Carcassonne Central
Carc Central Community => Strategy Guide => Topic started by: DIN0 on December 31, 2021, 07:29:11 PM
-
Happy new year Carcassonne players!
Allow me to present to you the first of my Carcassonne projects: Game notation for Carcassonne!
As you know, one of the ways to play Carcassonne is a competitive game or a tournament where two players face off against each other using the base game only. Several other highly competitive games have notation systems dedicated to record the games in the form of a transcript. And now Carcassonne has one too! These can later be used to reproduce the specific game perfectly. This has various uses such as studying and analyzing past games to get better, develop strategies against specific players or using the data contained in these games for statistical analyses.
Even non-competitive players can find a use in the notation - the system can be used to set up various scenarios for puzzles or quizes, or example situations. Some might simply want to immortalize memorable games.
This game notation system is meant to serve as a powerfull tool for the community. It can either be written in any text editor or on a physical paper.
The project consists of two parts, both of which can be found in the download section of the forum, 1) pdf file of game notation manual where the system is explained in detail, 2) excel file with a pre-written notation sheet.
The manual has many examples to help you grasp the system and also includes several transcripts of real games, last of which is a final of Slovak Championship 2021.
That concludes the introduction of project 1A. As a heads up for the future, project 1B will be an extension of this notation for the whole game - this however will not be released any time soon.
I hope you will enjoy the game notation and will try it for yourself! Details in the manual. You are welcome to ask any questions in this topic :(y)
Link to download page: https://www.carcassonnecentral.com/community/index.php?action=downloads;cat=1
Other members have contributed toward making the notation accessible to more people. Please check carlium's amazing game notation app: https://carliunterik.github.io/cgn/
ERRATA:
page 14: The second to last sentence of the third paragraph reads "comma", but should be "colon". The intent behind it is still readable based on examples and the symbol usage summary.
page 29: The tile list is missing two tiles cfff and frrr.
page 17: Figure 13 e, f, and g have their y coordinate shifted 1 tile.
page 25: In the trancript, turn 09. has a missing colon between tile orientation and code. Also coordinates for the turns 67. should be [-4;-1] and for turn 71. [-5;-1].
-
It’s a beautiful and impressive document. Looking forward to trying out the system. +1 merit from me.
-
Wow, ;D you have put an enourmous amount of work into this: describing a whole Carcassone game by finding a formula for player identity / tile placement coordinates / tile identity and orientation / meeple placement & point gain. Impressive !! :(y)
This is a great tool for the community +1 merit from me!
Congratulations!
Now I fear that people will start using things like B:[0;-1]N:crrf>N:K=+4B in their posts.... :o and I was already overwhelmed by CCCF and RRCC.... ;)
-
Great document!
It seems like yesterday when you started discussing this impressive notation back in July 2019... ;D
https://www.carcassonnecentral.com/community/index.php?topic=4350.0 (https://www.carcassonnecentral.com/community/index.php?topic=4350.0)
I'm intrigued with the outcome of Project 1B, since tiles can have weird configurations...
+1 merit from me.
-
I'm curious how it will deal with fan expansions.
Odoslané z SM-A202F pomocou Tapatalku
-
It’s a beautiful and impressive document.
Thanky you :) I hope you'll enjoy it!
Now I fear that people will start using things like B:[0;-1]N:crrf>N:K=+4B in their posts.... :o and I was already overwhelmed by CCCF and RRCC.
I know it might seem intimidating at first, but belive me, once you get into it, it becomes almost natural and has a nice flow to it ;D And yes, what you describe was partialy the goal, but consider this: It is way more consise to post a few lines of text or attach a single text document then to include loads of pictures with extra commentary and/or markings within the pictures themselves.
Great document! It seems like yesterday when you started discussing this impressive notation back in July 2019...
Thank you very much! Yes, you're right, that was the first conception of the system. It just took this long to get to writing the manual, because I have been preoccupied by projects 1B and even more so project 2.
I'm curious how it will deal with fan expansions.
This one only deals with the base game. Project 1B, which is the extended version of tournamnet notation deals with all of the Carcassonne material. This will be released in the more distant future. In fact project 2 will be released sooner. As to the fan expansions, it is not designed to support them (yet), but it is robust enough to handle even non-existing tile configurations, so if the fan expansion is just a bunch of crazy configurations, it can be included in the General notation (the name of project 1B).
-
This is great!
It shows a well thought out system, and effort put in to make it something useful for many players.
+1 merit from me for making this! It's things like this that boost the forum and lets people learn a new way if they want to
-
Thank you Willem, that was the intent - something for the community as a whole :(y)
I have been notified (thank you kothmann) of a typo in the manual on page 14: The second to last sentence of the third paragraph reads "comma", but should be "colon". The intent behind it is still readable based on examples and the symbol usage summary, but I will fix this soon.
-
Everyone is free to post their own games or usages of the system :)
In the meantime, here is something from me. I have been analyzing the highest possible score/s one can have in the base game. Of course I am not the first one to do it, but I have based my analylis on the highest scoring playing field possible which maximizes all the features. The picture is in the attachment (I am not the original creator, nor the owner).
Here is what I have found...
Total points for completed features:
FIELDS: 126
CITIES: 116
CLOISTERS: 54
ROADS: 52 (46*)
LEFTOVERS: 11
MAXIMUM FINISHED SCORE: 348 – includes three +2 roads : possible with Phantom (in which case +3 field = 351)
MAXIMUM POSSIBLE SCORE: 342 – *because not enough meeple placements, and 1 meeple per tile only
GAME POINT CONTENT: 359
GAME POINT POTENIAL: 360
The maximum score one player can realistically achieve in a tournamnet style game is 342 points! There is enough completed features to ramp this up to 348 points, however there simply isn't enough meeple placement opportunities to score them. One of the small roads is also impossible to score due to the need to occupy the fields located on the same tiles. Phantom would solve both of these problems, but that would no longer be a tournament game.
11 points are leftover in the unfinished features, which brings up the total game point content to 359. The ultimate theoretical game point potential is 360 - the potential here means the potential of the last city segment being finished and thus worth double the points. This of coure is not possible.
Where does the notation come in? I was interested in wether this game was physically possible at all even with the assumption that the opponents play in your favor and tiles come in in favorable order. I was not sure wether the points could be optimally gained by a single player.
After some deliberation I have found the answer to be YES! There are in fact multiple ways to achieve this. This is how I have realized the max amount of points one can realistically gain is 342 instead of 348.
I attach the transcript of the ideal game leading to the 342 score.
A challenge to all of you, should you take it, is to find all alternative games leading to this score and post the transcripts here.
For those of you who noticed in the manual that the notation can be applied to multiple player games, try finding playthroughs with 3, 4 or even 5 players which all work together to share as many features between as many players as possible to bring their cummulative scores to new hights! It would be interesting to see what player count stops being favorable for the purposes of increasing the shared features. :gray-meeple:
-
Everyone is free to post their own games or usages of the system :)
My daughter and I played a game today to give me a chance to use the notation system and it was unexpectedly wonderful...
We played a standard base game, except that we used the Wind Roses start tile, because it has the compass directions prominently displayed and also because I'm in the 73rd-Tile-Is-Missing (https://www.carcassonnecentral.com/community/index.php?topic=3357) club.
I decided to go old-school by using a pencil and lined paper, with the log spilling onto a second page. Below is the game log, as recorded during play, and a photo of the landscape and scoreboard, after we finished final scoring and removed all the meeples. My daughter was Red and I was Yellow: she beat me by one point, 99-98!
(https://drive.google.com/thumbnail?id=16-8mt9S7MZ053if-Rp--xBda8mRp8fJH&sz=w400)
(https://drive.google.com/thumbnail?id=15zOioe36vRNtSEdGyazZZgJCMdU8mnaq&sz=w400)
I'll offer some comments on the notation system in a follow-on post. But for now, suffice it to say that the system is easy and actually fun to use. After a few turns, my daughter (23) had picked up the basics of it and was helping by calling out the coordinates. It definitely slowed the game a little bit, but what's the rush?
Verification: After we finished, I decided to recreate the game using the log. I found a few minor mistakes (for some reason, I put "c" in place of "f" a few times at the beginning, including my first turn), but was easily able to recreate the game. There were several interesting plays and I felt like I got to reinforce some key tactical ideas during this process! For example, on my daughter's move 45, she originally was going to put a meeple on the road and then decided not to--good choice not to put three meeples where they can all be trapped by a single hole!
You might also have noticed in the original log that I occasionally wrote the score in the right margin. But during the recreation, something wasn't adding up. After an animated conversation, we concluded that my daughter, who had kept score, must have forgotten to give me 9 cloister points on turn 44! Luckily, I had also taken a photo of the landscape and scoreboard after turn 48, and it was easy to confirm that I was 9 points short:
(https://drive.google.com/thumbnail?id=15z8v3am2D4j0-mz3frWnR5Wa0BcwCq93&sz=w400)
We also found that she had failed to give me a point for my 1-tile incomplete city at the end, so the corrected final score was 108-99--I won! My daughter was quick to point out that I should temper my pride, since I had, after all, drawn all 6 cloisters!
Anyway, here is the log with corrections and a few comments in red.
(https://drive.google.com/thumbnail?id=16-PYFXg4WolzjutcPQmwFY0K9hICpAs5&sz=w400)
I can use every win I can get against my daughter, so I was really happy to have logged this game. I also really enjoyed the post-game walk-thru and strategy review.
Thanks, DIN0!
-
Great post, @kothmann. :(y) :(y)
+1 merit from me.
-
Amazing kothmann! +1 merit
I am happy to see people try it out for themselves :(y) I will review your game based on the transcript to see it all for myself :)
-
Let us know if they cheated... :P
-
I will :D if they did.
I was notified of one more editing mistake in the manual - the tile list is missing two tiles. I have included errata in the original post. A corrected version will be released at some point.
I should also mention that already some notation editing apps are in progress by other members :gray-meeple:
-
Let us know if they cheated... :P
Hahahahahaha! Well, it turns out that in turn 60, I originally placed the monastery at [-2;1], overlooking the perfect ffff hole at [2;-1]. When I drew another monastery on turn 62, I groaned because I had no meeples left and I then noticed the hole at [2;-1]. My daughter very graciously said, “you must have been distracted by the notation system, so you should change move 60 to fill the hole for 9 points, and then put the new monastery on [-2;1] on this turn.” Young people are so quick to assume that the brain just works as intended. :-[ Anyway, those 9 points were exactly the margin of victory! >:D
I should also say that I just now noticed that I switched from DIN0’s official “K = Knight” notation to the “N=kNight” notation of chess after the first turn! Super confusing here, because N=North. Sorry about that. Old dog learning new tricks… ::)
-
Ha ha ha!
Old dog? You are only 2 years older than me. :o
-
Looks great, will try to reproduce slovak final to see if I think the same as top players :)
couple mistake found in coordinates page 17, I write in red when It was previously a mistake
a) [0;0]N:crfr
b) A:[0;+1]S:ccff-J+>W:K
c) B:[-1;-1]W:frrr>E:R
d) A:[+1;+1]S:ffrr>W:F
e) B:[+1;0]N:frfr
f) A:[-1;+1]E:crrf>W:R=+8A
g) B:[+2;0]E:frrr=+4B
-
Go for it totor66 :D
Thank you for noticing! There is indeed a single tile shift in the y coordinate as you describe. Now included in the errata.
-
As I mentioned to DIN0, I like this project. I think this notation is simple and fits very well in tournaments and solo-challenge games.
I am not a pencil and paper person, so, I developed a very simple web tool, mostly intended for smartphones, to help us to notate a Carcassonne game (base only as intended in DIN0's proposal) and export a game in a simple text file.
https://carliunterik.github.io/cgn/
Hope it can be useful for some of you.
Best,
PD. next step perhaps is to create an interpreter, maybe!
-
Another great addition to the community's toolbox! +1 merit from me
PD. next step perhaps is to create an interpreter, maybe!
Do you mean that we then can see the play happen with the cards? This would be awesome!
-
Thank you again carlium for creating this amazing app! This will certainly encourage more people to try the notation for themselves and they will have a great digital tool to do it :(y)
An interpreter would indeed be an awesome tool as well.
-
next step perhaps is to create an interpreter, maybe!
“interpreter” = read the notation to create a game? It would be amazing to be able to jump to a certain point in the game. Even if you could just generate an image of the game and the score at that point, someone could build the game quickly with tiles and then play out possible endings….
Anyway, +1 merit from me. Thanks for sharing.
-
Yep. I was thinking of a program which where you could let the game play out like a BGA replay, but you would also have the ability to both advance and go back in the game by individual turns, and also to just jump directly into any turn by clicking on it in the notation.
This could also help the puzzle-makers set up their scenarios.
-
I am not a pencil and paper person...
I actually really enjoy the paper-and-pencil experience--reminds me of keeping score at a baseball game. ^-^
But, baseball is slow, so you have plenty of time to erase your log errors and then argue about what actually just happened (https://www.youtube.com/watch?v=Y_H5ofTzki8&t=41s), even if the official answer is incomprehensible. >:D Also, paper scoresheets are harder to share on the internet, which hopefully is part of the fun here.
Before seeing @carlium's nifty tool, I did what any mechanical engineer does when facing what looks like a coding problem: I created a spreadsheet. In this case, a google sheet, so it would be easy to share--check it out here (https://docs.google.com/spreadsheets/d/1Cj4aYxrArGK_yctc-7JKGsvsIWqjZ15gWDOlnOUTdOA/edit?usp=sharing). You should be able to save this as an Excel sheet from the File menu, under Download. I'm not sure if you can copy to a new google sheet? PM me if you have any trouble. I don't think this will work well at all on a phone. The iPad is okay, but I prefer using my laptop, which is really an unwelcome intrusion of technology into the middle ages. ::)
Each column contains one element of the turn, in what seemed to be an intuitive order to me, and @DIN0's notation is assembled automagically in the final column. Here is what it looks like in a 4-player game (there are templates for 2, 3 or 4 players, plus the 3 games we logged so far):
(http://drive.google.com/uc?export=view&id=16FvEWufHg2AEHxvHiV-YcSPX8aZ453Wy)
Things to note about this sheet:
- unlike @carlium's tool, there are no helpful graphics or menus, so you have to manually enter compass directions.
- The "orientation" is optional for unidirectional tiles (ffff, cccc, rrrr).
- The "direction" is optional for any meeple placement when there is only one feature of that type on the tile just placed. This helps to speed-up the logging.
- I added a "comment" field, which I append to the end of the line using the python comment character '#'. I have mostly been filling this in when reviewing the game to help me learn from what happened.
- I added a log of the end-of-game scores for each of the 4 types of features in the top left, and there is a "running total" in the top right of the sheet. We found this useful to double-check with the scoreboard while logging, improving both the log and the score on the board.
- In the {Role} column, the sheet won't accept anything except F, K, M, R. This is because I kept trying to put "N" for kNight, like chess.
- I leave caps-lock on, so the tile edges are entered uppercase, but the script converts them to lowercase to conform to the standard.
- You have to put a single-quote character (') before the + to indicate a shield--otherwise, google sheets thinks you are entering a formula.
- I started using a ^ to denote the presence of a monastery on a tile. Obviously not important with only the base game.
I entered my paper-and-pencil log (earlier post) to test the spreadsheet. Then we played two more games--one 2-player and one 4-player. Logging slows play a bit more in 2-player games; it is still a noticeable if small nuisance in 4-player games, too. But I really like being able to go over the game at the end! Also, in 2 of 3 games logged so far, there was a scoring error! :-[
Here is a photo of the final landscape and the complete log for the very exciting 4-player 60-tile game we played last night (blue has 4 too many points in the photo):
(http://drive.google.com/uc?export=view&id=16H1rfvvWJST4NaZHJWdLA4nWMWE733LN)
Player Name Points
A Lewis 66
B Garrett 67
C Andrea 62
D Bruce 69
Date: 2022-01-04
0. [0;0]E:cfrf #Wind Rose start tile
1. A:[-1;0]W:crrf>K #?! not take 4 points
2. B:[1;0]W:cfrr>K=+4B
3. C:[-1;1]W:ffrr>R
4. D:[-2;1]S:cfff>K
5. A:[-2;2]E:cfcf-J+>K
6. B:[1;-1]W:frfr>R
7. C:[0;1]E:ffrr
8. D:[-2;0]W:cccf
9. A:[0;-1]W:fffr^>M
10. B:[2;0]N:cccr>K
11. C:[-3;1]N:cfcf-S>N:K
12. D:[0;2]W:ccrr>K #I hate glom-across!
13. A:[0;-2]S:cfrr>R
14. B:[0;-3]N:crrr>K=+4B
15. C:[-3;2]E:ccrr+
16. D:[-4;2]S:crfr>K
17. A:[1;-2]E:ffrr=+6A+6B #friend!
18. B:[1;-3]N:ffrr>F
19. C:[-3;3]W:cccr+>K
20. D:[-5;2]N:frfr>R
21. A:[-3;0]W:cccr+ #risky-new city better?
22. B:[2;-1]N:cfff
23. C:[1;1]ffff^>M
24. D:[-4;1]N:cfcf-J+ #foe!
25. A:[-3;-1]W:crfr>K
26. B:[-1;-1]W:fffr^>M
27. C:[0;3]cccc+ #[-4;0] too risky
28. D:[-2;-1]ffff^>M
29. A:[3;0]W:cfff>F #nice farm glom
30. B:[1;-4]E:crfr>K #[-1;3]E = savage hole!
31. C:[-2;3]W:ccff-J+
32. D:[-6;2]S:frrr>S:F=+5D #big cities = bad farm?
33. A:[3;-1]E:cfff #no knight?!
34. B:[2;-4]S:ccrr
35. C:[-2;4]S:ccff-J
36. D:[-4;0]N:ccff-J=+20D #i thought game over!
37. A:[4;-1]W:ccrr+>K
38. B:[-1;-2]W:frfr>R=+9A
39. C:[-1;-3]S:crrr>E:R=+3B+2C #score=15-17-2-25
40. D:[-1;2]W:cccf+ #lucky! CCCF before CFCF-S
41. A:[-3;-2]W:frfr>R
42. B:[-1;4]E:cfff>K #glom, not future block?
43. C:[-4;3]W:cfcf-S>F
44. D:[3;-4]N:ffrr>SW:F
45. A:[4;0]S:cfrr=+8A
46. B:[0;4]S:ccrr
47. C:[-3;4]E:ccff+=+16C #nice!
48. D:[-1;3]E:ccff-S>F #only 3 turns left, so F
49. A:[-4;-1]E:ccff-S>F=+4A
50. B:[1;-5]E:frrr>R=+4B
51. C:[-1;-4]N:crrf>K=+4C
52. D:[2;-3]W:frrr>S:R
53. A:[-4;-2]W:ccff-J>K
54. B:[-3;-3]W:crrr>W:F=+4A #nice farm grab!
55. C:[-4;-3]E:crrf>K=+4C
56. D:[5;-1]rrrr>NW:F #last meeple for 3 pts?
57. A:[-5;-2]W:cfcf-S>F=+6A
58. B:[-3;5]W:cccf>F
59. C:[-2;-2]ffff^>M=+9B+9D #[-5;1]=gray gets no meeple
60. D:[-2;-3]N:frrr>SE:F=+9C #9B = being nice to blue :)
-
Great work kothmann! This is a very nice sheet optimized for real-time notation :(y) +1 merit
I discussed this with kothmann and some other people via PM, but I should make note of it publicly. The order in which the details about the game are ordered in the notation are a logical, esthetic and practical compromise between the many ways one can record notation. There are several different approaches and each has slightly differentpreferable order:
real-time participant (the style in which kothmann is recording games - he is simultaneously playing and recording)
real-time observer (the person writing is not playing)
writing notation off of video recording of past game (this is not real-time, but instead transcription of visually recorded game into a text form)
writing off of BGA replay, or similar feature (also not real-time, uses replay features of digital versions of Carcassonne)
The player count can also affect these things as seen in the recent post.
Kothmann's sheet is optimized for the purposes of the first two cases which makes real-time notation go even more smoothly :) and manages to maintain the formal notation at the end. Another great addition to the toolbox!
I do have a small suggestion to add and that is another column for discarded tiles. As far as I can tell, there currently isn't a way to enter this portion of notation into the sheet. You probably did not encounter any unplacable tiles, so there was no need for it in your games.
Keep up the good work!
-
Yes. Discarding of card is very important to reproduce game. Maybe is possible also to check JCZ save file which contains tiles / meeples / tokens placements their positions and it's features. What is cons for JCZ save file is that not contains scoring because it uses specific random algorithm which allows to continue game in random order but by placed tiles it always gets same results.
+1 merit for me to do this not easy solution in game like Carcassonne which has a lot of tiles combinations based not only on edge types bot also on inside (image tile CFCF can have two cities with one field, one city, one city with shield, one city with trade good token (beer), one city with shield and dragon symbol, two cities with catapult and two fields, one city with bazaar, one city with monastery inside city, two cities with hill and one field, two cities with vineyard touched so two fields, two cities with one field and mage and witch symbol, two cities with two fields and crop circle, one city with robber symbol, one city with wind rose WN, one city with wind rose W).
Uff. And and this was just official C1 releases.
C2 Includes also Two cities with two fields and circus place, two cities two fields and watchtower with city bonus, and one bonus, CutCassone two cities one field and garden.
Anyway @Meepledrone still wondering how to manage C3 on WiCa? MINDOK tells that BigBox 2021 is not new BigBox, that it's only reprint of 2017 version with updated graphics and with amazing rules :D
-
Anyway @Meepledrone still wondering how to manage C3 on WiCa? MINDOK tells that BigBox 2021 is not new BigBox, that it's only reprint of 2017 version with updated graphics and with amazing rules :D
Devir said the same on their last DevirTV show on Youtube... They commented about it since there was a lot of hearsay on social networks about Carc and Big Box 3.0 :o... So these sets would be just released as updated new reprints... Nothing fancy ... No fuzz about it... And they would be followed by new reprints of the expansions following suit... There you go. >:D
-
in game like Carcassonne which has a lot of tiles combinations based not only on edge types bot also on inside
The General Notation (project 1B) already accounts for everything you have mentioned. In other words, the tile reference is solved, it's the other things that need work. but as I mentioned, this extended notation won't be released anytime soon.
Project 2 takes priority as it is more practical and (believe it or not) even more ambitious than 1B.
-
Project 2 takes priority as it is more practical and (believe it or not) even more ambitious than 1B.
Expected release date for P2? We have high expectations! :)
-
Unknown. It doesn't have a well defined release date yet. There had been several previous release dates for that one, but it never quite worked out. And honestly that is for the better, because it allowed me to pack more things in :)
Alright I can give you very basic info about project 2. It is completely unrelated to Project 1A or 1B. It's completion doesn't merely depend on me finishing it, there are other factors beyond my control affecting the progress. In other words, unlike 1A and 1B, project 2 is not something I could theoretically put together in one sitting. I have to wait for certain things to transpire and to achieve few milestones before I can finish the first draft.
And another thing worth mentioning - when it comes to the "output" project 1A is static, meaning the manual and sheet has been released and that is largely it. I will of course make the promised corrected version where I implement all the errata and perhaps include few minor updates/changes, but overall the work on this one is done.
Project 1B as you know is General Notation, which by its very nature will require constant updates as the new material releases over the years.
Project 2 is similar in that regard, it too will be continually updated, so its first release won't be its final form. But there are few things it absolutely must have and that is what is currently in the works.
Hope that gives you at least a glimpse of the future.
-
Sounds good. We’ll patiently await the first release of 1B & 2!
-
I have added the link to the download page to the opening post, so anyone can find it more quickly.
-
Hello, I am back with some new stuff to demonstrate the power of the game notation :gray-meeple:
I have decided to make a series of game analyses based on the example transcripts included in the notation manual. I will try to publish one analysis per week. Note that most of these games are my games from BoardGameArena (BGA) and the performance has varying levels of quality. These analyses explore tactical decisions and potential improvements as well as the play of the opponents. This is game analysis 1/5 (as a heads up, we shall look at the Slovak championsip finals in the 5/5, so stay tuned :(y)).
The following transcript can be found on page 25 of the notation manual. Few minor mistakes have been found, each of them are now documented in the errata (in the opening post) and will be implemented once the updated notation manual is released. An interesting note to make is that while I was reconstructing the game for this analysis, I somehow totally forgot which player I was and so I originally looked at the game as if I were player A :D. After realizing I switched us around it felt like a real out of body experience. It may have helped me to look more critically at some of my moves. Nevertheless the following analysis is written with the correct player identity in mind, so you don't have to worry about it. Commentary is included only for noteworthy turns, not all of them.
Date of game: 19. 7. 2020
Player A: rayray243
Player B: DIN0
01. Player A opens with a decent sized city.
02. Seeing the potential, I launch an early attack.
04. I'm starting to set up for a take-over, while keeping up in point momentum.
07. Player A beats me to it and closes a city edge while preparing to sneak in their troops.
08. A proper defense on my part, simultaneous connection and cut-off.
09. Player A takes what's left of their invasion attempt and sets up an opportunity for further invasion for either player.
10. Since I draw the necessary tile I go for the offensive.
13. Player A has a very nice retaliation and barrs any further advancements in the city. Since we are both equal in points gained from this state, we both loose an interest in completing the city.
17. Player A gets a very nice road from the colateral.
19. Player A starts the first farm and it looks very promising. 9 points already with at least two near completion cities and plenty of room for growth.
23. Player A gives themselves a nice cloister, but provides me with building blocks for a great farm of my own.
24. is an interesting turn for me. I pretty much burn the tile while separating the two fields, but let's look at some other potential placements. I could claim the city at [+1;-2], but even with proper orientation the city has a lot of clutter around it and is not worth the risk. Alternative placement at [+2;-3] would no doubt promt Player A to direct the city into the unfinishable hole at [+1;-1]. The placement [0;+4]W:cccf+>S:F would ensure player A wouldn't steal the farm while unoccupied, and slightly inconvenience their city, but would result in farmer redundancy later down the game. The last thing I could have done is to use the fact I just separated the farms and place a farmer at >W:F. But since I had the meeple advantage at that point, I held out. Lastly, completing the big city would accomplish nothing, just wasted a turn while robbing me of the meeple advantage.
26. I try to work on that advantage and permanently block player A's robber...
27. ...but they retaliate.
29. Player A does not seem to care about my farm. I'm sure it won't haunt them later... >:D
30. If I can't block the road, let me join it.
32. This was probably the biggest offender on my part. I hoped using the last meeple would pay-off in this case because the city required only one tile to finish and this style of building is usually hard to attack. However I didn't realize the danger of the road branch comming from the west...
33. ...that's what I was talking about ::) we will be here for a while. Also a sneaky farm invasion - but not as sneaky as the one that's about to come later >:D
34. 35. and so begins the inadvertent (at least on my part) building of a big empty city. Advice: do not do this when you have no meeples.
36. Escape! But looking back, not choosing the obvious [-2;+1] placement was quite stupid.
39. Player A doesn't let up on my city.
41. Player A's farm invasion succesful, but will it be enough?
42. Building a farm potential for later. Player A does not have enough meeples to make use of it and they're preoccupied with other stuff. This will come in handy later.
44. Another escape.
45. Nice road for A!
47. Yet another block.
53. Things are starting to look bad over here in my city.
56 Finally! a lucky draw puts me back in the game, even though it joined the farms.
57. A good meeple management from player A, who finally took the big empty city at [-2;+1].
58. I was finally able to finish my city and make up for the point difference.
61. In the last 10 turns, this is where the real game begins! Player A finally finishes the first city with the advantage of taking the 6-point farm.
62. With newfound meeple power, I take the longest road I can find on the map.
63. Player A realizes the point lead I am taking and works toward neutralizing my additional points where possible, while maintaining the farm dominance. This however was their biggest mistake and I was about to capitalize on it.
64. I make the sneakiest attack on the big farm through the A's attempted road joining. The thing is there are no ffrr tiles left in the supply at this point. The only tile capable of the task is the last ccrr+.
65. This forces player A makes to block with the crrr tile, leaving me with the bigger chunk of the road and a hefty meeple advantage.
66. Sure enough, the very first tile I draw is ccrr+ which is now impossible to place in the intended position. I instead use my meeple advantage to take the pennant cluster at [+6;+6]. The tile pool is thinning...
67. Player A uses the ccff-J tile to extend their city. This could have been used to claim some of the empty cities with their last meeple. Even more importantly, what player A doesn't realize is they should be using it to try and block the farm invasion further because thanks to their block, there is now one more tile capable of joining the farms and that is fffr. The probability of its emergence is extremely high because of the ever-smaller tile pool (just 4 tiles at that point!).
68. Sure enough, there it is and I join the farms together taking away the huge amount of point difference and taking the cloister for good measure.
69. A's last minute cloister of their own.
70. I close a joined road neutralizing those points and place a farmer on the south field making use of those cities I mentioned earlier. All farm points are now equalized.
71. Nothing better left to do with final cfff than to add a single point to A's city. What would have been an optimal play, had they kept track of the tiles left in supply, was to use the ccff-J from turn 67. and place it like so [+5;+1]W:ccff-J, and then hope they would get the cfff tile in time to place it [+6;+1]N:cfff blocking my entrance completely.
At the end of the game, all the points left on the playing field are exactly equal between us, meaning the point difference player A was so actively trying to avert was maintained and I have won with the score 115:128!
Had they used their tiles as I suggested, player A would have been the winner with the score. 114:105 (one less point for A's city and +1 point for my would be cloister placement at [-1;-3].
This was a very interesting game. Both of us made mistakes, sometimes big ones, but clearly the most interesting and decisive play comes at nearly the end of the game. I was very pleased with the result especially because based on the BGA's ELO system, I was predicted to win only with 21% probability.
Alas, sneaky moves and knowledge of the remainig tile pool can result in a big win :(y)
I hope you enjoyed this analysis and will try to reconstruct this game yourself and go over it to see what I am talking about each turn :)
-
Wow, this is amazing. Thanks! No time tonight, but I’ll go thru it this weekend. +1 from me for such a great effort!
-
What a report! Nice readings
Odoslané z SM-A202F pomocou Tapatalku
-
Nice description. :(y)
I quite enjoy BGA's replay feature as you can literally watch any game replayed turn by turn if you have an account.
The game you describe so vividly above can be watched via this link for example: https://www.boardgamearena.com/gamereview?table=102298200
-
Thank you all for the kind words :gray-meeple:
I quite enjoy BGA's replay feature as you can literally watch any game replayed turn by turn if you have an account.
The game you describe so vividly above can be watched via this link for example:
Yeah, I like it too. Just keep in mind the point of these analyses is to reconstruct the game yourself and go along the comentary + suggest your own ideas opinions on the game ;)
Speaking of, has anybody had a chance to reproduce the game yet? I believe kothmann was about to get to it.
-
...has anybody had a chance to reproduce the game yet? I believe kothmann was about to get to it.
So, I started going through this game this morning, trying hard to take each tile without looking at placement information in the log and think about what I might do. I wrote comments after looking at what actually happened, but mostly without looking ahead at the next move. (A couple of times I misinterpreted the log and then had to backtrack.) This is super fun and interesting, though it does take time to give it so much thought. I had to stop part way through and didn't carefully proofread, because I actually have some work to do today. :-\ Hope to go through the rest of this game soon. I’ll post again when that happens.
Also important disclaimer: these are comments with the benefit of having minutes to think when needed ???, and having all the remaining tiles right in front of me. Many of the interesting decisions would definitely be tough to make in real time.
Thanks again for posting this. Your system is definitely a great tool for learning from games. Particularly if people are willing to add to the conversation. Hopefully someone will disagree with some (many?) of my ideas, so I can learn from my mistakes! >:D
03. I don’t like [0;-1]. This will leave a city cap that B can use to score a quick 4 points. If A plays [0;2]N:cfcf-S>N:K, the points for the south city cap might be shared with B, but there is a chance A gets all those points with a ccff-S, and at least A will share the points, instead of let B have them outright.04. Alternatives: [-1;0]S:crfr>S:K, to attack the other city. I also think [-1;0]Ncrfr>W:R is good. Gives away a 2-point city cap, but in exchange for avoiding a 3-point open road. I don’t understand “keep up point momentum”?
05. Early in the game there is no big rush to close 4-point cities? Especially when there are so many “free” road points on the landscape? I prefer [-1;0]W:crrf>N:R, to claim the 4-point road with a pretty easy close on the north side. [2;1]Wcrrf>W:K, tries to win back control of what will now be a huge city. This second idea is definitely risky, because if [1;1] becomes a hole, A will have 2 stuck meeples.
06. Good to make that open city cap harder to claim. But having only 1 meeple on the board so early seems bad to me, so again, maybe I prefer claiming the road with [3;0]E:crrr>W:R? (Lots of deployed meeples = lots of ways to get points on future turns?)
07. I thought [1;1]S:ccrr-S was an automatic play here. Both players get 1 tile added to their cities, but A keeps B from sharing the 3-net-tiles of A’s larger city. Is the threat of B attacking from the west side of [1;2] big enough to make this play foolish? Interesting…
08. There were a total of 16 tiles (5xccrr, 5xccff, 4xcccf, 2xcccr) that would have cut off A’s attempt to come back into this city. More evidence that A’s play on turn 7 was not strong?
09. “Player A …and sets up an opportunity for further invasion for either player.” Yes, this seems poor, because they give B first bite at the attack apple! That little football city was not vulnerable, so no rush to close. Again, get meeples on the landscape! Something like [0;-2]S:cfff>S:K would be good.
10. This is risky for the same reasons that I mention for A placing at [2;1] in turn 5: it isn’t too hard for [2;1] to become a hole here, and then B will have 2 stuck meeples.
11. Now A has a lot (20+) that will turn [2;1] into a ccrf-hole. B has to close that big city ASAP! If A plays [3;0]N:ffrr>W:R, they still have the same hole-making odds, but now they also claim the road—is this too risky because it might be difficult to fill in [-1;0] (this is obvious in the current game, after the next move).
12. Is “[-1;+1]” a typo that should be “[-1;0]”? Is the advantage of having only a single open end on a road really worth the extra 2 points this early in the game?
13. There are now 7 tiles that connect B to the big city, but still over 20 tiles that make [2;1] a ccrf-hole, so A could wait. On the other hand, if B gets any cfxx or ffxx, then they place this at [3;1] and make it very likely that B eventually connects and wins, so A playing for the big-city tie here is good? I guess so, but huge bummer for A not to have drawn one of the very many hole-making tiles here. Here's a photo showing the big pile of dangerous tiles that A didn't get (Yellow=A, Red=B):
(http://drive.google.com/uc?export=view&id=1AdIXyN1lO4JVnYxMJt4qCchU_DsLijx5)
16Interesting. My inclination would be to play [-2;1]S:cccr>N:K=+3B. But I think I might be too eager to start a large open city, and should be more cautious, like this. Hmm…
17. Finally! :-)
18. Did you think about [1;3]W:cfff>N:F? That field has 3 completed cities already and that big one will get completed when someone needs meeples, or just by you to grab the farm points? I know it is not well guarded from invasion at [4;0], and also still a bit early, but just curious for thoughts about farming at this point.
19. Haha, now a farmer. Playing the third city on an unoccupied farm is always such a pain, because you don’t want to give up those 4 points, but you know the opponent is likely to grab the farm. Yes, I like this farm, but I don’t love giving away a ccff+ tile to get it. Those turn into 8-point cities very quickly.
20. I think [0;4]W:cfcf-S>E:K is much better? Why leave the pennant city open for A?! Also, this creates another 3-city field with plenty of chances for A to grab it, and they now have the incentive to close the huge city and free up more meeples after playing another farmer.
22. Again, no big rush to complete here? That farmer seems pretty weak? Is the plan to connect with 2 more tiles to a bigger field? Could also try that with [3;-2]W:crrr>NE:F and then try to connect thru [4;0] and [4;1]? I thought about attacking with [-1;5]S:crrr>S:K, but then A can just play any rfxx tile at [-1;4] to make a ccfr-hole trapping one of each meeple, with a 2-point net gain. What about [3;-2]W:crrr>W:K, to grab the open city? Things are getting very complex
23. Agree this is a blunder. [3;3]N:ffff>N:M is one point less but much better.
24. Interesting! I agree that claiming a huge and wide-open city in the south is bad. And completing the big shared city is terrible, because you have the meeple advantage and don’t yet have the farm in the west, while A gets the farm points for that shared city. What about [-1;4]W:cccf+>W:K? There are still 3xcfcf-J, 2xcccf and no cfcf-S, so he can’t finish that city in the north without sharing with you. Furthermore, you have to be very worried that he gets a tile to play at [0;4] to claim the farm (even though he would probably have to share with you anyway). There are 13 tiles that allow him to place a farmer and grow his city. You would cut that down to just 5 tiles, and they all would connect you to the city.
(http://drive.google.com/uc?export=view&id=1AekG8svBmZbVySXc0uBv3RnlfO3zMGu1)
-
As promised, here are my comments on the rest of the game. Same disclaimers as before:
- I didn’t proofread for mistakes.
- This is much easier in hindsight.
- I don’t really know what I’m talking about, having rarely played competitive 2-player games, so I’m mostly hoping to learn from anyone who might disagree with my thoughts here.
I hope you will post more interesting games. This took several hours, but I think if I did this even once a month I would become a much better player. Again, thanks and another +1 merit for the system and the great example.
25. I like [-2;1]W:crrf>NE:F to secure the second farm and also give B an easy play at [-2;2] to connect to the field while also completing A’s monastery. But I guess A wants to conserve meeples with 5 monasteries out.
26. I expected [-2;2]E:ffrr to secure the field before A can get in. There are 8 tiles that play at [-2;0] to create a CRFR hole at [-1;0], so I guess that’s a fair alternative?
27. Nice defense, but now B can place a robber on a road tile at [-1;-2] with a great chance to share the large road. Also, the monastery is more difficult to finish now.
28. A steady stream of monasteries now would be a gold mine for B!
29. Wow, yeah, I don’t like how A has boxed themselves in here, while just giving the farm to B. I think closing the road at [-1;-2] would have been better here.
30. As expected. Interesting dynamic here. Now both players are down to 1 free meeple, and A has the better end of the road, so only B will close it? Note that if B has not placed cccf+ at [4;0] in turn 24, they could have now played [4;0]W:frfr>W:F and made invasion of the farm in the northeast very likely, but even if that failed, they still get at least 6 points from that farmer.
31. I was focused on finishing the monastery at [-2;1], but I guess there are still more tiles that do that job than there are that close this nice little 8-point city. Also B may want to play at [-2;1] to recover the robber at [-1;1]. I wonder how the game would have ended if this tile had come at turn 29. Presumably A would have put down the farmer while closing the city, leaving themselves with a free meeple. That is probably a huge turning point in the game.
32. Yeah, no great options here. Only 5 city caps left, so this might not be quick and there are 4 monasteries still to come and B needs a FFFx to finish the monastery and also now A gets another city? I thought maybe a throw-away at [1;6]?
33. A does care about the big farm after all! Now only 2 city caps and the monastery went from 6 possible completing tiles down to 2. This was a big shift.
34. I thought [1;6]W:cfcf+ would set up a play at [0;6] to make it tough for the farmer to join. Actually, as I look now, I see that for B to complete their monastery, they have to join the second A farmer into the northeast field, and then it can get around the corner at [0;6] to takeover the northwest field, too. Yikes.
35. I like [1;-2]N:cccf to make a big empty city in the south that can definitely never be completed, so not likely to be occupied anytime soon. This play creates a 4-point city that could still be completed so it is more dangerous.
36. Not [-2;1]S:ccff=+9A+4B? I think B needs to get a meeple back!? A has a meeple, so giving one more to A doesn’t matter as much? I suppose going from 2 to 5 tiles that complete the city in the northeast is good, but not A can place a tile at [4;4] with a northward facing city to try to invade and there are no more separating tiles to stop them! {Read the “stupid” comment after writing this. And I didn’t think about the simple road block that came with 39.}
37. Again I like [1;-2]N:ccff to make a permanently incomplete city. This dead city in the west is starting to get valuable!
39. Rough.
42. Seems risky to build a potential farm when A has more free meeples?
44. Still not [-2;1]?! Hmm. The “escape” can again be hindered, although an invasion is less likely than before.
48 Not [-1;0]?! At present, both players will get 5 at the end of the game, but A has a meeple, so can keep racking up small gains, while B has no meeples, so it seems to me like it is much better to finish that big road?!
50. Why connect to an existing road?! If A draws frrr, they can immediately score 3 at [-3;4], instead of 2 anywhere else? Playing [-2;-2]N:ffrr makes sure the big city in the west won’t close and doesn’t offer a quick 3.
51. Was this expected, instead of the quick 3 points? Was turn 50 an attempt to lure A away from this move?! Wow.
53. This feels overly defensive to me? I suppose A wants to keep their meeple, but in trying to prevent that city from closing, A has made it big and if B draws the last ccff+ tile, even open the city is worth 10 points! {OOPS! I assumed the placement here was W:ccff, to ensure that a lucky draw could not close the city in one turn!? Yikes, I think N:ccff was a big mistake?}
56. If turn 48 had been played at [-2;1], then presumably A would have placed a meeple on either turn 51 or 53, but this turn would have been worth 9 more points and B would have 2 meeples to 1 for A? Also, I’m really suspect of joining the farms, but I suppose B has to, in case A draws the other FFFR tile. Wow, really complicated situation.
57. I don’t know about placing the knight. Save the meeple to try to invade the big city in the northeast? {I’m confused—see OOPS at turn 53} Or get on the road in the east? Or a farmer in the south? This city is still unlikely to close.
58. This is a huge swing.
63. Yuck. Finish the big road with a farmer in the South for a guaranteed 6 points and prevent B from getting those 6. There are only 2 tiles to connect here and one tile that blocks, so this is a big goof.
64. This is sneaky and totally kills the attempt by A to get onto the road (A now has to play the crrr tile to block and can’t play the ccrr tile to connect). On the other hand, for this attack to work, B has to connect A to the big road. Playing [3;7]S:ffrr>NE:F has the same 2 critical tiles, but now if B gets the key tile, A doesn’t connect. So interesting and very clever…
65. Playing [5;0]W:crrr doesn’t give B the meeple back and creates a future block of the farm invasion with any of 4 tiles (2xffrr, fffr, ccrr+). But yeah, this is probably necessary.
66. Oh, wait! I’m just now seeing that turn 65 still leaves B with the possibility of FFFR at [6;2] to get into the field! Did you see this ahead of time?! Wow.
67. This is a blunder, for sure. [5;1]N:ccff>W:F gives A two ways to win. They draw the fffr tile to prevent B from joining the huge field, or they get the cfff tile to get a winning 3rd farmer in the field!
69. If B draws the cfff tile here, they can play [5;1]N:cfff>W:F and then hope to get the ffff and play it at [6;1] to win the field.
Congratulations on a great win!
-
Thanks for going over this kothmann! I will go over your comments today and post some thoughts :(y)
-
I definitely went a little overboard here. Thought it could be informative to see comments in (mostly) “real time”, meaning I don’t know what’s gonna happen. Once you have played it is hard not to think in terms of what you know is coming.
It could be really interesting to have a live chat where a group of people walk thru a game and discuss. But tough to organize with such a global network.
Anyway, I’ll look forward to your reply when you have time.
And maybe this weekend build a JCZ translator for your logs…(just the tile order not the play, so people can replay by themselves more easily.)
-
After the release of this project, it immediately occurred to me to do something along the lines of a grid map/play mat with coordinates in a [X;X] format.
However, in a accumulation of other projects (mostly related to Carcassonne :)) ) I kept postponing it.
I was thinking of doing something interactive where you could define the size, style and colors yourself, generating a print-ready PDF.
But I think I will just prepare few versions in different standard sizes (like A0) + way to big size that could be cropped as required and will offer help with doing so too.
Printing on paper/canvas or ordering printed tablecloth from aliexpress should work.
A very quick draft of what I mean.
(https://i.imgur.com/QKUBg2T.png)
-
Finally I have a clue how this position is calculated. As Web programmer I'm using system that [0,0] is on top left corner and [100,100] is bottom and right from this point.
So is uses classical Euclidean space, like regular axis x,y.
Ok. Thanks @Snearone. This grid can help :D
-
Ok @kothmann, I am finally here with my responses! The points to which I haven't included any comments here, I either agreed with, or simply didn't have anything interesting tto add.
03. I have to agree. The placement you propose would make more sense for A.
04. By point momentum I mean the effective meeple turnaround generating points. In this case it evens out A's small city. There is no guarantee that the big city will complete or that I will share the points, so keeping up in points elsewhere is important.
05. [-1;0]W:crrf>N:R is certainly a more "elegant" move to take, especially if A got a chance to score that extra city segment in the next turn (which did happen). I guess they were playing safe with the meeples and the next ccff-S tile went into reclaiming the big city.
06. I suspect at that point, we were both keeping our meeples ready for an all out fight for the big city, so we wanted to keep our meeple reserves dynamic instead of risking them being stuck on a road for a long time. It wouldn't be too difficult to prolong its closure in that board state.
07. It is kind of surprising in the hindsight - it would seem player A wouldn't be satisfied with the amount of point gain (+6)? I would close it for sure if I were in that position.
08. Yeah that really did not work out too well for A...
09. To be fair it is unlikely in these situations that the other player would create that chance for them unless they also wanted to further invade (such as now). Someone has to make that opening and that someone will always be at a slight disadvantage...
10. True. This came down to the psychological aspect of the game - based on A's previous play (especially turn 07) it seemed more likely that A would invest more meeples into the city instead of making it unfinishable. Of course they could have been baiting me, but a complex bait like this is quite risky on its own. But yes, I agree that it would be in A's best interest to block the city at that point.
12. Not a typo, I simply wanted to start a reliable road of some kind.
13. I agree with the assessment.
16. Yeah, I wouldn't take that city for that reason. I wanted to keep that road going longer if possible and most of the crossroads were still in the pile, so I took some small extra points.
18., 19. Yep, exactly the situation you describe :D I tried to hold out one more turn, but A was faster.
20. I tried to build new cities away from A's farm while preparing a new farm for me to claim. I agree the latter part is quite risky, but on the other hand I did have the meeple advantage at that point.
22. Yes, that was the plan. Again risky for sure, but it somehow created an interconnected series of pressures on both players which would lead to the farm extension. The immediate closure of small city and farmer placement kept my meeple advantage and also introduced a nearby end for my road. It also created the hole for a potential cloister, which could go to me, or (what really happened) go to A, but regardless, it would bring me closer to extending my farm. The cloister was especially attractive for A because of my road trajectory - from A's point of view, they wouldn't have to work toward completing their cloister.
23. All part of the (loose) plan >:D
24. I was certainly worried about A invading that farm before I was there, but ultimately this seemed better and I didn't want to spread my forces out in another city fight, especially in such an open city that would result from the [-1;4]W:cccf+>W:K move. And if I just placed it there without a meeple A would just connect and gain extra points.
25. I like that placement for A! That would make a lot of sense, had they had an extra meeple that is. 5 monasteries out? You meant 5 meeples right?
26. Yes I wanted that block, but it didn't work because of 27..
28. Where would you place them, [+2;+4]? And then what next? I guess if fffr came in time then [+2;+5]?
29. This was indeed bizzare on A's part. Yeah [-1;-2]E:frrr is definitely the optimal move here. Or at the very least get those +2 points from the road at [-3;+3].
32. Yeah that throwaway might have worked, or at the very least that would have been the best place to disperse those loose city points. But still this opportunity seemed good at the time, well except for that erroneous reading of the road comming from the west. It gave me extra points for the city in the end, but of course I couldn't know that then.
33. Big shift indeed! This was a great play from A.
36. Yes, just a straight up mistake. Should have finished that road and get the meeple back. Same with 44. later.
37. I think that was A's plan seeing as I had no meeples left to claim that city. I even contributed to making the empty city bigger :o
42. I don't thinks so in this case, A wouldn't waste the last meeple on a farm when they can instead take the full advantage of the meeple superiority and dynamically build up bunch of points, while I am stuck.
48. It might have been better, but at the time I was keeping that open for future as I wanted that farm.
50. Honestly I do not remember what was going through my head when making this move, but I doubt it was a concerted effort to prevevent move 51. There were still many tiles that A could have used for that purpose anyway.
53. Huh, I didn't notice that. Yes the W orientation would have been more effective.
56. Indeed! This was the only way out for me t this point, I desperatelly needed a meeple.
57. I would say the A never intended to finish the city and opted for the raw points from this byproduct of a construct. The didn't loose much of a meeple advantage either (at least at that point.
63. Agreed. I suspect A started o feel uneasy about the point lead I gained and thus wanted to prevent any further swings in point difference. But in doing so, they made the situation way worse.
64., 66. Yes, this I have seen in advance and was part of the strategy. That is not to say I would be able to replicate this 100% time in a similar situation. The tiles that played a huge role here were ffrr, ccrr, crrf and fffr. Ironically, for most of the game player A kept me dependent on exactly those few tiles, so I was always keenly aware of their numbers and it payed off big time right here. This is what I like about Carcassonne - even a series of suboptimal moves, or dubious decisions with consequences you have to mend, can lead to a circumstance which you can capitalize on later.
67. Yes that is a good point. A could have placed it there either in N orientation and join through cfff, or in W orientation to join through ffff, there were many options here, but A probably didn't have the necessary tile awarness at the moment.
Congratulations on a great win!
Thanks! I enjoyed this conversation and your suggestions. They have for sure enabled me to look for improvements for both players :(y)
-
Ok @kothmann, I am finally here with my responses! The points to which I haven't included any comments here, I either agreed with, or simply didn't have anything interesting to add.
Thanks! Interesting discussion once again.
In case anyone else wants to give it a try, I made a JCZ file with the tile order predefined, so you can follow the log or play the game anew, but with the same tile order. I've attached that file. I did this file conversion by hand (about 20 minutes), and I'm afraid my ambition to automate the process with a bit of code will have to wait. Maybe this summer.
(Note: CC won't let me upload JCZ file, so I changed extension to TXT. Just change it back if you download)
Anyway, thanks once more for your great notation system and the interesting detailed commentary on the game. Super enjoyable and educational for me!
-
Just to quickly let everybody know, the next game analysis will be delayed to the next week due to busy schedule on my side. But it will come, do not worry :(y)
-
Thanks looking forward to it!
-
Just discovered this thread and, all I can say is: Thanks DIN0.
+1 merit from me
-
You are most welcome! Thanks for the merit ;D
Sadly I have been busy lately, so my strategy commentaries are delayed significantly from the initial plan. And it doesn't seem to get any looser in the near months, so I have no idea when I will be able to get to it :'( That said they will return someday, along with new notation-related content :yellow-meeple: :orange-meeple:
-
I was going to ask if I can get permission to translate the notation system in other languages, then it dawned on me: the system, as it stands now, may come fairly easy to English speakers, but that won't be true for the rest of the world.
As I had in mind the groups that play Carcassonne competitively - they would be a natural beneficiary of a notation system - I think it would be good if it can be made language agnostic. The game itself did manage that
-
If you are referring to the manual, I would be open to arrange something with you (and potentially others) in the future for the translation to other languages. In fact I am planning one translation myself.
I you mean the system itself, that is in fact already language independent. It is true that certain symbols (letters) are derived from english forms of those terms, but they are only ever used in their short abbreviated forms, so it is just another symbol of the notation. You do not need to know the verbal origin of the symbols to use them, just when and how to use them properly. This is also the case in other notations such as the chess notation - the abbreviations are derived from english, but the system itself is international and language independent. I made my system the same in this regard.
What languages do you have in mind? :)
-
Chess notations comes in many flavours: you can specify Qc4 with the meaning of Queen moves from current position to c4, or Qc2-c4, meaning Queen moves from c2 (which is the current position) to c4. Back in the day, when I was playing chess competitionally in Romania, I was thought to write my same move as Dc4, D standing for the Romanian "Dama". Same principle as in English, but localized.
However, the same move can be easily "internationalized": c2-c4 (with or without dash). In notations that designate what moves, this would be a pawn move, but you can avoid specifying the figure with a language-specific letter by implying "whatever is on c2 moves to c4". No Q, no D, no nothing.
At the end of the day, using the "r" symbol for road, or using the "|" symbol (or any other) is just a matter of convention and agreement. But it does give an "associative advantage" to an Anglophone who can associate "r" with road. For a Czech the word, I believe, is "cest" - no "r" there. For a Romanian (go Marian Curcan, go!), the word is "drum". For a Greek it is "dromos", but spelled in a different alphabet, for a German (these guys invented the game) it is "Weg" - mandatory capital for nouns. I think this illustrates why I use "associative advantage" for English speakers.
My proposal is to use the principles proposed so far, but change a bit the symbols, trying to make them universal.
The other thing is that it would be easier to parse a game by a computer if the field separator is consistently the same.
Please allow me a few days to crystalize this vague comments into a concrete proposal and let's discuss from there.
As for translations... I think this notation can be useful to many globally. It has the potential to become the official way of recording games in championships, which means, in my mind, to make it available to the largest audience possible.
I hope these sound good.
-
I'm afraid I was not able to put my thoughts on paper in a concise way, so I ended up generating the attached pdf. Please consider it an invitation to discussion.
-
I'd created my own notation system two or three weeks ago before I realised this thread was here - doh!!
As expected, there are many similarities between the two notation systems. If I am correct, @DINO's system requires you to know or look-up the 'standard' orientation of a tile and then notate that tile with its placed orientation.
In contrast, my system requires that the faces be recorded from the perspective of the notator, starting from the left-most face.
To cope with different configurations of the features, such as splitter tiles and non-splitter tiles, my system makes two assumptions:
- city faces are assumed to be joined unless designated to be different, such as 'C' and 'C2'
- road faces are assumed to be separate unless there are exactly 2 road faces, in which case they are joined. Roads may be forced to be joined by giving them the same designation, such as 'R1R1R1F' being a roundabout junction, or forced separate with different designations such as 'RCR2C' for a city extender with roads either side
In the attached documentation of the notation I have also included a couple of notated games, including the final that won me the UK Championship on Sunday. I have not tried recreating a game from the notation yet, so I expect there will be a few mistakes!
If anyone wants it in Word.doc format I will happily make a link available.
Here is a snippet of that game to give you a flavour of the notation:
05/06/2022, ~15:45 BST, UK Championships Final, Real-time 15 min
A = wallaceprime (Chris Wallace) <yellow> 1st (104)
B = Uhome (Daniel Cheng) <blue> 2nd (103)
0 (0,0) FRCR
1A (0,-1) FRFRm
2B (1,-1) FFFFKm
3A (1,0) CFCmF
4B (2,-1) FRCmR
5A (0,1) FCmRR
6B (1-2) FFFFKm
7A (0,2) CCFC
8B (3,-1) CFC2mF +4
9A (3,0) RRRF
10B (4,1) CRRC
-
If I am correct, @DINO's system requires you to know or look-up the 'standard' orientation of a tile and then notate that tile with its placed orientation.
Actually, you don't need to memorize anything ;) The "standard" orientation you are referring to, is generated using the consolidated tile reference, which is quite easy to do on the spot.
As to using my notation for Carcassonne beyond the base game, stay tuned for Project 1B - the general notation. Project 2 will come out before then though.
-
Thanks for the clarification @DIN0 - I looked forwards to Project 2 :yellow-meeple:
I've just finished recreating a game from my notation and found loads of mistakes! My excuse is that I was notating from a YouTube video and I'm sticking to it! ;)
However, I've corrected that game - attached - and also added some additional notes with my thoughts about various moves and the tactics involved in that championship final if anyone's interested.
-
Just one suggestion. I like that farmer placement of field obedge with road.
Mark for farmer is nice abbr. for Fm (field meeole) but I will stay on that standardised form Fm on edge with only field, because people know tiles by its configuration.
This also works fine with base game tiles. When wanted to use on at least Large xpansion Inns and Cathedrals you will stack.
Imagine tile RRRR not from basic game but from I&C two separated roads, or CCCC with 4 small cities and field on the middle.
Actually when we are on basic game.
What about Abbot and Gardens.
Odoslané z SM-A202F pomocou Tapatalku
-
Thanks for your comments, @Bumsakalaka :)
The different forms of RRRR are easily accommodated with the road numbering system:
(https://www.wallaceprime.com/carc/CarcCentral/044_RRRR.jpg)
-
This is similar to mine suggestion but I used R1r1R2r2 or R1R2r1r2 you used RRR2R2 and RR2RR2 which is quire more readable.
Still issues CCCC tiles with field inside which can be CC2C3C4 and field and garden in C2 inside.
And what about CCCC with Cathedral?
-
HiG has unique tile IDs for each tile. Have you guys considered using that instead?
-
HiG has unique tile IDs for each tile. Have you guys considered using that instead?
In that case, you can make own list with IDs
1 = FFFF with monastery
2 = FFFR with monastery
etc.
But maybe this will solve this et least for official expansions
-
See @DIN0’s earlier post about the system design:
The order in which the details about the game are ordered in the notation are a logical, esthetic and practical compromise between the many ways one can record notation. There are several different approaches and each has slightly different preferable order:
real-time participant ([someone is simultaneously] playing and recording)
real-time observer (the person writing is not playing)
writing notation off of video recording of past game (this is not real-time, but instead transcription of visually recorded game into a text form)
writing off of BGA replay, or similar feature (also not real-time, uses replay features of digital versions of Carcassonne)
The player count can also affect these things as seen in the recent post.
A simple tile-index system is obviously better for a computer, but not at all practical for a real-time human logging system. I also considered recording without a separate orientation field, always just writing down the edges in order as placed on the landscape, like @wallaceprime, but after a bit of practice, I definitely preferred @DIN0’s convention. It really is quite simple to learn the tile names, as well.
I’ve invented my own additions to handle some expansions, but I tried to stay close to the intention of @DIN0’s base game protocol and I will adopt whatever system @DIN0 eventually publishes for this (Project1B). There is a lot of value to a common standard, so I think we should all try to adopt @DIN0’s convention unless we have a really compelling reason for alternatives.
-
HiG has unique tile IDs for each tile. Have you guys considered using that instead?
In that case, you can make own list with IDs
1 = FFFF with monastery
2 = FFFR with monastery
etc.
But maybe this will solve this et least for official expansions
It makes more sense to me use an existing standard rather than defining a new one.
-
Regarding unique ID's for each tile, I think this is perfectly fine for digital storage, such as BGA's tile_36 being the starter tile, but I don't think it works so well for manual notating where I feel that a system that doesn't require looking-up a tile works more easily. [edit: sorry, I've just seen this was addressed in a previous post]
-
Considering the inherent complexity of the task, I think that different approaches suit different mindsets. However, as long as any notation systems are fully documented, then it should be relatively simple (I think) to write a conversion utility to swap between systems. This would allow people to notate in whichever system they feel most comfortable with, but then share them in the format(s) that anyone would be familiar with.
Out of interest, which of the following are people intending to do?
- notate their games manually for personal reference
- notate their games directly onto a computer for personal reference
- convert their manual notation into a digital format (text, Word, Excel etc) for personal reference
- share an image of their manual notation sheet
- share their digital notation
-
Just to quickly add my own input, as @kothmann and @wallaceprime had said, the ID system is not ideal for human use, especially real-time.
When I was developing my notation system, I ran through numerous iterations of different solutions and/or permutations of individual notation elements. The aim was to make something ergonomic for human player, yet comprehensible for a computer, should someone try to develop digital tools utilizing the system (which did actually happen already).
After much deliberation, I ended up using the current system as a compromise between multiple specialized intended uses. Inevitably, there will be situations when a slightly different variant would be more optimal, but the current version is close enough to all of them. The trade off in these small optimizations is having a single global standardized notation system, which anyone can use and is supported by external tools, sheets etc.
I tackled all of this while keeping it as simple and readable as possible.
Of course this was (and still is) even more of a challenge in Project 1B. This one has to account for exponentially more information, while introducing as little complexity as possible.
I am preparing to release a short version of Project 1B, possibly in late summer - it will not be as comprehensive as the full thing but at the very least it will contain the full version of the extended tile reference. So the project release order at the moment is as follows:
Project 1B short version,
Project 2,
Project 1B full version.
I will not go into detail here but the extended tile reference is already complete and robust enough to cover all the existing tiles and even many potential new configurations. The symbology extensions are designed to be easily readable, logical, intuitive and low complexity.
-
This is similar to mine suggestion but I used R1r1R2r2 or R1R2r1r2 you used RRR2R2 and RR2RR2 which is quire more readable.
Still issues CCCC tiles with field inside which can be CC2C3C4 and field and garden in C2 inside.
And what about CCCC with Cathedral?
If the field existing between the four cities is too cryptic/hidden, perhaps this could be made explicit with a field suffix: CC2C3C4Fh (F for field, h for herbs) or CC2C3C4F<gdn> or CC2C3C4F<garden>.
In the same way that a cloister surrounded by 4 field faces is, in my version, FFFFK (with K being the cloister), perhaps CCCCK would be a memorable/logical cathedral, and KKKK an abbey. In this way, the K essentially represents any ecclesiastical building, its precise meaning relying on its context.
-
Out of interest, which of the following are people intending to do?
I have no personal interest in a notation system; I’m just thinking along with you guys. The unique tile IDs would definitely be more cumbersome for humans, but if you had a need for precise tile info I think it would be necessary. I think the feature-based notation that we’re all familiar with could serve as a filtering mechanism for tile IDs should someone have an interest in precise information. For example, entering CCCC into the filtering mechanism would display the tiles and corresponding IDs for all CCCC tiles.
Sent from my iPhone using Tapatalk
-
Fear not, the extensions I came up with enable to identify specific cccc tiles without the use of IDs ;)
-
Out of interest, which of the following are people intending to do?
- notate their games manually for personal reference
- notate their games directly onto a computer for personal reference
- convert their manual notation into a digital format (text, Word, Excel etc) for personal reference
- share an image of their manual notation sheet
- share their digital notation
I’m interested in all of these, but mostly would like a pencil-and-paper system that won’t slow play and can be easily shared. I implemented @DIN0’s system in a Google Sheet (https://docs.google.com/spreadsheets/d/1Cj4aYxrArGK_yctc-7JKGsvsIWqjZ15gWDOlnOUTdOA/edit) (details earlier post in this thread) that can be used in real time, but I do still prefer paper.
A system that I think would be very fast and easy would be to log only the coordinates of each tile, and the type and location of each meeple deployed, and then use software with a photo of the landscape, after final scoring, to complete the tile type and orientation. I would love to try this, but my python skills aren’t up to it…yet.
So the project release order at the moment is as follows:
Project 1B short version,
Project 2,
Project 1B full version.
We’re eager to see them!
-
I would love to try this, but my python skills aren’t up to it…yet.
I hope that's foreshadowing :D
I also felt the need to mention carlium's app, since the discussion is currently touching this. It can also be found in earlier posts of this topic, just like kothmann's sheets.
-
So the project release order at the moment is as follows:
Project 1B short version,
Project 2,
Project 1B full version.
We’re eager to see them!
Oh yes - very eager! :yellow-meeple:
-
A system that I think would be very fast and easy would be to log only the coordinates of each tile, and the type and location of each meeple deployed, and then use software with a photo of the landscape, after final scoring, to complete the tile type and orientation. I would love to try this, but my python skills aren’t up to it…yet.
Gosh, I imagine that would be incredibly difficult! I've had enough difficulties in C++ identifying tiles and meeples from screenshots of BGA games when everything is geometrically straight, let alone from real, imperfectly aligned games! One of my biggest aims has been to capture replayed games from BGA to build up a database of past games to analyse and hopefully data mine little gems of information, such as when is the most profitable time to start a farm.
Unless my web-scraping skills aren't up to it (a distinct possibility), I don't think BGA records meeples once they have been removed and scored so, unless you grab the information at precisely the right moment, you miss recording that aspect. Consequently, I have put my efforts into trying to get the computer to assess tile and meeple positions quickly enough from screen captures to map in real time what is happening. I got a fair way into this task, but the program was getting rather messy, so I am in the middle of a full rewrite. I've called my software WallyCarc :)
-
when everything is geometrically straight, let alone from real, imperfectly aligned games!
So that's why you have been making all those magnetic tiles! :D
a database of past games to analyse and hopefully data mine little gems of information
A data mining tool would certainly be useful. One of the reasons I started the notation in the first place was to enable people to study their games as well as games of others and possibly make some statistics.
-
identifying tiles and meeples from screenshots
I think a “clean” landscape would be much easier for image processing. Are you using openCV? Some other library?
unless you grab the information at precisely the right moment, you miss recording that aspect.
I had the thought yesterday that it would be fun to develop an AI that used only the final landscape plus the final score to generate a “most likely” game log. I think if you had the tile position sequence, this could be possible (that is, infer the meeple placement from final score). Not for me, but possible in the abstract. >:D
…so I am in the middle of a full rewrite. I've called my software WallyCarc :)
We’ll be super eager to see this, too!
-
I think a “clean” landscape would be much easier for image processing. Are you using openCV? Some other library?
I'm brute-forcing it in C++.
I take the following steps:
- capture the screen
- do a maze-walk to find the periphery of the playing area
- search for a couple of dark grey tile placement positions (so I can find the tile size)
- calculate the tile boundaries in the playing area
- reduce each tile position to a 5 x 5 pixel matrix leaving a small border to allow for pixel misalignment
- compare each reduced tile to a list of all 2116 tile/orientation/meeple combinations (just for the base game at the moment) to find the best match
Here's a composite image of all the 5x5 pixel combinations!
(https://www.wallaceprime.com/carc/CarcCentral/045_combinations.jpg)
It wouldn't be of any use at all for analysing photos of games played with real tiles on a table; I can't imagine players reacting to well to repeatedly being told that "you can't put the meeple on that side, it expects it on the other" or "well that doesn't work, your meeple is 2 mm away from the optimal position" and so on ;)
-
I originally tried looking at various locations within a tile to decipher whether it was city, road or field, but it got really tricky when a meeple almost occluded visibility of the small segment of road it was placed on, such as with a CCCR tile (so it couldn't tell if it was a farmer on a CCCF or a road meeple on a CCCR) or similarly with a RRRR being interpreted as a farmer on a RRRF tile.
I'm surprised I've got any hair left at all after all the hair-pulling that all those tweak/test iterations took! Comparing all 25 pixels seem to have worked flawlessly so far; I expect the slightly different layouts make enough difference for the matching to differentiate between them.
-
but it got really tricky when a meeple almost occluded visibility of the small segment of road it was placed on,
You could either make meeple substitutes, or even better clear elevated stands for meeples.
-
Here's a composite image of all the 5x5 pixel combinations!
WOW! This is fantastic! Thanks for sharing.
-
To find tile boundaries, I originally used a hue, saturation and value test to see if a pixel was a tile or not. Unfortunately, it got confused by meeples that hung over the edge of a tile; indicated with red circles below:
(https://www.wallaceprime.com/carc/CarcCentral/046_tile_pixels.jpg)
Rather than looking at every pixel, I used recursion to find the edges of the tile placement positions, where the function that tries to find a corner initially starts with a step size of 16 pixels, but then calls itself with half that step size when it overshoots. This way, if the edge was 63 pixels away for example, instead of checking all 63 pixels, it would only check at positions 16, 32, 48, 64, 56, 60, 62 and 63, so at just 8 positions. The tricky bit was tailoring the algorithm to ignore when meeples overhang the edge of tiles, looking ahead to see if it went back to being the background.
Ultimately, however, it was easier just to look for the dark grey placement positions:
(https://www.wallaceprime.com/carc/CarcCentral/047_placement_pixels.jpg)
I could then pick a couple and look for the smallest dimension and that was the tile size. I did have to go one step further, however, because BGA must calculate the corner positions in floating point, but then use them in integer form. This had the effect of sometimes causing tiles sizes to fluctuate between two integer sizes, such as between 54 and 55 pixels. This meant I couldn't rigidly use one size as, going across the playing area, the error would increase sufficiently to throw the tile boundaries off.
If any of you find this sort of stuff interesting, I'll create a new subject for it, but I know it's not everyone's cup of tea and I don't want to distract here from @DIN0's sterling work on his notation projects.
-
If any of you find this sort of stuff interesting, I'll create a new subject for it, but I know it's not everyone's cup of tea and I don't want to distract here from @DIN0's sterling work on his notation projects.
Please do. This definitelly deserves its own topic. This is some interesting stuff you are working on! I'll be watching the progress with great interest :(y)