What will the Go server of the future look like?

[Editor’s note: This is a guest article by Gabriel Benmergui of Kaya.gs. Go Game Guru isn’t involved in this project and is publishing this article as a public service.]

Many players who read this article will not have experienced Go in the recent dark ages.

Before the internet was invented, and used widely, the scattered (but still numerous) Go players had a real challenge in finding opponents and people to play, share and learn with.

Luckily for me, I did not experience that.

A little about me

I have been involved in the Go scene since 2003. I remember the wide-spread use of IGS, something that was a historic landmark for many players in the world.

Many players that had dropped Go altogether, due to the lack of players, could now find games in a daily basis!

I also remember how painful it was for me to sign up, for various reasons, which led me look for alternatives, like KGS.

I was a beginner back then (and the games where I was around 20 kyu are still there). The user experience at KGS was much better, but there was something more important that I liked about it. The community cohesion.

People were teaching a lot. There were go clubs (I was a member of Wings Go Club) and lots of chatting, as well as games. This was a key and still is a key determining factor in a successful Go server. Not all people log-in to just play a game or watch a pro game. There are a wealth of things to do with this game.

Back then there were some Go server wars, but in the long run the results were (in my opinion) consistent with the services provided. One result is that servers are fairly clearly divided into Asian and Western, which I think can be avoided. Asian servers have an enormous number of strong and able players, but they have almost no community activity, or if they do, they are absolutely unreachable.

The world changes quickly

The server wars happened almost 10 years ago. Back then many current technologies, that we now take for granted, didn’t exist yet. These servers have been around since before the birth of Youtube, Facebook, Twitter, Google Earth, and many more novelties like Ajax and the current state-of-the-art browsers – that are so much better than they were 10 years ago.

They were not built to anticipate these changes, and its impossible to blame them. Nobody knows what the future will give us. Today we know the future is uncertain more than ever.

In recent years WBaduk and Tygem have shown clear efforts to expand into the west, but their only competitive advantages are the number of strong players and that they broadcast live professional games.

They do seem to have a rich clients but somehow they are not so usable for Westerners. The social experience is so poor and also difficult for Westerners, due to language barrier. They haven’t achieved success.

Asian servers are huge and profitable, but also slow and heavy. They cannot really react to changes as fast as modern software companies do.

Still, looking at the current solutions for online Go is very short sighted.

I could never imagine that in the year 2050 we would still be using what we know today. So I started wondering, what features would the Go server of the future have and what would it look like?

Like many ideas one comes up with, I started by thinking about how we could improve what we have now. KGS with an internal training tool, like WBaduk has. Or WBaduk with a social spin. But even though those things would be cool, they would not be disruptive.

The next generation of Go servers

The next generation of Go servers has to be undisputably different. It’s not supposed to make other servers look worse, its simply supposed to be different. You don’t compare Facebook with Youtube.

Over time, I started writing down ideas that I thought would be awesome, thinking in a limitless fashion, as if anything could be done. Much of my inspiration came from personally experienced shortcomings.

  • Two years ago I ran an online Go school (Atsumi) and I had a large number of students. However, I didn’t have the tools I needed to make the big jump. I didn’t make enough revenue to live on it, nor could I devise a plan to do so. My biggest source of expenses was advertising myself through free lectures, and ads. I had no support network to grow as a business and make a living. You are absolutely alone. You can see that all the Go ventures out there fly solo – ASR, insei league, eidogo, gokifu, igolocal…
  • Also, many times I’ve met people who were curious about Go after I told them I was a player. But I had no resources to help them. Yes, playgo interactive is a nice site, but that’s a terrible solution if you think about it. I would like to share a link, and show someone in a super conversational way how to play. Telling someone to go to a website, download a client, log in as a guest, then find me there with x-nickname, then look for the game I open… It’s too many steps!

And then I started thinking about a key problem no community in the world is fixing in a systematic way: the creation of new players. One of a Go server’s top priorities has to be the generation of more players. It’s really only logical: more players will show more revenue eventually. Yet, no server is making such an attempt as the task seems too daunting to take.

These two things got me started on a wealth of features and, more importantly, the philosophy by the which Kaya was conceived several months later.

Gabriel Benmergui (left) and Patricio Reboratti from Kaya.gs.

All Go players share a bond through a Goban, even if we don’t know each other, we feel respect for someone that plays Go. It’s something personal and important.

Expanding the community of Go players

We should make a serious effort to expand the community. We should make a serious effort to pull the community together and synergize our efforts. What we build in this generation is going to help the next one.

I wouldn’t be here if IGS didn’t help the generation before me. And I wouldn’t be who I am without KGS. I want to give the next generation of Go players an even better environment.

Kaya.gs is not only intended to be a place where you can find players and opponents, its a place where you can enjoy the legacy of some of the most passionate players and contributors in the community. Its a venture aiming to help other ventures become real, interesting, professional and important, and hence making the whole community richer.

Thinking about what’s possible

When you start thinking about what is possible, only the sky is the limit. Things never done before, like tsumego competitions, or yose challenges (2 players get a 0.5 point pro game, play it out and compare their yose to the professional one), go shows and live TV, tournament news coverage and online organizational tools for Go communities. Online study groups, multi-video sessions, massive server games analysis to check out amateur joseki fashions. Fake betting, complex rank graphs, and much more.

Kaya.gs prototype mock-up.

The number of things we haven’t tried yet and the potential experience is so huge. We are not devoting enough energy to these matters as a community.

We picked the name Kaya because we want it to be organic, like the 10,000 year old tree that is used to make gobans.

We want to grow and expand and constantly be changing, and make a big forest. The ultimate goal is to improve how we experience Go, just like the tree does, by becoming a board.

Kaya.gs demo video

If there is one thing I would like the readers to take from this, it is to understand that we are trying to build a living, breathing and expanding space. One that constantly looks to improve the experience and also increase the number of players and activities related to this game.

Gabriel Benmergui – Kaya.gs – Kaya.gs blog


What’s your vision for the future?

How about you? What do you think the Go server of the future will look like? Leave a comment below to share your ideas.

You can read more about Kaya.gs at the Kaya.gs blog.

About Guest Writer

The article you've just read was submitted by a guest writer. Go Game Guru is always happy to accept guest articles from readers. If there's something you'd like to write about the game Go, why not let us know? You can leave us a short message below or send us a quick email. Please ensure that any articles you send us don't require a lot of extra work before they can be published, otherwise your article might be delayed.

You can follow Go Game Guru on Facebook, Twitter, Google+ and Youtube.


  1. I would say, join the big numbers. At the moment social media are big, Facebook seems unbeatable, so many million people communicate with each other this way. Also interest groups are present, you become a friend and get news in return. I know of watch companies that have 200.000+ friends, imagine the reach into our world. Go Game Guru does this already, great: I never fail to take a look after a message on my Facebook page, I replay all the commented games, try to solve the problems. It is free, no ties attached. The Facebook community around an interest like go could provide for the community, the fun. In any language: Google translate helps if necessary, in the future this could be automatic, and hopefully better. But some clever start-up marketing effort must be made to get those friends to the interest group: shout everywhere that you are there and that you provide a super service, like playing go with fitting opponents. Maybe the different go-servers can join in here.

    So, that is another thing to do, apart from communicating: playing games. I imagine that (in the future?) an automatic link to playing software is possible: one of the go-servers? You already can see which of your Facebook friends are online. Using the friend concept in one way or another could provide for opponents. You play a game, you may see actions from other friends too being still active on Facebook: flexibility and a low threshold should be key. No money involved, Facebook might take care of that.

    Following live games could work along the same concept. I guess this all would be realised well before 2050, why not 2015? If you put your mind to it, you can work this out. Then work, work, work. But communication is everything, go will then be one of the nice things to do.

    Kind regards,

  2. Although facebook is definitely a tactic, it too will in due time be replaced by something better, or succumb to its privacy holes. In any case one should not depend on it.

    From a user perspective, here’s what I’m missing out currently to make online go a great experience:

    1) I much prefer live games, even if they take much longer, because online I’m missing the facial expressions which make for a human atmosphere. The anonimity of playing “Luxorz001” who suffers from lag or escapes – I don’t know which – or says “gg & ty” instead of giving me a friendly smile, is putting me off. I prefer a club night 10 times, even if it’s always the same players.

    2) For discussion, it’s really a shame that Sensei’s and L19 are segregated. It’s incomprehensible to me that the (western) world has still not succeeded in creating its own reliable and complete wikipedia of techniques and we need awkwardly slow processes like me using gogod+kombilo or gobase to explore a technique, then post it on SL, then crosspost at L19 to generate some discussion.

    At first I was skeptical of yet another go server, but you are so right: when you think of it, there is so much that we could have and do not have today.

    Good luck!

  3. You must work very hard for a Go server to survive, and people have been very, very critical of kaya.gs. Even if the server is not a success, Gabriel’s vision and ideas about a technologically modern, social and beautiful Go site are essential to capturing a new generation of Go fans.

    I wish him and his collaborators all the best and will personally be contributing to the server in any way I can.

  4. Tami Jones says:

    By all means include Facebook, but please don’t make it necessary to have an account with FB in order to play on Kayago. Thanks!

  5. Tami Jones says:

    While I remember, is it possible to include webcam connection, so players can view each other if they wish? It’s possible to run skype and KGS next to each other, and do it that way, but if the operation could be built into the server it would help enormously to overcome the anonymity of server play and the problems that brings.

    • Gabriel Benmergui says:

      Actually we have in mind a service to include into our service that would solve us many problems at once. It is a private chat component, that has video/audio streaming and many other features like chat-translation through Google Translation Api or the sort.

      Because the audio/video connection is done peer2peer, it doesn’t go through the server, which means its absolutely limiless for us. (anyone can do it, and can try to show video to several people at the sme time, depending on their own broadband). In the last year Google plus outed with the Hangouts, and i haven’t looked even for a second if that is possible to use outside G+.

  6. I want as little to do with Facebook as possible. Tami, I don’t think he’s so stupid that he’d require a Facebook account to play on his server or any other external service.

    Though I would seriously caution against any annoying dual login schemes. There have been so many privacy breaches due to poorly made and/or malicious third party apps.

    On another note, I’m not at all a fan of the visual style he’s chosen but he has said the server will be quite modular and easily modified if I recall?

    One criticism people have had is naming it KGS. “[K]aya [G][S]” This seemed more like an innocent oversight and I doubt he’s going to change his mind. I would say that people are more likely to trust a .com address though.

    • David Ormerod says:

      Domain squatters (and the registrars that pretend to ‘reserve’ them for $1000 dollars or more) really have a lot to answer for. We went through quite a lot of names before we came up with gogameguru.com.

      Personally I agree that the potential for confusion with KGS is unfortunate and, overall, I think it would be better to change the name while they’re still early in the game.

      However, when you’re trying to register a .com and every sensible word is taken (because some ‘investor’ basically went through the dictionary with an automated program), it’s incredibly frustrating.

      In my opinion, domain names that are purchased and left idle should be taxed at a very low rate (based on the potential value of that domain name). It’s similar to what most governments do for unused commercial land. If you only owned a few domains and were doing (or planning to do) something constructive with them, it wouldn’t bother you. If you owned thousands and were just hoarding them – providing no utility whatsoever to the world community – the costs would start to add up… I guess the problem with that of course is that people would then just start building meaningless websites on them to avoid the tax… It’s a difficult problem.

    • Gabriel Benmergui says:

      This matter is a storm in a glass of water :). kaya.gs is an excellent domain becuse its short and easy to remember. .(dot)com domains are overly saturated and we would have had to select something like “www.kayagoserver.com” to have space, which is the same “issue”.
      This was debated many times and we will not change the name. There is a very easy short abreviation for those that worry about confusion: just call it Kaya, as we do.

  7. I so look forward to Kaya. For some reason no real time server got me to enjoy the game on a screen, yet. It might be unreasonable to hope that will be different with Kaya – but hey – maybe it will.

    • Gabriel Benmergui says:

      We are doing our best into creating the best online go-playing experience possible. Even our current state of the server, which is starting to get many features, is very incomplete with our current vision of it, and in the past 2 months that we have been working at it full time, we have had numerous new ideas.

      We plan to make a better experience than any other server as soon as possible, but we promise something even more important: that we will keep improving it constantly.

  8. TWITCH! Streaming Go is THE answer to spreading the game to more people. There are a few go streamers on twitch right now but not enough. More pro matches NEED to be streamed on twitch, and not only replayed inside clients. Eurogotv streams on Ustream right now, but if you ask me that is a mistake. The only people who watch those streams are people who already know about eurogoTV. Twitch has a lot more through traffic, and way more young people already have accounts on twitch to be able to chat. Twitch is an easy way for a new player to just click a link and watch the game being played in front of them, with opportunities to ask the streamer or other viewers questions about the game.

    Anyone serious about spreading the game of Go should be streaming their online matches on twitch, and streaming more live and relay tables on twitch. The Go community is small right now on twitch, but it is there and it would grow if more pro matches were streamed there (eurogotv would quadruple their livestream views overnight if they switched to twitch from ustream).