World Record Predictions for Year 2025

2x2- 0.38,1.15
3x3- 3.32, 4.87
4x4- 14.24,19.87
5x5- 26.34,37.56
6x6- 54.28,59.32
7x7- 1.49.33,1.58.49
OH- 4.59,7.32
FEET- 16.53,19.83
3x3 Blindfolded- 14.82,19.34
4x4 Blindfolded- 54.38,1.30.50
5x5 Blindfolded- 2.30.14,2.47.34
Multi Blindfolded- 55/56 in 58.34 minutes
Skewb- 0.76,1.98
Pyraminx- 0.67,1.45
Square-1- 3.43,5.86
Megaminx- 24.59,29.83
Clock- 2.34, 3.59
FMC- 16,19.66
 
Ok Lets do this
2x2: 0.33, 0.81 // I think that with the amount of competitions, someone is bound to get a solve that is just a sledge / half of a sune. When this happens, I think that 0.3 is possible. I think that when someone gets 4 solves that are cancellations into easy cases, sub 1 average is possible.
3X3: 2.96, 4.52 // Again, singles will be a lot lower than we predict because there will be more opportunity for lucky scrambles. However, the average will have a considerable drought at low 5, with the first sub 5 average being the biggest cubing moment in this time period. For a while in these 7 years the wr holder used roux
4X4: 9.99, 15.34 //A more optimised method that is superior to Yau will be invented, possibly involving solving 6 edges before 3-3 edge pairing. When a sizeable skip happens in all areas of the solve, this gives rise to the first sub 10 single.
5X5: 24.55, 30.42 // Yau will eventually take over for this event. The hardware for this - and big cubes in general - will be incredible, and max park will be dominant in a similar vein that feliks is today.
6X6: 49.25, 55.00 // The first unofficial sub 1 that happens in the next 2 years is something to remember. Around this time, better 6x6's are near-perfect, and people start learning tons of algs for this event, driving times down.
7x7: 1.24.53, 1.29.20// The hays7 continues to be dominant for a very long time. Similar to 6, this becomes more algorithmic.
3BLD: 9.35, 15.25 // People find more ways to solve a cube BLD, involving being a bit reckless about 2 twisted corners / flipped edges in memo phase. This means that when someone gets lucky, they get really good times, but it is much harder to execute consistently
Feet: 11.11, 13.57 // The change to averages makes this event a lot more consistent. As well, new efficient methods are invented, and something that we dont expect takes over as the best method (maybe Petrus?)
OH: 4.59, 5.80 // (Yes, I have used the wr for 3x3) The reason is that there will eventually be a scramble, with so many comps, that is similar to the one that max park had on his misscramble. Roux becomes the dominant method by far, with no one using CFOP in the top 5
Skewb and pyra: (Doing these together because of how similar the world records and the puzzles themselves are): 0.67, 1.55:
People begin to learn a lot more algs for pyra, with Last 4 edges + 1 flip being put into use as a one look method. kewb is also greatly optimised with more advanced one looking
SQ1: 2.5, 4.00: Most people here are putting the sq1 times as above 3x3. I disagree, as I think that fundamentally a SQ1 solve will only involve looking at the cube 6 times, and so will be quicker. I think that with luck, and somehow even more algs, this will be an incredibly fast event
Mega: 15.16, 19.46: This event gets treated more and more like 3x3. People learn full 2LLL, with some even learning some sort of COLL or WV. Doing two faces at the same time is also popular, with multislotting for every pair
Clock:Meh
FMC: 15, 20.00: Someone gets a perfect god average, using all sorts of techniques we don't know of yet. Someone gets an incredibly lucky solve
4BLD: 40.35, 51.33: People invent new memo techniques
5BLD: 1.55.25, 2:15.xy // This event propably has the longest way to go. I think that different cubes with different optimised hardware are released for this event to acommadate slice moves, and a bigger community helps drive records down
MBLD: Similar to 5BLD, a bigger community without just one dominant person helps this event become more competitive.

General predictions: A 5 day 2000+ comp takes place
Moyu, Qiyi and Yuxin are all in a place similar to Dayan and shengshou today, releasing cheaper puzzles, with a lot of non WCA ones as well
Cube stores start making their own puzzles
Feliks and Kevin will have not broken any sorts of records for 2 years
Max park will have broken over 70
Top 5 cubers all over 20
 
Burnsy101 said:
Ok Lets do this
2x2: 0.33, 0.81 // I think that with the amount of competitions, someone is bound to get a solve that is just a sledge / half of a sune. When this happens, I think that 0.3 is possible. I think that when someone gets 4 solves that are cancellations into easy cases, sub 1 average is possible.
3X3: 2.96, 4.52 // Again, singles will be a lot lower than we predict because there will be more opportunity for lucky scrambles. However, the average will have a considerable drought at low 5, with the first sub 5 average being the biggest cubing moment in this time period. For a while in these 7 years the wr holder used roux
4X4: 9.99, 15.34 //A more optimised method that is superior to Yau will be invented, possibly involving solving 6 edges before 3-3 edge pairing. When a sizeable skip happens in all areas of the solve, this gives rise to the first sub 10 single.
5X5: 24.55, 30.42 // Yau will eventually take over for this event. The hardware for this - and big cubes in general - will be incredible, and max park will be dominant in a similar vein that feliks is today.
6X6: 49.25, 55.00 // The first unofficial sub 1 that happens in the next 2 years is something to remember. Around this time, better 6x6's are near-perfect, and people start learning tons of algs for this event, driving times down.
7x7: 1.24.53, 1.29.20// The hays7 continues to be dominant for a very long time. Similar to 6, this becomes more algorithmic.
3BLD: 9.35, 15.25 // People find more ways to solve a cube BLD, involving being a bit reckless about 2 twisted corners / flipped edges in memo phase. This means that when someone gets lucky, they get really good times, but it is much harder to execute consistently
Feet: 11.11, 13.57 // The change to averages makes this event a lot more consistent. As well, new efficient methods are invented, and something that we dont expect takes over as the best method (maybe Petrus?)
OH: 4.59, 5.80 // (Yes, I have used the wr for 3x3) The reason is that there will eventually be a scramble, with so many comps, that is similar to the one that max park had on his misscramble. Roux becomes the dominant method by far, with no one using CFOP in the top 5
Skewb and pyra: (Doing these together because of how similar the world records and the puzzles themselves are): 0.67, 1.55:
People begin to learn a lot more algs for pyra, with Last 4 edges + 1 flip being put into use as a one look method. kewb is also greatly optimised with more advanced one looking
SQ1: 2.5, 4.00: Most people here are putting the sq1 times as above 3x3. I disagree, as I think that fundamentally a SQ1 solve will only involve looking at the cube 6 times, and so will be quicker. I think that with luck, and somehow even more algs, this will be an incredibly fast event
Mega: 15.16, 19.46: This event gets treated more and more like 3x3. People learn full 2LLL, with some even learning some sort of COLL or WV. Doing two faces at the same time is also popular, with multislotting for every pair
Clock:Meh
FMC: 15, 20.00: Someone gets a perfect god average, using all sorts of techniques we don't know of yet. Someone gets an incredibly lucky solve
4BLD: 40.35, 51.33: People invent new memo techniques
5BLD: 1.55.25, 2:15.xy // This event propably has the longest way to go. I think that different cubes with different optimised hardware are released for this event to acommadate slice moves, and a bigger community helps drive records down
MBLD: Similar to 5BLD, a bigger community without just one dominant person helps this event become more competitive.

General predictions: A 5 day 2000+ comp takes place
Moyu, Qiyi and Yuxin are all in a place similar to Dayan and shengshou today, releasing cheaper puzzles, with a lot of non WCA ones as well
Cube stores start making their own puzzles
Feliks and Kevin will have not broken any sorts of records for 2 years
Max park will have broken over 70
Top 5 cubers all over 20
Click to expand...

I reckon only 2x2, Pyra, Skewb, Clock and MBLD will happen. ^^ But I have a kind of feeling that Max Park will indeed be still dominant in 2025.
 
I understand that. However, that is why people made such unambitious predictions about 2020, all of which seem to have been surpassed. Some of them are over 50 seconds of the mark for 7x7. According to them a 35 s ao5 for 4x4 was pretty crazy. And, as the cubing community grows and hardware improves, I see no reason why that will not drive the record down. As shown in events such as MBLD in the past year, the record only increased when there were a few people competeing for it
 
Burnsy101 said:
Ok Lets do this
2x2: 0.33, 0.81 // I think that with the amount of competitions, someone is bound to get a solve that is just a sledge / half of a sune. When this happens, I think that 0.3 is possible. I think that when someone gets 4 solves that are cancellations into easy cases, sub 1 average is possible.
3X3: 2.96, 4.52 // Again, singles will be a lot lower than we predict because there will be more opportunity for lucky scrambles. However, the average will have a considerable drought at low 5, with the first sub 5 average being the biggest cubing moment in this time period. For a while in these 7 years the wr holder used roux
4X4: 9.99, 15.34 //A more optimised method that is superior to Yau will be invented, possibly involving solving 6 edges before 3-3 edge pairing. When a sizeable skip happens in all areas of the solve, this gives rise to the first sub 10 single.
5X5: 24.55, 30.42 // Yau will eventually take over for this event. The hardware for this - and big cubes in general - will be incredible, and max park will be dominant in a similar vein that feliks is today.
6X6: 49.25, 55.00 // The first unofficial sub 1 that happens in the next 2 years is something to remember. Around this time, better 6x6's are near-perfect, and people start learning tons of algs for this event, driving times down.
7x7: 1.24.53, 1.29.20// The hays7 continues to be dominant for a very long time. Similar to 6, this becomes more algorithmic.
3BLD: 9.35, 15.25 // People find more ways to solve a cube BLD, involving being a bit reckless about 2 twisted corners / flipped edges in memo phase. This means that when someone gets lucky, they get really good times, but it is much harder to execute consistently
Feet: 11.11, 13.57 // The change to averages makes this event a lot more consistent. As well, new efficient methods are invented, and something that we dont expect takes over as the best method (maybe Petrus?)
OH: 4.59, 5.80 // (Yes, I have used the wr for 3x3) The reason is that there will eventually be a scramble, with so many comps, that is similar to the one that max park had on his misscramble. Roux becomes the dominant method by far, with no one using CFOP in the top 5
Skewb and pyra: (Doing these together because of how similar the world records and the puzzles themselves are): 0.67, 1.55:
People begin to learn a lot more algs for pyra, with Last 4 edges + 1 flip being put into use as a one look method. kewb is also greatly optimised with more advanced one looking
SQ1: 2.5, 4.00: Most people here are putting the sq1 times as above 3x3. I disagree, as I think that fundamentally a SQ1 solve will only involve looking at the cube 6 times, and so will be quicker. I think that with luck, and somehow even more algs, this will be an incredibly fast event
Mega: 15.16, 19.46: This event gets treated more and more like 3x3. People learn full 2LLL, with some even learning some sort of COLL or WV. Doing two faces at the same time is also popular, with multislotting for every pair
Clock:Meh
FMC: 15, 20.00: Someone gets a perfect god average, using all sorts of techniques we don't know of yet. Someone gets an incredibly lucky solve
4BLD: 40.35, 51.33: People invent new memo techniques
5BLD: 1.55.25, 2:15.xy // This event propably has the longest way to go. I think that different cubes with different optimised hardware are released for this event to acommadate slice moves, and a bigger community helps drive records down
MBLD: Similar to 5BLD, a bigger community without just one dominant person helps this event become more competitive.

General predictions: A 5 day 2000+ comp takes place
Moyu, Qiyi and Yuxin are all in a place similar to Dayan and shengshou today, releasing cheaper puzzles, with a lot of non WCA ones as well
Cube stores start making their own puzzles
Feliks and Kevin will have not broken any sorts of records for 2 years
Max park will have broken over 70
Top 5 cubers all over 20
Click to expand...

I thought I was optimistic! I've got some free time and no cube, so I'll go through this.

2x2: Sounds about right.
3x3: I doubt we'll see a sub-3 official single, average sounds about right.
4x4: there is a possibility for crazy skips, but I have a hard time believing we'll see sub-10. I've had a few lucky solves with 3 extra edges solved, so I do 3-2 edge pairing, but I don't see how you could efficiently force that.
5x5: you're probably right that Yau has a higher ceiling. Redux is faster for me, but with better look ahead on f3e it's probably faster. Improved magnetic hardware might also make redux-roux a viable big cube method.
6x6: those times are hard to believe, especially inside the next 7 years, but I do believe we'll see at least a sub-1 single at some point. I am highly skeptical about the part about algorithmic solves. I could see algorithms extending as far as l2c and l4e, but probably not even to all cases for l4e.
7x7: same as 6x6, except I'd be surprised if we ever see a sub-1:30 single, let alone an average. I think we may be getting beyond the age of dominant hardware.
3BLD: it won't be me, but sounds good.

The cow I was watching while writing this up died, so I'll leave it there. I like the idea of specialized 5BLD hardware.
 
Burnsy101 said:
Most people here are putting the sq1 times as above 3x3. I disagree, as I think that fundamentally a SQ1 solve will only involve looking at the cube 6 times, and so will be quicker. I think that with luck, and somehow even more algs, this will be an incredibly fast event
Click to expand...

With current methods you could do it in 3 looks, CSP, OBL PBL. PBL. My guess is that Parity PBL will be like Parity is today, being annoying.
 
Let me explain a bit of this reasoning. Firstly, people seem to think that a sub-1:30 7x7 mean is impossible. However, just 1 YEAR ago, the WR was 18 seconds slower. With that reasoning, it would only take 2 years for the record to go down by 36 seconds, putting it under my prediction. Granted, this years was kind of a "freak year", if you willl, because of the introduction of magnets, but I see no reason why a year like that will not happen twice in the next seven years, and combined with progress, it could happen. 3x3 single:
The UWR now is roughly my prediction (2.99 to 2.96)
I would say that there are around 6 people in the world that could get that single, as currently 6 people have a faster average than what patrick ponce did when he got that single
In the future, there may be 100 people with potential to get that sort of single officially. Not 4 years ago someone with an 8.29 average was in the top 10, and now they would not get in the top 130.
Assuming 1/5 compete a week:
26 people doing 15 different scrambles a week for 4 years (lets say)
= 26 x 15 x 50 x 4 = 780000 different opportunities for that sort of solve. And I would say that the odds of a solve like that are 5% for an easy cross x 1/500 for a zbll skip or something like that (roughly)
This gives us around 8 solves that have this sort of potential, which might happen
 
Burnsy101 said:
Firstly, people seem to think that a sub-1:30 7x7 mean is impossible. However, just 1 YEAR ago, the WR was 18 seconds slower. With that reasoning, it would only take 2 years for the record to go down by 36 seconds, putting it under my prediction. Granted, this years was kind of a "freak year", if you willl, because of the introduction of magnets, but I see no reason why a year like that will not happen twice in the next seven years, and combined with progress, it could happen.
Click to expand...

I hope you're right. Here's why I think you may be wrong: for your estimates of 3x3 single you are relying on a lucky solve. The implicit assumption in that is that 3x3 hardware (and method) while maybe not yet perfect is very nearly optimized. Big cube hardware has traditionally lagged behind 3x3, and I don't think it's as close to perfect as 3x3 but it's getting darn good. To compare to some iconic 3x3s, we are well past the rubik's and Zhanchi stage with big cube hardware. Probably around Aolong or GTS1. Honestly a bigger advance than turning quality right now would be fully customizable stickerless shades to improve recognition. Big cube method is near optimized as well. It's possible to get a lucky scramble on a big cube, but the chances of having a lucky enough scramble to meaningfully affect times is vanishingly small.
 
So I thought I'd be methodical about this. I plotted the 3x3 single WR10 over time. Then this happened:
b1xXNTM.png


What was this explosion of sub5 solves in 2017?

And it's getting even crazier with 3x3 average, because it seems to actually speed up?
 
Okay, maybe time to do a set of my own:
(single, average)
2x2: 0.31, 0.89
3x3: 2.91, 3.98
4x4: 9.98, 14.61
5x5: 21.86, 29.85
6x6: 50.41, 59.64
7x7: 1:25.61, 1:29.98
3x3 BLD: 10.31, 14.51
3x3 FMC: 14, 18.67
3x3 OH: 4.59, 5.80
3x3 WF: 12.34, 16.91 //maybe unoptimistic
Clock: 1.98, 3.01
Megaminx: 19.54, 23.13
Pyraminx: 0.77, 1.54
Skewb: 0.78, 1.52
Square-one: 3.02, 5.46
4x4 BLD: 59.46
5x5 BLD: 2:29.54
3x3 Multi: 56 points
 
So I did something. I took WR10 since 2010 and let Libre Office interpolate it with an a*e^(bx)+c function where c is half of the current world record (in other words: a function that converges towards half of the current world record). This method is probably better at predicting what WR10 will be in 2025, but I guess it works as an upper bound.

These are the results:

* 2x2: 0.42, 1.04 - single sounds good, average WR will probably be lower

* 3x3: 3.72, 5.29 - sounds reasonable

* 4x4: 16.56, 19.67 - WRs will probably be lower

* 5x5: 36.02, 40.54 - WRs will definitely be lower, I'm counting on Max

* 6x6: 1:12.45, 1:18.26 - WRs will most probably be lower

* 7x7: 1:40.21, 1:44.59 - sounds ambitious, but doable

* 3bld: 12.51, 14.51 - single is possible, mean sounds insane

* 4bld: 1:11.16 - sounds reasonable as a WR, but not as a WR10

* 5bld: 2:44.03 - probably possible but insane

* MBLD: 37.67 - oh, Lol, this went horribly wrong (my approach was also pretty random). WR10 is currently 26 points. And apart from Maskow, the development hasn't slowed down yet. The rest of the world still has to catch up with the top4. My actual prediction is 44 points.

* OH: 6.32, 8.45 - sounds good

* feet: 17.49, 21.40 - Yeah, I think Daniel's 16.96 will probably stay. Average sounds good

* FMC: 17, 19.00 - Single (while using 9.5 as lower bound) sounds actually kinda reasonable. If I use WR/2 (12.00) as lower bound for average, I get a prediction of 16.67 - lol. So I used 17 as the lower bound and got 19 as a prediction. Still completely insane. My actual prediction is 21.67

* Pyra: 0.93, 1.76 - sounds good

* Mega: 28.17, 32.23 - Lol, average already got beaten. Single will be lower as well. The rest of the world needs to catch up with the top2.

* Skewb: 0.67, 1.30 - sounds a bit low, there isn't that much data because it's only been 4 years

* Squan: 5.26, 7.01 - Sounds a bit high. Maybe it overestimates because CSP is still rather new?

* Clock: 3.73, 4.81 - Lol, that was precise. But I think the single will fall. Average will probably also be lower.
 
Last edited:
You must log in or register to reply here.

Similar threads

Kiwi_Cuber
Speedcubing World Record Predictions for the End of 2022!
Replies
0
Views
1,011
Kiwi_Cuber
Kiwi_Cuber
D
  • Poll Poll
World Records single predictions for the year 3000
Replies
5
Views
1,351
abunickabhi
abunickabhi
L
World Record Predictions for the year 2030
9 10 11
Replies
202
Views
26,355
Caleb Valenzuela
Caleb Valenzuela
O
Theoretical World Record Prediction for Big Cubes
Replies
18
Views
5,631
Christopher Mowla
Christopher Mowla
That70sShowDude
World Record predictions in the year 2020?
22 23 24
Replies
476
Views
94,874
Drewe
Drewe
Back
Top