Chapter 4. The Rise and Fall of China's Esports Gaming Platforms

A chapter from ONE MORE WAY: The History Behind Chinese Esports, by BBKinG (Liu Yang). Translated from the Chinese original: 四 中国电子竞技对战平台兴衰录

First published in the author's Zhihu column on November 17, 2013: https://zhuanlan.zhihu.com/p/19615324

This translation was reviewed against the Chinese original on August 1, 2026. If you spot a mistranslation, a wrong name, or a factual error, please open an issue or send a pull request.

What was the first profitable business model in Chinese esports?

Some people say it was players competing for prize money, others say it was tournament organisers selling advertising. Looked at from different angles both answers are right, but if the test is stability and maturity, then I think the first profitable business model was the esports gaming platform.

The boom in esports gaming platforms was forced into being by the awkward gap years of the past decade and more.

Today, when we play League of Legends, we only have to click start inside the game and after a short wait we are matched with teammates and opponents. Chatting and making friends is easy. That is a gaming platform, and it comes built into the game itself.

Before the internet became widespread, though, most games were single-player affairs. When single-player games met the internet age and the games themselves could not change fast enough, third-party gaming platforms stepped in as a product of the transition. Nobody expected that transition to last more than ten years.

In January 1997, Blizzard of the United States launched its famous gaming platform, Battle.net. But in an era when computer skills and the internet were both only just getting started, most people in China found that simply logging in and finding somebody to talk to was hard enough, never mind being matched against an opponent.

Meanwhile Chinese players wanted more and more from their games, while China's own capacity to develop networked competitive titles lagged far behind (in fact only in the last few years has that begun to change). On top of that, foreign companies in those days were not about to cater to what Chinese players needed.

So Chinese IT engineers were forced to keep experimenting, and eventually they built one third-party esports gaming platform after another to serve China's networks and China's players.

Ourgame's card and board platform

In March 1998, three developers, Bao Yueqiao, Jian Jing and Wang Jianhua, put together 200,000 yuan and founded Beijing Lianzhong Computer Technology, known as Ourgame, in a small hotel in Shangdi, Beijing. Its main business was online card and board games. By 2000 Ourgame had 700,000 registered users and 9,000 people online at once, and by the end of 2003 its concurrent figure had broken clean through the 500,000 mark.

Ourgame's card and board games did not demand much of the network or the hardware, which suited the era, and the people behind it were the first in esports to taste success. Looking back now, though, Ourgame had nothing to follow up with. A card and board platform has a low barrier to entry and little core competitiveness, so when Tencent's QQ launched a card and board platform of its own it stripped Ourgame of its users almost at once.

The imitators of Blizzard's Battle.net

Battle.net, live since 1997, was a strategically far-sighted design. Blizzard had worked out how to bring single-player games and the internet together before the internet had properly arrived, and it had already seen clearly that the key to networked competition was the automatic matchmaking system, which pairs you with players of similar strength based on your record.

That one feature alone took many Chinese platforms a long time and a great deal of effort to copy.

Unfortunately for Blizzard, its Battle.net servers stayed abroad and did not come to China until 2008, which gave the imitators plenty of time to study and improve. It also meant that for almost the whole history of Chinese esports Battle.net remained a high-end, niche platform in China. That brand image had a great deal of influence on Blizzard's later strategy, which I will come back to.

Private battle.net servers everywhere in the StarCraft era

The arrival of FSGS made the number of local servers around China explode. FSGS stands for Free Standard Game Server, developed by the German company NETGAMES. It was third-party software that emulated Battle.net, Blizzard's gaming platform.

Between 1998 and 2001 there were a great many private StarCraft battle.net servers in China. Even an incomplete count came to nearly a hundred. Most were held back by their software and hardware and never grew large, but a few carried real influence: Beijing's 263 server, the CGOL server (AsiaGame) and DM (AsiaGame Dongmei), the Youxia server, the Beibei server, the Wanglian server and others.

Of these, 263 defined the team as a unit in its program, which did a great deal to push teams towards proper organisation and gave players a sense of belonging. Organised team communication helped esports grow. The CSA team really flourished on 263 in those days.

Then there is the famous AsiaGame server, a case worth going through properly.

AsiaGame ran free from 1999. Backed by relatively deep pockets, it equipped itself with high-specification servers and bought the advanced version of FSGS, which was not subject to the 999 limit of the Linux version and could support ladder games at full speed. Its stable matches drew the great majority of StarCraft players, and it very nearly unified China's StarCraft servers.

The enthusiasm of Chinese esports players was pushed to its first peak. Almost everybody believed that if things carried on this way, catching up with South Korean esports was only a matter of time.

In December 2001, after taking investment from the Hong Kong company Dongfang Meili (Oriental Charm), AsiaGame began pushing hard to charge for access.

We all know how that ended. Once the charges came in the players scattered overnight, and AsiaGame faded from view.

Let us talk about why. Start with the timing. I looked it up: Tencent launched the QQ membership feature on December 18, 2000, and because it had no way of collecting money it went nowhere for a long time.

In early 2001 China Mobile launched Monternet, its mobile internet service. Monternet collected payments through the phone bill under a twenty-eighty split, with the carrier taking twenty per cent and the internet content provider eighty. Tencent launched mobile QQ on it, built up paying members and added more value-added services. At its height Tencent accounted for seventy per cent of the content provider business on Monternet.

By the end of 2001 Tencent finally reached positive cash flow, and in March 2002 its concurrent users had only just passed three million.

Which is to say that Tencent only broke even at the end of 2001. The habit of Chinese players handing an internet company ten yuan a month also took Tencent many years to build. Talking about this period now, I imagine Taobao between 2003 and 2005 would want to cry too. So the missing payment channels, and the fact that the habit of paying had not yet formed, were the main reasons AsiaGame failed.

AsiaGame was the first esports gaming platform to try charging, and the first to get hurt for it. I have great respect for the pioneers of that era. Their courage and their willingness to explore were what drove esports forward.

The rise and fall of the Haofang platform

You cannot talk about Haofang without talking about its founder, Li Lijun. You will not find any background on Li Lijun on the internet, and it would not be appropriate for me to say much (this is the awkward part of writing contemporary history), but we can work things out from what is already public.

Shanghai Haofang Technology was a privately owned high-technology company founded in 1998. Its main business was software development, systems integration and network systems applications, and its annual output was steady at around 400 million yuan.

This was a company that could win the Shanghai Telecom upgrade contract between 1998 and 2001. Haofang was also the company that, in 2003, could deploy servers rapidly inside telecom equipment rooms all over China, something even the QQ gaming platform never managed (partly because Tencent kept hesitating over copyright questions). Bear in mind that the QQ card and board platform had just wiped out Ourgame. So "well connected" does not begin to cover it.

In 2001, after the Asian financial crisis, Li Lijun set out to make Haofang's products more varied, and in the second half of 2002 he acquired the Youxi Shanghaitan platform and used it to found Shanghai Haofang Online Information Technology. The platform had two developers at the time: Xu Xin on the server side, the Benteng who holds Haofang account number 01 and is now CEO of Qifan, and Zeng Fangfei, known as Laohu, on the client side, along with GAR and a few others on interface design.

Haofang's own technical strength was considerable, and the team it put together on the marketing side was not weak either.

At the end of 2002 there was a row among the senior people at CCSK, then the most active Chinese-language Counter-Strike community. On one side were DDM, who owned the site, and general manager Laomeng. On the other were Yang Chen, head of external liaison, later Haofang's marketing director and now running his own company, and Guo Zhuo, who ran the technical side, known as Xingxing, later Haofang's head of operations and now assistant to the president at Qifan.

DDM, Laomeng and Yang Chen were in Beijing at the time and Guo Zhuo was in Shanghai, so the argument was carried on in CCSK's internal forum. I had just been promoted to deputy head of the CCSK news group, so I could read all of it.

Why did Yang Chen and Guo Zhuo leave? They wrote a great many posts arguing with each other, and most of it was about old grievances. The accumulation of small grievances was one reason, but the real one was that nobody at the time could say clearly how esports was ever going to make money. It was like betting on high or low: before the dice cup is lifted, one person says high, another says low, and both can argue their case. Disagreements caused by that kind of information gap are hard to settle even looking back on it now. But people have to live, and they have to move forward.

So the result of that argument was that CCSK split in two. One group went to Beijing, where a woman investor was planning to set up a batch of companies with names beginning with E to sell the concept and go public, and CCSK became Esai. Yang Chen took Guo Zhuo and Zhang Ran with him to Shanghai Haofang Online, to work in marketing and technology.

Yang Chen's arrival brought the concept of the esports press corps to CGA. CGA was a new thing, though, and for a while CCSK's traffic was still much higher than CGA's.

Between 2002 and 2003 the Haofang platform did a great deal of marketing work.

For instance, Haofang signed a strategic cooperation agreement with CCTV5 and provided some of the content and staff support for the CCTV5 programme Esports World. K4, then CGA's editor-in-chief, Feiying, its tournament manager, and others spoke at length with the presenter Duan Xuan and the show's producers before every episode.

In 2003 Shuttle, the maker of XPC systems, invited the Swedish team SK, then at the height of their fame, to visit China, and CGA covered the whole trip. After the exhibition match at Metro City in Shanghai, hundreds of fans chased the team's coach down the street, which thrilled an SK side that had never been to China before. They simply had not realised that this country held so many esports fans. Back home they talked up everything they had seen, and that was when the rest of the world began to learn how hot Chinese esports had become.

Meanwhile Haofang rolled its platform out rapidly in telecom equipment rooms across China.

Haofang's timing was very good. Or rather, Chinese internet content providers were in a happy position at that point. Around 2002 the network operators were short of content. Great Wall Broadband, for example, had just started out and laid its network quickly, only to find once it was built that there was nothing inside it. Users still had to cross over to other networks to visit websites and play games, and that hop, as everyone knows, hurts speed badly. Given enough time, users would stop buying your broadband.

So to grab content, the network operators of the day handed out large amounts of free bandwidth to website and game operators, and Haofang got going just in time to catch that wave. Add its existing relationship with the telecom operator, and its server-side rollout was finished quickly. The man in charge of laying out the telecom channels was another Haofang veteran, Huang Chunyu (Rain, now CEO of Shanghai Youqu, which makes mobile games).

Several things drove the Haofang platform's rapid rise during the golden age of Counter-Strike esports.

1. It simplified the business of getting online and could handle a lot of complicated network situations automatically.

2. It had the money and the standing to deploy servers rapidly the length and breadth of China, which left its many competitors far behind.

3. The CGA website was strong enough to send out news teams to cover events at home and abroad, which brought players closer to the news.

4. The free Haofang platform and the CGA website were developed as a bundle, and each drove the gathering of esports enthusiasts.

All of this put the Haofang client on the desktops of internet cafés everywhere in China almost overnight.

Thinking back on the CGA days still makes me proud. Going out in a CGA work shirt was something to be proud of.

To this day I believe the CGA news group was the strongest of the esports era. Its core people were the editor-in-chief K4, JOJO, Feixiang de Helanren (the Flying Dutchman), Sfred, Baixiaosheng, BBC Zhang Hongsheng (the BBC of today's GamesTV, then head of CGA's Warcraft III section), Jkidd, bigxia, MagiCGoA, treebear, Shabusiren, Kevin, mint and others, all of them very capable people who could carry a job on their own.

That is how the Haofang platform and CGA reached their peak during the Counter-Strike years from 1999 to 2004.

Now for the decline of the Haofang platform. What, you think that came too fast?

Nothing to be done. History is like that, whether we like it or not, and the decline began quietly in 2004.

I have long talked about a five-year rule for Chinese esports titles. It sounds a little like superstitious fatalism, but it is a pattern I have observed for more than ten years. Counter-Strike ran from 1999 to 2004, which was exactly the fifth year, and in that year Counter-Strike began to decline while Blizzard's Warcraft III, released in 2003, was starting to take off after a year of growth.

At that pivotal moment of changing of the guard, in 2004, Shanda began its plan to acquire the Haofang platform. It looked at the time like a very promising piece of consolidation. Looking at it now, there is another and sadder story in it.

Why did the deal happen? Two reasons.

1. The Haofang platform ran free from 2002 until the beginning of 2004. Convenience, speed and the absence of charges brought Haofang enormous concurrent user numbers, and as I remember it Haofang had more than 900,000 users online at once by early 2004. But it made no money. There were no browser games then, esports players did not have the spending power they have now, payment methods were not mature enough and advertising revenue was unstable. Haofang was lost.

2. Shanda's rapid rise in 2003 showed the Chinese games industry a way forward, and by 2004 Shanda was preparing to list on Nasdaq in the United States. It had its eye on Haofang's concurrent user numbers, which were an important measure on the financial statements of the day.

With those two reasons in place, Shanda said to Haofang, quick, get in my bowl, and the thing was done.

The acquisition started in 2004 and took two years to complete. The process was too internal for me to say whether or not it was a bet-on agreement, but what I did see was that from 2004 Haofang started pushing out revenue features, the paying member bump feature among them.

Let me explain the background to that feature. The Haofang platform held its users in rooms, but a room could hold only a few hundred people at most, and everyone wanted to be matched with strong players quickly, so some servers were always full and hard to squeeze into.

The paying member bump feature was born out of that. Once a paying member entered such a server, the system would randomly bump out a non-paying user. It raised the paying rate substantially, but for the far more numerous non-paying users it made for a very poor experience.

What I heard at the time was that the order had come down from above.

My own reading of it was that the point was to raise the paying rate so that the platform could be sold at a better price later on.

And so, in a very short time, the Haofang platform's concurrent users fell from close to a million to around 300,000, before recovering slightly and settling at about 500,000.

How much did Shanda actually pay for Haofang?

Some people read 60 million US dollars out of Shanda's fourth quarter results for 2005. Others say Li Lijun spent 30 million yuan building Haofang and sold it for 30 million US dollars.

We will not dig into the exact figure, because for Haofang this was not a good beginning. Afterwards, Shanda's confused positioning of Haofang and the internal friction of company management left Haofang treading water for the next three to five years, which handed its many competitors their chance. (For this period you can read the answer by Yaoqi Linran, which is very detailed. He is Guo Zhuo, CGA's head of operations in those days, and one of the people who lived through it.)

The GG platform

The GG gaming platform has since been renamed Garena, and its Chinese name is Jingwutai.

The founder of the GG platform is Chen Ou, the man who went on to found Jumei. Small world.

In January 2006, the night before the second StarsWar began, I met Chen Ou, who was there doing technical checks. It was the first time GGTV, developed by the GG platform, was used for a live broadcast from a world-class event. Chen Ou back then was shy and smiled a lot without saying much, head down over his laptop, testing until very late.

The GG platform's technology was distinctive among the platforms of the day. Chen Ou himself said his technology was not complicated, yet the speed of its connections was remarkably good, and it performed especially well in matches between China and other countries.

2005 was when Warcraft III began its glory years. I had joined Replays.net by then and was working with its founder ZAX under IGE, the American virtual item trading platform, and WE, led by KinG, had also come under IGE.

Because Xu Yunbo, who ran IGE in China, was personally very fond of Warcraft III, and because esports was a more convenient way to promote virtual item trading in China at that time, and because IGE had healthy cash flow, the result was one of the most stable periods of growth in the history of Chinese esports. Sky's two consecutive WCG titles came out of that period.

That period was not so good for Chen Ou, though, who had only just come back from studying in Singapore. On one hand he wanted to carry on with his studies, on the other he wanted the GG platform to grow. But at that point even the Haofang platform, with the highest concurrent numbers of all, had no clear way of making money, so talks about investment in the GG platform dragged on without result.

So ZAX, the founder of Replays.net, joined forces with Chen Ou and the two of them set about the hard work of standing on their own. ZAX brought the users of the Replays.net forum, which had the highest engagement among Warcraft III players at the time, over to the GG platform, where they could log in directly with their forum accounts. GGTV, team and friend features and anti-lag and anti-cheat features followed, and all of them drew large numbers of users.

Even so the GG platform grew slowly. At one point Chen Ou planned to sell it to IGE, but Xu Yunbo felt the company's focus lay elsewhere and did not take it on.

And so Chen Ou took the GG platform to Singapore, raised investment and started his company there. That is today's Garena. Later some internal conflict arose and the old story of investors squeezing out the founder played out once again. Chen Ou sold off the shares he held and went to the United States to study at Stanford, then came back and started again, which is the prequel to another story, the one about Jumei.

The rise and fall of the VS platform

The success of the VS platform, and of the platforms that came after it such as Zhangmenren, came partly from the maturing of platform technology and partly from the fact that they were in a far better position than Haofang had been.

After WE.Sky took the WCG titles in 2005 and 2006, Chinese esports began to be accepted by more and more of the general public. When we went out to talk business in those days, we no longer had to work hard to explain what esports was.

On the payment side, Taobao, Alipay and the rest were fully mature by around 2007, and the first generation of esports players were approaching thirty, with far greater spending power. For the first time, esports looked genuinely promising.

In the Warcraft III era the VS platform very cleverly avoided a head-on fight with Haofang, choosing instead a brand strategy built around high-level professional players. It introduced the VS rating, a concept that reflected a user's level and appealed to players' urge to compare themselves with one another. It also seized on the fact that Haofang had many players but was slow, and went small and fast, which drew in a lot of Warcraft III's most committed users.

Life was not comfortable for the niche VS platform during the Warcraft III years, but unlike Haofang, VS held on until the good times arrived.

The VS platform launched in 2005 and, after two years of biding its time, ran into the rise of Dota in 2007. Its concurrent numbers climbed steadily, and at the height of Dota's popularity in 2010 VS reached a million users online at once, overtaking Haofang completely.

On revenue VS caught the good times too. Whether through the boom in internet retail advertising or the joint-operation model for browser games, the ways VS could turn the resources in its hands into money were all reaching maturity. With money behind it, the VS platform started investing in a game of its own, VS Three Kingdoms, reportedly to the tune of 100 million yuan.

The market response to VS Three Kingdoms was flat, however, and the VS platform's own technology did not change much either. On top of that, the popularity of the Dota ladder system launched by the up-and-coming 11 platform proved the old saying again: business is like war, and standing still means falling behind. The VS platform began its slow decline from 2010.

Since this piece is mainly about the years from 1998 to 2010, I will leave the history of the 11 platform aside for now. From the point of view of historical development, though, I can predict one direction.

Esports gaming platforms began with single-player games and grew with the development of the internet, meeting players' need to communicate about games. But their core matchmaking and communication functions really belong to the games themselves, and once competitive games moved online, developers began paying more and more attention to building matchmaking and communication into their own products.

To put it another way, I think the historical mission of the esports gaming platform in this transitional period is almost over. Gaming platforms will not disappear, but they will be pushed to the margins.