What about the league structure artificially restricts opportunities to play?

We could say, "all right, everyone in the ASR should play someone else in the ASR. And... Go!" But people need a little bit more structure and ritual to feel comfortable bothering each other about games. The exact size of the divisions is to balance ease of finding games, on the one hand, with giving a slight edge to people who win most of their games, on the other. (When the big complaint isn't "I can never get any games" it's "Everyone else is playing too many games, there's no way I can catch up even if I beat most of them.")

I think there was a time when stalkor was irrationally pessimistic about the future of the league on the basis of meaningless month-to-month variations in membership.

Yes there was a time when i took a break from managing (because i wasnt having fun anymore) to get my stuff together and do some personal things.

when i came back almost 3 months later i found the league had gone back from 400+ to about 200 players so my first reaction was a bit like "AAHHHH wth is happening!!" so i made this thread, which was a bit overreacting on my part in hindsight and i posted in this very thread too.

that said ez4u has an opinion about the league and how we try to reach our goals, which should be fundamental to how we do things in the ASR.

I still think we are doing a good job in making a competitive league with opportunities for every strength to get a chance of winning and creating a place for ppl to play serious games and learn from them.

Yes, we can do more about the learning part, but it's one of the hardest things to get someone to do lectures on a regular basis and get him/her to want to keep doing this (without getting paid for, since we have no steady income of course). Tabemasu tried but it didn't last for a long time.

Also the league is not made to replace all your games you play on KGS. You should play a league game when you can and want to, but if there is no one online to play you shouldnt wait endlessly for one to get online, you should just play someone else in the room or play somewhere else like the EGR. Also keep in mind it's a casual league with different kinds of ppl like players who play 4 games per month or play 60 games per month.

_________________
admin of the ASR league and KGS admin

Let me elaborate a bit. First my POV. I am one of the people in the Asian time zones. About a year ago I read about the ASR and joined the room to lurk and see how it worked. In Asian time it basically doesn't - not enough games.

When I wrote that the system artificially limits the opportunities to play, I meant that out of perhaps 300 league members, your serious games are limited to the other members of your class, say 20 people or so. The reason that I used the term artificial is that the assignment to classes does not appear to be designed in any way to promote finding games. This is despite the fact that it appears to be recognized as a serious issue. The September news says:
"Over one third of you guys didnt make the 4 game minimum but you guys still managed to play almost 1100 games, which makes me very proud considering it was harder to find a game."

Pride is nice, but wouldn't it be even nicer to announce that 2000 games were played and that over 90% of the members exceeded their 4-game minimum? If the answer is yes, then the question is what does it take to make the league structure more supportive of finding games. Since from the results charts it is clear that not everyone plays everyone else anyway, why are the class sizes what they are? If beta were a band of 40 players rather than two classes of 20 and if gamma were a band of 80 players rather than four classes, what difference would it make?

Right now it is about 9 PM Sunday evening in Japan, which I would consider prime time. There are five games going on. That is possibly the most I have seen at one time (then again maybe not, I haven't been checking for months). Five games means that only five (max. assuming no duplicates) of the classes have a game going on.

I would also say that the league as currently organized is not based on competition. It simply is not competitive for people more than a few ranks apart to compete in even games (unless someone is misranked! ) The actual results in the league are not based on improvement but mainly on activity. The intro states:
"...the points system strongly rewards activity – you get half as many points for losing as you do for winning, so you can easily achieve a good class position with a poor record just by playing a lot of games – the reason for this? We want people to improve as quickly as possible, and the motivation to play more games helps to achieve this..."

I assume that currently your won-loss ratio is pretty much determined by the makeup of your class - how many people in your class are seriously stronger than you are and how many are weaker. From there is should be a relatively simple matter to determine how much activity is necessary to advance. However, how does that support the idea of playing serious games? My idea of a serious game is one with an evenly matched opponent where the result has some meaning. In order to promote serious play, it would seem necessary to put people of a similar rank together. Of course this raises the issue of how to review the resulting games. You would like to have someone stronger to review with. The current idea in the league is that you do this by playing with the stronger players. However, as some people have commented, the stronger players do not necessarily find this interesting. One alternative would be to try to structure the league so that the stronger players come to play serious games against each other but then also hang out to review games with weaker players. I think that may be more interesting for both the stronger and weaker members. It may be possible to arrange and then announce when certain crowd favorites will be around. This could also end up causing more activity to center around such times.

Just a few thoughts...

_________________
Dave Sigaty
"Short-lived are both the praiser and the praised, and rememberer and the remembered..."
- Marcus Aurelius; Meditations, VIII 21

ez4u, I believe most ASR players like that there is a mix of ranks. I see the ASR mainly as a vehicle for me to play even games with stronger players, and get a review; then I reciprocate by playing even games with weaker players, and giving them reviews. When I play other 4kyu, I feel kind of silly... like I'm just playing for the points. If all the players in my division were my level, there wouldn't be much purpose in being part of the ASR. (You could say that I see it as a teaching ladder with the narrative form of a tournament.)

As you say, this screws over the strongest players. But they're already (relatively) strong at go, so I say unto ye, verily, ye have had your reward.

Careful, I don't think I said it screws over the strongest players. They are there by choice. However, I do think it will bore them and IIRC some have said so in their posts here as the reason they dropped out. At the other end of the spectrum, I see some people in the results table that are losing 19 out of 20 games. Now perhaps they are taking each of those games seriously and I am just lacking in imagination, but I really can't get my head around the mix of games produced by the current class structure as the best way of facilitating "advanced study". Of course I have only opinions, unsupported by real facts, about what might be done differently.

_________________
Dave Sigaty
"Short-lived are both the praiser and the praised, and rememberer and the remembered..."
- Marcus Aurelius; Meditations, VIII 21

Are these the opinions you mentioned earlier, or other ideas that you are coyly hiding?

Again (not to harp, but to get you to divulge more about your opinions), how will playing even games advance advanced study? It's really easy to get even games. Distressingly easy, in fact: if I get up to get a drink after clicking autoplay, my time is running down before I get back. These are slow games, using the same time settings as ASR. I don't get a review afterwards, but then, it's hard to get a review with even players even in the ASR. (Mostly because the loser is often fuming, but also because neither player feels obligated to the other.)

I agree that the stronger players aren't getting screwed over! I just didn't want to speak for them, since it would be rather self-serving to say they enjoy it. I can say, though, that I'm squarely in the middle and I rather enjoy my ASR games with players who are even beginnerer than I. They can be fun even if I'm poorly rested or stressed out, I can experiment with different ideas without feeling any pressure to be perfect, and I really like it when I get the sense they're learning something from the review.

I did not mean to imply that I have other opinions that have been held back. I only wanted to make it clear that regardless of what experience might back up my opinion, none of it is experience in the ASR itself, which is presumably the most relevant type.

On the second point, I think that advanced study requires serious games where the opponents put their current best effort into winning. I do not believe that uncompetitive games are serious. Actually of course that is just a theory of mine - I've certainly not tested how to successfully "do" advanced study! Although even games are simple to obtain, serious even games are not. I remember that my initial expectation for the ASR was something like what I am describing. Of course that was me applying my thoughts to someone else's reality - incorrectly as it turned out.

_________________
Dave Sigaty
"Short-lived are both the praiser and the praised, and rememberer and the remembered..."
- Marcus Aurelius; Meditations, VIII 21

Oh, okay. I agree, more or less... I actually posted a question a week or two ago,about whether people play to win or play for the best point spread when they're way behind. But this is mostly a problem once the gap gets to be too large, when it's hard to even think strategically about the game because the situation on the board is so lopsided. For me, at least, "best effort" has more to do with sleep and mood than the rank difference.

But you and Bill Spight should really have a debate on this - he (along with several other members of the forum, iirc) is a big proponent of "get whomped by stronger players in even games".

What I find to be most "serious" about the ASR games is the review. Getting a review from a player who is both much stronger than you and was your opponent is totally different from either one separately.

I have come to think that games would be taken more seriously if they were rated, and I now think ASR league games should be rated by default.

PS Your dog looks awesome Dave

Absolutely disgraceful that my wife and one of her friends took the most important member of the family to a local pet shop and photographed her wearing a some bear ears!

_________________
Dave Sigaty
"Short-lived are both the praiser and the praised, and rememberer and the remembered..."
- Marcus Aurelius; Meditations, VIII 21

Of course it is even worse that I then posted it on the net!

_________________
Dave Sigaty
"Short-lived are both the praiser and the praised, and rememberer and the remembered..."
- Marcus Aurelius; Meditations, VIII 21

I think you're mixing up more serious with more stressful. I really really hope that you don't implement this, as it would be a great deterrent to the stated goal of playing lots of games.

_________________
Patience, grasshopper.

i agree here, not only would it seriously get your rank stuck really quickly it would also ''scare off'' all [-] players in a heartbeat.

To get back to ez4u without going OT (poor doggie) the ASR is a community and to get things going the community must set this in motion (just like here on L19 or am i missing something?). Yes asia is really underrepresented in the ASR and i wish i could get them in there for an even more diverse and active ASR. So my proposal is then that you (blunt mode) do something about it (/blunt mode).
Im more then willing to promote lectures on the site and in the room if youre willing to give them because we need more of a study environment of high standard that motivates those strong players to like the ASR.

I compared other rooms to the ASR and i can still say ASR is one of the most active rooms that is not a national or standard room, so if you say the room is inactive, compare first.

_________________
admin of the ASR league and KGS admin

I may come under a lot of fire for this view, but I'll share it anyway (it's never stopped me before!) - Also, Dave, if you're reading this, I'd actually really value your opinion on it too.

The answer (if there is one) depends a lot on what the current goals of the ASR actually are. I'll suggest 4 possibilities, but I'm sure there are lots more:

1) A good excuse to play lots of games with a competitive feel

If this is the goal, reviews should be optional, and games should probably be a lot shorter, perhaps 10 mins + 5x0:20 byo. I'll keep this one short as I don't think anyone quite feels this is "the priority", but as daal mentioned the goal was playing lots of games (even though I'm sure this wasn't what he was implying) I thought I'd mention it.

2) A friendly environment to play and learn

By and large, the ASR seems a pretty friendly place, and it has its own KGS room, website, and forum (albeit here on L19). However, there's not a great deal of social interaction, and there isn't much ASR focus on making more happen. The number of players that participate on the forum is a tiny fraction, and the website has little access or social use, instead being used for rules, guidelines, hints and tips, and the league results.

The games have sensible time controls for learning and playing, and reviews are recommended at the end of the game, which helps in some respects to learning. However, I think if 2) is the goal, not enough is being done to encourage an interactive community that could interact explicitly on teaching and learning material as well as general social interaction. For example, threads in a dedicated ASR forum that are dedicated to studying josekis / fusekis, analysing when tesujis work and when they don't, regular L+D challenge problems for different levels etc. This would seem like a consistent way to develop further in the future, but finding a way of combining the existing league system successfully could be tricky.

3) A serious competition environment, with emphasis on learning as a side

I think this is probably the closest of the 4 goals with respect to how the ASR currently functions in practice. The games are suitably long, reviews are advocated by the administration, and the league system rewards both activity and performance. There's incentive to play games, and incentive to win, and I think that the league achieves 3) pretty well.

An enhanced online presence, perhaps on "featured games", or "featured players" might be interesting, with interviews (including questions on how to improve the league!) of some of the more active players. Stats would also be nice for many I think (total games played, total games won, total points earned, perhaps with the ability to break down both by tier and by month), as most competitive environments seem to benefit from this kind of thing (too many gamers are stats/numbers afficionados, especially in the Go world). It would also be possible to keep an "ASR performance rating" running alongside it with ELO evaluations of the players, so just as with the chess world, there'd by the champion (by competition), and the highest rated player (by overall performance), who are not necessarily the same.

4) A serious learning environment, with emphasis on competition as a side

I think this is probably the closest of the 4 with respect to ASR's officially stated goals. If the goal is to help players improve first, and offer a competitive environment second, I think there are a few ways of improving the current system.

As someone who has successfully taught quite a few students, I would have said that the greater the investment (in money, time, energy and/or stress), the greater the potential gain in ability. The more you invest into your improvement, the more focus you have, the more stress you have, and the more you remember of the pros and cons of your play and your opponent's play. If I was to have lots of time and money, so that I could play in the ASR (unrated), on KGS generally (for rated games), and paid tuition (with teachers), my concentration, focus and effort would be 1) teacher games, 2) KGS rated games, and quite a long way behind it, 3) ASR games. If we are genuinely keen on using the ASR as a platform for improvement in playing ability, this order seems somewhat perverse. I actually believe that, whilst playing stronger players and having them review is awesome and useful, simply playing long, serious, highly focused games against the same level opposition and reviewing it together afterwards can be really effective for improving and learning.

Rated games are more stressful, it's easier to make excuses to not play, and I feel 100% sure the ASR would lose a bunch of players, and a bunch of games per month if it was implemented. I also feel 100% sure that those that remain would focus harder, put in more energy effort, and have less need to play rated games in the EGR (allowing more time to focus on ASR games). I think this would help players to improve their play considerably more, and if 4) is the goal of the ASR, I think rated games makes much more sense than unrated games. Alex has taken a different route with the Insei league, and gets people's motivation and focus based on paying money, but I think making games rated will be more effective, and easier to find the players who feel in a position to play in the system. As with 2), the focus for online development would probably be ideally related around discussions of study material - the ability to post positions, moves, games, ideas, and discuss them, as well as in-depth analysis and ideas behind josekis and fusekis.



In summary, I actually don't feel particularly strongly, as it depends entirely on what the ASR feels is its goal. When I first joined, I was enamoured by the idea that this "wasn't just another league". The goal was a competitive environment (which I love), with serious games (which I love), and first and foremost an opportunity to really improve your game of Go. If the latter is now less of a priority than the former, "serious games" becomes "longer games" (which I love slightly less ), and much of the unique appeal it held for me disappears. I could go join the BGA individual league instead (where I know most of the players, so get a sense of community from it) if I wanted to simply have a competition with other Go players. On the other hand, if other people want it, and more importantly if the ASR administration feel that the competition should be the core of the ASR, then it's the right decision. However, trying to achieve both 3) and 4) equally I think will leave the ASR doing neither thing outstandingly well, which is worse than focusing properly on one of the two aims individually.

PS Anecdotal evidence of one: I'm a [-] player and it would bring me _back_

I think that your description 4: "A serious learning environment, with emphasis on competition as a side," best fits both the stated goals and the way ASR works in practice, though I would add the word "friendly" somewhere too. The reason that this model as opposed to your point 3 fits the current practice best is that the emphasis on playing lots of games, allowing people with lots of time to rise more easily, makes the competitive aspect of ASR a side benefit. The main focus is learning by playing and reviewing, and a friendly atmosphere is conducive to that. Rated games tend to be more competitive and less friendly.

Generally, KGS rated games are 1) not even games between players of differing strengths, and 2) don't end in a review. These are the aspects that make ASR unique. It is also the case that most KGS games are rated. In fact, ASR is one of the best ways to get a serious free game. This is for me a huge benefit. It allows me to free my mind of ratings concerns and simply concentrate on the game. This also reduces the anger at my own mistakes, and allows me to concentrate on learning from the review. I'm sure it would be great if I could direct my hate to my mistakes instead of myself when I lose a rated game, but I'm not the only one with this problem.

Whether playing and reviewing is the best way to improve is perhaps questionable - I've never had a teacher - but I see it as an important part of the mix. Rated games, as well as tsumego practice or lectures I can get elsewhere, but the regular game + review is something that I only get at ASR. I think this is what the league should focus on.

_________________
Patience, grasshopper.

I saw the last few posts but haven't had time to think them through yet. Meanwhile, I took the results chart for September and put the info into Excel. Here is a brief summary. I wanted to have a better understanding of the range of activity in the league. For my mental picture, I divided the number of games played by four to arrive at "games played per week" and grouped players as shown. The first line, for example, says that 17 people played five or more games a week on average (= they played 20 or more games total; that actual range was 20-31). Those 17 people represented 5% of the players in the league that month, grabbed 17% of the points scored, garnered 18% of the wins, suffered 18% of the losses, and accounted for 18% of all games. I have not drawn any conclusions yet, it is simply food for thought at this point.

September stats...

I also made a quick histogram of the number of games played...

_________________
Dave Sigaty
"Short-lived are both the praiser and the praised, and rememberer and the remembered..."
- Marcus Aurelius; Meditations, VIII 21

this is either totally wrong or im totally stupid. there were over a 1000 games last month and your histogram doesnt average out on 30+ games per day.


my stats for september


compare to stats for august


compare to stats for february (biggest league we ever had + most games ever in a month)


feel free to draw conclusions from this ez4u:)

_________________
admin of the ASR league and KGS admin

Oh, don't be so hard on yourself! Just notice the histogram is the number of people bucketed by the number of games they played. So: 43 people played no games (+0), 30 people played 1 game (+30), 22 people played 2 games (+44),..., and 1 person played 31 games (+31). It all adds up; it is just from running Excel's "histogram" analytical tool across the same list of players.

BTW - I think your method of calculating the average number of games understates the situation by half. In September, for example, although dividing the number of games by the number of players gives us 3.43, from the players' point of view they experienced playing an average of 6.86 games since each game had two players. That's why in my table I showed 2,152 games played instead of 1,076. It's true there were 1,076 wins, but there were also 1,076 losses!

_________________
Dave Sigaty
"Short-lived are both the praiser and the praised, and rememberer and the remembered..."
- Marcus Aurelius; Meditations, VIII 21

so yeah i am stupid i thought you did days on the horizontal line (DERPDIDERP) and yes i explicitly used the word GAMES everywhere instead of players to point that out. I really dont see the point in saying things in terms of players or points while this histogram is pretty nice to look at it doesnt say anything about what we should do about for example: the dropout rate.

btw its natural that for every win there is an opposite loss but still a player should play 4 games and the number under 4 games just says: a lot of ppl did not succeed in getting over 4 games, and this is true.

_________________
admin of the ASR league and KGS admin

I think we look at it a little differently. I try looking at the situation and want to concentrate on the way the people behave. So I am playing with the figures to try and see how they behave now. For example, another figure (not in my table) is the median number of games played per person in September, which was five games. That is only slightly over 1 game a week. Why do only half the people in the league play more games than that? The numbers will not tell us, nor will they tell us what changes to make in order to have different results. However, looking at the numbers in different ways may (only may!) lead us to think of different alternatives to try. And if something new is tried, having more ways of looking at the results may also help understand and evaluate the effect of the change.

One goal of the ASR is to promote playing a lot of games, right? But, how many is a lot? Do the organizers share among themselves a similar understanding of how many games that is? Do the playing members have a similar understanding? Right now (in September at least) 36% of the players achieved 8 or more games (my 2 games a week or more). Another 36% dropped out with less than 4 games. In the middle, the remaining 28% played 4-7 games. I assume that right now people would consider that reducing the dropout rate to 25% would be a more significant change than raising the "2+ games a week" rate to 50%. That is because the dropout rate is calculated and reported each month and is salient in people's minds. However, note that raising the 2+ rate by 14 points would probably require a bigger increase in the number of games played than reducing the dropout rate by 11 points. But which would be easier to achieve? Is it easier to get someone who played 2 games last month to play 4 this month or to get someone who play 6 to play 8 this month? I don't know. Probably no one does right now! They are not necessarily affected by the same issues.

_________________
Dave Sigaty
"Short-lived are both the praiser and the praised, and rememberer and the remembered..."
- Marcus Aurelius; Meditations, VIII 21