Exactly. The highest priority is and should always be promotion of Go to children and young people,
since they are the next generation. Focus on a US pro system is focusing on the top, which is completely backwards and upside down, IMHO.

Yertle -- Could you pm me or mail me (operations@usgo)? I can find out what happened there.

For scoping out what we want to do, that's the board. For people who figure out how we want to get there, well, that's me. That's kind of one of the reasons i started this thread...

People who want transparency, or action, from the board, would do well to remember that the board is not actually designed to get anything done. It's designed to decide on *what* to get done. The number of people who volunteer to decide what to do will always outnumber the people who volunteer to actually do it. They're both very important jobs, especially if you feel like things are going in the wrong direction. For my part, i think board membership would probably make me an alcoholic (hi Lisa! ) but that's me.


Re: ratings certificates. that's something Allan has told me to make a priority. pwaldron, all that work you did on it has not gone to waste -- i'm going to be picking that up very soon and putting it into place once usgo.org has migrated to its new home.

As for other general priorities, my priorities have been as follows:
1) get everything out of the way so jkk's membership db could go live (check)
2) phase out outdated elements of the website (working on it)
3) reduce our online costs & improve responsiveness to the website. (working on it)

and for a while in there i had #0 - help Lisa direct the go congress.

Once those priorities are done, i'm going to re-order the rest of my list, which includes things such as the ratings certificate program, an online pairing program, other things to add to the 'members only' sections of the website, re-tweak the ratings algorithm to be more responsive (i'm counting on pwaldron's brilliant work there to help me out again ,) and a few others besides. Thus again, this thread. What would you like to see?

One of the first things on my list is to try and get the "build environment", as it were, as clean as possible, so that someone can clone our git repository and get dev/local versions of the ratings system, member db, website, etc., up & running. This should let a lot of the potential volunteers write some features, and push them back to us with as little friction as possible -- i.e., put up or shut up.

For that matter, if anyone wants to help with that process, let me know

gurujeet -- in addition to what justin mentioned about the API, we're hopefully going to be extending the documentation on the aga wiki. Assistance is welcome on that one.


Lastly, and you can consider this an open question to flame on about, but:
if the AGA were to implement a separate, online-only rating system, how would you like to see it implemented? Should it be integrated with certain servers? should it allow uploaded sgf's, or scrape server archive pages where those sgf's are automatically placed? How would you like to see the ratings information presented?
Think mockups & wireframes of what you'd like it to look like from going to http://www.usgo.org to finding your online ranking. Thoughts? Feelings? Ideas?

still 30k at these boards,
-andrew

It's a very interesting idea - especially for those of us AGA ranked but who don't make many tournaments (a few years ago I know people talked about an East Coast/West Coast split in AGA rankings due to little cross-over).

*** The most important part is to get buy-in from those servers. ***

With this information, the AGA is poised to make the following things possible:

1. API available ranks across servers. For example, L19^2 could interface with the AGA API and provide real, up to date ranks for a number of different ranking systems (AGA, KGS, IGS, Tygem, etc). It'd also be nice to correlate various usernames and IDs. (AGA# 11987 = L19 Malweth = KGS Malweth)

2. Correlation of ranks across the board. AGA 1k = KGS ? = IGS ?.

3. Micro-adjustment of AGA (face-to-face) ranks based on online ranks.

4. Online AGA tournaments.

5. Online games (tournament games, perhaps) that affect face-to-face rankings.

The difficulty is time. I personally don't think Blitz games should be ranked by an organization that promotes 30 minutes per side (plus a minimum of :15/move byo-yomi). This is an added difficulty in parsing out time controls. If blitz games aren't ranked, what time control minimums are required for online play? When I automatch on KGS I get 25 minutes per side, which doesn't meet the AGA tournament regulation.

_________________

Of all the million things the AGA could do for its members, especially in the wake of the enthusiasm created by an apparently very successful congress, the only thing that stands out in the follow-up so far is a call/promise to "improve" ratings. Administrators giving members what they want can justify it just on those grounds, I suppose, but how do the members themselves justify this obsession with numbers?

First, if you are an amateur you are weak anyway, so a even a grade (as in "I'm 1-dan which is like a black belt in judo") is not likely to impress anyone much. If you want to measure your improvement, OK, but getting a spuriously accurate number like 2345 is meaningless. You may be the equiavlent of 2319 or 2678 the next day without having played a single game or studying, depending on your bowel movements, the bills that arrive in the mail, or finding a valuable antique in the loft. What's wrong with simple 1-dan, 2-dan, etc? They may not be finely calibrated, but you know you're improving anyway if you beat the guy who used to beat you, and being 2345 has no real value in predicting whether you'll beat a 2344 guy.

I've asked this question on the BGA forum before and never saw a satisfactory answer. The nearest I can get to an answer is that western go has a lot of number-oriented people. Liking numbers doesn't necessarily make them useful, though.

This is not idle sniping. I think a rational answer would benefit national associations and adminsitrators. If there is no rational answer, I think administrators should work instead towards putting time and resources into other areas of go development. For those who need their number fix come what may, they can e.g. use the graphs on kgs instead.

Your point about prioritization is well made, so I will just comment on this:
1. Those numbers do make a prediction, that if you play the 2344 guy 100 times, you should win about 50.
2. My understanding of "tweak" means something like a tiny formula change, not a complete overhaul. E.g., we won't be switching to the EGF's Elo style presentation.
3. The AGA rating system does seem a little sluggish; for example, my friend is 1.94 dan, wins something like 75% of his games, but his sigma is low enough that he never seems to move more than .01 per tournament. I'm certain that he would have at least a 50% win ratio as a 2d.

_________________
That which can be destroyed by the truth should be.
--
My (sadly neglected, but not forgotten) project: http://dailyjoseki.com

Please let us know when the code for the usgo.org site is available on the aga's github page, I'm probably not the only person that wants to take a look.

_________________
That which can be destroyed by the truth should be.
--
My (sadly neglected, but not forgotten) project: http://dailyjoseki.com

I'll give you two reasons for the Western go organization's obsession with . The first is that ratings appeal to one's feeling of making forward progress. So the organizations focus on something that their members have some interest in.

I submit that the other reason the organizations focus so much on ratings is organizational deficiency. The North American go organizations (Europeans as well?) do not seem able to bring together a team of people for a project. They aren't able to identify a concrete goal and bring together the necessary talent for the time required to see it to completion. Ratings are one of the things that a single person can do, and so that's where the people go.

OK, but I said that already. It the members I want to hear from. Feelings of progress based on spurious numbers are not rational - and as in the adverts for financial investments: be aware that what goes up can go down.



But 2-dan playing 2-dan makes, as near as dammit, the same prediction. The accuracy of the extra digits is entirely spurious given all the other things going on in a typical amateur's life. Your point about sluggishness in the system just makes the spuriousness clearer.

Pros in Japan get by without such a fiddly system. The Chinese have an Elo system that only matters four times a year, but I get the impression that less and less attention is paid to it. The Koreans have fiddled around with such systems, but that work has become otiose. In both cases, though, the reasons for the system had absolutely nothing to do with pros wanting to know if they were 2-points up yesterday, 3-points down today. The Chinese introduced their system as a way of trying to be fair (in the wake of many complaints) in allocating seedings and trips overseas. In other words, it matters for only a small percentage of players, as far as I know. The Koreans tried ratings as a way of drumming up interest among the public, along the same lines as golf's PGA rankings and tennis's LTA rankings. So far as I can see, the public there paid not a blind bit of interest.

So, this is a western amateur thing. Again, why?

daniel_the_smith -- We have some scrubbing of licenses, etc., to do, before we can put the whole shmeer up on github for all various & sundry. We do have a nonprofit organization account there, so if you'd like access, you should pass your github handle along to Alf or myself so we can add you as a member & you can get access.

Mr. Fairbarn-
The reason to have an "online only rating" is because for large swathes of the central US region, commuting to tournaments is just not an option. There are any number of people who see no value in an AGA membership because they see the top benefit of membership as being the ability to play in tournaments. Now, if you can have tournaments held across sites (e.g., 5 in this coffeeshop, 10 at this local club, another few in a library, etc), then you might see a lot more interest.

Secondly, i'd like to think that having a real, honest to god person associated with an online handle can help address a lot of the problems in online go. people playing with free pseudonymous accounts will have no problem escaping, making a new one, escaping, making a new one, etc.

Thirdly, i'd like to point out that ranks have consequences. It sucks when you've improved, your rank hasn't changed, and you get nothing but easy games at tournaments. It's just as bad when it's the other way 'round. Let's not forget the point of ranks, which is especially useful for us duffers -- so that we can play games that are challenging for both players.

Lastly, i'd also like to point out that i'm really trying *not* to get into a discussion about *what* should be done or *why* it's being done the way it is -- the board did that. I'm trying to get into a discussion of *how* it's going to get done, ask *who* is going to do it, and figure out *when* it's going to happen. Hope that helps. We could open another thread to debate it further, but in the meantime, i hope this might satisfy your immediate points.


Pwaldron-
I think your criticism is quite fair. I would ask a favor of you, though, to stop referring to it as "the AGA", and instead refer to the individual people involved. It's tougher to write sentences calling out individuals, but i ask you please to think of *who* it is who is doing all these things. We both know that there is no shortage of people willing to talk about the problems (e.g., the Board, or this forum ) but when it comes to *who* -- there's your problem. For instance, the board votes to put into place rank certificates. Who's going to actually connect all the dots? You got it 90% of the way there, and there it sits. Who did you expect to pick it up? Allan himself? Terry B, Chris K, Ken K? Which person from the AGA did you honestly see or expect to get it finished? Now i'm telling you i intend to pick it up and get it done and i hear this "the AGA can't identify a concrete goal and bring people together"... downright disheartening would be the beginning!

If that "they" refers to me, well, then i refer you again to the point of this thread, which is to a) get community input before setting a series of concrete goals and b) soliciting volunteers to help meet those goals. I don't know which "they" you are referring to, but i have identified a series of concrete goals, i'm trying to bring together the necessary talent (i'd say we have a hell of a good start!), and by god i intend to see it through

Can I get this quote as a custom forum title, please? Thanks.



I *really* enjoyed the YNOT last year. I think if that tournament repeats annually, it goes a long, long way towards establishing the AGA's online presence. Even though questions about what to do with results (they did not impact AGA ratings at all) are still out there, at least we were all there, playing online.

But this is not what I asked about nor does it relate to the point I was making.

There are ranks/grades (2-dan etc) and ratings (2345). I'm asking members why they make so much fuss about (I think) spuriously accurate ratings. Take away the fuss about that and the work on the AGA project becomes much simpler. You can have online ranks more easily than online ratings and at amateur level ranks are just as meaningful as ratings. KGS uses online ranks. UK meat-space tournaments use ranks. Possibly I've misunderstood and your system does after all also use ranks rather than ratings, but I've seen people talk about 1.95 dan, etc so the impulse to seek spurious precision is still there. Yet again, why? I understand the wish/need to have a rank. I don't understand, among people whose go strengths vary enormously from day to day through reasons nothing to do with go, the wish/need to have a rating, especially when it evidently involves so much effort by already hard-pressed administrators (which is why I think it belongs in this thread).

A few comments on AGA ratings, mostly directed at Mr. Fairbairn:

* They're published as dan/kyu ranks, even though internally the system uses an Elo-like scale.
* They're published with 5 or so decimal places.
* The first decimal place is somewhat important (e.g., officially, a 2.9 dan should take white, no komi vs. a 2.1 dan, but a 3.1 dan and a 2.9 dan should play even, 7.5 komi).
* No one I've met pays any attention to the rest of the decimal places, unless it's as a way of complaining about how slowly they're going up.
* You're not allowed to enter a tournament at a lower rank than your published one.
* You can enter at a higher rank if you can convince the TD you should.

_________________
That which can be destroyed by the truth should be.
--
My (sadly neglected, but not forgotten) project: http://dailyjoseki.com

As an AGA member who about 99% of the time is more or less inactive in the AGA (mainly due to geographic/logistic considerations) and would like to be more active -- Do you have any tasks/projects it would be possible to work on remotely that don't involve coding? Whenever I see calls for volunteers it's usually for some type of programming solution (which may just be due to the disproportionately large number of CS people who play go....When you have a group of people with hammers, the world looks like a box of nails...). Is there anything the rest of us could be doing (aside from recruiting more go players)?

I'm not sure those goals are mutually exclusive, because among the first to qualify for pro status may be players who are children today. It's a different way of focusing on children. Just, well, elite children at first. Longer team, more pros in the U.S. would mean more possibility for outreach of all kinds, though.

But like with anything, it's important to balance long-term stretch goals with shorter-term activities that provide more immediate benefit.

Yeah, it's a bit of a chicken-and-egg problem; having a pro system in place may actually give kids a reason to get strong in the first place (they would no longer have to travel overseas to "be" Hikaru). Something like 1/4 of the participants at congress were kids; whatever Paul Barchilon has been doing seems to be working.

_________________
That which can be destroyed by the truth should be.
--
My (sadly neglected, but not forgotten) project: http://dailyjoseki.com

Trying to threadjack a bit: what is the limiting factor in intensifying youth recruitment/education right now? Lack of plans, lack of volunteers, lack of resources?

Depends how you count success. The AGA has been pushing youth activities for well over a decade, even before Paul came along, The kids who were teenagers then should be dues-paying members in their twenties and thirties by now. But looking at the membership rosters those players are nowhere to be seen. Past Redmond Cup winners--James Chien, Timothy Law, Dennis Liang, Jonathan Wang, Mozheng Guan--aren't playing any more. In fact, if you look at all the Redmond Cup winners and runners-up and remove those who are still youth players, not a single one is still playing go. A retention rate of zero is not a success.

The catch phrase is always that the youth players are the AGA of tomorrow, but reality doesn't bear that out. The kids stick around as long as they are being subsidized and then vanish once the money stops.

Good point.

I'm not sure how to count that information, however. There's lots of stuff I did as a kid that I don't do anymore, and I don't consider that a failing of my parents or the organizations involved-- my interests/priorities changed. We can't force people to keep playing. Of course, if they still play on KGS and live in an area with some AGA presence, I would count that as strong evidence in favor of your point.

Also, does looking at only past winners bias things? I wonder how good the membership database is; can we see how many people who start with a youth membership go on to continue to be members after they grow up? If that number is as near 0 as you seem to think, then I'll agree we do have problems. But that still wouldn't tell us what the problems are. It may simply be that the "real" AGA sucks compared to the AGA the youth experience, which is probably true but is also, I think, just another facet of the widely acknowledged problem that the AGA has not sufficiently answered the question, "what do I get for my membership fee?"

_________________
That which can be destroyed by the truth should be.
--
My (sadly neglected, but not forgotten) project: http://dailyjoseki.com

I have to strongly disagree with this asked favor.

I firmly believe that the people who make up the AGA are well-meaning volunteers who are trying to do what they think is best for go in this country.

They are also doing a terrible job.

But the last thing I want is to start having individuals called out and attacked. While the recent efforts of the AGA fall short of what I would like, and as such it would be disingenuous of me to suggest that the individuals in the AGA deserve all of our support - they do not need to be singled out and attacked publicly either.

I believe, given pwaldron's concerns and experiences - the fact that he is still here making suggestions speaks volumes about his passion for go, and his character. The fact that he refuses to single out individuals speaks volumes about his compassionate restraint.

The more the AGA continues down a road of convincing itself that the only progress that can be made is online, the sooner it will arrive at its destination - not a shining City on a Hill, but rather a depressing suburb of KGS.

The above post may be the most insightful thing ever written about the AGA. It just needs one slight modification to be the perfect post...



There...fixed that for you.

_________________
Help make L19 more organized. Make an index: https://lifein19x19.com/viewtopic.php?f=14&t=5207