Man analyzed 1,950 episodes of Wheel of Fortune and found one 'divorce predictor' hiding in plain sight
Joey spent six years watching 'nearly 2,000 episodes' of the game show and examined episodes airing between 2010 and 2019
No one enters a marriage expecting it to end in divorce. Although there is no foolproof way to predict whether a couple will stay together, researchers and relationship experts have spent years studying the behaviors and patterns that may signal trouble ahead. For instance, a content creator, Joey (@joeytoks), watched thousands of episodes of "Wheel of Fortune" and developed a mathematical model that can act as a brilliant predictor of divorce. It took him at least six years to build this spectacular version of sartorial astrology, according to his April 2026 TikTok video, which gained over 394,300 views and 63,000 likes.
@joeytoks This took me 6 years to make btw #wheeloffortune #datascience ♬ original sound - Joey Toks
Mathematics of divorce
Joey spent six years watching “nearly 2,000 episodes” of the television game show and examined episodes airing between 2010 and 2019. He discovered that the male game players who didn’t use “complementary adjectives,” like "beautiful" or "wonderful," to describe their wives during their introduction were more likely, precisely three times more likely, to get divorced within five years than their “wife-praising counterparts.”
In the video, he shared the screengrab of a spreadsheet where he had outlined all the statistics. In one column, he wrote the total number of episodes: 1,950. The total number of male contestants was 2,855, while the total number of male contestants who mentioned their wives was 2,016, approximately 70.61%. The number of contestants who used a complimentary adjective was 1,660 (82.34%), and those who didn’t use a complimentary adjective were 356 (17.66%). Of those who used one, 91 (5.48%) were likely to get divorced within five years, and of those who didn’t, 55 (15.45%) were likely to separate within five years.
Another segment of the spreadsheet recorded the “big winners” who won $40,000 or more on the show. Of 185 (86.45%) of these big winners who used complementary adjectives, 13 (7.03%) are likely to get divorced within five years, and out of 29 (13.55%) who didn’t use complementary adjectives, 5 (17.24%) are likely to get divorced within five years. In a comment, he described that “long-running syndicated game shows are the best for data mining because they air hundreds of episodes a year, which leads to thousands total, so large sample sizes.”
Why does it matter?
While his project may seem like an idle task conducted for no particular purpose, psychologist John Gottman from the University of Washington, who has been an expert in predicting the survival rate and divorce in marriages, said, “Without the ability to predict divorce, we’ll never understand why marriages fail or work well.” He noted that although most studies have been unsuccessful in predicting divorces, 50 percent of all U.S. marriages end in divorce and men’s facial expressions and how they introduce their wives in conversations can predict it. Gottman could predict whether a couple would divorce with an average of over 90% accuracy, his website says.
What do viewers think about it?
No one knows what inspired him to venture into this highly mathematical project, but nearly people watched his video, and thousands of viewers were impressed by the statistics. @ghabee said, “Well, now I'm pissed off at my husband because he's never been on the show, so how am I supposed to find out how he'd introduce me?” @belladotcircle remarked that “this is investigative journalism.” @iancborsuk professed, “Society should be compensating you for this work!”
You can follow Joey (@joeytalks) on TikTok to watch more videos on his channel.
More on Scoop Upworthy
Therapist's viral video reveals how to predict if your partner will cheat on you
Share on Facebook
