Why Do They Muck with the Numbers?
If they are bad actors, we know why.
If you’d like these posts delivered directly to you, subscribe to the Decisions newsletter now!
From WTKR News 3
Pictures of four servicemembers killed relating from Operation Epic Fury; The Defense Department originally listed their fatalities but later removed them. The New York Times reported Thursday, citing three military officials, that the administration said the deaths were removed because they occurred after the declared ceasefire.
Sgt. Michael Emmanuel Swinton: Age 30, killed during a controlled detonation of a drone at Erbil Air Base in Iraq.
Pvt. Isabella Gonzales: Age 19, killed during an attack at Muwaffaq Salti Air Base in Jordan.
1st Lt. Tyler James Feehan: Age 25, killed during an attack at Muwaffaq Salti Air Base in Jordan.
Sgt. Angel S. Rampersad: Age 28, killed during the attack at Muwaffaq Salti Air Base in Jordan.
On Wednesday, the Pentagon's casualty website counted 18 American service members killed in the Iran war.
On Thursday, it counted 14. [1][2]
Nothing miraculous happened to these soldiers. Four names just disappeared from a government website overnight, despite the Defense Department putting out a press release about “the deaths are related to Iranian aggression.” [3]
According to reporting from the New York Times, three military officials said the four were removed because their deaths came after President Trump had declared a ceasefire back in April — even though the war, and the dying, obviously didn’t stop.
The Pentagon's own spokesperson offered a different explanation, calling it "temporary data disruptions" that would soon be corrected. Two explanations for the same missing four names, but neither one is "we made a mistake and we're restoring the count."
As Kai Ryssdal, longtime host of NPR’s Marketplace, put it: “Disgraceful. Simply disgraceful.” [4]
This is what happens when you already know the conclusion you want people to reach, and you adjust the measurement until the data show it, rather than reporting what actually happened and living with what it is.
Or more succinctly, hiding the truth to make yourself look better to others.
This week gave two more examples, each a different flavor of the same move.
The EEOC’s Republican majority voted 2-1 to end the 60-year-old requirement that private employers report workforce demographic data. [5] Chair Andrea Lucas argued the reporting itself risks encouraging discriminatory hiring decisions.
(…although those comments likely are focused on perceived discrimination against white men; Chair Lucas in a social media post on X publicly encouraged white men to come forward and report discrimination claims: “Are you a white male who has experienced discrimination at work based on your race or sex? You may have a claim to recover money under federal civil rights laws.” [6][7])
Whatever the merits of that argument, the effect is the same: if a number might someday produce an uncomfortable answer, retiring the number is not the same thing as fixing whatever it was measuring. We may not want to know the answer, so we just won’t capture the information.
Then there's the prediction-market story, which shows how people sometimes do want to know the information, only to take advantage of it for themselves.
The following is from Wall Street Journal reporting about White House insiders that seem to be profiting privately from information they have: [8]
Alarm bells went off inside the White House Counsel’s Office earlier this year when anonymous Polymarket accounts first bet five-figure sums that President Trump and Iran would agree to an initial ceasefire by the end of April.
Senior officials privately raised concerns that the users may have leveraged inside information to secure an easy payday on the prediction market, according to people familiar with the matter. White House lawyers spoke with staff, asking who from the hundreds of aides working inside the building could be making bets.
These lawyers concluded it would be nearly impossible to determine whether administration officials were behind the bets because Polymarket allows users to open accounts anonymously, the people said, even though authorities have successfully traced the proceeds of those wagers. Instead, officials sent a warning shot: a late March memo telling staff they were prohibited from using nonpublic information for financial benefit. Three anonymous Polymarket accounts later made over $600,000 correctly betting on the April ceasefire announcement, according to blockchain research firm Bubblemaps.
…
The latest example centers on Trump’s teleprompter operator, Gabriel Perez. The White House confirmed this week that he is under investigation by the Commodity Futures Trading Commission for making over $100,000 in wagers based on early access to presidential speeches.
A casino has to prove its house edge is fair, disclosed, and regulated. That's the deal that makes gambling legal: everyone loses to the same odds.
What we're describing here isn't gambling. The way the Kalshi and Polymarket apps work (I’ve written about these in the past… [9]) leaves the environment way too open to insider betting – those with far more certain information can place a wager and profit handsomely at the expense of everyone else playing by the rules.
The Atlantic reporter McKay Coppins said it best this week: “This industry is built on insider trading, and anyone who tells you differently is lying." [10]
The throughline? In each case, someone with authority over the numbers - casualty counts, demographic filings, market-moving information - treated the data as a lever to pull rather than a truth to report.
And in each case, the people harmed are the ones who took the data at face value, because taking data at face value is supposed to be reasonable. That's the whole point of publishing a number in the first place.
But when data are used to decorate an already-determined narrative, it’s only best for the storyteller.
Notes and Sources
[1] Helene Cooper, Lara Jakes, and John Ismay, “Number of U.S. Troop Deaths Suddenly Appears Lower on Pentagon Website,” New York Times, July 23, 2026, https://www.nytimes.com/2026/07/23/us/politics/iran-troop-deaths.html
[2] Robert Mackey, “Pentagon list of service members killed in Iran war removes most recent four,” The Guardian, July 23, 2026, https://www.theguardian.com/us-news/2026/jul/23/pentagon-service-members-killed-iran
[3] David Vergun, “Nation's Leaders Honor the Fallen During Dignified Transfer,” U.S. Department of War, press release, July 22, 2026, https://www.war.gov/News/News-Stories/Article/Article/4552653/nations-leaders-honor-the-fallen-during-dignified-transfer/
[4] Kai Ryssdal [kairyssdal.bsky.social], Bluesky, July 23, 2026, https://bsky.app/profile/kairyssdal.bsky.social/post/3mrdwl7zyh22x
[5] Alexandra Olson and Claire Savage, “US civil rights agency moves to end demographic data collection after 60 years,” AP News, July 21, 2026, https://apnews.com/article/eeoc-discrimination-dei-andrea-lucas-2f521d784a9d0376416499066dc9f178
[6] Claire Savage, “Head of workplace rights agency urges white men to report discrimination,” AP News, December 19, 2025, https://apnews.com/article/dei-white-men-discrimination-andrea-lucas-eeoc-2996e71763dd0fe4b7f377eb49036fbe
[7] EEOC Chair Andrea Lucas [@andrealucasEEOC], X, December 17, 2025, https://x.com/andrealucasEEOC/status/2001439099907961012
[8] Brian Schwartz and Philip Wegmann, “The New Washington Temptation: Inside Information and a Prediction Market Account,” Wall Street Journal, July 18, 2026, https://www.wsj.com/politics/policy/dc-government-prediction-market-betting-8134c3fc
[9] Mic Farris, “Terrible Decisions,” Decisions, December 14, 2025, https://www.micfarris.com/articles/terrible-decisions
[10] McKay Coppins, On the Line with Alicia Menendez, MS NOW, July 20, 2026.
Decisions with Mic Farris
Seek Truth. Honor Differences.
