will have averages, there will be pressure for more averages. (Although we might have organizers doing avg2/3 to enter averages in the db, and still awarding prizes to the three fastest singles, which we apparently agreed is allowed.) I don't think this should be introduced until BLD cubers have gotten a while to get used to this format. In particular, it will create necessitate different approaches to the two BLD world records in the category (more than any other puzzle), and so it's no longer the same single event it was before.
cubetalk (2009-01-09 22:30:24 +0000)
This is good if your doing it for the 3x3x3 If it was the 4x4x4 or 5x5x5, most avgs. would be dnf
Bob (2009-01-09 23:14:48 +0000)
Are you proposing this because you think that blindfold solves are too inconsistent from one solve to the next or because you want competitors to be more accurate?
cubetalk (2009-01-09 23:23:43 +0000)
[quote="Bob":3bhzpxdm]Are you proposing this because you think that blindfold solves are too inconsistent from one solve to the next or because you want competitors to be more accurate?[/quote:3bhzpxdm] I think with the higher order of cubes (4,5) it would be hard for people to do 3 solves in a row. I am sure they ARE people who can do this but how many are there? I don't believe that there are enough people out there to give competition to people who can do the avg. What i mean is that a record is set, it might be set for a LONG period of time.
Bob (2009-01-09 23:40:46 +0000)
[quote="cubetalk":1oavraz0][quote="Bob":1oavraz0]Are you proposing this because you think that blindfold solves are too inconsistent from one solve to the next or because you want competitors to be more accurate?[/quote:1oavraz0] I think with the higher order of cubes (4,5) it would be hard for people to do 3 solves in a row. I am sure they ARE people who can do this but how many are there? I don't believe that there are enough people out there to give competition to people who can do the avg. What i mean is that a record is set, it might be set for a LONG period of time.[/quote:1oavraz0] Sorry for the confusion. I was directing my question towards Edouard because he proposed it. I do agree with you that it would be ridiculous for 4x4 and up bld because the format for those is often best of 1 or 2 (partly because they take so long and partly because success rates are lower)
Edouard Chambon (2009-01-10 06:59:41 +0000)
[quote="Bob":nslha78k]Are you proposing this because you think that blindfold solves are too inconsistent from one solve to the next or because you want competitors to be more accurate?[/quote:nslha78k] Both. That would be like in the others events : for single you need to take risks and be lucky and for average you need to be consistenly fast. Asking 67% of success is not too much... And if nobody solved 2 cubes successfully, we order them by the best single. (actual ranking)
BryanLogan (2009-01-10 14:46:38 +0000)
[quote="Edouard Chambon":3vs6p2sy]And if nobody solved 2 cubes successfully, we order them by the best single. (actual ranking)[/quote:3vs6p2sy] Very bad. Let's say I have 1 solve left, and I have a 3:00 and a DNF for my previous two solves. What's my strategy? Well, it depends. 1) If the second place person has finished and has 2:30, DNF, DNF, then my strategy is to not DNF and I'll win. 2) If the second place person has finished and has 2:30, 3:00, DNF, then my strategy is to try and get below 2:30. So by being able to go last, I have an advantage. This is not good. While in theory this could occur for the events, the rate of DNF's there is so low if you have a DNF average, you won't win.
Edouard Chambon (2009-01-11 07:54:15 +0000)
BrianLogan, have you ever solved a cube blindfolded ? 1) With the actual regulations, being the last also gives a big advantage. And even more than if you have to calculate the means of other's times. In my recent competitions I knew what I had to do to be the first. 2) You are talking about a very specific case that won't occur very often. Most of the time, at least 3 person will have a mean.
BryanLogan (2009-01-11 14:19:38 +0000)
[quote="Edouard Chambon":29scq3we]1) With the actual regulations, being the last also gives a big advantage. And even more than if you have to calculate the means of other's times. In my recent competitions I knew what I had to do to be the first. [/quote:29scq3we] OK, so why try to advocate giving the last person another advantage? Going into a competition, you can see the competitors and get a rough idea of what you need to do to win. If you go last, then you get a specific number, but the strategy of "I need to be faster than that other person" still holds. Changing this advantage from "I need a 3:00 to beat his 3:01" to "I need a 9:59 to beat his 3:01" would be awful. [quote="Edouard Chambon":29scq3we] 2) You are talking about a very specific case that won't occur very often. Most of the time, at least 3 person will have a mean.[/quote:29scq3we] Really? Where are you getting this information? I'm glancing through the results and don't see this too often. Even at the French Open, you had 12 people without all DNF's and only one would have an average. Finals of Euro 2008, nope. US Open final wouldn't have had this problem.
Lucas (2009-01-11 22:31:26 +0000)
[quote="Edouard Chambon":mqhq4qdo]Asking 67% of success is not too much...[/quote:mqhq4qdo] Do you realize that if a competitor has a 67% success rate, they still have <45% chance of getting at least two solves? If you want a 2/3 chance of success in competition, you need over 90% accuracy. (Yes, I do know you're talking about "single solve success rate" at a competition. But that's not what competitors have to prepare for. Chris Hardwick is a good example of someone who does like this for his solves.)