Kalshi Reports $1.73 Billion in Esports Trading Volume: Four Quarters, Three Decelerating Rates, and a Metric Being Misread

Three in the morning in Binh Duong. The last CS2 match of the day ended hours ago, the casters are gone, and only the fan is still running. I stay behind with my data sheet — a six-year habit that started when I was thirteen, counting eleven set-piece goals by Becamex Binh Duong across fourteen V-League matches. Tonight I am counting something else: contracts.

Over the last four quarters, the prediction market Kalshi reported esports trading volume of $44 million, $173 million, $549 million and $967 million. Added together: $1,733 million. The announcement went out with a single headline — $1.73 billion in twelve months — and was immediately forwarded across forums as a health indicator for the entire esports industry.

I read those four lines ten times. The interesting part is the division nobody bothered to perform.

A contract exchange, not a sportsbook

Kalshi operates as an event-contract market licensed by the US Commodity Futures Trading Commission, under the designated contract market model. Users buy and sell binary contracts tied to the outcome of an event: which team wins, which title crosses a threshold, which election resolves. This mechanism differs from a traditional sportsbook in three structural ways, and all three directly shape how trading volume should be read.

First, price forms from an order book rather than from odds posted by a bookmaker. A bookmaker sets a price and carries the risk; an exchange only matches orders and collects fees. Second, a contract can be sold before the event resolves. A position bought at $0.40 and sold at $0.55 generates two volume entries, even though the total cash entering the system was recorded once. Third, an exchange does not need its own liquidity — it needs counterparties, and counterparties only appear when there is disagreement about probability.

Those three features combine into one consequence: trading volume on a contract exchange measures disagreement, not money wagered. A match everyone agrees on will produce near-zero volume. A match that splits the market will produce enormous volume, even if total cash in the system is a few tens of thousands of dollars.

This is the point most commentary on the $1.73 billion figure misses. They read it the way one reads a bookmaker's revenue, when it behaves far more like the turnover of a stock exchange. The same block of shares can trade ten times in a session and generate ten times its underlying value in volume.

Kalshi does not present the $1.73 billion as a claim about esports popularity. The announcement itself states plainly that volume counts contracts traded, and that trading in and out of a position raises volume without any additional cash entering. That is an important technical admission, buried in the middle of the piece.

For anyone in this trade, reading a volume table without reading the product structure behind it is exactly like reading possession stats without watching the tape. In 2026, in Kazan, Brazil held 57% of the ball and took 27 shots, five on target; Belgium took nine shots and scored twice. Read only the possession column and you reach the wrong conclusion about the match. The esports volume column sits in precisely that position.

Four quarters, three rates

Across four quarters of data, I found a growth formula being read backwards in the middle of a sales sheet.

Quarter-on-quarter growth: $44 million to $173 million is +293%. $173 million to $549 million is +217%. $549 million to $967 million is +76%. Three figures in sequence tracing a descending line, while absolute volume keeps climbing. This is the textbook shape of a growth curve entering a plateau phase — not of a market exploding.

A newly launched product always posts high percentage growth in its early quarters because the base is small. Doubling from $44 million is far easier than doubling from $549 million. The deceleration from 293% to 76% across three quarters is not necessarily a bad sign — it is the arithmetic consequence of scale. But it means one thing: anyone extrapolating the 293% line into next year is building a model on an ignition effect that has already burned out.

I learned this lesson during the pandemic. In 2026, when competitions were suspended, I collected data from 412 matches across four European leagues to compare home-win rates before and during the behind-closed-doors period. The rate fell from 46% to 39%. But what forced me to rewrite the entire analysis was something else: the decline was uneven over time. The first three rounds after stadiums emptied fell sharply; later rounds narrowed as players re-established habits. Averaging the whole period would have hidden the exact mechanism at work. The Kalshi volume table has the same problem: it must be read quarter by quarter, not as a twelve-month total.

And here the second division appears. Q3 2026 alone accounts for $967 million of the $1,733 million — meaning 55.8% of the entire twelve-month volume comes from a single quarter. The announcement is framed as an annual measure, but more than half its weight sits in the last three months. In other words: it was stale the moment it was published, because it describes an average state of a system that has passed through three very different phases.

The third division is even less common. Take Q3 2026 times four and the implied exit run-rate is roughly $3.87 billion a year — about 2.2 times the headline figure. That produces a notable paradox: the headline simultaneously overstates the past and understates the present. It overstates the past because the average drags three expired quarters along. It understates the present because it does not reflect the current rate. A metric that errs in both directions cannot serve as an input to any valuation decision without a technical footnote.

I tried rebuilding the series the way I build charts for athletics analysis: absolute line and percentage line on two separate axes. The absolute line rises almost vertically at the end. The percentage line descends steadily. They cross at exactly the point finance readers call the growth inflection. There is no mystery there — only a refusal to draw two axes.

What volume is, and is not

This is the most important part, and the most overlooked in every commentary I have read on the subject.

Trading volume is not revenue. Not gross profit for the operator. Not total money wagered. Not cash flowing into the esports ecosystem. It is the total value of matched contracts, and because contracts can be bought and sold repeatedly before resolution, it is a measure of turnover.

Picture a BO3 between two CS2 teams in a regional event. Suppose peak cash in the order book is $200,000. If price swings enough and enough participants enter and exit, recorded volume could reach $2 million — ten times over. Not one dollar of that $1.8 million gap exists as money wagered. It is the result of the same pool of capital changing hands repeatedly.

First consequence: trading volume cannot be used to infer the size of the esports betting market. Anyone performing that multiplication is comparing two different units. Traditional betting measures money on the table; a contract exchange measures how many times money crosses the table.

Second consequence: volume is highly sensitive to product design. Change contract duration, change fees, change settlement, add early-exit features — all of these push volume up without adding a single new user. An exchange can triple volume purely by making the product easier to churn.

Third consequence, and this is where I want to linger: volume measures disagreement, and disagreement can be manufactured. You cannot manufacture an audience's belief in a team. You can manufacture disagreement about a match by selecting the matches with the highest variance — short series, evenly matched teams, incomplete information.

That is precisely the choice the announcement describes: tracked markets are single-match markets across four titles — CS2, League of Legends, Valorant and Dota 2. Not tournament-winner markets. Not long-horizon markets. Single matches.

That choice is not neutral. It selects the highest-variance, highest-churn, thinnest-information product in the available spectrum. A tournament-winner market holds positions for weeks; a single-match market closes in two hours. Multiply by thousands of matches a year and you have a volume machine.

Put plainly: if anyone uses $1.73 billion to declare that esports is more watched than ever, they are using a metric designed to count attention at the highest frequency, across the highest-variance subset of matches, to infer a conclusion about all attention. That is a methodological error, not a data error.

Who generates volatility? Title structure

The four titles in the dataset are not equivalent in their capacity to generate tradable disagreement. And the announcement does not disaggregate them. This is the single largest analytical hole in the entire document.

Patch cadence determines uncertainty supply. Titles with fast, high-magnitude patch cycles generate more pre-match information asymmetry — roster news, last-minute meta shifts, rotation — and therefore more speculative volume per match. Titles with infrequent but large updates generate concentrated uncertainty windows around major patches. CS2 sits in the first group. Dota 2 patches are irregular but structurally deep. League of Legends runs a very steady two-week cadence, producing a continuous stream of uncertainty. Valorant sits in between.

Series length interacts with market microstructure even more strongly. A BO1 carries far higher variance than a BO5, because fewer decisive moments means luck weighs more. Variance is the raw material of a speculative product. A market tracking only single matches is quietly weighting itself toward short series, short-format events, and matchups too close to call.

There is a subtle consequence here that I have not seen raised. If volume concentrates in the highest-variance matches, then volume rises when predictive quality falls. A tournament with clear outcomes and dominant teams produces little volume. A chaotic tournament with parity and roster churn produces a lot. The metric can therefore move inversely to competitive quality. A great year for esports as a sport can be a low-volume year.

And without a per-title breakdown, the reader cannot know whether $1.73 billion is a CS2 phenomenon, a League phenomenon, or a diversified basket. That is the first question any sponsor would ask, and the only one current data cannot answer.

The event-spike model and the Q4 test

The announcement advances a structural thesis: strong underlying growth, plus sharp spikes around major events. The only supporting evidence is a soccer comparison — volume surging around the World Cup — transferred directly to esports.

The transfer is intuitively plausible but unverified with esports event-level data. There is no series showing how much volume rises during The International final week, or Valorant Champions week, or the Worlds group stage. All we have is an assertion.

If the spike model holds, then Q3 2026's $967 million is not a new baseline. It is a peak quarter containing the Dota 2 season climax, the Valorant Champions window and the CS2 circuit simultaneously, with League's Worlds build-up at the tail. Four event streams compressed into three months.

That makes Q4 2026 a falsifiable test, and it is the single most interesting thing to watch in this whole story. If Q4 falls back toward the Q1–Q2 trend line, the event-spike model is confirmed: growth came from the calendar, not from popularity. If Q4 holds near $967 million, structural growth is confirmed. Two outcomes, two entirely different worlds for the industry.

The author of the announcement places himself in exactly that position, committing to watch the next month. That is an honesty rare in this genre — an admission that the conclusion is pending.

There is a small detail I read over and over because it reveals more than what is written: volume "climbed almost every month" since tracking began. That "almost" entails at least one down month. The piece does not identify which, nor explain why. In a series this cyclical, the first down month is the most valuable data point in the whole set — it tells you where the break point sits. Skipping it is a systematic omission.

The geography of a metric

People blame speculative money, but I see a product structure bleeding at the second tier of the tournament system.

There is one sentence in the announcement I consider more important than the headline figure, and it is buried fairly deep: this data is for a largely US audience. One sentence, five seconds to read, and it changes how everything else should be understood.

Kalshi is a US-licensed exchange serving users within US legal reach. Users in mainland China cannot participate. Users in Korea are constrained by a tight licensing framework. Most of Southeast Asia sits outside the perimeter. So the $1.73 billion does not measure global esports attention. It measures the speculative attention America permits access to.

When the Chinese and Korean audiences vanish from the data sheet, the truth surfaces: this metric never measured global attention.

The first consequence is systematic title bias. League of Legends and Dota 2 draw most of their audience from regions a US-facing prediction market cannot reach. CS2 and Valorant have relatively stronger North American participation. A US-centric volume series therefore likely over-weights CS2 and Valorant relative to their global audience share, and under-weights League and Dota 2. That is the exact inverse of the global viewership ranking.

The second consequence is the absence of mobile titles. Honor of Kings, Peace Elite and an entire generation of Asian mobile titles are outside the tracked set. The most plausible explanation is not lack of interest — those are the largest mobile esports markets on earth. It is regulatory exclusion. A US-facing platform cannot serve the largest mobile esports user base, so that base does not exist in the data.

The third consequence is liquidity fragmentation across venues. The announcement notes Polymarket has long been home to most League of Legends trading. Polymarket is crypto-native and globally oriented, while the $1.73 billion comes from Kalshi. Two venues split the market, and the price for the same match can diverge between them. Neither has a clean view of total attention.

And here is where the story touches Vietnam, in a way people in the industry should notice.

Vietnamese audiences watch CS2, League of Legends and Valorant at among the highest intensities in the region. Domestic leagues hold stable live audiences. But Vietnamese viewers do not appear in the $1.73 billion series, because they cannot legally access a US event-contract exchange, and because most domestic matches run in Southeast Asian time slots — hours a North American-facing platform naturally under-weights.

The practical consequence is very concrete: if a tournament organiser in Vietnam wants to use trading-volume data to price a sponsorship package, they are using a metric that cannot see their own audience. And if an international sponsor uses that metric to compare regional value, they will systematically underprice Southeast Asia — not because the audience is small, but because the instrument does not reach it.

The integrity surface: the player as a node

The announcement is about the market, but the market's underlying asset is individual and team performance. The gap between those two layers creates risks the piece never touches.

Single-match markets create per-match, per-player financial incentive structures. Historically, the second and third tiers of esports are where match-fixing enforcement concentrates — precisely because liquidity is thin and incomes are low. Thin liquidity means a small stake moves the price. Low income means the financial incentive is large relative to earnings. Together those are the textbook conditions for integrity risk.

A contract exchange does not create match-fixing. But it does create a liquidity channel for anyone holding inside information. And single-match markets have a feature that distinguishes them from tournament-winner markets: match-level inside information has immediate value and a very short shelf life.

Roster changes, stand-ins, a player withdrawing at short notice — these are the highest-value non-public information in all of esports, and they are typically announced close to match time. Anyone holding that information before announcement has a direct edge on a single-match market. The announcement does not address any information-handling rules.

There is a governance gap worth naming. Publishers like Riot Games and Valve hold near-absolute integrity jurisdiction within their ecosystems — they ban players, suspend teams, prosecute fixing. But they have no clear jurisdiction over third-party trading venues. The exchange sits outside the publisher. The publisher does not control the exchange. Risk accumulates in the space between.

At the same time, there is a countervailing commercial incentive nobody wants to state: low-tier matches are where mispricing is widest, and therefore where exchange margins are highest. The incentive to list low-tier matches is directly proportional to integrity risk. That is a real trade-off, and the announcement does not surface it.

The money does not go anywhere

This is the most important correction in this entire analysis, and I want to state it plainly.

None of the $1.73 billion flows into the esports ecosystem.

Clubs receive no share of trading volume. The announcement identifies no integrity fee, no data-licensing payment to clubs, no revenue-share mechanism. The only stated commercial linkage is that Kalshi supplies data for organisers, publishers and sponsors to sit alongside viewership data. That is a data-supply relationship, not a revenue relationship.

The difference between those two is larger than it looks. In a revenue relationship, money flows from one party to another and can be split. In a data-supply relationship, what is exchanged is access to information, and the value is created on the user's side — the side selling rights, selling sponsorship slots, selling inventory.

The real value is optionality, not cash. The plausible monetisation path is indirect: if trading volume can validate attention, and validated attention can support sponsorship pricing, media-rights negotiation and slot valuation, then the benefit accrues to whoever sells attention — organisers, publishers, and indirectly clubs. But that is an inferred thesis, not a demonstrated mechanism. And it only holds if the metric actually measures what it claims to measure — which the preceding sections have put in doubt.

For practitioners in Vietnam, this deserves emphasis. A domestic organiser negotiating sponsorship may increasingly be handed a figure now common in international commercial decks: esports trading volume grew by such-and-such percent. The immediate questions should be: volume from which exchange, for which match set, for which viewing region, and in what unit. If the answer is single-match contract volume on a US exchange, then that figure says nothing about Vietnamese audiences — and nothing about anyone's revenue in this industry.

The contrarian read

The story currently being told is a demand story: esports is more popular, more people want to bet, and surging volume proves it.

I think the true story is the reverse. Three consecutive quarters of triple-digit growth came from the supply side, not the demand side. They came from a financial product becoming more legally accessible, more churn-friendly by design, and easier to understand at the interface. The same pool of esports fans, with the same intensity of passion, can generate ten times the volume if the product allows in-play trading rather than pre-match betting only.

If that holds, the Kalshi volume curve measures the quality of Kalshi's product more than the health of esports. And it carries an uncomfortable corollary: when other exchanges launch equivalent products, Kalshi's volume can fall without anything bad happening to esports. Conversely, when a rival launches, volume can spike without a single new fan.

I want to push this one step further, because it is where few dare go. The esports industry has spent a decade learning to price attention with ever more sophisticated metrics: concurrent viewers, average watch time, engagement rates, equivalent advertising value. Each new metric promised greater precision than the last. Prediction-market volume is now being offered as the next one — an indicator supposedly harder to fake because it is tied to real money.

But a metric tied to real money is only hard to fake when that money is constrained. A metric counting contract turnover is very easy to inflate, in exactly the way stock-exchange volume can be inflated by high-frequency trading without a single new investor. We are preparing to replace one inflatable metric with another, except the new one is harder for outsiders to verify.

And there is a larger blind spot on the club side. If trading volume becomes an input to sponsorship pricing, clubs have an incentive to make their matches harder to predict — rotating rosters, withholding information, manufacturing uncertainty. Those behaviours help volume and hurt competitive quality. The industry has no mechanism to recognise that trade-off, because it has never had to face it.

One more thing unsettles me about how this story is being told. Four titles, four entirely different competitive structures, aggregated into one total. No breakdown. In every sport I have analysed, aggregating events with different statistical properties into a single total is the signature of a metric designed to impress rather than to understand. You do not add 100m sprint times to 200m breaststroke times and call it a delegation's performance. You separate them, because each event has a different distribution.

Same here. A CS2 market has fast patch cycles and a strong North American community. A Dota 2 market has a season concentrated around one event. A League market has historical liquidity on a different exchange. A Valorant market is mid-expansion. Add those four together and you get a total that describes none of them.

What to watch next

There is a very cheap, very clear test arriving in the next few months, and it will say more than any analysis.

If Q4 2026 prints somewhere between $400 million and $600 million, the event-spike model is confirmed and the entire structural-growth story needs rewriting. If it holds near $900 million or above, the growth thesis stands and the industry has reason to start treating this as a permanent data layer. Both outcomes are useful. The only useless outcome is continuing to circulate a twelve-month figure that was half-expired on publication day.

For practitioners here — in Binh Duong, in Ho Chi Minh City, in Hanoi — there is one concrete thing worth doing now: start measuring what we can measure ourselves. Concurrent viewers by time slot. Retention across tournament rounds. Sponsorship value per thousand viewing hours. Those metrics are less glamorous than a billion dollars, but they measure exactly what we are selling, and nobody can take them out of our hands.

Six years ago I sat counting eleven set-piece goals by a V-League side, and a young coach called to argue with me for an hour. He did not argue about the number. He argued about whether I had counted the right thing. That remains the most correct argument in this profession, and it is waiting here again, between a very large volume table and a very small question: what are we counting, and for whom?