2/17/26 – Early voting starts today. I hope everyone votes. Electing people who address the problems of their constituents is essential. But sometimes, it’s hard to recognize who those people are because of political deception using trolls, bots and “engagement farms.” These invisible influencers can:

• Undermine our understanding of issues
• Alter our perception of reality, much like a drug would
• Rob us of the power to make informed decisions
• Trick us into voting against our own self-interest.

How? By flooding social media with false information disguised as comments from legitimate users, they make it appear as though there is a groundswell of opposition to a particular candidate.

My First Clue

Several weeks ago, I posted what I thought was an innocent story about Congressman Dan Crenshaw. It detailed how a Federal grant he secured for improvements to the Kingwood Diversion Ditch kept the project from being killed by Democrats in Harris County Commissioners Court.

Out of nowhere, the post was flooded with off-topic comments criticizing Crenshaw and promoting Steve Toth. Toth is running against Crenshaw in the new Texas Congressional District 2.

My post never mentioned Toth. But comments on the post ripped apart Crenshaw, a war hero who has dedicated his life to public service and improving public safety.

Toth’s trolls were as relentless as a pack of rabid jackals. I spent hours trying to respond to their inane comments before I gave up … exhausted. Since then, I’ve been investigating what happened. It goes way beyond the childish defacing of an opponents election signs … which Toth also does.

Online, bots and trolls can be programmed to bomb posts and pages with scripted comments that wear down humans and crowd out legitimate discussion. They often work together to create the impression of widespread discontent about political opponents. This can influence people who may not follow politics closely. They sense the majority has concerns about a candidate. So, they vote for his/her opponent.

Engagement Farms

I also learned about “engagement farms.” Engagement farms coordinate employees who control multiple accounts and are directed to:

• Comment
• Dislike
• Share
• Gang up on others
• Attack opponents
• Artificially inflate visibility.

They are often paid by the volume of interactions. ChatGPT cites publicly advertised engagement-farm service rates ranging from $50 to $300 per 1000 comments. Now you know why trolls often use one word comments such as “Hogwash”!

Unlike bots, engagement farms generate authentic human interaction. That makes detecting them harder. But they all use similar techniques to slime an opponent’s reputation.

How to Recognize When You’re Arguing with a Troll, Bot or Engagement Farm

Below, I use the word “troll” to apply to all three categories. Their tactics and goals are similar. Their hallmarks include:

• Intentional provocation (not just disagreement):
  • They post inflammatory, extreme or deliberately offensive claims to trigger outrage
  • When others try to respond constructively, they steer toward conflict
  • They seek antagonism
  • Comments amplify conflict rather than resolve it
  • Trolls label someone as corrupt, inept, or bad without evidence.
• Bad-faith argumentation:
  • Ignoring responses
  • Misrepresenting others’ positions to make them easier to attack
  • Moving the goal posts once counter-evidence appears
  • Trying to muddy the analysis
  • Nitpicking technicalities to derail substantive discussion
  • No amount of evidence ever convinces a troll.
• Circular discussions
• Anonymity and disposable identities:
  • Trolls often operate from new accounts with no or little history in them
  • The accounts have no or minimal personal information
  • These allow lying without jeopardizing the reputation of the troll
  • If blocked, new fake accounts quickly replace old fake accounts.
• Derailing topic threads into unrelated controversies
• Hijacking legitimate discussion with ideological provocation
• Repeating previously debunked claims to restart conflict
• Rapid “pile on” behavior indicates coordinated activity
• Minimal debate among those piling on
• Unlike someone who simply vents, trolls typically:
  • Re-engage repeatedly after pushback
  • Continue, even when ignored, to escalate intensity and reignite arguments
  • Seek visibility, not persuasion
  • Take extreme, minority positions to provoke conflict
  • Cast doubt without proving allegations
  • Spew misinformation
  • Cite obscure blogs or misleading statistics
  • Present partial truths without context
  • Mix legitimate data with distortions.
• Claiming false affiliations (I’m a lifelong Republican, but that guy…)
• Pretending to be local when they may be overseas
• Selective skepticism:
  • They apply high evidentiary standards to opponents, but…
  • Accept vague, anecdotal evidence that supports their side
  • Ignore contradictory facts.
• Coordinated behavior:
  • Multiple accounts repeat similar phrases
  • Synchronized messaging
  • Identical sentence structures
  • Specific rhetorical constructions, such as:
    • Unusual word choices
    • Semantic duplication, even when wording changes slightly
    • Copy-pasted paragraphs with minor edits
    • The same metaphors, accusations and claims without citations or context (He’s a RINO!)
    • Reused rhetoric
    • Name-calling
• High outward aggression, low inward disagreement.

Why It Matters

Trolls use intentional, bad-faith provocation to disrupt legitimate discourse or elicit emotional reactions. Trolls want to mislead. Their disruption denies people the information they need to make voting decisions based on facts.

Political trolling also increases polarization and creates public distrust of government. People begin to think, “All politicians are corrupt.”

Taking Trolling to Next Level

“Engagement Farms” pay people to perform online tasks that influence your behavior. They began a decade or so ago by paying people to submit fake product reviews. In politics today, they are used to:

• Damage and harass opponents
• Simulate grassroots outrage
• Influence undecideds.

Engagement farm attacks:

• Feel personal
• Appear socially validated
• Create a strong perception of widespread consensus.

They seek to make you doubt. In the end, unless you really know the contestants first hand, you may follow an imaginary crowd. They make it hard to discern fact from fiction. And that is happening in the Crenshaw/Toth race.

Exploiting a Regulatory Gap

Social media comment-based attacks often avoid regulations on traditional advertising. For instance, they don’t require disclosure of the sponsor. This regulatory gap is significant.

Super PACs can often legally spend unlimited, untraceable amounts on engagement farm services.

While hiding behind the First Amendment, they wear down and exhaust humans trying to interact legitimately.

Bob Rehak

In the digital age, they amplify digital discontent like hecklers at a political rally. Enforcement actions against them are rare, difficult to prove and occur long after the election is over…if ever.

I personally believe nothing coming from the Toth camp. I’ve caught them in too many lies. I have endorsed Crenshaw, a man I have known for eight years and who has delivered hundreds of millions of flood-mitigation dollars for the Lake Houston Area.

Caution: Disclaimer

Campaign supporters acting individually may use some of the tactics outlined above. Some may even support Steve Toth without pay.

Posted by Bob Rehak on 2/17/26

3094 Days since Hurricane Harvey

The thoughts expressed in this post represent opinions on matters of public concern and safety. They are protected by the First Amendment of the US Constitution and the Anti-SLAPP Statute of the Great State of Texas.