Friday, 11 November 2016

Judging system idea (credits to Bula)

Bula's suggestion: To have separate judges for each criteria aspect, rather than each judge judging all criteria. For example, taichi and HAL would judge only execution, I would judge only difficulty, fel2fram or RPD would judge only creativity.

For effectiveness and presentation: it is probably fine if all judges graded them, as they do not require as specialised focus/knowledge/high skill level to evaluate, and are not weighted as highly as exec/diff/crea.

Whereas specialisation is old idea applied to many areas of life (medicine, manufacturing etc), this is probably first time I recall it mentioned to me (but I would not be surprised if it was discussed in passing by other spinners before). Nonetheless, credit goes to Bula for stating this idea, when I write this post.

Note: this post discusses his suggestion and related aspects rather than judging in PS in general (as applicability of number scoring to artform like spinning or whether PS can/should even be judged, the deficiencies in judging criteria regarding finesse and fine finger movements for execution, how strictly the original vs creative borderline should be applied, what difficulty is considered 'unbelievable', more than enough for pages and pages etc).

As skill levels in PS (in refinement, technical material, trick/linkage range, difficulty and variations) have increased hugely since typical judging allocations of 'each judge scores all categories' method, which has been used since WT07, specialisation is definitely needed to ensure proper evaluation of various aspects of spinning. Whereas most arts (performance types like music, ice skating etc) or sports do not use specialised judges for different criteria areas, it is quite clear in current PS environment (and has been so since around 2009) that level gap between good spinners' areas of focus compared to areas they do not focus on has increased, and that existing materials shown in PS have also increased enormously.

As spinners focused on creativity probably have better idea of existing concepts, they will have better assessments of creativity and be less likely to think old variations are new. Similarly, spinners who are very focused on technical material aspects probably underestimate the weight of errors in execution or deficiencies in finger positioning/finger movements; whereas spinners who do not focus on using unusual/hard variations are unlikely to be able to separate something that is truly difficult from something that is just a bit hard.

Of course, Bula's system also means each judge has less work to do (as they do not have to judge in all categories), which reduces workload and should increase consistency in scoring within each judge due to lower mental fatigue. As judges selected for each category are specialists/focus on those aspects, there will probably also be less variability in scoring and higher accuracy (i.e. no more menowa WT15 R3 9/10 execution from PSH judge saying combo was dynamic so score was justified, or something like that).

Largest 'problem' with this system is that it ends up narrowing perceptions (reducing open-minded approach to exec/diff/crea) to what the selected group of specialists think, which may decrease innovation. However, I feel small group of specialists in each area should give a better result than having -coughcough- 'strange ideas' contributed from large group of non-specialists.

Another issue with this system is determining which spinners are specialised enough to judge each category (but I feel spinners generally have good idea of who has good control, who is creative, who has hard material - because while individual preferences may be different, community gravitates towards admiring small group of a few dozen or so good spinners who are great at their respective focuses).

There's some spinners who will make combos that will confuse everyone, like fel2fram (who no doubt has top level creativity and difficulty, but his execution is hard to evaluate because we do not really have 'ideal image' that his direction of spinning should appear like), but such out-of-the-box geniuses will always go beyond existing logic.

Tuesday, 1 November 2016

This is power V5 commentary

Hardest by fair margin and most filming time dedicated for this compilation, way more than any solo or previous ‘this is power’ videos. Just some comments about material used and frustration involved.

Estimation of success rate based on raw clips of attempts, which have one try every 4-6 seconds, so 10 min raw clip to get 1 successful draft → 100-150 attempts. As comparison, most of my 9 year's solo's combos could be landed once every 5 min or so (50-70 attempts) after my hand was fully warmed up.

0:00-0:08 - half tap aerial > pun kan with coffee packet, I find it easier to start pun kan on strange objects (like inktube, toothpick etc) this way than usual ta start. Just a funny introduction clip, kind of like solo introductions.

0:09-0:20 - [fl ia - spiderspin 2.0] x 4 and fl ia aerial 1.0 combo, added extra rotations to spiderspin and aerial variations, very fun and far harder. Usually fl around - topspin and aerials have 0.5 rotations, I have vid from 2 weeks ago of [fl ia - ss 1.0] x 15 (below) and I’d say it’s at least as hard as ordinary [fl ia - ss 0.5] x 100 (at least you’d need to master the technique to point where you can do 100+ perfectly and have very good idea and feeling of pen’s position and rotations). Aerial with 1.0 rotations rather than 0.5 doesn’t give as large difficulty boost, but I’d estimate [fl ia - ss] aerial 1.0 x 10 is about as hard as aerial 0.5 x 20.

Compared to rest of clips in V5, this one was easy to land (about 5 minutes excluding warmup time, so 50-80 attempts roughly).

0:21-0:27 - [knee bounce - spiderspin] combo, fairly difficult and more of ‘gimmick’ depending on some luck and calmness (skill needed to align hand for spiderspin, but luck needed to have pen spinning in horizontal plane and rising right up from knee bounce). A significant upgrade on the ta rel - knee bounce combo at 0:19-0:28  in ‘THIS IS POWER V3’ (below) from 2013.

Took about 10 minutes to land, 200 attempts roughly (as messups often happen on first knee bounce in 1 second so higher tries per time)

0:27-0:37 - [ex spiderspin - fl around spiderspin] x 2 fall variation, this was really hard to record, final segment not as controlled as I wanted. While I had mastered [fl ia - ss], [fl ta - fl ia - ss], [fl ma - ring ss] and [index mid mirr pp rev - ring ss] to high level while training them in past 2 ½ years, I am still developing way to do [mid ring mirr pp rev - pinky ss] and [fl ra - pinky ss], so I kept messing up on them. Easily one of the hardest clips in this video.

Took 95-100 minutes (most of a Saturday morning) recording to land twice, so 400-500 attempts per successful take.

0:38-0:44 - ringbak pop spin and pun kan rev - bakspin combo, inspired by vine clips from katts [JEB]. I had another take with 1.0 bakspin after pun kan rev, but control of ringbak pop in it was worse, so it was rejected. Very fun to try this sequence and many variations possible.

As I’m not that good at pun kan rev - bakspin yet, took 20 minutes to get decent take, so 200-250 attempts.

0:45-0:56 - [mirr pp rev - ring ss] and [fl ra - pinky ss] combo with bonus [fl ma - ring ss] aerial 1.0, using all my favourite craziest variations of ring and pinky ss. Transition from ring ss to fl ra pinky ss is strange due to ring finger movement. I had another 2 drafts filmed with the fl ma ring ss aerial 1.0 leading into fl ra pinky ss (instead of fl ma ring ss in the actual video), but control was worse overall so they were rejected. Difficulty of this sequence with aerial 1.0 - fl ra pinky ss nearly as high as the ex ss fall variation clip.

For the aerial 1.0 - fl ra pinky ss clips, landed twice in 60 minutes, so 350-400 attempts. The selected clip with aerial 1.0 - fl ma ring ss was easier, landed once every 10-15 min so roughly 120-160 attempts.

0:57-1:17 - 1p2h power + rev power combo with [mirr pp aerial - palmspin rev], [fl ta rev - ss rev], RH mirr pp - LH pun kan, and antigravity ‘hai tua’ variation. Wanted to try some new rev power on RH and linking LH palmspin fl ta cont in middle of combo, far harder than the 1:01-1:22 1p2h fl combo in my 8 year solo (below). Took way longer to land this than I should have as computer crashed 5 times 40-60 minutes into recording and wore my mind out even more T_T. Very frustrating to film this as left hand kept messing up the palmspin fl ta due to RH’s bakriser being slightly imperfect, even though I could do this exact transition consistently by itself.

Used an entire afternoon and early evening (4 hours) to get to half-passable control, after I’d calmed down. Roughly 30 min to land successful take, maybe (hard to specify as there were many 3 or 4 drafts with worse exec that I really didn’t want to use). Roughly 400 attempts per take I suppose.

1:18-1:24 - 2p1h [fl penaround aerial - fl penaround]? Feels strange to not have feeling of mod on hand. Not as hard as you may expect after you know what feeling and movements you need.

Took 10-15 min to learn and get x 3, got the x 4 2-3 min afterwards, so 100-150 attempts.

1:25-1:29 - [ss - pun kan] x 3. After ction's WT13 R1 [palmspin - pun kan], I suggested doing it with ss instead with sekai, who filmed a nice x 2 of it (below). Very hard to control, mastering ss variations and pun kan helps only a little bit as it’s based more on eye/pen/hand co-ordination level. Landed x 3 three times, used most controlled (comparatively) one.

Roughly 10 min each, so 120-150 attempts.

1:30-1:39 - fl combo with LH ‘assist’: RH [mirr pp rev - ring ss] - LH spiderspin rev - RH [fl ia ss - mirr pp rev ring ss] inspired by conversation I had with ohzers (old UPSB member) back in WC12 times, ta rel - wrist bounce - fl ia ss, and ction’s WT13 R5 finisher 2 arm pun kan idea. Quite happy with how this one turned out.

15-20 min per successful take, so 200-250 attempts.

1:40-1:45 - octuple pd ta, upgrade of 9 year solo’s 2:01-2:05 septuple pd ta with better control. Hard to teach people to do this trick.

8 min, so 80-120 attempts.

1:45-1:49 - 2p2h RH [fl ta - ss] x 4 LH [fl ta - palmspin] x 4, failed final one on either hand half a dozen times or so. Had to focus on left while keeping right on autopilot, but left does palmspins slightly slower than right’s spiderspins, so it felt strange.

15-20 min, so 200-250 attempts.

1:50-1:59 - fc fl combo with crossed index+middle fl indexmiddlearound - ring ss, technique needs improvement to reduce extent of ‘cheat’ aerial used. Very very hard to do as pen keeps getting stuck after fl around (far harder than fl ma - ring ss or fl ra - pinky ss). This was far harder than fc fl combo in 9 year solo which did not actually use the crossed fingers for spinning.

Practised this for hour practising before day I filmed this clip, so roughly 100-200 attempts.

2:00-2:07 - [fl ma ring ss 1.0 - fl ma ring ss 0.5] x 2 combo, adding 0.5 rotations to ring ss in fl ma ring ss is ridiculously hard. Difficulty jump for fl ma ring ss 1.0 vs fl ma ring ss 0.5 is bigger than fl ia ss 1.0 vs fl ia ss 0.5 (IMO bigger than jump between fl ia ss 1.5 and fl ia ss 0.5 even)

This clip as hard as [fl ma - ring ss] x 20-25 for me. Took 10 min ish, so 100-150 attempts.

2:07-2:16 - pun kan x 13 clip from before, probably one of hardest tricks to develop technique for.

2:17-2:20 - plush toys of Reimu and Marisa from Touhou project.