{"id":323,"date":"2026-08-07T22:58:29","date_gmt":"2026-08-07T22:58:29","guid":{"rendered":"https:\/\/t1x2bettingpro.advision-dev.com\/?p=323"},"modified":"2026-08-08T14:41:48","modified_gmt":"2026-08-08T14:41:48","slug":"23-million-trades-prediction-market-pricing-study","status":"publish","type":"post","link":"https:\/\/1x2bettingpro.advision-dev.com\/index.php\/2026\/08\/07\/23-million-trades-prediction-market-pricing-study\/","title":{"rendered":"What 23 Million Trades Reveal About Sports Prediction-Market Prices"},"content":{"rendered":"\n

Prediction markets are often described as if their prices translate directly into probabilities.<\/p>\n\n\n\n

A contract trading at 60 cents appears to suggest a 60% chance that the event will occur. A 25-cent contract seems to imply 25%. That interpretation is intuitive, convenient and central to how many users compare prediction markets with traditional sportsbook odds.<\/p>\n\n\n\n

But a large 2026 study of sports trading on Kalshi suggests the relationship is more complicated.<\/p>\n\n\n\n

Researcher Niusha Moshrefi<\/strong> analyzed approximately 23 million sports moneyline trades<\/strong> and found evidence that prediction-market prices can be reasonably calibrated during parts of a contract\u2019s life while developing systematic distortions in other situations. The study also identified an important difference between individual event contracts and multi-event combinations: contracts resembling parlays were priced above the value implied by combining their underlying individual markets.<\/p>\n\n\n\n

The findings do not prove that prediction markets are fundamentally inefficient. They show why traders and bettors should be cautious about treating every quoted market price as a precise estimate of real-world probability.<\/p>\n\n\n\n

For a market-focused betting audience, that distinction matters. Price, probability, liquidity and transaction structure are related, but they are not interchangeable.<\/p>\n\n\n\n

The Study Analyzed Roughly 23 Million Sports Trades<\/h2>\n\n\n\n

The research was released on arXiv in July 2026 and focused on sports moneyline trading conducted through Kalshi.<\/p>\n\n\n\n

In the paper, Moshrefi examined approximately 23 million individual trades<\/strong>, creating a large dataset for studying how prices behaved as sporting events approached resolution. The analysis looked at whether market prices were calibrated to eventual outcomes and how those relationships changed throughout the lifecycle of a contract.<\/p>\n\n\n\n

The full research paper is available through the sports prediction-market pricing study<\/a>.<\/p>\n\n\n\n

Calibration is an important concept because a market price is useful as a probability forecast only if similar prices produce corresponding outcomes over large samples.<\/p>\n\n\n\n

For example, if contracts priced around 70 cents truly represent events with approximately a 70% chance of occurring, then events in that price range should eventually settle in the affirmative roughly seven times out of ten.<\/p>\n\n\n\n

Perfect calibration is unrealistic. Markets respond continuously to new information, participants disagree about probabilities and short-term prices can be affected by liquidity.<\/p>\n\n\n\n

The more relevant question is whether deviations are mostly random or whether identifiable patterns consistently push prices away from the probabilities suggested by the underlying information.<\/p>\n\n\n\n

Moshrefi\u2019s results point toward the second possibility in some areas of sports prediction trading.<\/p>\n\n\n\n

Market Prices Were Not Equally Reliable at Every Stage<\/h2>\n\n\n\n

One of the study\u2019s most useful findings is that calibration changed over the life of a contract.<\/p>\n\n\n\n

Prices could be relatively well calibrated during intermediate stages of trading. That supports the broader idea that an active market containing many participants can aggregate information reasonably effectively.<\/p>\n\n\n\n

But the relationship became less stable near settlement.<\/p>\n\n\n\n

That is significant because intuition might suggest the opposite. As a game approaches its conclusion, uncertainty usually decreases. A team leading comfortably late in the contest should have a higher probability of winning than it did before kickoff.<\/p>\n\n\n\n

More information, however, does not automatically guarantee more efficient pricing.<\/p>\n\n\n\n

Late in an event, market structure can change rapidly. Liquidity may become concentrated around extreme prices, fewer participants may be willing to take the opposite side, and small timing differences can matter more as the remaining uncertainty approaches zero.<\/p>\n\n\n\n

A contract trading near settlement is therefore operating under different conditions from the same contract hours earlier.<\/p>\n\n\n\n

The study provides a useful warning against viewing every point on a prediction-market chart as equally informative.<\/p>\n\n\n\n

A Market Price Is Not Automatically a True Probability<\/h2>\n\n\n\n

The attraction of prediction markets comes partly from their simplicity.<\/p>\n\n\n\n

<\/p>\n\n\n

\n
\"Why<\/figure>\n<\/div>\n\n\n

<\/p>\n\n\n\n

A yes contract priced at 65 cents can be mentally translated into a 65% probability. Unlike American sportsbook odds, users do not need to convert -150 or +175 into an implied percentage before comparing outcomes.<\/p>\n\n\n\n

That simplicity can create false precision.<\/p>\n\n\n\n

A quoted contract price is ultimately the price at which participants are currently willing to trade. It reflects supply, demand, information, liquidity, transaction costs and the structure of the market itself.<\/p>\n\n\n\n

Probability is an estimate of how likely the event is to occur.<\/p>\n\n\n\n

Those concepts can converge without being identical.<\/p>\n\n\n\n

Consider a hypothetical market in which a contract trades at 60 cents. If the contract is highly liquid, information is widely distributed and buyers and sellers are competing aggressively around the current price, 60 cents may provide a useful approximation of the market consensus.<\/p>\n\n\n\n

If the contract is thinly traded and only a small amount of money is available at 60 cents, the same displayed price may contain less information.<\/p>\n\n\n\n

The distinction resembles traditional sportsbook analysis. The posted line is not simply a mathematical forecast. It also reflects market-making decisions and the behavior of customers.<\/p>\n\n\n\n

Our guide to what makes a bet good value<\/a> explains the same underlying principle: the relevant comparison is between the probability implied by a price and a defensible estimate of the event\u2019s actual probability.<\/p>\n\n\n\n

Prediction markets change the format of the price. They do not eliminate the need for that comparison.<\/p>\n\n\n\n

Liquidity Influences What a Price Can Tell You<\/h2>\n\n\n\n

A displayed price becomes more informative when substantial trading can occur around it.<\/p>\n\n\n\n

Imagine two sports contracts both showing 55 cents.<\/p>\n\n\n\n

The first has thousands of participants, substantial orders on both sides and a narrow gap between available buy and sell prices.<\/p>\n\n\n\n

The second has little activity and only a small quantity available near 55 cents.<\/p>\n\n\n\n

It would be difficult to argue that the two prices contain exactly the same amount of information.<\/p>\n\n\n\n

Liquidity affects how easily new information becomes incorporated into a market. In a deep market, participants who believe the price is wrong can trade meaningful amounts against it. Their activity helps push the price toward a new equilibrium.<\/p>\n\n\n\n

In a shallow market, informed participants may not be able to deploy enough capital to correct the price without moving it dramatically.<\/p>\n\n\n

\n
\"Liquidity<\/figure>\n<\/div>\n\n\n

<\/p>\n\n\n\n

This is especially relevant to sports prediction platforms because liquidity can vary considerably between events. A nationally televised NFL game may attract far more activity than a lower-profile contest.<\/p>\n\n\n\n

A market can therefore be structurally identical while producing different levels of price efficiency depending on participation.<\/p>\n\n\n\n

The Parlay Finding Is Particularly Important<\/h2>\n\n\n\n

The study\u2019s analysis of multi-event contracts may have the most direct relevance for users accustomed to sportsbook parlays.<\/p>\n\n\n\n

Moshrefi found that combinations involving multiple sporting outcomes were systematically priced above the value derived from combining their individual component markets.<\/p>\n\n\n\n

The difference also tended to increase as more legs were added.<\/p>\n\n\n\n

That is an important result because a multi-event contract can be valued mathematically from its component probabilities, assuming the events are sufficiently independent.<\/p>\n\n\n\n

Consider a simplified example.<\/p>\n\n\n\n

Suppose two independent outcomes are each estimated at 50%.<\/p>\n\n\n\n

The probability that both occur is:<\/p>\n\n\n\n

0.50 \u00d7 0.50 = 0.25<\/strong><\/p>\n\n\n\n

A fairly valued contract on both outcomes occurring would therefore correspond to approximately 25% before considering transaction costs, market structure and uncertainty in the underlying estimates.<\/p>\n\n\n\n

Adding another independent 50% event reduces the combined probability again:<\/p>\n\n\n\n

0.50 \u00d7 0.50 \u00d7 0.50 = 0.125<\/strong><\/p>\n\n\n\n

That produces 12.5%.<\/p>\n\n\n\n

The probability falls rapidly as additional requirements are added.<\/p>\n\n\n\n

If the market price of the combined contract consistently exceeds the probability implied by the component markets, buyers of that combination are paying a premium.<\/p>\n\n\n\n

The study indicates that this effect was systematic rather than appearing only in isolated examples.<\/p>\n\n\n\n

More Legs Can Magnify Pricing Differences<\/h2>\n\n\n\n

Multi-leg products are attractive because they transform several modest probabilities into one larger potential payoff.<\/p>\n\n\n\n

That structure is familiar from traditional sportsbooks.<\/p>\n\n\n\n

The difficulty is that every added leg creates another condition that must be satisfied. Even when individual selections appear likely, the probability that all of them occur can decline quickly.<\/p>\n\n\n\n

Prediction-market interfaces can make this less intuitive because contracts may be displayed as simple dollar prices rather than traditional parlay odds.<\/p>\n\n\n\n

A trader might see several plausible events grouped together and focus on the individual likelihood of each one instead of the combined probability of all outcomes occurring.<\/p>\n\n\n\n

The 23-million-trade study suggests that this behavioral dynamic may have measurable pricing consequences.<\/p>\n\n\n\n

<\/p>\n\n\n\n

Market Structure<\/th>Main Pricing Question<\/th><\/tr>
Single sports contract<\/td>Does the price reasonably reflect the probability of one outcome?<\/td><\/tr>
Thinly traded contract<\/td>Is there enough liquidity for price discovery?<\/td><\/tr>
Near-settlement contract<\/td>Is the displayed price still well calibrated?<\/td><\/tr>
Two-event combination<\/td>Does the combined price match component probabilities?<\/td><\/tr>
Large multi-event contract<\/td>How much additional premium appears as more legs are added?<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

The important point is not that every combination is mispriced.<\/p>\n\n\n\n

It is that increasing complexity should produce more scrutiny, not less.<\/p>\n\n\n\n

Sportsbooks and Prediction Markets Express Prices Differently<\/h2>\n\n\n\n

Traditional sportsbooks and prediction exchanges do not build markets in exactly the same way.<\/p>\n\n\n\n

A sportsbook generally posts prices and accepts wagers against its own market. The operator can adjust odds in response to information, betting activity and liability.<\/p>\n\n\n\n

An exchange-style prediction platform matches participants taking opposing positions in contracts, although the precise mechanics, fees and liquidity structure depend on the platform.<\/p>\n\n\n\n

Sportsbook odds also contain margin.<\/p>\n\n\n\n

If a two-sided market lists both outcomes at -110, converting each price into implied probability produces approximately 52.38% per side. Together, those implied probabilities total more than 100%.<\/p>\n\n\n\n

That difference represents the sportsbook\u2019s built-in margin before other considerations.<\/p>\n\n\n\n

Prediction-market users sometimes assume that an exchange structure eliminates equivalent pricing friction.<\/p>\n\n\n\n

The new research suggests a more cautious interpretation.<\/p>\n\n\n\n

Even where a traditional sportsbook margin does not appear in the same format, pricing inefficiencies can emerge through liquidity conditions, transaction structure, trader behavior or the valuation of complex contracts.<\/p>\n\n\n\n

Changing the market mechanism changes where pricing friction may appear. It does not necessarily eliminate it.<\/p>\n\n\n\n

Near-Settlement Prices Deserve Special Attention<\/h2>\n\n\n\n

The study\u2019s calibration findings also have implications for live sports trading.<\/p>\n\n\n\n

During an NFL, NBA, MLB or soccer game, probabilities can move quickly. A touchdown, goal, red card, home run or turnover can dramatically change the expected result within seconds.<\/p>\n\n\n\n

Prediction markets respond by repricing contracts.<\/p>\n\n\n\n

As the event approaches settlement, prices can move toward zero or one dollar. At that point, remaining upside becomes small while the consequences of an unexpected reversal remain significant.<\/p>\n\n\n\n

A contract priced at 95 cents can appear almost certain.<\/p>\n\n\n\n

But the difference between 95% and 98% probability is economically meaningful even though both numbers feel psychologically close to certainty.<\/p>\n\n\n\n

Someone paying 95 cents for a contract that is actually worth 90 cents is not making a minor error simply because both estimates indicate the same team is very likely to win.<\/p>\n\n\n\n

The event can be correctly predicted while the price is still poor.<\/p>\n\n\n\n

This is one of the central lessons of value-based betting analysis: identifying the likely winner and identifying a favorable price are separate tasks.<\/p>\n\n\n\n

Favorite-Longshot Bias May Have a New Form<\/h2>\n\n\n\n

Traditional betting markets have long been studied for favorite-longshot effects, where customers may systematically value low-probability, high-payout outcomes differently from favorites.<\/p>\n\n\n\n

Prediction markets can display related behavioral patterns through a different interface.<\/p>\n\n\n\n

Instead of choosing a +600 underdog, a user may buy a low-priced yes contract. Instead of backing a large favorite at negative American odds, the same user may pay 80 or 90 cents for a high-probability outcome.<\/p>\n\n\n\n

The psychology remains similar.<\/p>\n\n\n\n

Small contract prices can make unlikely outcomes appear inexpensive. High contract prices can make likely outcomes appear safe.<\/p>\n\n\n\n

Neither interpretation is sufficient without comparing price and probability.<\/p>\n\n\n\n

The study\u2019s evidence of systematic discrepancies in multi-event markets reinforces the possibility that format does not remove behavioral bias. It may simply change how that bias appears.<\/p>\n\n\n\n

Market Efficiency Does Not Require Perfect Prices<\/h2>\n\n\n\n

The study should not be interpreted as proving that prediction markets fail.<\/p>\n\n\n\n

A market can be useful and reasonably efficient while still containing systematic biases.<\/p>\n\n\n\n

Traditional financial markets, sportsbooks and betting exchanges all contain periods when prices respond imperfectly to new information. Efficient-market concepts generally describe how difficult persistent mispricing is to exploit after accounting for costs and competition.<\/p>\n\n\n\n

That distinction matters here.<\/p>\n\n\n\n

Discovering that certain contracts have historically been overpriced does not guarantee that a trader can profit from the pattern going forward. Other participants can learn from published research. Platforms can change their products. Liquidity can improve. Fees can change the economics of taking the opposite position.<\/p>\n\n\n\n

Market behavior itself evolves once participants recognize a pattern. (Prediction markets take center stage in latest round of quarterly earnings reports<\/a>)<\/p>\n\n\n\n

This is why the study is more valuable as evidence about how prediction-market pricing works than as a betting system.<\/p>\n\n\n\n

The Research Has Important Limits<\/h2>\n\n\n\n

The findings also need to be interpreted within the scope of the dataset.<\/p>\n\n\n\n

This is research based on Kalshi sports trading rather than every sportsbook, exchange or prediction platform.<\/p>\n\n\n\n

Different platforms may have different participants, liquidity levels, fee structures and contract rules. A pricing pattern detected on one platform cannot automatically be assumed to exist at exactly the same magnitude elsewhere.<\/p>\n\n\n\n

Sports themselves also differ.<\/p>\n\n\n\n

A highly liquid NFL moneyline behaves differently from a small market on a less-followed event. Pregame trading differs from live trading. A single-outcome contract differs substantially from a multi-event combination.<\/p>\n\n\n\n

The paper is also a research preprint distributed through arXiv. Its methodology and results can be examined directly, but readers should distinguish a large empirical dataset from a universal law governing every future prediction market.<\/p>\n\n\n\n

The strongest interpretation is narrow and useful: across approximately 23 million analyzed sports trades, the researcher found identifiable patterns showing that quoted prices did not behave as perfectly calibrated probabilities under every market condition.<\/p>\n\n\n\n

Prediction Markets Still Offer Valuable Information<\/h2>\n\n\n\n

None of these findings make prediction-market prices irrelevant.<\/p>\n\n\n\n

Quite the opposite.<\/p>\n\n\n\n

A liquid contract provides another observable market estimate that can be compared with sportsbook moneylines, betting exchanges, statistical models and other probability forecasts.<\/p>\n\n\n\n

Disagreement between markets can itself be informative.<\/p>\n\n\n\n

Suppose a sportsbook price implies roughly 58%, a prediction market trades near 63% and an independent model estimates 60%.<\/p>\n\n\n\n

<\/p>\n\n\n

\n
\"Near<\/figure>\n<\/div>\n\n\n

The relevant analytical question is not which number is automatically correct.<\/p>\n\n\n\n

It is why the estimates differ.<\/p>\n\n\n\n

The prediction market may be reacting faster to new information. The sportsbook may be protecting against liability. The statistical model may not yet include an injury update. One market may simply be temporarily inefficient.<\/p>\n\n\n\n

Market comparison becomes more useful when prices are treated as evidence rather than truth.<\/p>\n\n\n\n

Twenty-Three Million Trades Strengthen a Simple Betting Principle<\/h2>\n\n\n\n

The most important lesson from the study is not specific to Kalshi.<\/p>\n\n\n\n

A price is an offer, not a promise of probability.<\/strong><\/p>\n\n\n\n

Prediction markets make probabilities easier to visualize because dollar-denominated contract prices resemble percentages. That convenience can obscure the fact that markets remain shaped by liquidity, participant behavior, transaction mechanics and product design.<\/p>\n\n\n\n

Moshrefi\u2019s analysis of approximately 23 million sports moneyline trades found that calibration was not constant throughout the life of a contract. It also found systematic overpricing in multi-event combinations relative to the values implied by their individual legs, with the discrepancy increasing as combinations became larger. (The vibes are young male vibes’: Why prediction markets attract a certain type<\/a>)<\/p>\n\n\n\n

Those findings provide a useful counterweight to the idea that prediction markets automatically reveal the \u201ctrue probability\u201d of a sporting event.<\/p>\n\n\n\n

They may provide an informative consensus estimate. Sometimes that estimate may be exceptionally strong.<\/p>\n\n\n\n

But a 60-cent contract is still something being bought and sold at 60 cents.<\/p>\n\n\n\n

Whether it is actually worth 60 cents remains the analytical question.<\/p>\n\n\n\n

<\/p>\n","protected":false},"excerpt":{"rendered":"

Prediction markets are often described as if their prices translate directly into probabilities. A contract trading at 60 cents appears to suggest a 60% chance that the event will occur. A 25-cent contract seems to imply 25%. That interpretation is intuitive, convenient and central to how many users compare prediction markets with traditional sportsbook odds. […]<\/p>\n","protected":false},"author":3,"featured_media":324,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"om_disable_all_campaigns":false,"footnotes":""},"categories":[8],"tags":[],"class_list":["post-323","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-betting-markets-analysis"],"acf":[],"aioseo_notices":[],"_links":{"self":[{"href":"https:\/\/1x2bettingpro.advision-dev.com\/index.php\/wp-json\/wp\/v2\/posts\/323","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/1x2bettingpro.advision-dev.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/1x2bettingpro.advision-dev.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/1x2bettingpro.advision-dev.com\/index.php\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/1x2bettingpro.advision-dev.com\/index.php\/wp-json\/wp\/v2\/comments?post=323"}],"version-history":[{"count":1,"href":"https:\/\/1x2bettingpro.advision-dev.com\/index.php\/wp-json\/wp\/v2\/posts\/323\/revisions"}],"predecessor-version":[{"id":328,"href":"https:\/\/1x2bettingpro.advision-dev.com\/index.php\/wp-json\/wp\/v2\/posts\/323\/revisions\/328"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/1x2bettingpro.advision-dev.com\/index.php\/wp-json\/wp\/v2\/media\/324"}],"wp:attachment":[{"href":"https:\/\/1x2bettingpro.advision-dev.com\/index.php\/wp-json\/wp\/v2\/media?parent=323"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/1x2bettingpro.advision-dev.com\/index.php\/wp-json\/wp\/v2\/categories?post=323"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/1x2bettingpro.advision-dev.com\/index.php\/wp-json\/wp\/v2\/tags?post=323"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}