Can You Make Money on Prediction Markets?

Layerhub data shows 87.3% of wallets that traded on Polymarket did not profit. Only 12.7% made money. Of those, most made less than $100. Only 2,138 wallets cleared more than $1,000 in profit.

Those numbers cover the full history of the platform through late 2024, across hundreds of thousands of active wallets. If you are wondering whether prediction markets are a realistic path to profit, that is the answer. Not for most people.

The Data

The 12.7% who profited is a misleading number on its own. That bucket includes wallets that made $2 in profit after dozens of trades. It includes people who got lucky on a single market and never traded again. The median profitable wallet made somewhere in the range of $20 to $50 across its entire lifetime on the platform.

The distribution is highly skewed. A small number of sophisticated traders captured a disproportionate share of the total profits. Layerhub's breakdown shows that the 2,138 wallets clearing more than $1,000 in profit represented roughly 0.8% of all wallets that ever traded. That group almost certainly accounted for the majority of total winning volume.

This is not unique to Polymarket. It mirrors what happens in poker rooms, sports betting markets, and any other domain where informed participants trade against less informed ones. The money flows from the many to the few. Knowing which group you are in before you deposit is the most valuable research you can do.

One more thing the data does not show: survivorship bias in the profitable group. Some portion of those winners got lucky and would lose over a longer sample. Distinguishing genuine skill from a good run requires sample sizes most traders never accumulate.

Why Most People Lose

The first problem is adverse selection. On Polymarket, you are not playing against the house. You are trading against other participants, some of whom are professional researchers, data scientists, and former quants who spend hours building models for individual markets. When you buy a position, someone is selling it to you. That seller either knows something you do not, or is making the same mistake you are. Statistically, the people on the other side of your trade are more informed than you are.

The second problem is overconfidence. Every retail participant believes they are the smart money. The political bettor who "follows the news closely" thinks they have an edge over casual observers. They do not have an edge over the professionals who are also tracking the same news, plus polling data, plus historical base rates, plus model outputs. Overconfidence in your own research is the most common way to lose money on prediction markets.

The spread is the third problem, and it is underappreciated. Thin markets have wide bid-ask spreads. If you buy YES at 55 cents and the true probability is 50 cents, you are already behind before the event even happens. You need the market to move significantly in your direction just to break even. On low-volume markets, spreads of 5 to 10 percentage points are common. That is a significant headwind on any trade.

Emotional trading on political markets deserves its own mention. People bet on elections, policy outcomes, and political figures with their preferences rather than their analysis. Someone who strongly supports a candidate will consistently overestimate that candidate's chances. The market prices in actual probabilities. Bettors who bet their beliefs rather than the true odds are donating to better-calibrated participants.

Fees compound the problem for small accounts. A 2% fee on a $10 trade is 20 cents. If your edge on that trade is 3%, your expected profit is 30 cents. After fees, you are netting 10 cents on a trade that requires research, capital allocation, and correct judgment. The math only works at scale.

Compare this to poker. Somewhere between 70 and 80 percent of poker players lose money over the long run, depending on the rake structure and the player pool. Prediction markets are structurally similar: zero-sum (minus fees), where the winnings of the skilled come from the losses of the unskilled. The difference is that in poker, you can see your opponents and read their behavior. On Polymarket, you have no idea who is on the other side of your trade.

Who Actually Profits

The most-cited example in prediction market circles is a trader known as "Domer." His public wallet shows over $2.5 million in profit generated on roughly $311 million in volume. That is approximately 0.8% profit margin on volume, which sounds small but represents an extraordinary absolute return given the capital base.

Domer has described his approach publicly. He is a former poker player who calls prediction markets "poker in slow motion." The key phrase is worth sitting with. In poker, you make hundreds of decisions per hour under time pressure. In prediction markets, you can take days to research a single position, model the probabilities carefully, and size your bet according to your confidence. The slower pace rewards different skills than casino gambling.

The skills he describes as central to his edge: deep research on specific domains, strict bankroll management, and the emotional discipline to hold positions through short-term adverse movement without panic-selling. These are exactly the skills that separate long-run winning poker players and sports bettors from losing ones. They are not common.

Research means more than reading the news. It means understanding base rates (how often does this type of event resolve the way you think?), identifying where the market consensus is likely wrong, and being honest about the limits of your own knowledge. Most retail traders skip this step entirely and trade on vibes.

Bankroll management means never putting a significant percentage of your capital on a single trade, sizing positions according to your actual edge rather than your confidence level, and having rules that protect you from ruin. Most traders set no position limits and go large on markets where they feel certain. Certainty is usually a signal to bet less, not more.

Emotional discipline means sitting on a losing position that your model says is still correct, without selling just to stop the psychological discomfort. It also means taking a profit when your target probability is reached, not holding because you "have a feeling" the trade will go further. Most people cannot do this consistently. They exit winners early and hold losers too long, exactly the opposite of what works.

The people who profit on prediction markets share a profile that overlaps almost entirely with the profile of a winning poker player or sports bettor. That is not a coincidence. The underlying skill set is the same.

How Returns Compare to Other Crypto Betting

The key distinction between prediction markets and casino games is that positive expected value is theoretically possible on a prediction market if you are skilled enough. At a casino, it is not. The house edge is fixed, built into the math, and no strategy overcomes it over time. At Polymarket, if your probability estimates are consistently better than the market's, you will profit.

That distinction matters less than most people think, for one reason: most participants are not skilled enough. The effective edge against the average Polymarket retail trader is not zero. Based on the Layerhub data, if 87.3% of wallets lost money, the aggregate loss rate across those wallets represents the average cost of participating. That implied average edge against a losing player is somewhere around 3% per dollar wagered, after fees, based on the distribution of outcomes.

There is something to be said for the 1% house edge you can verify versus an unknown edge against anonymous counterparties. A provably fair crash game at Stake tells you exactly what you are paying. The game publishes its algorithm, its seed chain, and the formula. You can calculate your expected loss per dollar with precision. On Polymarket, you cannot know your true edge or lack of edge without an honest, long-run accounting of your own trades against calibrated probabilities.

This is not an argument that casino games are better. They are not, if you have genuine prediction skill. It is an argument for being clear about what you are actually doing. Calling yourself a "prediction market trader" because you bought some political contracts does not make you Domer any more than sitting at a poker table makes you a professional player.

For the full comparison of expected returns across formats, see the prediction markets vs crypto casinos breakdown. The house edge calculator lets you model the long-run cost of any fixed-edge game for your specific bet size and volume.

Practical Advice for Beginners

Start with $50 to $200. This is not a minimum deposit recommendation. It is a ceiling for your first 30 to 60 days. The purpose is to learn the mechanics, track your decision-making process, and find out honestly whether you have an edge, without meaningful financial damage if the answer is no.

Stick to markets where you have genuine domain knowledge. If you work in healthcare, trade FDA approval markets. If you follow a specific sport obsessively, consider sports outcome markets. Do not trade on political markets because you "follow politics." Everyone follows politics. The edge in those markets is held by people with formal forecasting training and access to aggregated data sources, not by people with strong opinions.

Check the spread before buying anything. On the Polymarket interface, the difference between the displayed price and your actual fill price is your immediate loss the moment you enter the position. On thin markets, that spread can be 5 cents or more. A 5-cent spread on a contract you expect to move 8 cents in your direction means your risk-adjusted profit is less than you think.

Do not trade on emotion. This sounds obvious and almost nobody does it. If you notice yourself wanting to buy a position because the news makes you feel certain, that feeling is a warning sign rather than a trading signal. Certainty is the emotional state most correlated with overconfidence. Step back, model the probability independently, and compare it to the current market price. If your number does not differ from the market by more than the spread plus fees, there is no trade.

Set a loss limit before you start and treat it as binding. Decide in advance the maximum you will lose before stopping for the month. Write it down. The people who blow up their accounts on prediction markets almost universally did not do this step.

Track every trade. Record the market, the price you paid, your stated probability at the time, the resolution, and your profit or loss. After 50 trades, you will have enough data to ask whether your probability estimates are actually calibrated. Most people who do this honestly discover they are worse than they thought.

Read the how prediction markets work guide before depositing. Understanding how market making, liquidity, and resolution work will prevent a category of beginner mistakes that have nothing to do with your predictive skill.

If what you actually want is low-edge entertainment without the research burden, that is a legitimate preference. A 1% house edge crash game at a provably fair casino gives you a clear, known, verifiable cost to play. The fastest withdrawal casinos list is useful if fast access to winnings matters to you. Neither option gives you a path to Domer-level returns, but neither requires you to be a professional forecaster to understand what you are paying.

The honest summary: prediction markets offer the possibility of positive returns for skilled participants. The data shows that fewer than 1% of wallets reach meaningful profitability. The skills required overlap substantially with professional poker and sports betting. Most people do not have them, and most people are not honest with themselves about whether they do.

FAQ

Can you make consistent money on Polymarket?

A small percentage of traders do. The Layerhub data puts it at 0.8% of wallets clearing more than $1,000 in profit across the platform's history. Consistent profitability requires research depth, bankroll discipline, and emotional control that most participants do not apply systematically.

Is Polymarket legal in the United States?

Polymarket is geo-blocked for US users following a 2022 CFTC settlement. US-based traders who use VPNs to access the platform take on regulatory risk. The legal status of prediction markets in the US remains unsettled, and that situation may change.

How do prediction market fees work?

Polymarket charges approximately 2% on trades. This fee is embedded in the spread on most markets. Thin markets with low liquidity have additional implicit costs from wide spreads, meaning your effective cost per trade can be significantly higher than 2%.

What is the difference between prediction markets and sports betting?

The mechanics are similar: you are buying contracts that pay out if an event occurs. The structural difference is that sports betting takes place against a bookmaker with a fixed margin, while prediction markets match you against other traders. Both formats favor skilled participants over casual ones, but the path to edge is different. Sports bettors model outcomes relative to bookmaker lines. Prediction market traders model outcomes relative to crowd probability estimates.

Are prediction market winnings taxable?

Yes. In the US, prediction market winnings are treated as capital gains or ordinary income depending on how long you hold the position and how the IRS characterizes the activity. The same principles that apply to crypto gambling taxes apply here, though the specific treatment of prediction market contracts is still evolving in IRS guidance. Consult a tax professional if your volume is significant.

Should I start with prediction markets or crypto casino games?

That depends on what you want. If you have genuine domain knowledge and want to apply analytical skill, prediction markets give you a path to positive EV that casino games do not. If you want entertainment with a known, fixed cost and no research requirement, a provably fair casino with a verified house edge is more transparent about what you are paying. The mistake is treating prediction markets as entertainment, since the cost of playing without skill is higher than most casino games.

Last updated: March 2026. Profitability data from Layerhub analysis. Individual results vary.