The Structure and Style of a Dogma Community: Influencer profiteering, conspiracy theories, and organized engagement in the Sushant Singh Rajput discussion

Anirban Sen and Joyojeet Pal

The Structure and Style of a Dogma Community: Influencer profiteering, conspiracy theories, and organized engagement in the Sushant Singh Rajput discussion

Anirban Sen and Joyojeet Pal

The barrage of anti-Bollywood Twitter trending is closely tied to a small group of dedicated Sushant Rajput fans. Their networks, language and articulation of fandom, and the temporal patterns around their social media activity give valuable insights into organized trolling on Twitter in India. But the consistent messaging from the community also gives us unique insight into how a small community with a dedicated belief in an otherwise debunked notion can have an outsize effect on a broader ecosystem.

The community itself, often referring to itself as SSRians, is diverse and those engaging in messaging around Sushant Rajput over time have evolved significantly. The core community of those engaged in messaging now (mid-2022) treats the death of Rajput in the same way that religious communities treat dogma. There can be no evidential argument on it. It is also not clear if any explanation for the events will be satisfactory for the community. The nature of doubt over the events leading to his death are central to the existence of this community and to the value it brings such as fraternity and a sense of common purpose. In this paper, we examine the community and its messaging using the Twitter feed of a curated set of 173 “core member accounts” of the SSR community. 


The core community include some who were dedicated fans before his passing, but several who got interested only after his death. The core group of fans exhibit phenomena of personal devotion that has been described as obsessive disorders in psychology, as well as sociological characteristics of participatory engagement with culture in fandom studies. Several studies have shown that fan groups in India for film stars like Rajinikanth, Salman Khan, Vijay, etc. have such characteristics of intense, almost cult-like devotion on both personal and collective levels. With the SSRians, we see a coming together of elements of fandom around the individual and an obsession around an event – his passing. The dogmatic belief in his death as being engineered, and by extension, of a broader set of conspiracies to target outsiders is elemental to being part of the community. However, unlike fan communities of Indian film stars, which celebrate the body of work by the individual, the ‘SSRians’ tend to focus on the person and his characteristics, rather than on his canonical works. In this, the action around SSR is possibly closest to the community of people dedicated to engaging with theories around the death of Diana Spencer, the former princess of Great Britain.

Graphical user interface, text, application

Description automatically generated

Figure 1: Tweet from SSRian user comparing the death of Sushant Rajput to those of several others whose deaths are still discussed as conspiracies

While the group currently referring to itself as SSRians on various social media is part of a core dogma community, a community whose sole existence on social media depends on an event or a set of events, and who do not generally engage in posting about any other event(s). That community itself has not stayed constant – various Twitter handles have joined, left, or been renamed since the actor’s passing. The longer history of the group and its discourse has been influenced by messaging from mainstream media channels, which milked the story in the months following the actor’s passing, but also several celebrities, politicians, and social media influencers who found that the story got traction online. Work by Zainab Akbar and others shows that television networks and their online channels, particularly Republic and TimesNow, had a sustained engagement in the story from prime-time viewers, and at the same time, politicians used the story as a means of attacking the state government in charge of the investigation.

However, there also exists another set of profiteers who have benefited from the SSR story. These are people whose social media following is largely or primarily driven by the fact that they engaged in theories relating to the death of Sushant Singh Rajput. These accounts coordinated messages, arranged hashtags, and sent instructions to be retweeted by others in their networks. Many of these profiteers have high “in-degree” networks, which are calculated by the number of people who engage (follow or retweet) them, but low “out-degree” networks, which are calculated by the number of people that they engage.

Graphical user interface, application

Description automatically generated

Figure 2: The first messages to demand retribution days after the actor’s death from actor Kangana Ranaut and influencer Sonam Mahajan 

A misinformation community also typically has a few large-scale influencers who nudge the issue along. In the SSR case, this was the parliamentarian Subramanian Swamy (@swamy39) who had over 85 tweets that were significantly engaged and was retweeted over 650k times overall on SSR-related messaging. Swamy led the demand for a CBI probe into the actor’s death, suggesting that the investigation by Mumbai police was incomplete or incompetently carried out through a series of tweets. More importantly, Swamy validated the community by referring to them as “warriors.” Swamy also added fuel to the direct vitriol aimed at Rhea Chakraborty, Rajput’s girlfriend who was both viciously trolled online and arrested, arguably in an environment highly charged by a trial by public opinion.

Graphical user interface, text, application

Description automatically generated

Figure 3: Messages from Subramanian Swamy encouraging the formation of a #Warriors4SSR community, and subsequent messages promoting conspiracy theories

A look at Swamy’s engagement with the SSR story however also gives us some insight into his astute understanding of what is a newsworthy story. First, Swamy did not start tweeting about the subject for a month into the actor’s death. He hopped onto the bandwagon only when the story was already catching steam from a mix of influencers and general public engaging the story in July 2020. He continued to tweet about the subject till early October 2020. However, in the first week of October 2020, both a Maharashtra government report on the use of bots in the SSR social media outreach, and the study by Zainab Akbar were released, showing the level of media and political complicity in the misinformation around the case. A day after the release of the report, Rhea Chakraborty was released from jail, and we see the drop in the average retweet value of the story

Chart, box and whisker chart

Description automatically generated

Figure 4: Monthly engagement of @swamy39 with SSR-related tweeting showing that the account stopped engaging it as the public engagement fell

Other influencers, not as directly complicit as @swamy39 in suggesting wrongdoing contributed to the overall movement by asking questions that threw shade on the investigation – these included actors with large social media footprints like Sonu Sood (@sonusood), Akshay Kumar (@akshaykumar), Paresh Rawal (@SirPareshRawal), and Simi Garewal who also asked for CBI probes. Such one-off engagement by mainstream stars like Sood and Kumar brought very important legitimacy and mainstream attention to the narrative than accounts like those of @Swamy39 who carry the weight of being distrusted for their past sensationalist messaging on Twitter.  Simi Garewal, however, skirted the space between a casual interest and active engagement and propagation of misinformation, as she became one of the first few influencers to add the Disha Salian angle to the SSR case. The story, entirely fabricated, was that that a nude Ms. Salian was raped and murdered as part of a larger conspiracy. This was debunked in August 2020 — Salian’s parents confirming it and had repeatedly requested that people desist from involving her in the conspiracy theories. Nonetheless influencers including Garewal and BJP politician Narayan Rane continued to make and amplify these claims.

Graphical user interface, text, application, chat or text message

Description automatically generated

Figure 5: Tweets from actor and television presenter @Simi_Garewal presenting conspiracy theories about the death of Sushant Rajput

We see in figure 6 that several influential accounts (larger bubbles) have a massively outsized impact on the discourse, both in scale and in the number of messages (Swamy for instance, posted more messages in the study period than Sushant Singh Rajput’s own sister). We also see that many other influential accounts including media channel @republic, BJP spokesperson @ippatel, character actor @shekharsuman7, and anchor @pradip103 were key drivers of misinformation. We find that the majority of the dogma community sits on the upper end of the figure – ie higher number of posts about SSR, but relatively lower overall impact as measured through retweets earned.

Figure 6: Bubble plot of Tweets (log-scaled on Y axis) of key players and the engagement to their original tweets on Sushant Singh Rajput

We plotted the dates when major influencers tweeted about Sushant Rajput. We did this by plotting the total daily retweet to the messages from the top five influencers on any given date (thus the cumulative retweet count of the top five messages on any date) and we find that the overwhelming majority of high-engagement messages are early in the news cycle, with a huge drop in engagement by October. This is also when Subramanian Swamy stops engaging the subject.

Chart, histogram

Description automatically generated

Figure 7: Graphical representation of total traffic of the top five tweets on Sushant Singh Rajput by day, showing n of Retweets (y axis) by date (x axis)

It is interesting nonetheless that although the peak of overall messaging was in the hours following the actor’s death, the actual increase in highly viral tweets came about a month after his passing, and the high-intensity tweeting continued into mid-September 2020. This is because ion the day after his passing, many online expressed grief and shock, and it was only after a week didthat the flutters of murder emerged from the initial discussion on mental health. While other work has shown the political influence in the murder narrative, particularly as a means to attack the Maharashtra government, there is also a strong element of Bihar politics in claiming Sushant Singh as a son-of-the-soil, both to impact the North Indian voter bank in Maharashtra and also to influence the voters in the Bihar legislative elections in 2020. Besides Swamy, the head of BJP in the state, Sushil Modi, threw shade on the Maharashtra government, and sitting BJP MP claimed at a campaign rally that the murderers of Sushant were being protected.

Graphical user interface, text, application

Description automatically generated

Figure 8: Tweets by BJP politicians using the Sushant Singh Rajput case as a poll issue in Bihar, and opposition and journalists calling it out

Network Characteristics

We studied a core retweet network of 173 user accounts that we identified as repeatedly tweeting about Sushant Singh Rajput as most of all content they put out (see methodology section below for more discussion on how these were identified). In this network, an edge between two accounts is indicative of at least one retweet between the two. We found 13990 edges between the accounts (nodes in this network) with a network density of approximately 0.49, which indicates that the network is significantly dense. This indicates that the network is comprised of a significant number of “core” users who are closely-knit and tend to retweet each other. The network characteristics are also very useful in understanding which users are drivers of the discourse. We see in Figure 1 the network of users exhibiting an eigenvector centrality of >=0.8 (i.e., users who are highly influential in terms of the engagement received. A user in this network has a high eigenvector centrality if they are retweeted by other highly influential users). The width of an edge in this network denotes the number of retweets between two users. Even in this network of users with high influence, we observe the existence of a hierarchy – users like @Khushi4justice (since banned) and @ApurvaU21 exhibit a lot of engagement (them being retweeted by several users in the network as can be seen from the thick edges incident on these accounts). These accounts are the ones that have the most network effect on the discourse, however, as we see later, the account from which a lot of the original tweets emerge is @nilotpalm3. This suggests that the network is a well-oiled machine, with alpha accounts that generate the content, lieutenants that have the ability to push the content to a large number of nodes – soldiers, which in turn do the work of spreading the content. All but one node on this figure ranks above the threshold of a ‘bot’ on botometer.

Figure 9: Graphical representation of the key players in the SSRian network for the study period (excluding banned accounts at the time of the study)

A consistent pattern of accounts in misinformation networks is the use of alternate handles. A significant number of accounts, including all the accounts that were central to the network’s output, had either advertised alternate accounts (on their profiles) or were known to have similarly named alternate accounts that were brought into action in case the primary account got banned. For example, a key influential user, a local influencer called Nilotpal Mrinal (who identifies himself publicly) has the handles @nilotpalm3, @nilotpalm6, and @nilotpalm1, the first account being the main handle of the user. The two other major accounts with massive footprint such as @ApurvaU21 doubles with @ApurvaU23 and a possible number of banned accounts, while @PiyaliBH had at least 7 similarly named accounts.

Graphical user interface, text, application, email

Description automatically generated

Figure 10: Planning tweet for coordinated activity, requesting users to post their back-up account IDs in case of bans

Another pattern consistent with coordinated activity is highly variable activity from various accounts. Unlike typical Twitter handles, which tend to tweet at a relatively stable pattern, even if they have bursts of tweeting followed by periods of silences, we find that the several of the SSRian accounts in the dogma community go from not tweeting at all to tweeting aggressively, more than 2/3rds of the core SSRian community tweeted over 100 times a day, over a third tweeted over 200 times a day about Sushant Rajput during the timeline of this study. These high-tweeting days often have multiple accounts that go from silence to aggressive activity on the same days, and these also tend to be days when the core accounts send out messages that go viral. The sudden jumps in tweeting and retweeting activity can be seen in figure 2, which shows aggregated number of tweets and retweets over time by the core community of users considered in this study.


Several of the major accounts and their variants score highly on the Botometer (including the accounts associated with Piyali and Apurva) – which doesn’t necessarily confirm that they are bots, but suggests strongly that they behave like bots, or are unusually single-minded, with little else to do in the day except messaging about SSR. The account ApurvaU21, which does appear to have consistently tweeted about SSR right from his death, unlike most other accounts, averaged over 100 tweets on a daily basis, while the combined output of Piyali amounted to over 500 tweets daily, averaged over the study period. We found several instances of individual accounts that put out in excess of 500 tweets in a single day about SSR, several putting out more than 2000 tweets in single days, despite not tweeting at all on other days. That several of these accounts are astroturfing is obvious.

The behavior of these accounts has a few consistent patterns. First, of the 173 dogma community accounts, 150 were above the threshold of what botometer considers a bot, with only nilotpalm3, the core account from which most outgoing messages flowed, below the experimentally decided threshold of what counts as a bot. We visualized the accounts that were identified as bots, Figure 11 below gives us some insight into their behavior.

Chart, scatter chart

Description automatically generated

Figure 11: Each of the core SSRian accounts visualized for its active days, total number of tweets, and proportion of original tweets to retweets. 

We see that as an SSRian stays on Twitter longer, the number of original tweets reduces, and the retweeting activity increases. We also see that a small number of accounts that have a very high impact in terms of overall retweets tend to be the ones that have a higher tweet to retweet ratio. But we also see that the frequency of messaging is so intense, that even if there were individuals who actually messaged with as much frequency and with the kinds of content that they engage, they would closely resemble trolls. In fact, even the alternate account of Nilotpal, nilotpalm6, scores above the bot threshold on botometer. Much as it is clear that many of these accounts are likely to be bots, they also exhibit patterns that can be explained in other ways – one of the most influential accounts @Hemant36182804 followed the consistent high tweeting style and had a consistent pattern of tweeting between 9-11 am IST, and thereafter between 3-5 pm IST on weekdays, with some sporadic activity on weekends, suggesting the use of a workplace computer.

Only a small number of handles consistently originate the tweets that go viral, and the nilotpalm3 account is at the heart of virtually all activity. The suggestions of coordinated and bot-supported activity is also seen in platforms’ regulation of the SSRian community – at least on one occasion, several accounts went offline overnight. As we see below, there are several precipitous drops in activity around the overall tweeting related to SSR, suggesting that the community comes together to systematically tweet in spikes that are totally unrelated to the news cycle. 

Figure 12: Date-wise breakup of the number of tweets and retweets in the entire network 

We see in figure 12 that as the overall mainstream community reduces engagement in the issue, the core SSRian community gradually rises at first. However, the rise in activity increases precipitously after January 2021. There is a spike in messaging on Feb 21, 2021, when the community started trending a “Jan Kranti 4 SSR” tagline. The mobilization seems to have energized the community, and we can see in Figure 12 that there is a new trajectory of tweeting. However, this rise in tweeting is misaligned with the overall interest on Twitter on the subject, since we see much higher engagement from the core community as compared to the overall retweets. This is even more obvious when we turn to the mainstream media coverage of the subject in Figure 13.

Figure 13: Daily number of tweets from five media sources and by the number of tweets from the overall core SSR community

The line signifying the n of tweets from the core community is the same in figures 12 and 13. We plotted the number of tweets from the key news channels covering the issue – @republic, @aajtak, @TimesNow, @ZeeNewsEnglish , @ZeeNews – each of these channels covered the subject extensively during the early phase of the issue,  and are generally considered right-leaning. We find that there is little or no interest in the subject in the mainstream media, as seen through the drop in media coverage, since October 2020 – exactly as we see in Figure 7 with the influencers’ engagement with the subject. As the mainstream media sources reduce their engagement in the subject, it also changes what gets retweeted within the network. While past work has already shown that the television channels engaged in much of the inflammatory speech and conspiracy theory mongering in the first months after his death, we see that even mainstream news disappears from the networks, as it sinks deeper into a discourse that is almost entirely self-contained and recursive.

A means of examining abnormal patterns in social media tweeting is tweeting in fits and starts, but with high output volume. We classified all the core accounts based on if they started tweeting at the beginning of the Sushant Singh issue online or started later, and if they went through a period of extended silence (two months or more) or tweeted consistently. Based on this classification, we find that different types of users show different tweeting behaviors in the core community. Some users start tweeting early in the period considered but stop midway due to their account getting banned (e.g., @_ADITICHOURASIA in the following figure). Some users start late and continue tweeting. Some others keep tweeting during the entire timeline of event, and these are generally the drivers of community activity (e.g., @ApurvaU21 in the following figure).


Description automatically generated with medium confidence

Figure 14: Visualization of the daily tweeting pattern of four sampled accounts, based on when they started tweeting about SSR, and whether their accounts were suspended at any point by Twitter

Conspiracy theories

SSR community shows signs of typical conspiracy communities, which research in psychology shows tend to be around epistemic concerns, existential drivers, and social drivers. Research shows that during crisis periods, acceptance of conspiracy theories spike. The intersection of the COVID-19 lockdown created conditions that enabled some of these – there was a high degree of uncertainty due to lockdown, a spike in household consumption of news, and an overall sense of existential fear.

The research on misinformation categorizes drivers of misinformation into three categories. The first, epistemic concerns, are among the most basic drivers of misinformation since they are primarily a struggle with the inability to make sense of an event. The lack of clear resolution drives misinformation, and people want to be able to explain what is going on. Innuendo by influential accounts and media channels help through doubt and confusion, increasing the drive for epistemic uncertainty. However, over time, the slow progress with the case, despite several clean chits from investigation agencies suggests that institutions have allowed this community to thrive. It is important that the case has still not been closed, allowing this sense of uncertainty to continue. Epistemic concerns are essentially fueled by speculation – as many unanswered possibilities that appear can add more possibilities for threads of misinformation. We see in figure 15 a series of threads of theories started by various individuals — @Apurva21, the long-standing member of the SSR community was the first to throw doubt, followed by digital influencer @soumyadipta, but it was Subramanian Swamy’s 26-point breakdown of reasons for doubt which drove up the conspiracy theories, especially in August and September of 2020, when insinuations were being rapidly amplified – and the tweet from BJP spokesperson bringing attention to an unconnected person is a case in point.

Figure 15: Examples of epistemic drivers of misinformation, tweets that propose speculative possibilities for an event

The second type of contributors to misinformation spread are existential drivers, which are typically around fears related to one’s group. This makes people willing to trust conspiracy theories. Idea of systematic discrimination vs. outsiders, and peoples’ self-identification as outsiders in some setting. Several viral messages used the notion of Biharis as outsiders in Mumbai, harping on the difficulty of ‘making it’ in an India where nepotism is rife. These messages suggest that one is at risk, and that what happened in the case of SSR is something that could happen to you. While the majority of what we see as existential misinformation tends to be Bihar centric, there is also a general sense of the individual as being at the mercy of a nepotistic system. Here, as within some of the other threads of misinformation, there is an initial spark thrown by a commentator not necessarily seeking to trigger conspiracy theories, but end up doing it. The case here is of journalists Shekhar Gupta and Rahul Kanwal starting conversations about outsiders in Bollywood, which triggers a series of quoted tweets that push speculation. 

Graphical user interface, text, website

Description automatically generated

Figure 16: Examples of existential drivers of misinformation, tweets that promote the idea that one is at fundamental risk

The third type of contributor to conspiracy theory is that around social drivers. These relate to engaging in misinformation because it aligns with one’s social worldview, as it relates to one’s group identity. Here, one attempts to show oneself, or one’s social group, as superior by proposing misinformation about another. When the social driver is group-oriented, it usually presents a social antagonist, which could be a collective such as people of a certain religion, ideology, or gender identity – they could be based on values such as “women who eschew religious orthodoxy”, crafted as antagonists for #MenToo activists, or “Muslims” crafted as antagonists in the immediate post Tablighi Jamaat discourse around COVID-19.

In the SSR case, nationalistic language was incorporated into the social drivers of misinformation, presenting the murder of Sushant Singh Rajput as something perpetrated by a group of defined antagonists, mainly leading members of the Bollywood industry – often Muslim male stars, presented as a cabal hand in glove with a broader conspiracy against the national good. We see for instance several tweets which hint at an involvement between the SSR case and mafia don Dawood Ibrahim, or a Pakistani conspiracy.

Graphical user interface, text, application

Description automatically generated

Figure 17: Examples of social drivers of misinformation — tweets where the conspiracy theory acts in support of one’s perception of a social group

An example of this is the unusual intersection of revenue officer Sameer Wankhede with the SSRian community. The officer, who came to national headlines for the very public arrest and extended detention of Aryan Khan for a case that was eventually showed to have no evidence. The reputation of Wankhede as a hero, waging war against a Bollywood cabal was strengthened by the SSR community. We see a massive number of trending hashtags with Sameer Wankhede in them, and an intersection between the SSR community and the mobilization against Aryan Khan.

The key mover of the entire discourse continued to be the community kingpin, Nilotpal Mrinal, who photographed himself with Wankhede at the peak of the Aryan Khan controversy, proposing a linkage between the issues. The nexus between the two, and the connections between these and a broader ecosystem of attack on Bollywood was also brought to fore by journalist Sudhir Suryavanshi, as we see in the figure below.

Graphical user interface, text, application

Description automatically generated

Figure 18: Tweets on Nilotpal Mrinal, one of the key movers of the SSRian community

Coordination Strategy

One strategy we see since early 2021 is that of coordinated hashtagging. Here, the SSRian community has taken a page out of political tweeting, where the community takes a pyramid structure, and a small number of accounts at the top of that who decide what has to be talked about by the community as a whole.  For much of the first year, the community used hashtags to communicate, whereas by mid 2021, the strategy had moved to taglines where an exact string of text would be cut and pasted into every message that went out. This was ostensibly to avoid the attention of algorithms. We see examples of the community being explicitly instructed to avoid hashtags and use taglines instead. Ironically, this also suggests that those getting instructions are humans, not bots.

Figure 19: Instructions to the SSRians to use an exact string of text as a tagline in place of a hashtag

The move to the tagline also shows how closely the community has gamed the twitter algorithm to continuously make the waves. There are a number of spike periods, however, the most significant of all is the spike on his death anniversary on June 14, 2021. The event was planned in advance, with instructions sent out to the entire community on when they needed to be online, what they needed to do, and how much. Messages were also sent out periodically, through the trending date, to encourage users to tweet more aggressively, as we see in Figure 20.

Figure 20: Instructions to the SSRians to use an exact string of text as a tagline in place of a hashtag

Events like the “Sushant Justice Matters” trending eventually help give the community something to find common cause in. The community, through a discursive process, planned a major event to commemorate the actor’s anniversary, and in fact succeeded in reaching their target. They were able to get influential accounts like @ErayCr and @ishkarnBHANDARI (both of which gained online following largely due to their engagement with the SSR issue) to use the exact tagline that the community wanted to trend.

Graphical user interface, text, application, email

Description automatically generated

Figure 21: Influencers using the same hashtags as the core SSRian community for the June 14, 2021 trending event

What are the implications

It is important not to dismiss the SSR community as a group that is purely engaged in conspiracy peddling. The community shares characteristics with other groups that spread misinformation or extreme speech, such as political hate speech groups, but unlike groups that are driven by a broad-based ideology, the SSR community has a shared sense of fraternity based on its shared belief. Although the community coalesced around Nilotpal and a few key figures at various points, the nature of belief in an idea that has quickly fallen out of favour with the mainstream has also meant that there is a lot of tension around the leadership of the community. Running an online astroturfing operation of this kind requires a certain amount of centralization that is not always easy to pull off, especially in a community whose foundational belief is that of suspicion. The community is beset by bickering between leaders, and those who decide when to make things trend online. The times of greatest coming together are often negative events such as organized trolling of those that are seen as antagonistic to the core cause.

The continuing targeting of the Bollywood community is ongoing, including the systematic marking down of film ratings on platforms like IMDB. The nexus between the Aryan Khan and SSR cases is also important, because the SSR community has been weaponized several times, and this case is important particularly because it highlights the possibility of building a community around one idea, throwing a series of concepts as the driver of discontent, and then incorporating new issues as extensions of that discontent. 

While there is clear evidence of systematic manipulation by key players in the community, there is clearly also a significant component of people in the SSR community who are genuinely convinced of the conspiracy and broadly agree with a number of threads that emerge from it. They exist well outside of Twitter – on Telegram and WhatsApp channels, have YouTube meetups and discussions. While a small number of users like Nilotpal Mrinal may be the primary movers of the community for a majority of its first two years (as also discussed in different news pieces previously), there are new players now, and there exists a constant churn in the community. The community is generally volatile – while it has been effective at consistently attacking a few select targets, it has significant internal differences over leadership and direction. Despite its history of mobilizing against certain causes, there is not always (or even typically) agreement outside of the death of Sushant.


We started with a set of 10 most active seed user accounts that were engaged in tweeting about Sushant Singh Rajput (commonly known by his initials, SSR), referred to henceforth as the seed users. These users were sampled from the results of another study on conspiracy theories around the SSR case \cite{akbar2020anatomy}. We next collected the user details for their one-hop friends on Twitter. We applied two levels of filtering on these one-hop friends of the seed users: (A) We removed users who had less than a threshold number of followers in the seed set and the previous hop combined, and (B) We checked if the profiles of these users and a small set of their previous tweets (from the day of data collection) contained at least one keyword/hashtag from a set of 20 popular keywords/hashtags related to the SSR issue (e.g. “ssr”, “sushant”, “rajput”, “#justice4sssr”, etc.), and removed those with no such keywords/hashtags in their profiles or previous tweets. We applied a similar method for the two hop friends of the seed users. The remaining two-hop users (after applying the filtering steps as indicated above), along with the users in previous set contained a total of 1532 users (excluding the seed users). By iterating this process twice, we see that this set includes an overwhelming majority of individual accounts retweeted in the SSR case. We did a ground truth analysis of this by tracking all the handles ever mentioned in tweets by this community,  and found that the hypothesis holds true. While we did find a certain number of handles that were highly mentioned, these were overwhelming in one of three categories —  the first was general influencers such as India’s prime minister, various influential public officials, and members of the press, whose attention was sought by the tweeting, the second was influencers who were being attacked by the SSR core community for alleged complicity in his death such as other film stars or politicians, and the third were influencers who had at some point or the other expressed sympathy for the actor or his fans in a tweet. We left out direct members of the star’s family or his close associates active on Twitter since they are not part of a dogma community that has coalesced around fandom and the event as specified in our research.

We found that a significant number of users in this set post a small number of original tweets, while posting a much higher number of retweets. There exist a handful of users who tend to post both a high number of original tweets and retweets. Our assumption is that the users who post a high number of tweets might be the influential users in the community, who generate and drive the narrative related to the SSR issue. On the other hand, the users who exhibit a high number of retweets act as information spreaders, i.e., they retweet these highly influential users to spread the message generated by them.

We first selected only those users from the aforementioned set who retweeted at least one tweet authored by a user belonging to the set. This was to ensure that the user is interested to an extent in the group of users that we sampled. Finally, to extract users who are active participants in the dogma community, we selected a set of core users from the resultant set, who post a minimum threshold of tweets and retweets during the event timeline. We selected the top 80 percentile of the user population, i.e., we weeded out users who posted less than a threshold number of tweets (including retweets) on the SSR issue during the timeline considered (14 June 2020 – 20 April 2022), a total of 612 days. The resultant set including the seed users was used for further analysis. In essence, these accounts are what could be called a core community that is heavily invested in engaging material related to the Sushant Singh Rajput case.