It was the fourth quarter of the Pinstripe Bowl at Yankee Stadium on Dec. 27 when Penn State defensive end Dani Dennis-Sutton darted between two hapless blockers and sacked Clemson quarterback Cade Klubnik. Or as the game announcer described it: “Dennis-Sutton wraps him up and turfs him!”
This was good news for Penn State fans, en route to a 22-10 win. It might have been even better for the traders on prediction market platform Kalshi who’d wagered that the announcer would say the word “turf.” At 3:18 p.m., right after Dennis-Sutton’s sack, the market’s odds on “Yes” surged from 25% all the way up to 99%, indicating a surefire win.
But Kalshi ultimately resolved the turf market to “No,” ruling that the announcer’s utterance didn’t count. Many traders complained, but the rules were the rules: According to Kalshi’s policy at the time for so-called mention markets such as this one, “tense inflections” on verbs didn’t count. (For the market to have paid out, the announcer would have had to use “turf” as a noun or say something like, “He sure did turf him.”)
There’s a term in the prediction market community for a bet that resolves unfairly on a technicality: a “rulescuck.” (We’ll tell you the full etymology when you’re older.) The turf rulescuck — one influential trader classified it in the subcategory of “verbcuck” — was just one of many disputes Kalshi users have debated on the platform’s Discord channel in recent months. Should homophones count? What about words spoken as part of proper nouns or URLs? How about mispronunciations, acronyms, abbreviations, compound words and foreign-language terms?
In early March, Kalshi tried to clear up the confusion by releasing a new set of rules for mention markets. The guidelines, which stretch to seven pages, settled some long-running disputes: Verbs that are spelled the same as plural nouns do count. (Too little, too late for the “turfs” gang.) Acronyms (“AI”) count except when uttered in their expanded form (“artificial intelligence”). Plurals count if the given word is singular, but singulars don’t count if the given word is plural. Numbers count only when spoken exactly, not as part of a larger number.
On the surface the rules document is a curious, even amusing, glimpse into a flavor of lawyerly hairsplitting. But it also reflects the broader challenge prediction market platforms face in their attempt to reduce reality to a series of yes-or-no bets: Kalshi and its rival Polymarket pitch themselves as sources of truth, but the truth is often hard if not impossible to establish. Many of the best-known disputes involve angels-on-a-pinhead-type debates: Did Cardi B “perform” at the Super Bowl? Did Ukrainian President Volodymyr Zelenskiy wear a suit? Did the US “invade” Venezuela?
The language disputes are even narrower, and Kalshi’s new rules are its latest attempt to set the terms. But linguists interviewed for this story say the English language isn’t easily wrangled to such ends. To them, Kalshi’s rules lacked consistency and foundation, all but guaranteeing more Discord debate.
Consistency vs. Ambiguity
Kalshi doesn’t attempt to explain or justify its language rules. Karlos Arregi, a linguistics professor at the University of Chicago, says they appear to be arbitrary rather than based on a particular theory or philosophy of language. “This looks like the kind of rules you’d have in a game like Scrabble,” he says. “It’s obvious to me these rules were not done by a linguist.”
Rivka Levitan, a professor of computer science and linguistics at Brooklyn College CUNY, describes them as “more legalistic than linguistic.” A good set of rules, she suggests, ought to prioritize logical consistency. Instead, Kalshi’s rules seem to draw lines for no clear reason. They allow for certain types of inflections (such as -s for plural) but not others (such as -ed, -ing, -er, or -est). If the “strike” word — that is, the one listed on the mention market — is “veteran,” then “veterans” counts. But if the strike word is “veterans,” then “veteran” doesn’t count. This is unambiguous, but it’s also illogical. Similarly random-seeming are rules that allow deviations from the strike word when it comes to meaning but not form: If traders bet on a football announcer mentioning “wind,” for instance, their bet still hits for “the clock winds down.” But if the strike word is “run,” then “ran” or “running” doesn’t count.
Then there’s the distinction between open and closed compound words. If the strike word is “fire,” then “fire station” counts, as does the hyphenate “fire-resistant,” but “firefighter” doesn’t. This doesn’t appear to be based on any linguistic theory, Levitan says, but rather on the convention. Kalshi relies on Merriam-Webster and the Oxford English Dictionary; if either spells the term as an open compound or hyphenate, then it counts.
Alejandro Lopez Lira, a professor of finance at the University of Florida who studies mention markets, suspects the rules’ guiding principle is the ease with which they might allow outcomes to be decided automatically. “They’re trying to make rules that computer code would be able to easily catch, because that would save a lot of cost,” he says. That would explain why homonyms (“bass guitar”; “bass fishing”) count but homophones (“write”; “right”) don’t. It also explains the preference for open compound words and hyphenates over closed compound words, says Gregory Guy, a professor of linguistics at New York University. The former are easier to detect automatically, since they’re separated by spaces or hyphens, whereas closed compounds would require parsing letter by letter.
Asked about the thinking behind the new rules, a Kalshi spokesperson said: “It is important that we are proactive about making sure that our rules are as clear and well-defined for our traders as they can be. Clear rules create both a meaningfully better user experience and produce better price discovery outcomes.” The spokesperson didn’t respond to questions about whether the rules were designed for ease of automatic parsing or whether the company had consulted any linguists.
No Perfect Rulebook
Despite Kalshi’s drive for clarity and specificity, traders are already pointing out that ambiguity remains. On one recent Discord thread, they debated the rule on portmanteaus, which aren’t allowed unless they contain the strike word “as a distinct, unaltered element.” According to Kalshi, if the strike word is “MAGA,” then “MAGA-nificence” — a neologism uttered by Jimmy Kimmel — counts. But this definition left traders wondering whether “Obamacare” is a compound word combining “Obama” and “care” or a portmanteau of “Obama” and “health care.” Or what about “Trumpification” — should it be spelled with a hyphen or not? Merriam-Webster and the OED are silent on this question.
When it comes to mispronunciations, the rules can be outright subjective. “Minor unintentional mispronunciations [of the strike word] COUNT if context makes intent clear,” the rules say. “Severe mispronunciations” that make the word “unrecognizable,” however, don’t. Left undefined is who determines intent, what constitutes “severe” and to whom the word is “unrecognizable.”
“That’s not going to get rid of any argument,” Levitan says. That’s especially true, she points out, when you factor in dialects and the accents of non-native English speakers. For example, according to Kalshi, “Mandani” counts for “Mamdani” — this one has caused problems before. But what about, say, “Ryan” for “Brian”? Or John Travolta’s infamous botching at the 2014 Oscars of Idina Menzel’s name as “Adele Dazeem”?
To be fair, the rules are likely designed with the average trader in mind, not linguists. And by this standard they arguably succeed. “In the context of this kind of market, being clear is perhaps more important than linguistic consistency,” says Ken Adams, a lawyer and author of A Manual of Style for Contract Drafting. His only linguistic quibble: The verb structures in the document itself are “a little erratic.” For instance, it uses “shall”and “will” interchangeably. “That suggests to me that whoever compiled the rules felt they should sound legal but was half-hearted about it,” Adams says.
How would the linguists themselves design the rules? Levitan says she would “allow for more inflections besides plural -s, such as -ed, -ing, -er, and -est,” wouldn’t distinguish between hyphens and spaces, and wouldn’t allow homonyms or homographs. Arregi says he’d eliminate the rule about mispronunciations, since it’s too subjective.
But, he points out, there’s no perfect rulebook: “I do think that this particular set of rules is arbitrary, but that’s just how rules in a game work.” And linguists probably wouldn’t be the best people to design them, Arregi adds, given their frequent disagreement over the basics: “The question of what a word is — it’s actually a very complicated question.”
