Worlds 2026 Semifinals Odds: 8 Teams, 4 Seats

Go looking for Worlds 2026 semifinals odds on Polymarket and you will not find them: the exchange that has taken 9.49 million dollars on this tournament has no semifinal market. It does not need one. Eight teams are left, each plays one best-of-five, and the winner is in the semifinals — so the semifinal price is the match price. We read all four boards off Polymarket's public Gamma API at 21:15 UTC on 2 November, took best bid and offer on every contract rather than the midpoint, pulled Kalshi's separate semifinalist market, and checked both against the bracket Riot drew after the Swiss stage closed on 31 October.

The short version: the derived board is the cleanest-priced thing at this tournament, the exchange that lists the question outright is the looser of the two, and the real mispricing is not in any contract — it is between the two halves of the bracket.

The board this one replaces, with ten contracts still fighting over four Swiss seats, is in Worlds 2026 quarterfinals odds, and the Swiss maths behind these eight teams is in Worlds 2026 format explained.

Worlds 2026 semifinals odds: all eight teams priced

Eight teams, four best-of-fives, four seats. Quarterfinals and semifinals run 3 to 8 November at the Credit Union of Texas Event Center in Allen, Texas, every match starting 16:00 CST (22:00 UTC); the final is 14 November at Barclays Center in Brooklyn. Every knockout series is best-of-five under Fearless Draft.

Team	Region	Swiss	Quarterfinal	Date	Polymarket	Kalshi
Gen.G	LCK	3-0	v LYON	4 Nov	91.50	91.50
T1	LCK	3-1	v Karmine Corp	6 Nov	81.50	82.50
Bilibili Gaming	LPL	3-2	v Dplus KIA	5 Nov	58.50	59.50
Invictus Gaming	LPL	3-1	v Top Esports	3 Nov	53.50	55.50
Top Esports	LPL	3-1	v Invictus Gaming	3 Nov	46.50	43.50
Dplus KIA	LCK	3-0	v Bilibili Gaming	5 Nov	41.50	41.50
Karmine Corp	LEC	3-2	v T1	6 Nov	18.50	18.00
LYON	LCS	3-2	v Gen.G	4 Nov	8.50	8.50
All eight					400.00	400.50

Four advance from eight, so the honest target is 400 points, not 100, and Polymarket hits it exactly. That is structural rather than skilful: each quarterfinal is one market with two outcomes, so the prices are complements and each pairing sums to 100.00 whatever anyone believes. The cost shows up at the offer, where all four boards are quoted exactly one point wide: buying all eight semifinalists at the ask costs 404.00 dollars to collect a certain 400, a 4-point overround or 1.00 percent.

The Swiss board this replaced carried 5.65 points across ten contracts and the title board carries 2.65 on one seat, so per seat this is the tightest complex at the tournament. Our no-vig calculator strips a margin like it out of any board.

Kalshi quotes the market Polymarket does not

Kalshi lists Worlds 2026 Semifinalists outright — eight separate Yes/No contracts, one per team, resolving on whether that team qualifies. Nothing forces them to agree, which makes the board a test rather than an identity.

Team	Polymarket mid	Kalshi mid	Kalshi spread	Difference
Gen.G	91.50	91.50	1.0	0.00
Dplus KIA	41.50	41.50	3.0	0.00
LYON	8.50	8.50	1.0	0.00
Karmine Corp	18.50	18.00	6.0	−0.50
T1	81.50	82.50	3.0	+1.00
Bilibili Gaming	58.50	59.50	1.0	+1.00
Invictus Gaming	53.50	55.50	1.0	+2.00
Top Esports	46.50	43.50	1.0	−3.00

Three of the eight agree to the basis point, five are inside a point, and the board sums to 400.50 against Polymarket's 400.00. Kalshi's independence shows in the pairings: Top Esports plus Invictus Gaming reads 99.00, Dplus KIA plus Bilibili Gaming 101.00, Karmine Corp plus T1 100.50, and only Gen.G plus LYON lands on 100.00. Each must be exactly 100, because one of the two plays on, so Kalshi carries a point of contradiction on three of four series. Polymarket cannot make that error, and tells you nothing by not making it.

The one disagreement worth money is Top Esports. Kalshi has Invictus Gaming 12 points clear of its league-mate; Polymarket has it 7 clear — a 5-point gap on the margin of the only quarterfinal between teams with identical 3-1 records, and the series carrying the most money on either exchange. Kalshi's board has traded 54,866 contracts with 29,848 still open, 12,179 in the last 24 hours: real, but a ninth of the Polymarket series.

What going 3-0 was actually worth

Riot's draw rule is narrow: the two 3-0 teams go on opposite sides of the bracket and each faces a 3-2 team, and everyone else is paired at random. Gen.G and Dplus KIA were the 3-0 teams, and the draw handed Gen.G the LCS runner-up and Dplus KIA the deepest 3-2 team in the field. The pools then price out almost on top of each other: 41.50 to 91.50 for the two 3-0 teams, 46.50 to 81.50 for the three 3-1 teams, 8.50 to 58.50 for the three 3-2 teams.

Dplus KIA is the only undefeated team quoted as an underdog in its own quarterfinal, and it is an underdog to a side that lost twice: Bilibili Gaming at 58.50 is priced above two of the three 3-1 teams. The Swiss record is a seeding device, not a rating, and this is the clearest demonstration the tournament has produced — the same achievement worth a 91.50 quarterfinal to one team and 41.50 to the other, a 50-point spread earned entirely at the draw.

It also explains the shape of everything below. Gen.G's half holds the LCS team and two LPL sides priced under 11 to reach the final; T1's half holds Bilibili Gaming and the other 3-0 team. One side is much heavier, and the market has not finished pricing it.

The two halves of the bracket are 5.20 points apart

The semifinal pairings are fixed even though the teams are not: the 7 November semifinal takes the winners of Top Esports v Invictus Gaming and Gen.G v LYON, and the 8 November semifinal takes the winners of Dplus KIA v Bilibili Gaming and Karmine Corp v T1. That gives the grand-final board a constraint as hard as the 400-point one: exactly one team comes out of each half, so each half must sum to 100.

Half	Teams and grand-final prices	Sum	Error
7 Nov semifinal	Gen.G 74.00, Top Esports 10.70, Invictus Gaming 9.60, LYON 3.15	97.45	−2.55
8 Nov semifinal	T1 47.50, Bilibili Gaming 32.50, Dplus KIA 19.00, Karmine Corp 3.65	102.65	+2.65
Whole board	all eight	200.10	+0.10

200.10 against a true 200 is the best-priced number at this tournament — a tenth of a point across eight contracts that are not linked, since unlike the quarterfinal boards these are independent Yes/No markets with nothing enforcing the total. And it hides the error completely: the halves are 5.20 apart, so the aggregate is right for the wrong reason. One side is 2.55 too cheap, the other 2.65 too rich, and they cancel.

Which side is wrong is a judgement; the direction is not. The 8 November half carries both remaining LCK sides plus Bilibili Gaming, and it is the half quoted above 100. Karmine Corp's 3.65 does most of the damage: 18.50 to beat T1 and 3.65 to reach Brooklyn implies a 19.73 percent semifinal win, against 44.60 to 56.78 for the three teams it could meet there.

Three boards, one team, and what dividing them says

Each of the eight is priced three times: to reach the semifinals (its quarterfinal board), to reach the grand final, and to win. Divide the second by the first and you have the market's view of a semifinal nobody has drawn yet; the third by the second gives its view of the final.

Team	Semifinal	Grand final	Title	Implied semifinal win	Implied final win
Gen.G	91.50	74.00	47.50	80.87	64.19
T1	81.50	47.50	20.50	58.28	43.16
Bilibili Gaming	58.50	32.50	15.25	55.56	46.92
Invictus Gaming	53.50	9.60	5.40	17.94	56.25
Top Esports	46.50	10.70	3.20	23.01	29.91
Dplus KIA	41.50	19.00	8.40	45.78	44.21
Karmine Corp	18.50	3.65	1.75	19.73	47.95
LYON	8.50	3.15	0.65	37.06	20.63

The implied semifinal column makes the bracket visible without anyone naming it. Invictus Gaming is 53.50 to win its quarterfinal and 17.94 to win its semifinal; Top Esports is 46.50 and 23.01. Whichever survives 3 November runs into Gen.G or LYON on 7 November, and Gen.G is 91.50 to be the one waiting. Gen.G's own 80.87 is the mirror image — the only team in the field whose semifinal is priced as a formality.

Two numbers sit awkwardly. LYON's implied 37.06 is the second-highest in the field on a team quoted 8.50 to get there at all — what dividing a 3.15 contract by an 8.50 one in half-point ticks produces, so noise rather than a view. Top Esports' implied 29.91 is the lowest final win rate of the eight, and that one is a real statement.

The best-of-five arithmetic: cubes, handicaps and sweeps

Each board also prices the individual games, giving a second route to the match price. If games were independent at probability p, a best-of-five win is p³(1 + 3q + 6q²). They are not independent — Fearless Draft guarantees it, since every series drains the champion pool — so the gap is a measurement, not an error.

Series	Per-game price	Best-of-five model	Quoted match price	Gap
Dplus KIA v Bilibili Gaming	DK 47.50	45.32	DK 41.50	−3.82
Gen.G v LYON	Gen.G 81.50	95.30	Gen.G 91.50	−3.80
Top Esports v Invictus Gaming	TES 48.50	47.19	TES 46.50	−0.69
Karmine Corp v T1	KC 30.50	16.98	KC 18.50	+1.52

Three of four quote below the independent-games model, the expected direction: compounding a per-game edge three times over-rewards the favourite, and real series regress. The exception is Karmine Corp, the biggest underdog, where the model's 16.98 sits below the quoted 18.50 — a premium for upset optionality independence will not generate.

The cleaner tell is in the game prices. On all four boards games 1, 2 and 3 are quoted flat to the half point. Game 4 is where conditioning appears, and on Gen.G against LYON it is 16 points lower — 65.50 against 81.50 — because a fourth game exists only in the branch where LYON has already won one. On Karmine Corp against T1 it moves further, 45.00 against 30.50, on a single dollar of volume. Game 4 is the only contract here that admits the games are not independent.

Series	−1.5 quoted	−1.5 model	Sweep quoted	Sweep model	Any sweep quoted	Any sweep model
Top Esports v IG	IG 33.50	33.53	IG 16.00	13.66	25.50	25.07
Gen.G v LYON	Gen.G 83.50	84.18	Gen.G 54.50	54.13	57.50	54.77
Karmine Corp v T1	T1 64.50	64.29	T1 30.50	33.57	39.50	36.41
DK v Bilibili Gaming	DK 25.00	27.60	DK 12.00	10.72	26.00	25.19

A minus 1.5 handicap pays only on a two-game margin — 3-0 or 3-1 — and a minus 2.5 only on a sweep. The minus 1.5 column matches its own per-game price to within 0.68 of a point on three of four boards, and to within 0.03 on Top Esports against Invictus Gaming. The any-sweep column is the inverse of the over 3.5 games total, and agrees to within 0.43 on that board and 0.81 on Dplus KIA against Bilibili Gaming. The handicaps and totals are coherent with the game prices in a way the match-winner contracts are not.

Where the money is, and the board that is two-thirds handicap

The four boards hold 625,161 dollars of traded volume and 1,384,305 dollars of resting liquidity, 453,411 of that volume inside the last 24 hours. The grand-final prop that depends on them has taken 159,659 dollars in its whole life. The series are the forecast; the props are the lag.

Series	Board volume	Resting liquidity	Match winner	Handicaps	Handicap share
Karmine Corp v T1	$192,237	$236,702	$65,804	$122,411	63.7%
Top Esports v Invictus Gaming	$179,949	$508,186	$148,983	$21,884	12.2%
Gen.G v LYON	$135,815	$320,205	$62,815	$37,314	27.5%
Dplus KIA v Bilibili Gaming	$117,160	$319,212	$88,308	$2,293	2.0%
All four	$625,161	$1,384,305	$365,910	$183,902	29.4%

The distribution is the interesting part. Karmine Corp against T1 is the most lopsided series at 18.50 against 81.50, and has taken the most money of the four — 63.7 percent of it on the handicaps rather than the winner. That is what a market does when the outcome is nearly settled and the margin is not: 73,536 dollars on whether T1 sweeps, 48,875 on whether it wins by two, against 65,804 on who wins at all. Dplus KIA against Bilibili Gaming is the mirror case at 2.0 percent, because when you do not know who wins there is no reason to bet how.

Top Esports against Invictus Gaming carries the heaviest book, 508,186 dollars resting, with 82.8 percent of its volume on the straight winner. Two teams on identical Swiss records priced 46.50 against 53.50 is the simplest question here, and the one the money likes most.

Every one of the four boards carries the wrong date

Check this before you trade any of them: all four quarterfinal markets state the wrong date in their own description, and the error is an exact reversal of the calendar. The boards are labelled Quarterfinal 1 to 4 and their descriptions give 3, 4, 5 and 6 November in that order. The bracket runs them backwards.

Board	Description says	Slug says	Machine field says	Actual date
Quarterfinal 1: Dplus KIA v Bilibili Gaming	3 November	2026-11-03	5 Nov 22:00 UTC	5 November
Quarterfinal 2: Karmine Corp v T1	4 November	2026-11-04	6 Nov 22:00 UTC	6 November
Quarterfinal 3: Gen.G v LYON	5 November	2026-11-05	4 Nov 22:00 UTC	4 November
Quarterfinal 4: Top Esports v Invictus Gaming	6 November	2026-11-06	3 Nov 22:00 UTC	3 November

The machine-readable start time is right on all four and agrees with Riot's bracket; the description and the slug are wrong on all four. Mostly that is cosmetic — but not for the postponement rule, which each market keys to its own stated date: a postponed series must restart within 14 calendar days of "the originally scheduled start", and a miss resolves 50-50 whether or not the match is later played.

So the backstops are misaligned both ways. Dplus KIA against Bilibili Gaming starts 5 November with a 17 November deadline — 12 days of cover, not 14. Top Esports against Invictus Gaming starts 3 November with a 20 November deadline, 17 days. Riot has not postponed a Worlds knockout series in the modern era, so the risk is small; but the contract you own is not the one the text describes.

The Western props and a 167-dollar market that is right

Two Western teams reached the quarterfinals for the first time since 2021, and they landed in opposite halves — Karmine Corp with T1, LYON with Gen.G — so both can reach the final. That makes the Western props checkable against the boards above.

Prop	Quoted	Implied by the boards	Gap	Volume
A Western team reaches the grand final	7.80	6.69	+1.11	$29,531
A Western team wins Worlds	2.85	2.34	+0.51	$13,468
A Western team wins a knockout game	81.00	83.38	−2.38	$0
Gen.G and T1 meet in the grand final	32.00	35.14	−3.14	$9,389
A first-time champion	22.00	20.31	+1.69	$1,232
T1 wins a fourth straight title	20.50	20.50	0.00	$167

Both traded Western props are rich by about a point, the ordinary shape of a long shot. The untraded one is more interesting: "a Western team wins a knockout game" has taken zero dollars and quotes 81.00 between a 72 bid and a 90 offer. Both Western teams fail only if T1 sweeps Karmine Corp and Gen.G sweeps LYON — sweeps the market prices itself at 30.50 and 54.50. Multiply them and nobody wins a game 16.62 percent of the time, so the prop is worth 83.38.

Gen.G and T1 meeting in the final is the clearest disagreement. Opposite halves means the events are independent, so the joint probability is the product of the grand-final prices: 74.00 times 47.50 is 35.15, or 35.14 de-vigged. The market quotes 32.00 with the offer at 33.00, more than two points below what the single-team contracts imply. The smallest market listed is the one with nothing wrong with it — T1's fourth straight title is the same event as T1 winning Worlds, quoted 20.50 on 167 dollars against 20.50 on 676,629. For the player version, see our Worlds 2026 MVP odds.

The region board agrees from the other end. Quoted, it sums to 99.70 while the title board sums to 102.65, so they look two points apart on Korea — 74.50 against an aggregated 76.40. De-vig each against its own total and the gap collapses: 74.72 against 74.43 on the LCK, every region inside 0.30 of a point, the LCS matching to 0.02. The league breakdown is in Worlds 2026 winning region odds.

How we would read the Worlds 2026 semifinals odds

Start by accepting that the 400.00 is free. Polymarket's semifinal board cannot be incoherent, so it earns no credit for being coherent, and the test that bites is the one it cannot pass by construction: the two halves of the grand-final board, 97.45 and 102.65 where both must be 100.

Then let the deep markets lead. Four series hold 625,161 dollars against 159,659 on the prop that depends on them, so when the two disagree the series is the forecast — the Western knockout-game contract being the live example at 83.38 implied against 81.00 quoted.

Treat the handicaps as primary and the match-winner contracts as derived, which inverts how they are usually presented — and watch game 4 rather than games 1 to 3. Last, keep the scales apart. These eight teams are a certainty to fill four semifinal seats, and the two Western sides are 2.85 percent combined to win the thing. Live numbers for every board sit on our Worlds 2026 hub, and the full title board is in Worlds 2026 winner odds.

RiftOdds earns a commission if you open a Polymarket account through our link. It costs you nothing and never changes the numbers above.

Related reading

• Worlds 2026 quarterfinals odds: 4 seats left
• Worlds 2026 grand finals odds: 16 teams, 2 slots
• Worlds 2026 winner odds: every team priced
• Worlds 2026 format explained: the Swiss stage maths
• T1 Worlds 2026 odds: the four-peat priced
• How to use Polymarket: bids, offers and spreads

Frequently asked questions

What are the Worlds 2026 semifinals odds right now?

Read at 21:15 UTC on 2 November 2026: Gen.G 91.50, T1 81.50, Bilibili Gaming 58.50, Invictus Gaming 53.50, Top Esports 46.50, Dplus KIA 41.50, Karmine Corp 18.50, LYON 8.50. Those are the four quarterfinal boards, because winning a quarterfinal is what reaching the semifinals means.

Does Polymarket have a Worlds 2026 semifinal market?

No. There is no team-to-reach-the-semifinals board, and the four best-of-five quarterfinal markets carry the same information. Kalshi does list the question directly, as Worlds 2026 Semifinalists, and its board sums to 400.50 against Polymarket's 400.00.

When are the Worlds 2026 semifinals?

7 and 8 November 2026 at 16:00 CST (22:00 UTC) in Allen, Texas. The 7th takes the winners of Top Esports v Invictus Gaming and Gen.G v LYON; the 8th takes the winners of Dplus KIA v Bilibili Gaming and Karmine Corp v T1.

Why do the eight semifinal prices add up to 400 percent?

Because four of the eight advance, so the honest target is 400 points rather than 100. Each quarterfinal is a single two-outcome market, so its prices are complements and the total is 400.00 by construction. The real margin is the one-point spread: buying all eight at the offer costs 404.00 to collect 400.

Who is favourite to win Worlds 2026 after the quarterfinal draw?

Gen.G, at 47.50 percent on a title board summing to 102.65, with T1 at 20.50 and Bilibili Gaming at 15.25. Gen.G is also the only team whose semifinal is priced as a formality: 74.00 divided by 91.50 implies it wins that semifinal 80.87 percent of the time.

Why is Dplus KIA an underdog after going 3-0 in the Swiss stage?

Because the draw decides the quarterfinal, not the record. Riot's rule puts the two 3-0 teams on opposite sides against 3-2 opponents; Dplus KIA drew Bilibili Gaming and Gen.G drew LYON. At 41.50 it is the only undefeated team priced as an underdog.

Sources. All prices, spreads, volumes and resting liquidity are best bid, best offer, book midpoints and traded volume read from Polymarket's public Gamma API at 21:15 UTC on 2 November 2026, and will have moved since. The four quarterfinal boards are Top Esports v Invictus Gaming, Gen.G v LYON, Dplus KIA v Bilibili Gaming and Karmine Corp v T1, alongside the grand-final, winner, winning-region, Western knockout game, Gen.G and T1, Western grand-final, first-time-champion and T1 four-peat props. Resolution rules, the 14-day postponement clause and the previous-winner list are quoted from those markets' own descriptions. The comparison board is Kalshi's Worlds 2026 Semifinalists, read from its public API at 21:20 UTC. The bracket, draw rule, Swiss pools, start times and venues come from Riot's LoL Esports and the 2026 World Championship record. All de-vigged figures, board and half sums, implied columns and models are our own calculations. 18+, please gamble responsibly.