How Does the “Water Army” Work, and How Does It Turn a Film or TV Show into a “Hit”? | QDaily Anthropology

This is the classic story of a market where everyone does wrong and, in the end, the standards are lost.

Two films based on Journey to the West, Monkey King: Hero Is Back (《西游记之大圣归来》) and Never Expected: Journey to the West (《万万没想到：西游篇》), grabbed most of the limelight in China’s film market over the past year. The first set a mainland box office record for an animated film, at 956 million yuan, and revived a long-dormant conversation about domestic animation. The second was widely expected to do well, yet met an “unexpected” chill during the year-end holiday season.

Box office aside, one more thing links the two films. While Hero Is Back was a hit in theaters, a new term appeared: “zilaishui” (“tap water”), meaning people who liked the film after seeing it and recommended it entirely of their own accord. Its counterpart is the “water army” (shuijun): paid online posters hired to do the same job. Before Never Expected opened, someone noticed a crop of positive reviews of it in the Douban comments on Detective Chinatown (《唐人街探案》), and suspected the “water army had flooded the wrong film.”

Because of these two incidents, the water army became one of the film industry’s hottest topics over the past year.

People define the water army differently, and use it in very different ways. Film companies with little confidence in their own work line up a water army in advance to create favorable public opinion. Some, “forced” by competition, try to use a water army to smear rivals. Others say they need a water army for “legitimate self-defense” so that rivals cannot smear them. In their eyes, water armies come in black and white. On the other side, every faction of the water army claims to have professional ethics and to know black from white: they simply meet clients’ needs, and what clients want roughly divides into “a positive view and a negative view.”

These clients accuse one another. Some who gain an edge quietly enjoy it, as their work becomes a big “hit” and an overnight sensation. Most people, however, seem to have become victims. Quite possibly, even those quietly enjoying their gains now will become victims later.

This is the classic story of a market where everyone does wrong and, in the end, the standards are lost. Today, whenever the box office (or ratings) of a “hit” is very good or very bad, people more or less assume that someone “hired a water army” or that it was smeared by one. Users’ demands keep multiplying and becoming more specific and finely segmented, so the water army’s methods for inflating ratings and word of mouth have kept pace. How did all this happen, and how did things sink this low?

The contest between online communities and the water army

The water army hides on almost every major social platform, sometimes in several waves, sometimes with “opposing” sides appearing at once. More and more people watch films (and TV dramas), yet it is harder and harder for them to find “real” conversations about what they watched.

Douban Movies and Mtime, China’s two main film-fan communities, have long been regarded as the places moviegoers trust most. Douban Movies’ ratings were for a time accepted as the industry authority, and many people use them as a guide to what to watch.

Yang Xuehai, a Zhihu user, is one who relies heavily on Douban Movie ratings. He says that most of the time he decides whether to see a film by its Douban score. Above 8.5, there are few he hasn’t seen, so he watches without hesitation; from 8 to 8.5, some may have slipped past him, so he either watches straight away or bookmarks it; from 7.5 to 8, he reads the reviews before deciding, and also checks the score on the foreign site IMDB; from 7.0 to 7.5, he considers it if it is a domestic film, and for foreign ones he checks IMDB and skips the film if the score isn’t high; below 7.0, he basically ignores it.

In the past few years, however, Douban has repeatedly been criticized for “becoming a battleground for the water army.” Douban Movies had never responded to the doubts and complaints. The latest “water army controversy” broke out again at the end of 2015. On December 18, Douban’s founder Abei (Yang Bo) published a long article, “Eight Questions About Douban Movie Ratings,” answering some of the questions netizens had raised. They included: What is the deal with the water army? Can Douban ratings be gamed? Who sets Douban ratings? Why is the Douban rating of the film or TV drama I like / hate / made / directed / invested in abnormally low / high? (View the original)

This time, perhaps, things had reached the point where a response could not be avoided. Abei wrote that recent changes in the film and TV industry had made the company feel directly that keeping a low profile might no longer be wise.

In the past two years, as new media flourished and more open platforms such as Weibo, WeChat, online ticketing and self-media rose, Douban Movies has come to mean less to film fans. Whether its ratings and reviews can keep their reputation for impartiality and fairness has become crucial. Abei admitted frankly that the water army exists, but said Douban ratings are hard to move.

The implication: Douban Movies has always faced the risk of a water army, but that does not mean Douban, as the platform, has not “fought” it, for example by building defenses and using better algorithms and technical means to identify the water army hidden behind all manner of “fancy flooding and score-inflating” tricks.

Most of the time, open online platforms and the water army are evenly matched, and the latter, working in the dark, often has the upper hand. As the saying goes, the devil is always a foot taller than the priest.

A former water army insider told QDaily that the bar for inflating Douban Movie ratings has indeed risen a lot: the platform has defenses, and the water army has new attacks, and as long as the client can pay, it can still be done. Some of his friends still take Douban jobs, he said, and the asking price for “Douban Movie ratings” has risen to 60 yuan per rating, which is no exaggeration.

According to Abei’s article, one defense against water army attacks is that “ratings from new users registered within the past three months are generally treated as invalid.” From this it is not hard to infer that prices rose because the water army has to steal accounts, or use other, costlier means, to get accounts whose ratings count.

The following example may offer some clues.

What does it look like when Douban ratings are gamed to death?

On December 10, a netizen, @Shijie Renmin de Lili, posted on Weibo: “I don’t use it often, but I suddenly noticed something wrong with my Douban account. After logging in I found I had given five stars to The Vanished Murderer, Bad Guys Must Die and The Legend of Mi Yue hahahahahahahaha.” It was the 10th day after The Legend of Mi Yue began airing. Another netizen, @Wo Ai Dingding forever, found the same day that his Douban account had been stolen too, and complained: “Someone used my account to post five or six Weibo posts about The Legend of Mi Yue. Are the ratings so low that a water army has shown up?”

At the same time, a large batch of “water army” Douban accounts registered on December 10 were dug up. They had one thing in common: their home pages had little content, and under “Films I’ve watched” there was only one title, The Legend of Mi Yue. All had given the drama five stars. These accounts were exposed in a discussion thread called “Water Army Report Roundup”.

Screenshots of newly registered water army accounts

A Douban user, “One-Quarter Sea,” also supplied screenshots of suspected stolen accounts (“Because of the NetEase Xcode incident, 163 and 126 email addresses are basically all unsafe now. Click to view”). He told QDaily: “The ID in this screenshot went straight into the discussion area and traded insults with netizens. When it isn’t your own account, you really don’t care.”

In another thread, titled “Those who take others for fools will come to an interesting end,” “One-Quarter Sea” dug into how a water army was hired to inflate The Legend of Mi Yue’s ratings, complete with screenshots and data analysis. By his account, he watched the drama’s water army inflate its score on Douban, and Douban’s back end also appeared to have been altered; how the manipulation worked is unknown, as is who hired the water army. By his observation, within 12 hours the absolute number of 1-star ratings for the drama fell, while the absolute number of 5-star ratings rose by more than the total number of new ratings.

Few ordinary viewers follow a drama’s reputation this closely, so some commenters accused “One-Quarter Sea” himself of being the “black water army” sent by the drama’s “enemies.” (According to someone who sells “water army” services, a studio that wants the effect to look realistic usually hires one wave of posters to “praise” and another to “bash.” In this case, however, the praise was plainly one-sided.)

QDaily tracked down “One-Quarter Sea,” who told us the whole story. He is a student. He followed The Legend of Mi Yue because he had watched Nirvana in Fire and Empresses in the Palace, and the drama’s producers promoted it with gimmicks such as “grand vision, grand scale,” “serious historical drama” and “Nirvana in Fire + Empresses in the Palace,” which piqued his curiosity. But he did not enjoy watching it, so he checked the Douban rating. “At first there were mostly 3-star and 2-star ratings. Suddenly someone in the Douban Movies mobile app group said, the water army is here. We watched a wave of 5 stars sweep through all at once. So I took screenshots.” He then posted a report thread, and from then on, every time he opened Douban he found a pile of new notifications.

He also found that the water army inflating The Legend of Mi Yue’s ratings was keen to hold the score at 6.0. For several days it hovered between 5.9 and 6.1, until it fell below 6.0 and they gave up and let it slide. “One-Quarter Sea” guesses that Douban staff later cleaned out most of these newly registered accounts, and that the timing may have tracked the staff’s working hours. “I remember that on the 19th, the reported accounts on the fourth page of my thread all seemed to have been banned. I suspect the administrator on shift that day was particularly ruthless.” (Laughs)

A banned account’s page

Another netizen, Muzi Xinxin, also wrote a Douban post after discovering the water army: “Why Can’t I Stay Silent? The Strange State of Chinese TV Dramas, Seen Through The Legend of Mi Yue”

Reports like these, made by netizens on their own initiative, are common. Earlier, one netizen claimed that the good reviews of Hero Is Back might also have been the work of a water army. “Monk Comes Down the Mountain grossed under 400 million yuan and has over 30,000 Douban ratings; Tiny Times 4 grossed under 400 million yuan and has over 30,000 Douban ratings; Hero Is Back, three days after release, had just passed 100 million yuan at the box office and has nearly 50,000 Douban ratings. How many of those people gave it 5 stars without having seen it?”

Many people followed up on “One-Quarter Sea’s” Douban posts with reports of their own, which shows that some water armies can indeed breach a platform’s defenses. For now, at least, they can still “roam freely,” slipping in wherever there is a gap. But “One-Quarter Sea” has noticed that the water army comes in different tiers. “My feeling is that the people inflating The Legend of Mi Yue are fairly low-level, while those who specialize in film reviews are more skilled. I hear they now split the work so that short reviews and long reviews have different people and different prices. It is unrealistic for Douban to stamp out the water army by technical means, but it can enlarge the pool of users behind each rating, so that the water army’s ratings have little effect.”

This is only one way the water army operates; in practice its tricks are many and varied.

When do films (and TV dramas) use a water army?

In general, a film whose content is decent does not use a water army from the start. Water armies are mostly deployed within three days to a week of release: if word of mouth is poor, the producers try to find one for emergency support. Examples include The Last Supper (《王的盛宴》), The Banquet of Hongmen (《鸿门宴》) and Back to 1942. Because The Last Supper had a low Douban score, its publicists tried to rescue it by publicly claiming it had been smeared by a “water army,” and said that to fight the “black water army’s” misdeeds they had no choice but to hire a water army of their own to “whitewash” the film.

This looks like legitimate self-defense, but it is still, in essence, an improper way to compete. The industry and the audience reach their own verdict, especially when the same thing happens again and again; even if a film wins at the box office, its reputation among viewers will not be shifted by “water army work.” This year, Chronicles of the Ghostly Tribe (《九层妖塔》), directed by Lu Chuan, grossed more than 600 million yuan, but its Douban score is only 4.3 and its word of mouth is poor. Gossipy netizens claimed that in the second half of its run, large numbers of water army posters again appeared on the film’s official Weibo account and on Lu Chuan’s, posting “good reviews.”

On “marketing haters,” “One-Quarter Sea” also told QDaily that The Legend of Mi Yue’s planner Cao Ping and screenwriter Wang Xiaoping have both called Douban a gathering place for “trolls,” and that Deng Chao, of Rogue Angel (《恶棍天使》), has likewise accused netizens of being “marketing haters.” He felt this matched what Abei described: water armies on both sides probably exist, but Douban has already done its best to reduce their weight; the drama’s side uses them to attack the authority of Douban ratings, and even suspects every netizen who gives a bad review of maliciously gaming the score, which turns black and white upside down.

Does every film use a water army? In fact, the projects that use water armies heavily are mostly low-budget productions of poor quality to begin with. A person who has worked in film marketing for more than two years said: “Directors and producers are generally blindly confident about their own films. No one thinks their own film is bad, and big-name directors have even less to worry about. To be honest, sometimes we produce these effects mainly to please the producers, not for the audience.”

How does the water army work? How does it protect itself?

One view holds that “water army” services are a standard item in film marketing. To make them sound less “gray,” the industry usually calls them “word-of-mouth marketing.” Three kinds of parties generally have a water army on hand: film producers, film marketing companies, and specialized water army services.

The first two generally do not keep a water army of their own. As promoters of a film, steering public opinion is part of their routine service, and their staff use their own spare Douban accounts, or Weibo and Maoyan accounts, to steer opinion in the film’s favor. They know better where the film’s selling points lie, and many viewers never watch trailers or read film news. They will not call this “water army work,” describing it instead as “normal marketing to get the film out there better.” This kind of “water army” usually starts working during the project’s early promotion and keeps going, on and off, until the film is released or the show has finished airing. (Some TV dramas air in several rounds.)

Another, more “formal” kind of water army team usually works alongside publicity and marketing staff, and its campaigns are shorter.

According to a film publicist, hiring a water army usually takes up a small share of a film’s total promotion spending. Producers generally give a marketing company about one-tenth of the total promotion budget, and the water army gets about one-fifth of what the marketing company is paid. On a typical media marketing budget of about 300,000 yuan, the water army would get only about 30,000. Other Weibo marketing includes topic marketing, event marketing and paid “advertorials” written by big-name influencers. “Frankly, this work is a pain to coordinate, full of fiddly details, and the price is very flexible; 20,000 or 30,000 yuan can basically take care of a project. Publicity firms aren’t foolish enough to do all the dirty, tiring work themselves.”

A water army member who recently posted positive Weibo comments for the stars of several hit films told QDaily that a typical film or TV job lasts about 2 weeks, and that their latest job was 4,000 comments in 2 days. The more “professional” among them save screenshots once the work is done and bill by volume.

It is not only individual films and shows that want this. Some celebrities, protecting their personal brand, also buy a lot of water army services. The same member revealed that someone in the trade had just done similar work for Liu Tao and Yang Zi, each of whom had a new production showing: The Legend of Mi Yue and Legend of Chusen (《诛仙》) respectively.

Most water army workers look out for themselves, or rather, they know how to protect themselves. They all want an easier living and prefer fast money. To avoid fighting platforms like Douban, which can only be breached with sophisticated techniques, most of them, when a client has no firm requirement, advise the client to start with Weibo, WeChat and Baidu Tieba.

One Taobao seller of “water army” services, for example, showed some reluctance about inflating Douban scores. “Jobs like inflating Douban ratings, we sometimes take depending on the situation. It is complicated, and not many people can afford the price.” He also said frankly that his team was used to fast money, and that following, reposting, adding followers and flooding comments on Weibo and Tieba alone already brought in plenty, so there was no need to take on thankless jobs.

The work is “thankless” because the “soldiers” under him now all work by phone; few use computers. Many of their usual throwaway accounts are on Weibo and Tieba, which are convenient; they do little on sites like Douban and Maoyan, and clients rarely ask for Douban in particular. Working there is also more trouble for these “water army” members: they must download the app again, register new accounts, and it may still not work. Above all, as the boss, he runs dozens of WeChat groups, large and small, and would have to “coax” members one by one into downloading the app; unless it is a long-term client, they generally can’t be bothered with these relatively niche services.

“Most of our members don’t have old Douban accounts. If you pay 10 yuan each, with a minimum of 1,000, I can definitely inflate it for you, and there are discounts for long-term clients. It’s just a matter of switching platforms and throwaway accounts. As for the ‘lines’ in the reviews, we practice them every day; we have plenty of experience with all kinds, whether it’s snark, gossip or paid plugs,” the Taobao seller said.

Another reason is that, compared with Douban and Mtime, platforms such as Weibo, WeChat and Tieba offer better value for money.

A veteran of film marketing told QDaily that most clients (film producers) ask for their film’s ratings to be inflated, but the platforms his firm mainly works on are Weibo (clients ask to get onto the trending list or into the top few on the topic chart), the Baidu Index (clients ask for one peak a week), the Weibo Index and the like. Inflating scores and reviews on Douban Movies and Mtime is relatively hard, and the cost and return are basically out of proportion. They have tried it on some projects and found it poor value; it works better to have big Weibo accounts push advertorials and create topics, then bring in a water army to flood the comments, which produces a more visible effect.

On this point, most of the water army agree with marketing and publicity staff. “Many film and TV companies and film marketing firms still treat Weibo as the main battlefield, especially when it comes to building a celebrity’s personal brand; they barely consider other platforms. Weibo has big influencers, media and audiences, and it is so open. If you want to create topics or get reposts, all these interactions are visible, and it’s good value too,” said a water army insider.

The “water army” trade has blurry boundaries, and a wide gap between rich and poor

Search Taobao for keywords such as “water army” and “fan boosting” and you will find hundreds of sellers. Some ask what you need and then quote a price; others simply send you a public price list. Every kind of service comes in standard and flexible packages; for Weibo, WeChat, QQ Zone, video sites and every social network you can think of, someone takes the work, at clearly marked prices.

Here, for example, is a mid-range price list: 20 yuan for 1,000 WeChat reads, 5 yuan for 100 likes, 1 yuan per repost on WeChat Moments (with a screenshot of the repost provided). Sina Weibo is somewhat pricier: 40 yuan for 100 reposts, and 40 yuan per repost by a verified account or “expert.” A repost by a verified account with under 1 million followers can generally be had for under 500 yuan (varying widely with fame), and a repost generally costs 60 to 100 yuan less than an original post. For accounts with several million followers the prices vary a lot, from 600 yuan to one or two thousand.

A team claiming to be the most professional water army service offered a package price and said its quote was the highest. “We price according to the time and process of our operation. Rest assured, my team is spread across the country, the administrators are all university graduates, and our score-inflating content is very professional. For one account, the full ‘Weibo comment + repost + like + top comment’ service is 3.5 yuan, and with a minimum of 1,000 it’s 0.5 yuan cheaper.”

So-called “formal” water army teams fall roughly into two kinds. One is a temporary team of three or five to a dozen or so people; some reportedly cut prices to the lowest in the trade to make fast money, earning the popular name “50-cent party” (0.5 yuan per post), but if orders dry up, such teams fall apart easily. “I turned the 50-cent party around and trained them into an 80-cent party. There should be at least 30 or so professional teams like ours, but not every one has income beyond water army services. Teams like that are generally unstable, their people don’t care much about the work, and they do it sloppily.” This insider told QDaily that the “water army” is a shoddy industry at the very bottom of the food chain, and that to keep his team alive he depends on many other lines of business: he runs media accounts for clients, sells through WeChat, and resells software.

Some more “professional” water armies run teams across the country with a dozen or so WeChat groups, uniform management standards and a ranked hierarchy (for example, “special forces,” a “general headquarters where the group administrators sit,” units handling training and administration, and execution squads numbered 1 to 10), each with its own code name and clear duties.

Overall, the people doing this work are mostly students, or married women at home without jobs. The head of one “water army” outfit claimed that what he runs is not a WeChat sales scheme and not a pyramid scheme, but a way to help tens of thousands of full-time “mothers” in China’s second- and third-tier cities find work at home in the future. By his team’s pay scale, an ordinary worker earns 600 yuan on average, a group leader 2,000 yuan, and a mid-level administrator 4,000 to 6,000 yuan; those earning more than 10,000 yuan a month are all core executives, usually kept to about 5 people. (Staffing ratios differ with the size of the outfit.)

How much does the “water army” really affect the box office?

It has to be said that a sizable number of people do choose which film to see by reviews and ratings. Some video sites and smart TV boxes even set up sections such as “Douban high-score collections.” These do give the box office a boost, but how big a boost has no measurable standard to this day.

Sometimes, in fact, people do not simply watch only high-rated films, and odd things happen. After Dwelling in the Fuchun Mountains (《富春山居图》) opened, for example, reviews were almost uniformly bad, yet some viewers went to the theater precisely to judge for themselves “just how bad it is,” and the film ended up grossing nearly 300 million yuan.

Behind all this lies the “distorted” state a market falls into once its rules are broken. In fact, very few people end up laughing last; most seem to have become victims, not only producers but even actors. Andy Lau, who starred in Dwelling in the Fuchun Mountains, was criticized by his fans; at a press conference, when asked about it, he stood up, said “I’m sorry” to the audience three times, and added: “Poor choice of films is my biggest flaw! I overestimate my own judgment, and always assume that any script will get better once it’s in my hands. I’ll do better next time.”

After the release of The Last Supper, as the “water army scandal” spread, Qin Hong, the boss of Xingmei, publicly said he regretted investing in the project, and that he would “not invest in (Lu Chuan) again.”

Weibo screenshot

“Although the water army keeps up with the times, using cleverer techniques to adapt to the new era, and although it takes on platforms that have defenses against it, it is a very small link in the marketing chain, and its final effect on any battlefield is still very small. Among publicists who have hired water army services, more regret it than not, and even those who hired them feel it looks fake, because the overall quality of the water army is low,” said a person who handles publicity for entertainers.

A film and TV industry insider agreed. However professional a review-inflating “water army” is, he believes, its methods are still fairly mechanical. In the past two years the film industry has boomed, more and more people go to the movies, and they have more ways to choose a film. Plenty of people pick a film by its score, but that is certainly not the only standard. For producers or publicists, rather than using a water army, it would be better to put effort into proper marketing materials, such as polished trailers, stills and merchandise.

In his article, Abei admitted frankly that the water army exists, but said that for the past two years Douban’s rule toward it has been that “any rating that can be identified as abnormal does not count at all,” whether high or low, from the powerful or the humble, of whatever color.

What has Douban done? Abei’s article reads: We call rating driven by the urge of “damn it, I won’t accept this, I’m going to push this average up / pull it down” “abnormal rating,” that is, rating whose purpose is to directly interfere with the average score. In my own impression, “abnormal rating” falls roughly into four types: registering / buying accounts to inflate scores, registering / buying accounts to deflate scores (this one has puzzled me too), celebrity fan clubs “attacking Douban,” and die-hard users’ counter-campaigns to “defend the fairness of Douban ratings.” There must be others, such as performance art.

Abei also said that the water army’s effect on Douban ratings is actually not that great, and that what small effect it has is short-lived and isolated, because so many people rate normally, and because fighting score inflation has long been part of Douban Movies’ daily work, which many colleagues have quietly carried on with the help of more and more software.

At the same time, Abei admitted that there is no once-and-for-all solution to the water army problem. So what can people do to lift their film’s Douban score? His answer: “I’ve already covered score inflation: it is less and less useful. So I really don’t know what you can do apart from making a good film.”

Cover images: weibo.chinatimes.com, xiaoxiongbizhi