My week in numbers (was "Tiny steps towards shodan")

[RBerenguel]

Absolutely, here it is! I stopped recording when I realised I had some wrong order on the middle-left area... There was a misplaced stone that made any subsequent plays there pointless :S So the endgame is missing, which may be an interesting part, specially in this game.

[RBerenguel]

16:42 Namii: I suggest watching Japanese games from 80s to 90s
16:42 Namii: Try to check at least 500 games before Summer
[...]
16:42 Namii: so you have 3 months
16:42 Namii: Then solve tsumego for 2 hours each week, feel free to distribute it as you like
[...]
16:42 Namii: so it becomes 8 hrs per month, 24 hrs by Summer
16:42 Namii: Then play at least 10 games (in addition to NGA games) per month
16:43 Namii: so it comes to 30 games by Summer, + NGA games
16:43 Namii: so I guess around 40-50 games in total
16:43 RBerenguel: blitz counts as games?
16:43 Namii: Ah, blitz is 0.5 games
16:43 RBerenguel: hehe ok
16:43 Namii: Hmm, I guess 500 pro games is a lot
16:43 Namii: Go for 100 instead
[...]
16:43 Namii: So about 1 per day
16:43 RBerenguel: aha
16:44 Namii: Oh, 100 per month I meant, so 300 before Summer
16:44 Namii: Summer begins on June 1
[...]
16:44 Namii: So if you keep this ~5-7 pro games / week, 2 hours of tsumego / week, 2-3 games / week (+ NGA games)
16:45 Namii: then I will call you Iron Man Ruben for the entire June+July
16:45 RBerenguel:
16:45 RBerenguel: challenge accepted

So Namii put forward this "challenge." Now I re-read it, I'm unsure about the number of pro games, so I'll aim for ~ 300 (3 per day) and in any case I'll do way more than 100 by summer. As for tsumego, I'm splitting it into 15' chunks (regardless of amount of problems done in this time.) For pro games I take around 10-15 minutes to go through one. I look for interesting patterns, like josekis I haven't seen (many,) middle game "joseki-like" I've seen rarely, endgame tesujis, direction of play, tenukis. This kind of things. As for games, I'll mostly play blitz, since it's a particularly blind spot for me, and it's easier to fit in a busy schedule (10 games per month is 20 blitz games per month, which is 0.6 games per day)

Week in review (since Wednesday 26th Feb, when I started logging it, Sunday not counted):

• 1h15' of tesuji problems from ishi no renkaku tore ningu nihiyakunanajiyuu maikomi igo bunko shiri zu, a cutting-and-connecting book with 270 tesujis. From now on, 270-cut. I'm on problem 79 (I'm taking ~1 minute per problem so far, it seems.)
• 30' of "endgame" problems from igo totsukun go kakeru go godouban jiyoutatsuhou ("How to improve on a 5x5 board", from now on 5x5.) I'm on problem 16, taking almost 2 minutes per problem.
• 4 blitz games (that's a +2 to game count,) lost all. 3 giving 4-5 stones, one even. From the handicap games I got a clear improvement in playing calmer moves and waiting for the opponent mistake (lost the first 2 by resignation, the third one by 1.5 points.) From the even game, to don't play stupid moves in a joseki.

Checked the following pro games (one-line annotation and reference from SmartGo Kifu included):

• Ishi Kunio - Ishida Yoshio 1980-00-00a: Middle game "joseki" for sabaki
• Kobayashi Koichi - Takemiya Masaki 1980-01-10f: Nothing specially interesting
• Cho Chikun - Rin Kaiho 1980-01-10b Big post-joseki fight and resignation
• Hashimoto Shoji - Ishida Yoshio 1980-01-10a Miskill/misread at the end (?), resignation
• Kobayashi Koichi - Cho Chikun 1980-01-11b How to neutralise centre thickness
• Rin Kaiho - Fujisawa Hideyuki 1980-01-16a: Mutual settling fight followed by attack without territory
• Shiraishi Yutaka - Fujisawa Hosai 1980-01-17b: Counter-example of getting 3rd line territory in exchange of a wall (Fujisawa did, won).
• Rin Kaiho - Kudo Norio 1980-01-24b: Rin lost a huge dragon misreading a ko threat, resigned
• Sakata Eio - Shiraishi Yutaka 1980-01-24e: A territorial game in fuseki, followed by reduction.

Forgot to say, I'm picking Japanese games starting from 1980 where at least one player is 9d and the other is at least 7d. Since there are many games, I'd rather start with the top pros, if I ever ran out of games I can see all others.

Code: Select all

|    | ∆   | =  |
| TT | +2h | 2h |
| GC | +2  | 2  |
| PG | +9  | 9  |

Since I started the counts in February, it's hard to tell month-wise if I'm on schedule or not. I have used 4 days (W,T,F,S,) so:

• 2h of tsumego
• Count of 2 games (0.5 games per day)
• 9 pro games (2.25 per day) (technically almost on schedule, but 2 of these pro games were technically checked "today," since I went to bed late...)

Let's see how the following week goes!

[RBerenguel]

Two problems from igo totsukun go kakeru go godouban jiyoutatsuhou. They are not hard ( the second has a minor trick, but most of this section aren't specially hard... but there are 3 sections) but it shows the kind of things to keep in mind when solving these small board situations. It's a great book, I'm very glad I got it.

$$B Black to win
$$ +-----------+
$$ | . . . . O |
$$ | O O O O O |
$$ | X X X X O |
$$ | . . . . . |
$$ | . . X . X |
$$ +-----------+

Click Here To Show Diagram Code

[go]$$B Black to win
$$ +-----------+
$$ | . . . . O |
$$ | O O O O O |
$$ | X X X X O |
$$ | . . . . . |
$$ | . . X . X |
$$ +-----------+[/go]

$$B Black to win
$$ +-----------+
$$ | . O . . . |
$$ | . . . O . |
$$ | O O O X X |
$$ | X X O X . |
$$ | . X X . X |
$$ +-----------+

Click Here To Show Diagram Code

[go]$$B Black to win
$$ +-----------+
$$ | . O . . . |
$$ | . . . O . |
$$ | O O O X X |
$$ | X X O X . |
$$ | . X X . X |
$$ +-----------+[/go]

[RBerenguel]

This was hard to read

$$B Black to win
$$ +-----------+
$$ | . . . . . |
$$ | . O X O O |
$$ | O X O O . |
$$ | . X X X X |
$$ | . . . . . |
$$ +-----------+

Click Here To Show Diagram Code

[go]$$B Black to win
$$ +-----------+
$$ | . . . . . |
$$ | . O X O O |
$$ | O X O O . |
$$ | . X X X X |
$$ | . . . . . |
$$ +-----------+[/go]

[RBerenguel]

Refer to this in case you don't know what this is about.

• 30' on the 270 book
• 45' on the 5x5 book
• 45' on the NGA assignment (with re-solving)

2h total.

Checked the following pro games

• Cho Chikun - Hashimoto Utaro 01-24d: Interesting fights all over
• Kato Masao - Takemiya Masaki 01-24c: Fights and attacks by Kato
• (9x9) Go Seigen - Miyamoto Naoki 68-8
• (9x9) Miyamoto Naoki - Go Seigen 68-9
• Kataoka Satoshi - Fujisawa Kazunari 2014-02-27 (Judan prelim): White had two weak groups and won anyway
• Fujisawa Hideyuki - Rin Kaiho 01-30a: Weird game
• Kobayashi Koichi - Sakata Eio 01-31a: Heavy fighting game with many struggling groups
• Kajiwara Takeo - Ushikubo Yoshitaka 01-31b: Huge loss in several trades
• Takemiya Masaki - Hane Yasumasa 01-31c: Very interesting game with several neat moves
• Rin Kaiho - Fujisawa Hideyuki 02-05a: Huge side vs. rest. Interesting endgame play
• Tajiri Yuto - Ida Atsushi 2014-01-09 (Chisato cup): Seemed passive by black, who lost
• Kudo Norio - Sakata Eio 02-07a: Nothing remarkable
• (9x9) Ishida Atsushi 4d - Kurahashi Masayuki 1d 1987-10-10a
• O Rissei (5d) - Sakata Eio (9d) 02-13c
• Kudo Norio - Otake Hideo 02-13b: Interesting shimari invasion against a hoshi+shimari
• Fujisawa Hosai - Takemiya Masaki 02-14d: Takemiya can be a real killer
• Ishii Shinzo - Sato Masaharu 02-14c: Weird cap by black, also got a carpenter square on the board
• Kato Masao - Kobayashi Koichi 02-14b: Again, a cap I didn't see right. Slightly less weird
• Takao Shinji - Yuki Satoshi 2014-3-4 (Judan 1): Not specially violent
• Hane Yasumasa - Rin Kaiho 02-14a: Pincer-jump-jump-attach-extend-pincer joseki
• Fujisawa Hideyuki - Rin Kaiho 02-20a: Group dies, Rin loses
• Takemiya Masaki - Hane Yasumasa 02-21d: Mistake by Hane in an endgame hane XD
• Kishimoto Kazuo - Hashimoto Yoshimi 02-11c: Tenuki in joseki to play enclosure
• Haruyama Isamu - Hisai Keishi 02-21b: Weird side-corner invasion
• Kobayashi Koichi - Ishida Yoshio 02-21a: Interesting pincer-jump. Koba killed a group to get life at the end, +R.
• Otake Hideo - Ishida Yoshio 02-24a: Interesting hoshi-side invasion
• Cho Chikun - Hashimoto Shoji 02-25b: Cho playing 3-3 a lot lately
• Miyazaki Ryutaro - Rin Kaiho 2014-3-6 (Meijin Pre) White kept the pressure, on endgame a B group died. W was ahead anyway
• Sato Sunao - Tsuchida Masamitsu 02-27b: Didn't understand anything
• Cho Chikun - Hashimoto Shoji 02-25b: Nothing remarkable
• Rin Kaiho - Fujisawa Hideyuki: Heavy fighting

In parentheses, 9x9 pro games I replayed.

Code: Select all

|    | ∆        | =      |
| TT | +2h      | 4h     |
| GC | 0        | 2      |
| PG | +28 (+3) | 37 (3) |

• 2h of tsumego
• Count of 2 games (0.5 games per day) :tmbdown:
• 28 pro games (4 per day)

[RBerenguel]

Refer to this in case you don't know what this is about.

Played 6 blitz games. I still see blitz as too random: some of the games I won I was ~40 points behind (essentially because right now I suck at blitz, in the sense of making good moves) but the opponent timed out.

• 4 even (3 wins, 1 loss)
• 2 handicap (getting 3 & 4, won 1, lost 1)

Tesuji/tsumego:

• 15' on the 270 book
• 45' on Segoe's tsumego dictionary
• 15' on the 5x5 book

1h15 total.

Checked the following pro games. I asked Namii and he told me 5-7 per week were enough, so this week I cut it.

• Otake Hideo - Awaji Shuzo 02-28d: Many pincers, failed carpenter square, group died
• Kuboichi Shuchi - Hayase Hiroshi 02-28c: Nerve-wracking fight and semeai
• Sakata Eio . Shiraishi Yutaka 02-28b: Pincer joseki
• Cho Chikun - Takemiya Masaki 85?-01-30a: Impressive kill & use of center thickness. Saw fuseki comments in J. Hop's Youtube
• Kobayashi Koichi - Cho Chikun 02-28a
• Otake Hideo - Kato Masao 03-05a
• Yamashita Keigo - Iyama Yuta (6th game Kisei Final 2014)

7 games in total

Code: Select all

|    | ∆        | =      |
| TT | +1h15    | 5h15   |
| GC | 3        | 5      |
| PG | +7       | 44 (3) |

• 2h of tsumego 1h15
• Count of 2 games (~0.5 games per day) 3 games
• 5-7 pro games 7

[RBerenguel]

Refer to this in case you don't know what this is about.

Played 2 blitz games. Both giving 2 stones, won 1 on time and lost the other by 3.5
Played 1 simultaneous game with Su Yang (JeffChang aka Finnish8d, one of the NGA teachers)

• 2 handicap (giving 2, 1w 1l)
• 1 handicap simul with Jeff, getting 6 lost big

Tesuji/tsumego:

• 30' on the 270 book
• 15' on Segoe's tsumego dictionary
• 30' on the 5x5 book
• 15' Gokyo Shumyo
• 30' on several books/files (commute, usually)
• 15' (probably more) reading game situations in one of the watched games this week

2h15 total.

Checked the following pro games.

• Honda Kunihisa - Kawamoto Noboru 03-06e Close, hard-fought game
• Rin Kaiho - Shiraishi Yutaka 03-06b
• Antti Törmannen - Fredrik Blomback (Pandanet Team Tournament, I read things here)
• Hsu Chiayuan - Iyama Yuta 2014-03-22 (1st Japanese Championship Winners)

4 games in total

Code: Select all

|    | ∆        | =      |
| TT | +2h15    | 7h30   |
| GC | 2        | 7      |
| PG | +4       | 48 (3) |

• 2h of tsumego 2h15
• Count of 2 games (~0.5 games per day) 2 games
• 5-7 pro games 4

The count for March proper is:

• 5h30 of reading practice
• 5 games
• 39 pro games (+3 9x9 pro games)

In the coming week (granted, Monday of the week after is still March) I need to add 5 games, 2h30 of tsumego practice and no pro games (I did too many at the beginning.)

[RBerenguel]

Refer to this in case you don't know what this is about.

Played 10 blitz games. 3 getting 3 stones (won two), 2 giving (2 and 3 resp, won the first lost the second) and 5 even (won all 5.) So this week I had a 80% WR. Managed to make (barely but surely) 3k on my blitz account.

• 10 blitz

Tesuji/tsumego:

• 30' on NGA assignment
• 15' on GGPFDP-Joseki1-3
• 15' on GGPFDP-OMG1-7
• 1h on 270
• 15' Segoe

2h15 total.

Checked the following pro games.

• Yamashita Keigo - Iyama Yuta (1st Japanese Championship of Tournament Winners)
• Kobayashi Koichi - Choi Chikun 03-06a
• Takemiya Masaki - Hashimoto Utaro 03-06g
• Sakata Eio - Ishida Yoshio 03-08b

Also watched 0.5 of a video by Dwyrin on how to review pro games.

4 games in total

Code: Select all

|    | ∆        | =      |
| TT | +2h15    | 9h45   |
| GC | 5        | 12      |
| PG | +4       | 52 (3) |

• 2h of tsumego 2h15
• Count of 2 games (~0.5 games per day) 2 games
• 5-7 pro games 4

The count for March proper is:

• 7h45 of reading practice
• 10 games
• 43 pro games (+3 9x9 pro games)

So, as long as I do 15 minutes of tsumego today or tomorrow, I'll be on track. I think it's doable, specially since I have to deliver the tsumego assignment on Thursday and I'm still missing a couple.

[RBerenguel]

Refer to this in case you don't know what this is about.

Played 4 blitz games, now as 3k. Ended 50/50, one game won giving 4 stones (to a 7k,) another won against a 1d (getting 3 stones,) another lost to a 1d (getting 3) and another lost to a 2d (getting 4.) I wanted to play 2 more on Friday, but couldn't

• 4 blitz

Tesuji/tsumego:

• 45' on NGA assignment
• 30' on 270
• 45' on GGPFDP-J1-3

2h total.

I wanted to do more than 2h this week, but due to some personal stuff I couldn't Let's hope this coming week, I want to compensate the bad week I had a on March.

Checked the following pro games (I didn't write the names of the 9x9 games with very legible handwriting, expect mistakes).

• (9x9) Katsuma Shori - Kori Toshio
• (9x9) Goto Shungo - Yakata Koichi
• (9x9) Yamanouchi Masaki - Ueki Yoshio
• (9x9) Yamada Kimio - Sumi Shinsuke
• (9x9) Yamada Kimio - Shigeno Yuki
• (9x9) Maeda Ryo - Minematsu Masaki
• (9x9) Saito Tadashi - Kitimoto Kazuo
• (9x9) Izumo Tetsuya - Araki Ishi
• (9x9) Yamada Noriyoshi - Ishida Atsushi
• (9x9) Yamada Noriyoshi - Enda Yoichi
• Kato Masao - Rin Kaiho 03-11a
• Kato Masao - Otake Hideo 03-12a
• Cho Chikun - Takemiya Masaki 03-13c

10 9x9, 3 19x19

4 games in total

Code: Select all

|    | ∆        | =       |
| TT | +2h      | 11h45   |
| GC | 5        | 12      |
| PG | +3 (10)  | 55 (13) |

• 2h of tsumego 2h15
• Count of 2 games (~0.5 games per day) 2 games
• 5-7 pro games 3, (10)

Finally, March is done. I'll just wiggle the numbers so I know what is left for "April" (since I did 1h of tsumego on the 30th March, but I'll reallocate 45' for April)

• 8h of reading practice
• 10 games
• 43 pro games (+3 9x9 pro games)

Thus, first week april is 1h45 so far.

[RBerenguel]

Refer to this in case you don't know what this is about.

Didn't play this week, was pretty hectic. Next week I'm on holidays, and I plan on (at least) doing more tsumego and checking pro games (probably a lot of 9x9 in a real board to explore variations at my leisure). I'll have to recover blitz play when I'm back to normal connectivity, next week I'll probably be mostly offline.

For next week's tsumego intensive "bootcamp" (I'll also work on finishing my PhD and a redesign for the Nordic Go Academy website, if I have the time) I'm bringing:

• Segoe-Seigen Tesuji Vol 1
• The book on 5x5 problems
• Fujisawa's Dictionary of Basic Tesuji Vol 1
• Graded Go Problems for Beginners Vol 3

I'll essentially try to solve GGP3 2 times (I have never done it in its entirety! Seriously! And I've had it for 8 years already!!,) the first time with an allocated time of max 1 minute per problem (at most would take me 7 hours) and the second run would be with half that, 30 seconds per problem. I'll mix in "theory" from the DBT1 and when I eventually get bored of so many tsumego, do some tesuji and 5x5 problems. Yeah, I know, it's quite a lot for this amount of time I have... But on my previous holidays (shorter!) I managed to solve 540 (Lee Chang'Ho tesuji and tsumego, several volumes, several times) without having much "off time" (I was visiting the International Shoe Museum most of the day, so I only had late afternoons for it.)

• 0 blitz, 0 normal games

Tesuji/tsumego:

• 1h on 5x5
• 1h on Dictionary of Basic Tesuji V1 (from now on, DBT1)
• 15' SST1 (Segoe-Seigen Tesuji Vol1)

2h15 total.

Formally, the DBT is not a tsumego book, but it has "problems" (although not seeing the answer can be tricky) and some reading is required anyway to understand most of it. So I'll count it here, if I didn't I'd end up not reading it ever.

Checked the following pro games:

• (4) The first 4 games of Book 3 of JF's The life and games of Honinbo Shuei
• Ishi Kunio - Chino Tadahiko 03-13b

5 19x19

4 games in total

Code: Select all

|    | ∆        | =       |
| TT | +2h15    | 14h00   |
| GC | 0        | 12      |
| PG | +5       | 60 (13) |

• 2h of tsumego 2h15
• Count of 2 games (~0.5 games per day) 0 games
• 5-7 pro games 5

April is relatively on track. Correct tsumego timing for now (this week I'll probably triple the numbers, though, so I can slack a little later if needed) and bad in games played. I'll try to catch up on the week after.

[RBerenguel]

I have changed the journal title, since lately I'm just dumping numbers. Here's this week's (which were holidays for me) data.

----

This week I was on holiday, and decided to train a little more my go reading. I wasn't sure about which book to bring, had a lot of candidates: 1001 life and death problems, Graded Go Problems for Beginners Vol 3, any amount of Lee Ch'ang Ho Life and Death or Tesuji books, Seigen-Segoe tesuji books, a couple Chinese tsumego books I have... Finally settled for Graded Go Problems for Beginners, Vol 3, which has 421 problems. I've had this book for 8 years, and never, ever gotten farther than page 20 or so (I got this and an opening book – In the beginning, I think – in exchange for my copy of The Endgame from a former go playing friend.) I re-bound the book, covering the ugly cover with a plain, dark grey paper and film. Now it's a sturdy, dull book. I hated that cover.

My original plan was to just do all the problems, twice, giving me 1 minute for the first time (at most) and 30 seconds for the second run. What I didn't consider is that tallying so many times is incredibly time consuming, a significant overhead to an already large undertaking (421 problems at 1 minute per problem can easily end up being close to 7 hours). In the end, I have just made one run of the book, But I logged more or less all times used, to gauge reading improvement in the near future. Here comes the time and % data.

For the first two batches (which comprise Sections 1 and 2) I logged the problem time, whether I was right or wrong and if I passed 60 seconds, the problem was marked as wrong without further reading. The times are cut to the full seconds figure unless the tens of second was higher than 90, where I rounded up for no special reason.

From these two first batches I analysed some data, splitting wrong problems (and also right ones) into very wrong and "just" wrong classes. This made "problematic reading check" afterwards somewhat easier.

First batch:

Covering problems 1-64 (I was tired and stopped short of a few problems to finish "real" Part 1, which ends on problem 86)

• Total problem solving time: 16 minutes, 17 seconds
• Average time per problem: 15 seconds
• Median time per problem: 13 seconds
• Mode time per problem: 15 seconds
• First quartile: 9 seconds
• Standard deviation per problem time: 10.25
• Percentage of correct problems: 82.8%

The problems where the time taken is larger than the average AND are wrong are one class of very wrong problems, the other class being problems where the time taken is lower than the first quartile and the problem is wrong.

The first type covers problems where my reading is totally mistaken: either I can't see the solution (so, time>=60 and directly wrong) or I convinced myself a wrong solution is actually right. In a game this results in bad time efficiency, since I'm taking more time than usual to just answer wrongly.

Problems of the other class include instant responses where not only my intuition, but my shape sense and "instant reading" are very wrong. This is time<=Q1 AND wrong. In a game setting this is essentially a brain fart: for life and death or tesuji I may not answer instantly in a game with long time settings, and re-read (probably catching the mistake). In a blitz game though, this may be a game losing move.

My reasoning is that improving in this set of very wrong problems (which is smaller than the set of wrong problems) will yield better results than checking all wrong problems and analysing why I was wrong: it can just be a minor hiccup in my reading.

• \>avg&wrong: 3 problems
• \<=Q1&wrong: 1 problem
• Percentage of very wrong problems: 6.25%
• Percentage of very wrong problems among wrong problems: 36.36%

Likewise, problems where time taken is less (or equal) than the first quartile and are right are considered "very right," since my intuition and reading where spot on and fast.

• \<=Q1&right: 16 problems
• Percentage of very right problems among right problems: 30.1%

Second batch:

Covering problems 65-169 (end of Section 1 and the whole Section 2)

• Total problem solving time: 36 minutes, 31 seconds
• Average time per problem: 20.8 seconds
• Median time per problem: 15 seconds
• Mode time per problem: 60 seconds
• First quartile: 9 seconds
• Standard deviation per problem time: 10.25
• Percentage of correct problems: 72.1%
• \>avg&wrong: 18 problems
• \<=Q1&wrong: 6 problems
• Percentage of very wrong problems: 23%
• Percentage of very wrong problems among wrong problems: 80%
• \<=Q1&right: 26 problems
• Percentage of very right problems among right problems: 34.6%

The fact that the mode is 60 seconds means I timed out in "many" problems. Actually the mode in this data sets is not that relevant anyway. I'm somewhat surprised about the difference between the percentages of very wrong within wrong problems in the two batches. Since the average time and percentage of wrong problems are worse in this second section, it seems the problems are on average somewhat harder (the two batches were done in different days with roughly the same day load, so the results should be independent of my own mental tiredness). I would expect very wrong problems to be roughly the same ratio. I think this hints at a problem when solutions are somewhat deep: I'm expecting solutions to have a certain depth, and I'm reading no further. If I can't see a solution at "that" depth I keep circling around wrong solutions until I time out or I miss a refutation of my shorter, wrong line which I then accept as right.

Third batch:

This one covered Section 3, which is "just" life and death problems (124 problems). For this third batch I was somewhat tired of recording times and just timed each page, each page (except the last, having 4) had 6 problems.

• Total problem solving time: 34 minutes
• Average time per page: 97 seconds
• Median time per page: 91 seconds
• Mode time per page 113 seconds
• First quartile: 72 seconds
• Standard deviation of time per page: 31.6
• Percentage of correct problems: 92.7

Another surprise here: this section seems far, far easier than the rest (or I was already reaping rewards from doing so many problems before.). I had long stretches of 6/6 pages and none of the problems I miss had any "stupid" surprises, I just plainly misread or miscounted. Worth noting, some problems from this section are already in the previous two sections.

Fourth (and final) batch

This one covered Section 4, another set of only life and death problems (128 problems.) Again, each page had 6 problems except for the last page, which had only 2. My reading was slightly less sharp (or so I felt, seeing some of the mistakes I did) than in the previous batch, but I caught several "repeated errors" from this section. Again, I counted per-page times. I wasn't specially motivated to finish this batch, I only did it because it may be the last chance during holidays. So, spent ~1h30 to finish it (adding the time needed to check answers, find the reading mistake if it wasn't in the "wrong" diagram and record the time taken per page.)

• Total problem solving time: 51 minutes
• Average time per page: 140 seconds
• Median time per page: 122.5 seconds
• Mode time per page: NA
• First quartile: 98 seconds
• Standard deviation of time per page: 62
• Percentage of correct problems: 71%

I had quite a few "shape katteyomi," where I saw a clear shape defect in the target group and falsely assumed it worked to kill. I also had a stupidly hard time with "easy" problems where the shape is almost completely settled and just a simple move is needed, the kind covered in 1001 LD problems. I think I need to tackle this book next.

Final results:

Total time for the book: 2h 18 minutes (give or take a couple minutes due to rounding mistakes)
Percentage of correct problems: 79.3 %



Clearly this book was not challenging "enough" (Bill Spight suggested in a previous comment in this log that the best is aiming for books with ~50% failure rate) but this revisiting of the basics is always good. I also managed to find some areas to improve and work on in problem solving, and now have an interesting data set to use as baseline of reading improvement.

[RBerenguel]

Refer to this in case you don't know what this is about.

I did a lot of problem solving this week (see previous post), but no games (well, 2 NGA games but these don't count) and no pro game reviewing either.

• 0 blitz, 0 normal games

Tesuji/tsumego:

• 1h15 on DBT1
• 2h15 on GGPFB1 (finished completely, 79% correct)

3h30 total.

Code: Select all

|    | ∆        | =       |
| TT | +3h30    | 17h30   |
| GC | 0        | 12      |
| PG |  0       | 60 (13) |

• 2h of tsumego 3h30
• Count of 2 games (~0.5 games per day) 0 games
• 5-7 pro games 0

Although I spent a lot of time on problem solving, a big chunk was tracking the times and analysing the numbers (and checking the correct answers) afterwards. So although I probably invested more like 7 or 8 hours on tsumego, it only shows as 3h30 of reading.

This week I need to do a lot of blitz to catch up. And I also need to finish the NGA tsumego assignment, so probably April will either end with less games played than I should, or with tons of extra tsumego done. Go figure...

[RBerenguel]

Refer to this in case you don't know what this is about.

Oh crap, I'll definitely miss the mark on games for this month.

• 2 blitz (2 losses :/), 0 normal games

Tesuji/tsumego:

• 1h45 on GSJ1
• 30' on Weiqi Life and Death Drills V1 (from now on WLDD1)
• 30' on a Chinese book I won in last year's NGA Summer Camp Tournament, it's upstairs... Will look the name later

As you see, I got (temporary) hold of that Chinese book. So far it's a pretty good one.

2h45 total.

Checked the following pro games:

• Fukui Masaaki - Ronald Schlemper (Masaaki is the author of that 5x5 problem book I like so much, so I made an exception) 03-18a
• Rin Kaiho-Cho Chikun 03-20b
• Sato Masaharu-Sakata Eio 03-20d
• Takagi Shoichi - Ohira Shuzo 03-20c
• Chino Tadahiko - Hane Yasumasa 03-27a
• Takemiya Masaki - Ishida Yoshio 03-27d
• Kobayashi Koichi - Fujisawa Hosai 03-27c

I also memorised (well, still unsure about how well I remember it completely) one of the 9x9 games between Go Seigen and Miyamoto Naoki. Working on the other one now.

Code: Select all

|    | ∆        | =       |
| TT | +2h45    | 20h15   |
| GC | 2        | 14      |
| PG |  7       | 67 (13) |

• 2h of tsumego 3h30
• Count of 2 games (~0.5 games per day) 0 games :tmbdown::tmbdown:
• 5-7 pro games 7

Severely trailing in playing :/

[RBerenguel]

Last weekend (and still today somewhat) I've tweaked the code of sgftopng (from the awesome sgfutils set of programs) to generate "neat" goban images for the Nordic Go Academy and "plain" game images for Antti's blog (still unsure if he will use them or not.)

The changes involve using real goban, stones and shadows instead of drawn graphics (for the "neat" one) and a plain black and white goban for the plain one. There's a minor randomisation of stone placement for the neat one, can be activated or deactivated at will. I also forced the font to be Inconsolata (originally Times New Roman.) For these examples I used jgoboard's image assets, but it could be made to work with any set of board and stone images. Since image size is more or less hardcoded in sgftopng's code, I followed the same method (I'm lazy and I just wanted it to work) and just generate very big images taking advantage of jgoboard's 1000px big board. There's probably some tweaking needed to make number placement better, but I stopped once it was good enough, because, well, good placement is a b***h.

I'll post more details and the modified source in due time.

Here is a sample of how a problem looks like (I posted this and a harder one to NGA's social media accounts, btw, started to take advantage of this )

Coordinates are optional, of course.

P1-temp.jpg (74.83 KiB) Viewed 14601 times

Here is a sample game to show some numberings in place (in hide tags since the image is generated very big on purpose.) Has a very small random stone placement active:

game.jpg (291.49 KiB) Viewed 14601 times

And here you can see the "plain" format (plain overrides random placement):

plain.jpg (195.21 KiB) Viewed 14601 times

[RBerenguel]

Refer to this in case you don't know what this is about.

Last week I missed reporting... Meaning I kind of lost track. I had slightly more than 2 hours, but since I forgot to record how much, let's say 2 hours. On that week I played 3 blitz, losing 2.

• 3 blitz (2 losses :/), 0 normal games

Namii told me that I seem to be making mistakes in easy problems but I am able to solve complex tsumego... And that I should work through all 900 problems in the Elementary level of Cho's Encyclopedia of Life and Death. To do so, I converted the TeX sources into 1-problem-per-page and used imagemagick to convert the PDF into a set of images for each page, imported this into Anki and moved it to my iPad. So instead of working through the book in force-of-will batches, I'm just using Anki to dose them slowly, while showing them with a certain regularity (remember that since they have no solutions, I have to decide whether I "know" the solution and how well do I know it.) An advantage is that Anki tracks my time. In the beginning of the week I worked on the first 120 or so before doing this Anki process, taking roughly 30 minutes. This and Anki will give the totals for the week.

Tesuji/tsumego:

• 2h (generic, for the week before last)
• 107'+30' = 137 min for Cho's encyclopedia (say, 2h15).

Checked the following pro games:

• Go Seigen against (I think) Liu Changhua, a game appearing in Go Seigen's collected games but not in SmartGo Kifu (at least fuseki search didn't match it)

Code: Select all

|    | ∆        | =       |
| TT | +4h15    | 22h30   |
| GC | 1.5      | 15.5    |
| PG |  1       | 68 (13) |

• 2h of tsumego 2h, 2h15
• Count of 2 games (~0.5 games per day) 1.5,0 games :tmbdown:
• 5-7 pro games 0,1

Severely trailing in playing :/ I'm still somewhat ahead in pro games due to the huge amount I did in February.