Key takeaways
- Pick'Em opens 10 October at 4pm PDT and the Play-In window locks on 15 October. Four windows in all, each locking before its stage starts, with scores finalised after the final on 14 November.
- Only one of the four windows has a liquid market behind it. The outright winner board has traded 416,301 dollars and the winning-region board 622,536. The two markets that match the Swiss window hold 181 dollars of liquidity between them.
- Gen.G are 32.50 percent to win it, which is 31.03 after you strip the overround. Then Hanwha Life and Bilibili Gaming at 15.75, T1 at 14.80, and a 25-point cliff down to Dplus KIA.
- Exactly two of sixteen Swiss teams finish 3-0 and exactly two finish 0-3. That is forced by the pairing rules, not a guess, so the base rate on each is 12.5 percent — and a coin-flip team hits 3-0 exactly 12.5 percent of the time.
- The title price is the wrong number for the Swiss picks and we can say by how much. Scaled to the field it implies Gen.G go 3-0 62 percent of the time, which needs an 82 percent per-game win rate. Nobody wins 82 percent of games at Worlds.
On this page
- The four Pick'Em windows and what each one asks
- The champion pick: every priced team on the board
- Two markets, one question, and they agree
- Polymarket opened the Swiss markets and nobody is trading them
- What a 3-0 and a 0-3 actually cost
- The Play-In pick, and the team that does not exist yet
- Crystal Ball: where the market cannot help you
- A week of drift, and whether to lock early
- How we would fill it in
- Related reading
- Frequently asked questions
Every page you will find about the Worlds 2026 Pick'Em gives you the same six dates and then stops. The best of them, Insider Gaming's, runs about 550 words, has no tables, and tells you to check Polymarket for the odds — without ever reading what Polymarket says.
The dates are the easy half. The hard half is which of your picks has a real number behind it and which is a guess wearing a percentage sign, and a window you miss cannot be reopened once its stage has started.
So we pulled all three relevant Polymarket boards through the Gamma API at 16:30 UTC on 8 October 2026 — the outright winner market, the winning-region market, and the two Swiss-record markets that opened this week — took best bid and best offer on every contract rather than the last trade, and checked the volume behind each. The short version: one window is nearly free, one is well priced, one has a market that looks perfect and is completely empty, and one no market will ever touch. Team-by-team pricing is in Worlds 2026 winner odds; the stage rules are in Worlds 2026 format explained.
The four Pick'Em windows and what each one asks
Riot splits the game into four prediction sets, each opening after the previous stage resolves and locking before the next one starts. The dates below come from Riot's own Worlds 2026 primer, and the right-hand columns are ours: the Polymarket market that covers the same question, and the money actually standing behind it.
| Window | Opens | Locks | What you pick | Matching market | Volume |
|---|---|---|---|---|---|
| Crystal Ball | 10 Oct, 4pm PDT | 15 Oct | Tournament-wide meta calls | None | n/a |
| Play-In | 10 Oct, 4pm PDT | 15 Oct | Which of 4 teams reaches Swiss | Winner board, partial | $61,399 |
| Swiss | 18 Oct | 23 Oct | Who advances, who goes 3-0 and 0-3 | Two dedicated markets | $0 |
| Knockout | 31 Oct | 3 Nov | Bracket through to the champion | Winner board, full | $416,301 |
Two of those rows are worth staring at. The Knockout window is the one with real money behind it, because "who wins Worlds" is the question prediction markets have always priced and 416,301 dollars of turnover has gone through 18 contracts to answer it. You can lift that pick almost directly.
The Swiss window is the opposite, and it is the one that will decide your score. It is the only window where you are asked to pick 3-0 and 0-3 teams, Polymarket has opened markets with exactly those names, and the volume column reads zero. We come back to that below, because the gap between "a market exists" and "a market has a price" is the single most useful thing on this page.
Note also that Crystal Ball and Play-In share a window, opening and locking together, so that is one sitting rather than two.
The champion pick: every priced team on the board
Eighteen of the nineteen Worlds teams carry their own contract on the outright winner market. Every figure below is a book mid, the midpoint of best bid and best offer, which is the only price you could actually transact around in both directions.
| Team | Region | Bid | Ask | Book mid | No-vig | Decimal |
|---|---|---|---|---|---|---|
| Gen.G | LCK | 0.32 | 0.33 | 32.50% | 31.03% | 3.22 |
| Hanwha Life Esports | LCK | 0.16 | 0.17 | 16.50% | 15.75% | 6.35 |
| Bilibili Gaming | LPL | 0.16 | 0.17 | 16.50% | 15.75% | 6.35 |
| T1 | LCK | 0.15 | 0.16 | 15.50% | 14.80% | 6.76 |
| Dplus KIA | LCK | 0.057 | 0.070 | 6.35% | 6.06% | 16.5 |
| G2 Esports | LEC | 0.052 | 0.053 | 5.25% | 5.01% | 20.0 |
| Anyone's Legend | LPL | 0.048 | 0.051 | 4.95% | 4.73% | 21.1 |
| Invictus Gaming | LPL | 0.024 | 0.025 | 2.45% | 2.34% | 42.7 |
| Top Esports | LPL | 0.013 | 0.014 | 1.35% | 1.29% | 77.5 |
| Team Liquid | LCS | 0.010 | 0.011 | 1.05% | 1.00% | 100 |
| Karmine Corp | LEC | 0.007 | 0.008 | 0.75% | 0.72% | 139 |
| LYON | LCS | 0.006 | 0.009 | 0.75% | 0.72% | 139 |
| Movistar KOI | LEC | 0.003 | 0.004 | 0.35% | 0.33% | 303 |
| Team Secret Whales | LCP | 0.001 | 0.002 | 0.15% | 0.14% | 714 |
| Cloud9 Kia | LCS | 0.001 | 0.002 | 0.15% | 0.14% | 714 |
| LOS | CBLOL | 0.000 | 0.002 | 0.10% | 0.10% | 1000 |
| CTBC Flying Oyster | LCP | 0.000 | 0.001 | 0.05% | 0.05% | 2000 |
| MVK Esports | LCP | 0.000 | 0.001 | 0.05% | 0.05% | 2000 |
| Book total | 104.75% | 100.00% | ||||
The 104.75 percent total is the overround. One team wins, so a perfect book sums to 100, and the 4.75 points of excess is the market's cushion spread across eighteen contracts. The no-vig column divides it out, and that is the column to use for a Pick'Em, because you are not paying a spread — you are choosing the most likely outcome.
The shape of the board is four teams and then a cliff. Gen.G, Hanwha Life, Bilibili Gaming and T1 hold 77.33 points of the no-vig 100 between them, and the fifth price, Dplus KIA at 6.06, is a fifth of Gen.G's. For the knockout window that is close to an instruction. The one absence is structural: Brazil has two slots and only LOS has a contract, because the second is still being played for.
Two markets, one question, and they agree
Before using the board you should check whether it is internally coherent, because a board that disagrees with itself is a board you cannot lean on. Polymarket runs a separate winning-region market with 622,536 dollars of volume, more than the team board. Each region contract and the basket of that region's teams describe the identical event, so they should cost the same.
| Region | Seeds priced | Team basket | Region mid | Raw gap | Basket, no-vig | Region, no-vig | Gap |
|---|---|---|---|---|---|---|---|
| LCK (Korea) | 4 of 4 | 70.85% | 67.50% | +3.35 | 67.64% | 67.74% | -0.10 |
| LPL (China) | 4 of 4 | 25.25% | 24.50% | +0.75 | 24.11% | 24.59% | -0.48 |
| LEC (Europe, EMEA) | 3 of 3 | 6.35% | 5.65% | +0.70 | 6.06% | 5.67% | +0.39 |
| LCS (North America) | 3 of 3 | 1.95% | 1.60% | +0.35 | 1.86% | 1.61% | +0.26 |
| LCP (Asia-Pacific) | 3 of 3 | 0.25% | 0.25% | 0.00 | 0.24% | 0.25% | -0.01 |
| CBLOL (Brazil) | 1 of 2 | 0.10% | 0.15% | -0.05 | 0.10% | 0.15% | -0.06 |
Read the raw columns and Korea looks 3.35 points mispriced, which is the kind of number that gets written up as an arbitrage. It is not one. The region book sums to 99.65 and the team book to 104.75, so the two are quoted on different scales, and comparing them before normalising compares a cushion to no cushion.
Normalise both and the disagreement disappears. Every region agrees to within half a point, Korea to within a tenth. Two separately traded books, built by different people betting different amounts, land on the same view of the field.
That is the licence to use these prices for your bracket. The alternative every other Pick'Em page offers is somebody's ranking, and a ranking cannot tell you that fifth place is worth a fifth of first. We took the region board apart on its own in Worlds 2026 winning region odds, where Korea was 68.50 a week ago.
Polymarket opened the Swiss markets and nobody is trading them
Here is the finding that changed how we would fill this in. Polymarket has two markets named, almost word for word, after the hardest Pick'Em question on the sheet: Worlds 2026: Teams to Go 3-0 in Swiss Stage and Worlds 2026: Teams to Go 0-3 in Swiss Stage. Nineteen contracts each, one per team, open and accepting orders.
| Market | Contracts | Quoted bid | Quoted ask | Spread | Book sum | Volume | Liquidity |
|---|---|---|---|---|---|---|---|
| Teams to go 3-0 | 19 | 0.02 | 0.98 | 96 pts | 950% | $0 | $181 |
| Teams to go 0-3 | 19 | 0.02 | 0.98 | 96 pts | 950% | $0 | $184 |
| Winner (for contrast) | 18 | varies | varies | 0.1-1.3 pts | 104.75% | $416,301 | $485,455 |
Every contract on both boards is quoted two cents bid against ninety-eight cents offered, which is not a price. It is the widest legal quote, posted so the market exists. Gen.G and CTBC Flying Oyster carry the identical 50 percent mid. The books sum to 950 percent on a question where the answer is exactly two teams out of sixteen, so they are nine hundred and fifty points of nothing.
This is worth knowing precisely because it looks so useful. A market titled after your exact question, listing your exact teams, showing a tidy 50 percent — and 181 dollars of liquidity behind the whole board. If those markets fill up before the Swiss window locks on 23 October they become the best input you will get. Today they tell you only which teams Polymarket thinks are at Worlds, which includes FURIA.
So the Swiss picks have to be built, and the next section builds them.
What a 3-0 and a 0-3 actually cost
Start with something the format gives you for free. Sixteen teams enter the Swiss stage, round one pairs them into eight matches, and from round two the pairings are drawn inside each record group. So eight teams reach 1-0, four of those reach 2-0, and two of those reach 3-0. Exactly two teams finish 3-0 and exactly two finish 0-3, every time. It is arithmetic, not a tendency, and Riot's quarter-final draw rule depends on it: the two 3-0 teams are each drawn against a team that qualified at 3-2.
Two slots across sixteen teams puts the base rate at 12.5 percent on each. That is the number to beat, and it is more useful than it looks, because of what the path to 3-0 is made of. Rounds one and two are best-of-one. The round-three match at 2-0 is a best-of-three, because it decides qualification. A 3-0 is therefore two coin-flip-format games and then a series.
| Per-game win rate | Wins the Bo3 | Goes 3-0 | Goes 0-3 |
|---|---|---|---|
| 50% | 50.0% | 12.5% | 12.5% |
| 55% | 57.5% | 17.4% | 8.6% |
| 60% | 64.8% | 23.3% | 5.6% |
| 65% | 71.8% | 30.3% | 3.5% |
| 70% | 78.4% | 38.4% | 1.9% |
| 75% | 84.4% | 47.5% | 1.0% |
| 80% | 89.6% | 57.3% | 0.4% |
The first row is the check that the model is sane: a team that wins half its games goes 3-0 exactly 12.5 percent of the time, which is the structural base rate to the decimal. Nothing was fitted to make that happen.
Now the obvious shortcut, and why it fails. Scale the no-vig title prices across the sixteen-team field and multiply by the two available slots, and you get Gen.G at 62 percent to go 3-0, Hanwha Life and Bilibili Gaming at 31.5. Run those back through the table and Gen.G's figure needs an 82 percent per-game win rate and Hanwha Life's needs 66. The first is not a thing that happens at Worlds.
The reason is that a title price already contains three best-of-five knockout wins, where favourites convert most reliably. The Swiss stage is where they convert least, because two of the three games on the 3-0 path are single games. The title board ranks the field well and sizes the Swiss picks badly, and it fails in a known direction: it overstates the top.
Our read, with the method stated so you can disagree with it: put Gen.G at 70 to 75 percent per game against this field, which is 38 to 48 percent to go 3-0, and the rest of the top four at 60 to 65, which is 23 to 30. Invert it for 0-3: the three cheapest contracts on the board, all at or under 0.10 percent, are the teams a 40 to 45 percent per-game rate puts near 20 percent to go winless.
The Play-In pick, and the team that does not exist yet
The Play-In window is the cheapest points on the sheet and the one most people will get wrong by overthinking. Four teams go to the Riot Games Arena in Los Angeles from 15 to 18 October, it is a double-elimination bracket in best-of-fives, and one team advances.
Three of the four are confirmed: Karmine Corp as the LEC third seed, Cloud9 Kia as the LCS third seed, and MVK Esports out of the LCP. The fourth is Brazil's second seed, and on the day Pick'Em opens it still does not exist — the CBLOL Split 2 Grand Final is on 10 October.
The winner board gives you the ranking directly. Karmine Corp are 0.72 percent no-vig to win the whole tournament, Cloud9 0.14 and MVK Esports 0.05. Those are title prices, not Play-In prices, but they are conditioned on surviving this bracket, so the ordering carries: Karmine Corp are priced at roughly five times Cloud9 and fourteen times MVK. We priced this bracket properly, with best-of-five maths, in Worlds 2026 play-in odds.
Pick on 10 October and you are picking before the fourth team exists. Karmine Corp is the market's answer and nothing arriving before the lock would change it.
Crystal Ball: where the market cannot help you
Crystal Ball is the set of tournament-wide calls — the kind of question that resolves on aggregate statistics rather than results — and it is scored after the final. It is also the one window where we have nothing to offer you, and it is worth saying so rather than inventing a method.
Polymarket lists one market in the neighbourhood, Most Kills at Worlds 2026, with 105 player contracts. It has traded 692 dollars in total and most of its contracts sit at a flat 0.10, which is a default and not an opinion. There is nothing to read there.
Riot has not published the 2026 scoring weights either. Several pages ranking for this keyword present a points table as fact; we could not find it on any Riot property, and the primer carrying the dates does not carry the categories. Read the scoring off lolesports.com once the window opens, not off an aggregator. Treat Crystal Ball as a free guess and spend your time on the Swiss sheet.
A week of drift, and whether to lock early
Each window stays open for several days, which makes "when to submit" a decision rather than a formality. The winner board's one-week price changes say how much waiting is actually worth.
| Team | Book mid now | One-week change | Direction |
|---|---|---|---|
| Gen.G | 32.50% | -3.00 pts | Shortening favourite, softening |
| Bilibili Gaming | 16.50% | -2.50 pts | Biggest faller in the top four |
| Dplus KIA | 6.35% | +2.05 pts | Biggest riser on the board |
| Hanwha Life Esports | 16.50% | +1.50 pts | Rising |
| T1 | 15.50% | +1.50 pts | Rising |
| Anyone's Legend | 4.95% | -1.25 pts | Falling |
| G2 Esports | 5.25% | 0.00 pts | Flat |
| LCK region contract | 67.50% | -1.00 pts | Korea softening as a bloc |
Korea's four teams have reshuffled by two to three points each inside one week while the region contract moved one, which tells you the market is arguing about which Korean team rather than about Korea. Dplus KIA gaining 2.05 points off a 6.35 base is a third of their price in seven days, the largest proportional move on the board.
The practical read: for the knockout window, which does not lock until 3 November, waiting is clearly right — the whole Swiss stage resolves first. For the Play-In and Crystal Ball window closing on 15 October there is nothing to wait for. For the Swiss window closing 23 October, wait, and check whether those two empty 3-0 and 0-3 boards have found any volume by then. A live price there beats everything on this page.
How we would fill it in
Knockout window: pick inside the top four and take Gen.G as champion. They are 31.03 percent no-vig against a field where nobody else clears 16, and two independent books agree on the shape. Their 3.00-point slide this week is worth noting, not acting on.
Play-In: Karmine Corp. Five times Cloud9's price, fourteen times MVK's, and the fourth team is not known when the window opens. This is the one pick where the market's answer and the obvious answer are the same.
Swiss 3-0: build it, do not lift it. Two slots, sixteen teams, 12.5 percent base rate, and the title board overstates the top of the field — Gen.G at an implied 82 percent per game is the tell. Two of the top four, sized off the per-game table rather than off the title price.
Swiss 0-3: this is where the points are. Two slots again, and the bottom of the board is far more separated than the top: three teams priced at or below 0.10 percent to win the tournament against four at 14.8 percent or better. Favourites are hard to rank, no-hopers are easy.
Crystal Ball: guess. No market, no published scoring weights, and the pages offering a points table could not show us where it came from.
One caveat on all of it. These are prices from 16:30 UTC on 8 October, and the Play-In result alone will move the Swiss picture. Re-read the board when each window opens rather than filling the sheet now, and remember a market price is an implied probability, not a forecast anyone owes you. 18+, and please gamble responsibly.
Related reading
- Worlds 2026 winner odds: every team priced
- Worlds 2026 winning region odds: Korea at 68.5%
- Worlds 2026 format explained: Swiss stage maths
- Worlds 2026 play-in odds: 4 teams, 1 slot
- Worlds 2026 qualified teams: all 19 slots tracked
- LoL Worlds 2026: location, dates and schedule
Frequently asked questions
When does the Worlds 2026 Pick'Em open and lock?
Crystal Ball and Play-In picks open on 10 October at 4pm PDT and lock on 15 October. Swiss picks open 18 October and lock 23 October. Knockout picks open 31 October and lock 3 November. Scores are finalised after the final on 14 November, with rewards targeted inside 48 hours.
Who is favourite to win Worlds 2026?
Gen.G, at a 32.50 percent book mid on Polymarket as of 16:30 UTC on 8 October 2026, or 31.03 percent once the 4.75-point overround is divided out. Hanwha Life and Bilibili Gaming follow at 16.50 each, then T1 at 15.50 — four teams holding 77 of the 100 no-vig points.
How many teams go 3-0 in the Worlds Swiss stage?
Exactly two, and exactly two go 0-3. Sixteen teams are paired inside their record group each round, so eight reach 1-0, four reach 2-0 and two reach 3-0. The base rate on either pick is therefore 12.5 percent per team, and Riot's quarter-final draw assumes it.
Can I use Polymarket odds to fill in my Pick'Em?
For the knockout bracket, yes: the winner market has traded 416,301 dollars across 18 contracts and a separate region market agrees with it to within half a point. For the Swiss records, no — Polymarket opened markets with exactly those names, but every contract is quoted 0.02 against 0.98 on zero volume.
Which teams are in the Worlds 2026 Play-In?
Karmine Corp, Cloud9 Kia and MVK Esports, plus Brazil's second seed, decided by the CBLOL Split 2 Grand Final on 10 October. It runs 15 to 18 October at the Riot Games Arena in Los Angeles, double elimination in best-of-fives, and one team advances.
What does the Worlds Pick'Em scoring look like in 2026?
Riot had not published the 2026 point weights at the time of writing. The official primer carries the window dates but not the categories or scoring, so several pages presenting a points table are presenting something we could not source. Check lolesports.com once a window opens.
Does a Pick'Em entry cost anything?
No. It is free and needs a Riot account with a summoner name. Some regions are excluded and reward eligibility varies, so check Riot's own note before relying on a prize. Nothing on this page involves staking money.
Sources. All prices are best bid, best offer and book mids pulled from Polymarket's Gamma API at 16:30 UTC on 8 October 2026 and will have moved since: the Worlds 2026 winner market, the winning-region market, and the two Swiss-record markets. Pick'Em windows, stage dates, venues and reward timing from Riot's Worlds 2026 primer; the tournament is run by Riot Games. Field size, slot allocation and Swiss pairing rules checked against the 2026 World Championship record. Best-of-three and 3-0 path probabilities are our own calculation from the stated per-game rate.