New Backgammon teaching/learning site?

[sita]

Hi all,

I plan to start a new backgammon-site, the BTL (Backgammon-Teaching-Ladder).

The basic idea is that participants of the ladder volunteer to comment on games of other (weaker) players, and receive commentaries of their own games in return. You need some volunteers at the top of the list who will not be able to sent in games, because there are no stronger players, of course. I am quite sure that a commentor will benefit from doing the comment itself - if you want to tell others about the game, your concepts, etc., it also helps to clarify these things for you.

This idea is not completely new, it exists already for the ancient board-game go. The archive of the GTL is aproaching 6500 reviews, have a look there to get an idea what a teaching ladder means.

I could set up such a teaching ladder also for BG, and maintain it. But before I start working out the details (which would be a bit work), I wanted to hear your opinions. I see some problems here, which might it make questionable if the BTL could be a success, like the GTL:

- .sgf is far from beeing a standard file-format for Backgammon games. I really wouldn't like to maintain any other format, though, because I am quite familiar with sgf, and there exist already an excellent tool for working with these files (namely sgfc).
- Most Backgammon players know how to use bg-programs, but to my observation nearly nobody knows that it is possible to annotate moves with gnubg. Furthermore, the way gnubg supports these anotations is way beyond the standard of go-programs.
- We have already the bots who tell us our mistakes. They tell us only which moves are wrong, and not why they are wrong. Anyway, this limits the benefits a BTL could have.

I am intersted in your opinion to the following questions:
- Do you anyway think that ths BTL might be a good idea?
- Could you imagine to participate in the BTL as a volunteer (i.e. reviewer)?
- Last, but not least: I need someone helping me to write all the necessary texts. I could write them myself, but I need at least a native speaker, or someone speaking english better than me, to find all my spelling/grammar mistakes.

sita

[wintom]

Thank you sita!  I think this is really a brilliant idea!  I would really like to participate in the project. I am neither a native speaker nor at the top of the ladder, but I am really interested in the format. 

wintom

[Hardy_whv]

Sounds very interesting and promising indeed 

Three questions:

• Commenting whole games (or even matches) is quite a time consuming job. You want games to be commented or single positions? (I actually think - no, KNOW - that I dedicate not enough time in analyzing my own matches already ...)
• How will you set that up? Via some forum software?
• What kind of help do you require? I am happy to help, if my time permits.

Hardy 

[sita]

Commenting whole games (or even matches) is quite a time consuming job. You want games to be commented or single positions? (I actually think - no, KNOW - that I dedicate not enough time in analyzing my own matches already ...)

I think commenting whole games/matches makes more sense for this approach. You can already use a normal forum to discuss positions. The advantage of the whole match is that the commentor can spot the really big conceptual mistakes the player has. Even so they might lead to comparatively small equity-loss in a given position, pointing them out will help the player to a deeper understanding of the game.

How will you set that up? Via some forum software?[/li][/list]

I plan to use the same software as the GTL, so just follow the link to get an impression.

• What kind of help do you require? I am happy to help, if my time permits.

I mainly need reviewers, of course. In the process of setting it up I will have to enhance sgfc. Thus, a good C-Programmer with a decent knowledge of sgf, and backgammon could help there - but I fear I have to do that myself.  Beside writing new texts, the backend need some very simple changes (mainly integrating a new rating-system). If it turns out that it is more work, I could also need a php-programmer.

At the moment I am a co-maintainer of the GTL. Since I assume that there will be much less traffic on the BTL, I think I can handle it alone. If it turns out that there is a lot of traffic, I might need another maintainer there, too.

sita

[Hardy_whv]

I had a look at the Go-Site. Very interesting. So you need to include GNU Backgammon as a viewer or do you want to programme one yourself? I fear, I can't help you on the software side, though I studied computer science ... but thats 20 years ago now and I am out of that business since long.

But I am happy to help with anything else.

Do you have sufficient webspace available?

Hardy 

[sita]

So you need to include GNU Backgammon as a viewer or do you want to programme one yourself?

Neither this, nor that. I plan to make them available for download in sgf-format. But you can of course configure your browser, so that it opens sgf-downloads automatically with gnubg.

sita

[sixty_something]

sita, i think this is an excellent idea .. i've been documenting some of my own games and thinking of ways to enable multiple commentators at all skill levels to contribute to that commentary with both analysis and questions of there own .. i've also been using GNUbg to do this .. so, it seems we've been thinking along the same lines

to annotate games can indeed be quite time consuming .. i started out trying to document every move and cube decision .. frankly, it became absurdly detailed and time consuming .. while documenting *everything* may have some benefit in that it provides for seeing into the big picture of overall game strategy rather than the immediate tactics of the position, it can become almost as exciting as "watching paint dry" .. watching a Picasso painting dry or a Nack Ballard tournament game might be well worth it, but few matches, imho, deserve that close an examination

therefore, i decided to *only* evaluate moves and cube decisions GNUbg marks as doubtful, bad, and very bad .. this further provides the ability to quickly page through an annotated match using the vcr buttons above the GNUbg game record panel (the ones with the question mark) to go to the Next marked move or my preference the shortcut keys (Shift+Ctrl+Page Down)

this makes even a lengthy match quick and easy to review without the added "noise" from unmarked acceptable decisions .. additionally, the flow and sense of games is never lost as the user always has the ability to look at every move in any game at any time .. those moves just have no comments associated with them other than GNUbg's already excellent metrics .. further analysis, annotation, and attempts to share with others has led to other observations which i'll save for later

i'll make my annotated matches available for you or others to see with the caveat you remember these are first draft attempts at moving toward a useable shared annotation tool .. some of the comments and ideas may be good ones; others may deserve to be forgotten

if you'd like to see them, just message me here or at fibs .. you can also send me an email from my profile link here .. either way include your email address and i'll reply with a sample set of my annotated .sgf files attached .. of course, you'll need GNUbg to open and view the files or an ability to convert that format .. i'll be happy to share them and enjoy the feedback both laughter and cheers ..

i'd especially like a few viewers of these files to follow my suggested instructions and add comments of their own which can become the new current master .sgf file for that match .. this would further test whether enabling multiple commentators to comment and/or ask questions within the commentary panel may indeed be practical or useful

again, great idea, sita .. i also look forward to seeing the Go site .. it's another one of my favorite games, but i've never been able or willing to find the time to really learn it .. finding other Go players outside college communities was never an easy thing to do .. the internet, of course, makes that excuse history

sixty

[sita]

In fact, I do not so much think along the lines of gnu error-rates. The equity loss of a decision says little about the importance of the mistake regarding the comprehension of the game.

An example: Say I would annotate a game of a beginner, who still makes huge stacks, and tries not to leave a single blot if possible. I will try to find situations, where he could leave indirect shots to enhance the probability of making key-points in his next roll. Maybe I will tell him something about the probability of being hit, and compare it to the probability of making a big improvement in his next roll. I will focus on these mistakes, because that's what he has to learn. Maybe he made somewhere a huge blunder, not switching points in his home-board with a double. Switching might have increased the probability of scoring a gammon by a large amount. It is a huge blunder - but it is not what the beginner has to learn. He has to learn to take small risks to build key-points and primes.

That is what I mean, when I say "overall picture". Not commenting every single move. Instead finding the next step the player has to learn - searching for examples in the match, and explaining these.

Writing this I am wondering whether a gnu analysis might be even hindering a good comment...

sita

[sixty_something]

yes, i know what you mean, sita .. sometimes the wealth of information gnu provides is just overwhelming, especially for new users

i have a different approach which is much more driven by equity loss .. imo, blunders like stacking can come out by focusing on such wasted equity because gnu will almost always mark those moves anyway

my focus in analysis has shifted entirely toward collecting my own moves marked very bad by gnu for analysis both after play and later by collecting these as saved positions .. over thousands of analyzed games and accumulated summary stats maintained separately in Excel, i've found such errors have an average equity loss around -0.40

correcting just moves and cube decisions like that can and has made and enormous difference in my level of play .. such narrowly defined focus may seem too limiitng at first glance .. but there appears to be a steady trickle down effect on my personal play which makes sense and is measureable .. admittedly, i am not coming at it from the perspective of what may be best for a beginner .. certainly there are some differences in teaching and learning foundation skills and more advanced techniques

regardless, when first learning to use gnu to analyze games, i'd look at every single move and attempt to analyze why my choice, even my opponents choice was not the same choice gnu made .. while interesting and productive for a while .. this appraoch is incredibly time consuming and painstaking .. much too much for almost anyone with a life outside of gnu and fibs - which recently has been limited for me

thus, approaching the many opportunities for learning available in any match by narrowing the focus to where the greatest losses occur and then working back to grand strategic decisions once error rate reaches an acceptable level seems to me to be an approach that is both very doeable and perhaps most efficient

this all clisely parallels another learning path i'm pursuing regarding the review of collections of problem sets in gnu .. having annotated problem sets would make that even better .. again, all this can be done using gnu .sgf files as the foundation

however, it works out in the end, i think you've got a good idea here and we are on the same wavelength .. perhaps, just coming at it from different perspectives .. there are always many paths to the same destination .. no one better or worse particularly .. just different styles for different learners and teachers alike

[sita]

I just had an intersting discussion with sixty-something, and want to share the essence with the community.

We came to the conclusion that it might be enough to have a new forum here on fibs-board, where we can start threads with questions to positions, or even attached sgf-files. Other people can read it, answer the questions, look into the files, comment on moves, upload the files again, etc. Maybe all else we need is a text describing how to use this special discussion-board.

This should be enough for a first try, in my opinion. If we see that we have really a lot of traffic, there is still the option to change to something "better".

sita

[Hardy_whv]

Okay, then we could ask webrunner, whether he would support a seperate board called "BG School" (or something similar) under "Backgammon". Let's see how much interest there is.


Hardy 

[webrunner]

Just say the word and i will set it up

[socksey]

I need at least a native speaker, or someone speaking English better than me, to find all my spelling/grammar mistakes.

I'm your girl for spelling/grammar mistakes. 

I like the idea of keeping it all here on fibsboard.  Being in one place is quick, and easier to navigate. 

socksey



Life began with waking up and loving my mother's face." - George Elliot

[dorbel]

I'm your girl for spelling/grammar mistakes. 

I like the idea of keeping it all here on fibsboard.  Being in one place is quick and easier to navigate. 

Possibly not as hot on grammar as she seems to think!

[webrunner]

There is a lot of facilities here on fibsboard that can aid in this, like the posting of articles, forum and of course.. there already is a steady amount of readers here.
I am willing to add stuff if possible so just say the word.

[socksey]

Possibly not as hot on grammar as she seems to think!

I may not be grammatically and spell perfect, but I am fairy good, and have volunteered.     However, if you think you are better qualified to edit, you may certainly volunteer for the job, dorbel.   

sockey



More than anything I must have flowers, always, always. -Claude Monet

[PersianLord]

this is a great idea. i'd be delighted to participate.

[cutecub00]

Great idea, .. i also look forward to seeing the Go site .. it's another one of my favorite games, but i've never been able or willing to find the time to really learn it ..
advertising?