請選擇 進入手機版 | 繼續訪問電腦版

Cubing-TW.Net|台灣魔術方塊社群|

 找回密碼
 立即註冊
搜索
熱搜: 3x3x3 WCA 小丸號
查看: 7126|回復: 5

[SQ-1] Square-1的進階技巧與高級解法 [複製鏈接]

Rank: 3Rank: 3

發表於 2015-4-19 14:18:53 |顯示全部樓層
本帖最後由 jimmy49684 於 2017-5-2 21:41 編輯

嗨嗨我是短Q,Square-1現任台灣單次及平均NR。

單次:
Video ID:cdOFE2xw4qo
(last edited in 5/2/17)

平均:
Video ID:3Q002w7sF9c
(last edited in 5/2/17)

這篇文章主要是針對已經學會如何復原整顆Square-1,但卻不知道如何進步的人,給予步驟上的建議或指導。
另外我也會附錄高級解法,引進的目的是希望有朝一日各位玩家們可以練習到爐火純青的境界。

如果你是完全不會解Square-1卻不小心點進來的新手,可以參閱郭大的Square-1解法教學,我也是從這裡一步一腳印的練習才有今日的成果。

依照基礎解法,大概練習兩、三個月到半年左右可以達到sub40,看個人練習狀況,若無法達到這種秒數代表練習量不夠,或許對公式還不熟悉,觀察也不夠快。一般來說,復原時間在60秒至40秒左右的玩家,主要的問題在於手速的不足以及觀察的能力還不太夠,另外也很有可能是你的Square-1方塊有問題,如果方塊很緊的話建議調一下螺絲鬆緊,要不然就換一顆,因為你可能年輕的時候買錯顆了。

市面上最佳的魔方格SQ1:
Video ID:IGMMJV75qH0
(last edited in 3/10)

會解Square-1的玩家一定知道,Square-1一共分成以下步驟:

1. Cubeshape (回正 or 復形)
2. CO (角方向)
3. EO (邊方向)
4. CP (角位置)
5. EP (邊位置)
6. Middle Layer (中間層)



關於以上步驟最詳盡的網站首推Lars Vandenbergh's CubeZone,建議學習步驟是先學習EO,再背CP,因為這兩個步驟都是公式,也都比較簡單容易記憶。接著大概在sub30左右可以開始學習Cubeshape的最佳解,最後才是精進EP,因為這是cost & performence最低的地方。

接下來就針對每一個步驟做深入說明以及個人經驗分享。

Video ID:Gkvl1xQhTu0


1. Cubeshape (回正 or 復形)
A 學會拼六星的技巧和六星的五條公式
B 有一定的複形經驗,已經不用每次都拼六星了
C 系統學習過蜘蛛網,看到任何情況可以知道複形全程的動作是怎樣的,但不一定是最優的路線
D 複形路線的優化
E 個別情況多方向
F 個別情況從複形的最後兩三個「/」裡觀察出角歸層的情況(特殊例子)
G 能從打亂的狀態可以看出複形後的角歸層是什麼情況(要將所有情況完全的公式化,而且列出每個情況的複形前後的角塊位置變化,似乎是人類不可能達到的境界,而且短時間的的觀察沒可能實現)
H 甚至能看出角歸層之後棱歸層的情況(這個當我沒說過)
I 能從打亂的狀態看到最後棱順序的單雙數(是單數還是雙數,怎麼看,完全沒想通,單雙數後面說),從而選擇更優的一條複形路線(這個也當我沒說過)

關於復形的唯一指名網站:
由還原步數分類: SQ1 shape by depth
由形狀種類分類: SQ1 shape by type



Advanced Cubeshape (optimal)
這邊的學習有點類似F2L,並不能拘泥於公式的解法,而是理解公式背後的原理,Square-1因為會變形,所以可以記住各種形狀之間的關係以及變化,更有利於學習最佳化的回正解法。
一開始接觸請先從兩步或三步復原開始辨認,一邊記憶圖形一邊理解每個步驟的下一步會遇到的情況。

Video ID:pAr0b3ffI1k


Video ID:emeODPxL0Vk


Singular Orientation
當技術比較純熟一點時,可以強迫自己復形時,中層一定左邊拿小塊,如此一來就不用換面,減少復原時間和regrip的不方便性。

Parity Cubeshape
簡單來說,就是在觀察時期就判斷出最終EP時會不會遇到parity的情況,詳細說明請參考Parity Cubeshape
另外觀察和編碼方式可以參考Square-1 Blindfold,這個建議看看就好,有興趣的玩家可以點進去讓大腦刺激一下,真正在速解上的CP值很低,目前不流行這樣的解法,但不確定未來有沒有可能成為主流。

2. CO (角方向)

A 踏踏實實地學會五種情況對應的公式(會二階色先可直接跳過)
B 出現Square-1整體解法的分歧,a不管什麼情況直接做公式b每次都把頂層顏色的角放到上面
C 個別情況從複形的最後兩三個「/」裡觀察出角歸層的情況,不用觀察下層甚至從複形的最後兩三個「/」裡看出角歸層的情況(跳過複形後的觀察)
D 從角歸層最後一「/」大概猜到棱歸層的情況
E 角歸層時同時完成棱歸層(當我沒說過)

Square-1的CO就跟2x2的Ortega有點類似,將頂層和底層的顏色弄成一樣,但Square-1又更簡單一點因為它不用調整方向。
這邊的建議只有一個,將頂層顏色湊到頂層,底層顏色就堆到底層,如此一來就不用在最後中心層調整時煩惱。
雖然我知道大貓和李炳良好像是Color Neutral,也就是哪邊顏色簡單就湊哪邊,但我個人習慣頂層和底層分開。

AO (dealing with CO & EO simotaneously)
這個想法我跟大貓聊過,也丟到Speedsolving論壇發問過,實際情況是學習的實用性不高,因為case非常多,很難瞬間判斷完又要執行公式。我有嘗試要學習一些簡單的case但我還沒時間ww。

AO 參考網址



3. EO (邊方向)棱歸層XD

A 踏踏實實地學會七種情況對應的公式
B 對每個情況選出最適合自己的公式(b的有一條公式不能用Lars的)
C 某些情況可以多方向處理
D 某些情況可以多公式處理(有些公式會做上下面顏色調換,反正把上層顏色放回上層理論上是不會虧本的)
E 重點了,能從棱歸層觀察的時候直接看出角順序是什麼情況(這個正常人都能練出來的,關鍵是總結出每個棱歸層公式對角塊位置的影響;此步乃角順序的輔助步驟,不必強求每次都將角順序預判得一清二楚,別撿了芝麻丟了西瓜)

秒數大約在30~40秒就可以開始學習Full EO,一開始可以挑順手的公式學習,但有興趣的玩家可以一開始就學習不動角公式。

不動角EO(預判CP)
Square-1 角位置预判教程

所謂的不動角公式,顧名思義就是做完公式後角位置不產生變化的公式。不動角公式的應用可以使預判變得更為輕鬆,減少停頓時間,加強還原的連貫性具有很重要的作用!不動角公式比一般公式長一點,但實際應用中卻能保持還原的連續性,算是一種犧牲步數獲得速度的方式!



如果功力夠的人可以選擇再多背動角的EO,可以增加選擇CP的彈性,甚至因此跳CP。

4. CP (角位置)角順序wtf

A 踏踏實實地學會八種情況對應的公式
B 個別情況多方向處理
C 對棱順序有一定預判,只有一層角需要調整角順序的能觀察到不用調整那層情況
D 同一情況多公式處理(也是上下層顏色問題)
E 做角順序是能預判出棱順序的情況(當我沒說過)
F 用PLL同時做好角和棱(當我沒說過)

As you (should) know, all the standard CP algorithms are basically ortega PBL's. They use the square-1 as a turn-limited 2x2.



(預判EP)
parity CP

Video ID:y8KOpnjC6hY


5. EP (邊位置)

A 用最基本的幾條公式疊加出所有情況
B 學會十來條基本的公式,看到每種情況都能在做之前預先想到怎樣疊加
C 觀察棱順序的同時觀察棱順序後的調整並且在公式的最後一步將調整加進去
D 部分公式的多方向處理
E 學會所有情況對應的公式,部分情況可以將疊加公式化、手法化


Z cases are removed because of hard recognition

EP one look
建議7個EP
手法或背法
相似的圖案放一組

EP相同手法的公式or依照類型分組

Inspection of Cubeshape

U+D <= 180 degree in total

From CO to EO

Choose Your EO (Knowing how they affect the corners)
這個技巧有點tricky,需要大量經驗才有辦法體會。當你Square-1玩久了,CO會遇到哪種EO大概都瞭若指掌,因此瞬間依據CO的狀況選擇CO的解法。例如有時候CO結束的EO會遇到換三邊或換單邊,如果中層沒有flip的時候,我可能會在在做CO時選擇我EO要做換三邊;反之,如果中層flip時,我可能會選擇強制遇到換單邊的EO,然後在EO調整我的中層。

From EO to CP

(預判CP)無停頓銜接
EO和CP一起處理

From CP to EP
Preserving block to force good ep(預判EP)lookahead

From EP to Middle Layer

Adjustment of Middle Layer

33 regular
EP公式最後的AUF

The difficulty of the square-1: fingertrick, cubeshape, parity
Fingertrick

SQ的公式符号解释与基本手指分工
SQ手法心得(修改版)

Polish Style:

Western Style:

給example

秒數分段:breakdown
Breakdown Suggestion

simon 2, 1.5, 2.5, 2.5, 3.5 roughly.

before EP 7.5~8.0
EP 4.5~5.0

cubeshape approximate 2.0

    For those who might curious about my learning process, you learn how I get fast and what I've learnt between every official average PBs.

    I started to do Square-1 because there was a competition called Taiwan Summer Open 2012 which held Square-1. I thought it was cool and began to learn those algs, after a couple of months' practice, I average around a minute and then competed in the competition. It became more interesting once I went further solving a Square-1, kinda like "Oh, somehow I can solve a Square-1 fast even though I didn't practice a lot".

    After half a year, I've learnt full EO and CP, also advenced cubeshape as well, but most important above all, I learned four EPs which were "3+,2;3-,2;2,3+;2,3-". Since I've averaged 20s, I held a Square-1 competition myself, and I approved my PBs to 18.30s single and 20.67s average, which was both top 100 at that time.

    The road from 20.67 to 16.50 was a dramatic story. I just did little practice before the competition and didn't improve that much, but somehow I got lucky from the first two solves. I might have gotten the NR, but 大貓 set a new record at 16.44 and beat me by a small margin.    It was a long journey from 16.50 to 14.90, though I still keep on practicung but actually I didn't learn new thing about Square-1. After Taiwan Summer 2014 and Cross-strait 2014, I gained some experience from former average NR 大貓, who gave me suggestion instead of algorithms. It really works after some pratice, so I decided to hold another Square-1 competition, Master Challenge 2014. Due to reletively easy scrambles, I finally got the average NR and won the first prize, cheers!

    I thought it was the end of the adventure because I didn't even expect that excellent official ranking initially. As you may know, I'm 13.94 average right now. The difference is the new hardware I bought from 阿良 after Master Challenge 2014. After the long usage of my former Cubetwist Square-1, it couldn't meet my expectation for solving, so I decided to change my main to Calvin's  Square-1. The new cube just astonished me in terms of its fluency, and gave me a lot room for improvements. When I was in Kaohsiung Open 2015, I had the confidence that I can broke the NR, then I achieve the accomplishment.

    Despite of sub14 official average, I still wasn't safisfied. I was still seeking for the best I can do, striving for greatness and persuing perfection. I started to practice through Plustimer, a scramble program authorized by the WCA offical scramble generator - TNoodle0.9.0. I also downloaded every scrambles that was used from official WCA competitions, in order to be used to TNoodle-generated-scrambles. Recently, I tried to solve a Square-1more consitently and precisely, and it worked since the times were decreasing. At last, I finally wanted to learn some EP algorithms which I named them "wtf EPs", you know what I mean from the literal meaning.





附件: 你需要登錄才可以下載或查看附件。沒有帳號?立即註冊

Rank: 6Rank: 6

發表於 2015-4-19 22:49:32 |顯示全部樓層
未看先推
短Q短Q愛尻尻

Rank: 8Rank: 8

發表於 2015-4-20 12:21:58 |顯示全部樓層
Looks like I've improved faster than you in the beginning haha
Anyway, really thanks for your advice and algs, thus I can get around 3x now.
Also big thank to DaMao, the tutorial he made for how to learn solve the cubeshape is just incredibly awesome!

Rank: 1

發表於 2015-4-22 15:14:56 |顯示全部樓層
看完發現我還在A級幼幼班

Rank: 3Rank: 3

發表於 2015-4-28 22:46:21 |顯示全部樓層
....................................

Rank: 8Rank: 8

發表於 2015-5-4 14:39:08 |顯示全部樓層
好大一隻貓

申請友鏈|Archiver|手機版|Cubing-TW.Net|台灣魔術方塊社群|

GMT+8, 2017-10-24 06:27 , Processed in 0.034354 second(s), 13 queries .

Powered by Discuz! X2

© 2001-2011 Comsenz Inc.

回頂部