A friend once told me he had turned $500 into $1,800 on Polymarket over a long election night, and showed me the screenshots to prove it. Six weeks later he was down to $200. He had not done anything stupid in particular. He had just kept trading after the easy money was gone, on markets where he had no edge, against people who did. That is the shape of most retail prediction-market careers, and it is the honest place to start.
The pitch is seductive. Real-time odds, public information, a market that pays you for being right about the news. If you read the news closely already, surely you can monetise the reading. The problem is that the people on the other side of your trades have also read the news, and many of them have built spreadsheets, models, and Discord servers around it. Making money is possible. Making it consistently is hard. This guide walks through the realistic edges, the realistic frictions, and the realistic returns, so you can decide whether the time is worth it.
The first honest question: are you trading or punting?
There is a useful distinction the professional traders draw, and it is worth borrowing on day one. Punting is taking a position because you have a view on the outcome. Trading is taking a position because you believe the price is wrong relative to the probability. They look identical on the order book. They produce wildly different long-term results.
Punting can be fun. It can also win for a while. But if you do not separate "I think Labour will win" from "I think 67 cents is too cheap for Labour to win," you will eventually drift toward the markets you find emotionally interesting rather than the markets where you have an edge. That is the road to my friend's $200 balance. The good news is that the same platforms support both behaviours; the bad news is they cannot tell you which one you are doing.
Where the edges actually live
The edges that work for retail traders cluster in a few specific places, and they are smaller than the YouTube thumbnails suggest.
The first is information speed. When a verdict drops, a result is called, a candidate concedes, or a central bank statement lands, prediction-market prices take seconds to minutes to fully adjust. If you are watching the source feed (the court livestream, the official scoreboard, the central bank press release) and not the news aggregator, you can occasionally trade ahead of the slow money. This edge is real but narrow. It rewards preparation, fast fingers, and being awake at unsociable hours.
The second is niche expertise. If you genuinely know more than the median trader about a specific corner of a specific sport, a specific regulatory process, or a specific subculture, the markets in those corners can be inefficient for long stretches. A reality-TV superfan trading reality-TV markets has a real shot. A cricket statistician trading cricket markets has a real shot. A generalist trading election markets against political professionals does not.
The third is price differences between venues. Polymarket and Kalshi sometimes price the same event differently. Our guide to Polymarket arbitrage walks through the mechanics; the short version is that the gaps are smaller, the friction is higher, and the windows are shorter than the textbook suggests, but the strategy is real.
What is not on this list: "trading the news." By the time you have read a headline and opened the app, the price has moved. Reading the news is the price of admission, not the edge.
The mechanics nobody walks you through
Knowing where edges live is half the job. The other half is the unglamorous mechanics of execution.
Start with the spread. On a thin market you might see Yes bid at 42 and offered at 47. If you buy at 47 and immediately sell, you crystallise a five-cent loss before anything has happened. That is a 10% drag on a coin-flip position. The spread is your first opponent, not the other traders. Liquidity matters more than people realise; our explainer on what liquidity is in prediction markets covers why some contracts will eat you alive on entry alone.
Then there are fees. Kalshi charges trading fees that scale with price and contract count; Polymarket charges on the crypto rails (gas, conversion, withdrawal). Neither is ruinous, but both compound. A 2% round-trip cost on a strategy that wins 53% of coin flips is the difference between profit and a slow bleed.
Resolution is the third trap. Every market resolves according to rules written before you arrived, and those rules are sometimes ambiguous in ways that matter. The Polymarket UMA dispute window, the Kalshi resolution sources, the precise wording of "by" versus "before" a date: any of these can flip a winning ticket into a losing one. Our piece on how prediction markets actually resolve is worth reading before you put real money down, not after.
And then taxes. Prediction-market winnings are taxable in both the UK and US, on different rules, and the platforms do not generally do the paperwork for you. The headline rate is not the point; the recordkeeping is.
What a realistic return actually looks like
The number that gets quoted on Twitter is the 10x night. The number that nobody quotes is the median outcome over a year.
A disciplined retail trader with a genuine edge in a specific niche, working perhaps five to ten hours a week on research and execution, might realistically aim for somewhere between break-even and a modest annual return on staked capital. That is not nothing. It is also not a salary. The traders making real money are running serious volume, often with custom tooling, sometimes with team structures, and almost always with edges that scale beyond the niche.
This matters because it sets the right comparison. If your bankroll is $20, even a brilliant year leaves you with $26. If your bankroll is $20,000, the same edge produces something more like a part-time income, but you have also taken twenty-thousand-dollar risk on platforms with regulatory and counterparty considerations. The edge does not change. The proposition does.
The people who lose money tend to lose it in two ways. Either they stake too large a fraction of their bankroll on a single trade and get unlucky, or they trade too often on too-thin edges and bleed out through spread and fees. Both are mathematical, not emotional, problems. Both have mathematical solutions: smaller position sizes, fewer trades, more patience.
Bankroll, position sizing, and the bit nobody enjoys
The single most boring piece of advice in this guide is also the most important. Stake small fractions of your bankroll on individual positions, especially early on. Two to five percent per trade is a reasonable upper bound when you are learning, and even professionals rarely go beyond ten percent on a single high-conviction position.
The reason is not that prediction markets are uniquely dangerous. The reason is that your estimate of your own edge is almost certainly too high in your first six months. You will see a 60-cent Yes, believe the true probability is 75%, and feel certain you have found alpha. Sometimes you will be right. Often you will be miscalibrated, and the only protection against miscalibration is position size. A blown-up account teaches you nothing because you cannot keep trading; a slowly drawn-down account teaches you everything because you can see which categories of trade are bleeding you.
Keep a journal. Date, market, entry price, your stated probability, position size, outcome, exit price. After fifty trades you will know whether you have an edge. Before fifty trades, you are guessing.
Where most people should actually start
If you are reading this and have not yet placed a trade, the cleanest first step is not to chase a strategy. It is to fund a small account, place a few exploratory trades on markets you genuinely understand, and watch how the order book moves. Our guide to placing your first prediction market trade covers the literal mechanics; this guide covers what you do after the first week.
For UK readers, the available platforms are narrower than the US offering, and the regulatory picture is its own thing; the legal alternatives to Polymarket for UK users is the right place to start there. For US readers, Kalshi's CFTC-regulated structure is the cleanest on-ramp, and Polymarket's regulated US access through QCEX is the second.
Whichever venue you choose, the discipline is the same. Stake small. Track everything. Trade only the markets where you can articulate, in one sentence, why the price is wrong. If you cannot finish that sentence, you are punting, and that is fine, but call it what it is.
Whatever you choose, this is where iPredicta is built to help. The platform at app.ipredicta.co tracks these contracts side by side across Polymarket, Kalshi, and the regulated UK venues, so the where-am-I-trading question is settled before you fund anything. It also publishes the analysis behind each market move, so you can see where a price is coming from before you decide whether to take the other side. That context is the whole point of the platform. It is also, in the end, the difference between a trade and a punt, which is the one distinction this guide keeps coming back to. If you would rather have that analysis land in your inbox than chase it down yourself, sign up to the iPredicta email list at ipredicta.co.
Frequently asked questions
Can you actually make a living from prediction markets?
A handful of people do, but it is rare and almost never starts that way. The traders earning a full income tend to run six-figure-plus bankrolls, have specialised edges (sports modelling, fast-news execution, cross-venue arbitrage), and treat it as a job with hours, journals, and risk limits. For most retail traders, a realistic ambition is to make the activity self-funding rather than salary-replacing. The honest framing is closer to a serious hobby with occasional good years than a career path. If you are choosing between learning to trade prediction markets and learning a marketable skill for a salary, the salary is almost always the higher-expected-value choice.
Which platform is best for beginners?
For US-based beginners, Kalshi is usually the cleanest start because it is CFTC-regulated, takes US dollars directly, and handles tax reporting more straightforwardly than the crypto-native alternatives. For UK readers, the picture is more constrained, and Smarkets or Betfair Exchange are the workable on-ramps for the event-style markets that overlap with prediction-market territory. Our Polymarket vs Kalshi comparison goes into the trade-offs in detail. The general principle: start on the platform where the funding rails, the tax position, and the legal status are simplest for your jurisdiction, not the platform with the most markets.
How much money do I need to start?
You can start with as little as the platform's minimum deposit, which is typically a few dollars or pounds. Whether you should is a different question. Very small bankrolls struggle because fees and spreads consume a meaningful fraction of every trade, and because position sizing discipline is hard to apply when each position is already tiny. A reasonable learning bankroll sits in the low hundreds: large enough that two-percent position sizes are still meaningful, small enough that losing the lot is an education rather than a disaster. Treat the first bankroll as tuition. The goal of the first few months is to learn whether you have an edge, not to make money.
Is prediction market trading the same as gambling?
Legally it depends on jurisdiction; mechanically the distinction is real but blurred. Gambling typically describes a wager against a fixed-house counterparty, with the house's edge built into the odds. Prediction markets are peer-to-peer order books where prices come from other traders, and a skilled trader can theoretically extract value from less-skilled ones, similar to poker rather than roulette. That said, the practical experience for a casual user is closer to gambling than to investing, especially without a defined edge. Our explainer on prediction markets versus sports betting covers where the line actually sits and why it matters for tax and regulatory treatment.
What is the most common mistake new traders make?
Trading too often, on too many markets, with positions that are too large for their bankroll. The mistake feels different in the moment (a hot tip, a strong conviction, a chance to win back yesterday's loss) but the underlying pattern is always the same: the trader stops asking whether the price is wrong and starts asking whether they want the outcome to happen. The fix is structural, not motivational. Limit yourself to a small number of markets you can articulate an edge in. Cap position sizes mechanically. Keep a journal. The traders who survive their first year are not the ones with the best instincts; they are the ones who built rules and stuck to them when their instincts disagreed.