SpeedRun setup for all Speedsolving competition events

[abunickabhi]

There are a total of 35 events on the Speedsolving weekly competition as of 2021. They are 2x2 to 7x7, 2BLD to 7BLD, MBLD, 2-4 relay to 2-7 relay, OH, Feet, FMC, SpeedFMC, Megaminx, Clock, Pyra, Skewb, squan, FTO, Curvy Copter, Mirror Blocks, 15 puzzle, Redi Cube, Master Pyra, Mini Guildford, Kilominx and Rescramble.

I had first attempted a speedrun of all the SS events (except FMC,SpeedFMC,Curvy,Feet,7BLD) on 13th October last year, with a time of 3:45:30.62. Original post here
I used LiveSplits to time myself, similar to speed runs in video games. The speedrun was tiring but there were a lot of mistakes that I did that I do not want to do this time. In the first attempt, I did not do FMCs as I forgot about it, FTO was not introduced, I was not comfortable solving curvy copter, I gave up on 6BLD midway and I did not do feet as I had to change my camera setup, also my OBS layout was a bit offsetted which made the split times not clearly visible.

I am going to attempt another speedrun this month, and there are a few things that I am planning to do which will make this speedrun a bit more efficient:
1) Not do a livestream, as it is requires extra focus to read the chat, and handle the OBS.
2) Do 4 attempts out of the 5 for sighted events, as I will be get an average DNSing my 5th solve, unless I DNF any previous solve. Also for 2BLD-7BLD, just do one attempt and try for a success, as these events are ranked for single.
3) Try to get better in solving curvy copter, and order FTO in India. (It is not available here).
4) Try to do 5 attempts of SpeedFMC and 3 attempts of FMC, faster by typing a linear solution, and doing the linear solution on the cube OH. (Using a paper pen for these attempts will waste a bit of time, and it is better to just type in the solution).
5) Scramble the first solve of every event, so as to save the scrambling time for the first solve, and made a speedrun a bit easier on me. Don't worry, I do have multiple bigcubes, so I can setup relays and bigBLD easily.
6) Also have no inspection on all the sighted solves to save a bit of time there (even for 2x2 and Skewb)

If I do not DNF any solves, the number of solves that I will have to do and the estimated time of doing the entire 4 solves in the event are listed below.
The list in the order of the execution of the speedrun:

Doing BLD events first as the brain is fresh
2BLD
1 solve
Time estimate 10 seconds

3BLD
1 solve
Time estimate 25 seconds

4BLD
1 solve (2 solves if 1st one is DNFed)
Time estimate 2:30

5BLD
1 solve (2 solves if 1st one is DNFed)
Time estimate 5:00

6BLD
1 solve
Time estimate14:00

7BLD
1 solve
Time estimate 30:00

MBLD
2 cubes attempt as its the fastest
Time estimate 1:10

Feet
4 solves
Time estimate 7:00

Doing FMCs next as brain is still fresh and want to do non spam TPS stuff,
FMC
3 solves
Time estimate 15:00

SpeedFMC
4 solves
Time estimate 20:00

Rescramble
4 solves
Time estimate 3:00

Clock
4 solves
Time estimate 2:00

Pyraminx
4 solves
Time estimate 48 seconds

Master Pyraminx
4 solves
Time estimate 4:00

Skewb
4 solves
Time estimate 40 seconds

Redi Cube
4 solves
Time estimate 2:00

Kilominx
4 solves
Time estimate 4:00

Megaminx
4 solves
Time estimate 8:00

Mirror Blocks
4 solves
Time estimate 6:00

15 puzzle
4 solves
Time estimate 4:00

2x2
4 solves
Time estimate 20 seconds

3x3
4 solves
Time estimate 1:00

OH
4 solves
Time estimate 1:30

4x4
4 solves
Time estimate 4:00

5x5
4 solves
Time estimate 8:00

6x6
4 solves
Time estimate 14:00 (will take mini breaks in between solves, to not stress my wrists too much)

7x7
4 solves
Time estimate 24:00

Events that I might fumble in as I rarely practice them,
Square-1
4 solves
Time estimate 6:00

FTO
4 solves
Time estimate 20:00 (no idea as I have not attempted on a real puzzle)

Curvy Copter
4 solves
Time estimate 20:00 (no idea as I have mess up the last layer)

Relays will be faster as the cubes are already scrambled before the speedrun starts,
2-4 relay
Time estimate 50 seconds

2-5 relay
Time estimate 2:10

2-6 relay
Time estimate 4:55

2-7 relay
Time estimate 9:10

Mini Guildford (10 puzzles)
Time estimate 6:10

Total number of solves: (7+ 3 + 22*4 + 5) = 103 solves
Total estimated time: 14848 seconds or, 4 hours and 4 minutes


Do let me know what you guys feel about this ruleset I am following? Is scrambling one solve for each event beforehand logical or will it not be a complete speed run? Is doing 4 solves instead of 5, logical and be considered as a complete speed run? If I DNF 6BLD or 7BLD (there is only 1 attempt), will the speed run still be counted?

 

[xyzzy]

My $0.20:

4) Try to do 5 attempts of SpeedFMC and 3 attempts of FMC, faster by typing a linear solution, and doing the linear solution on the cube OH.

The rules of those events are to use pen and paper; typing the solutions in is absolutely cheating, not only in the speedrun sense, but also in the weekly competition events sense. (But also you only need to do four speedFMC attempts anyway.)

5) Scramble the first solve of every event, so as to save the scrambling time for the first solve, and made a speedrun a bit easier on me. […]
6) Also have no inspection on all the sighted solves to save a bit of time there

Pre-scrambling definitely seems fine to me. I'd even go as far as to say you should be allowed to pre-scramble everything, not just the first solve… although that would require a lot of cubes. (Failing this, I would also allow having someone else scramble while you're doing the solving, and/or using scrambling robots.)

One issue with pre-scrambling is inspection, though; it's easy enough to look ahead to future cubes to solve (e.g. the two seconds of dead time during OLL parity can be spent searching for pieces). Then again, that's still much less than 15 seconds, so maybe it doesn't matter.

Is doing 4 solves instead of 5, logical and be considered as a complete speed run?

Enough to get a result would be my benchmark, so yes. This also ties in to the next question…

If I DNF 6BLD or 7BLD (there is only 1 attempt), will the speed run still be counted?

Per my previous remark, I'd say no, but obviously 6BLD and 7BLD have very low success rates and that would make this speedrun almost impossible. And yet we don't want to allow "complete" DNFs where you just do five random turns and stop the timer.

Speedrunning is about optimising things to their logical conclusion, not to "uphold sportsmanship" or to "follow the spirit of the rules", and the rules should be written while taking that into account. We shouldn't decide now that "DNFing 6BLD and 7BLD are okay" and then complain later when someone does a speedrun that doesn't even include serious 6BLD/7BLD attempts.

A compromise is to instead add time penalties based on the number of unsolved pieces. (Only for 6BLD and 7BLD!) E.g. 10 minutes penalty per unsolved piece (you get to choose the orientation, so e.g. this counts as six unsolved pieces), with a maximum of twenty unsolved pieces allowed before the whole thing is a DNF. You're still incentivised to do a proper 6BLD/7BLD attempt, since it takes much less than 10 minutes to memorise an extra letter and do an extra half of a commutator, but it no longer punishes DNFs as harshly. (The exact time penalties / maximum allowed unsolved pieces don't have to be the example values I'm giving here. I don't do bigBLD, much less verybigBLD, so I don't have a good feel for what makes sense.)

---

Actually, speaking of time penalties, should +2s (going over inspection, one move off, etc.) also be added to the raw speedrun time? I could see arguments both for and against this.

For: Do we really want to count a speedrun where every cube ends up one move away from solved just because that was 0.1 s faster? That seems to be violating the spirit of the weekly comps.

Against: The speedrun category is primarily for optimising real time, not solving time or even penalty-adjusted solving time. If skipping the last move in a solve is faster (which it probably isn't—you still need to click the +2 button on the website), then you should absolutely be doing that.