Built from what collector users captured whenever they opened the Champion Duel screens in game (July–August 2026, 30 events across Seasons 1–6). For qualifiers we only used group captures we could place relative to the closing time; for semifinals, the 15 events where all 128 semifinalists were captured; for matches, the 1,621 that had a decided winner. Season-by-season schedules are in the Champion War archive.
Champion Duel is the most personal event in Last War. Your alliance can’t fight for you and neither can your server — it’s your squad against someone else’s, one on one. Maybe that’s why so many people go in hoping that a lucky run in the knockouts will carry them. I did, my first time.
The data says otherwise. Most of the outcome is already set by qualifier score, and everything after that mostly confirms the qualifier order.
The format in one paragraph
Each event has 16 servers. Qualifiers split players into 16 groups of 100, and the top 8 of each group (128 players) go to the semifinals. Semifinals are 16 groups of 8, the top 2 of each (32 players) go to the knockout tournament, and the tournament is one-on-one elimination. All 30 events we collected used exactly these numbers.
The 8th-place cut is two-thirds of 1st — and below it, it’s packed
Score distribution of the 285 qualifier groups captured within 12 hours of closing:
| Season | Groups | 1st place (median) | 8th place | 8th ÷ 1st | 9th ÷ 8th | 100th ÷ 8th |
|---|---|---|---|---|---|---|
| Season 2 | 42 | 33.3M | 21.9M | 0.66 | 0.993 | 0.88 |
| Season 3 | 58 | 38.5M | 24.8M | 0.64 | 0.989 | 0.87 |
| Season 4 | 20 | 40.8M | 26.6M | 0.65 | 0.990 | 0.87 |
| Season 5 | 38 | 41.4M | 27.5M | 0.64 | 0.981 | 0.84 |
| Season 6 | 127 | 44.5M | 28.6M | 0.64 | 0.990 | 0.87 |
The 8th-place cutoff rises with season (21.9M → 28.6M), but as a fraction of 1st place it barely moves: 0.64–0.66. Whoever tops your group effectively sets the bar for everyone else in it.
The more striking part is below 8th. 9th place sits at 98–99% of 8th, and even 100th is at 84–88%. The gap from 1st to 8th is wide; the gap from 8th to 100th is tiny. Dozens of people are crammed within a few hundred thousand points of the cutoff, which is why the last day or two decides who gets through.
Rankings flip until the last half-day
How well does a mid-qualifier rank predict who actually advances? We split captures by how long before the close they were taken (12 events, 287 group captures).
From a rank taken 2–12 hours before the close, 8th place advanced 88% of the time and 9th only 10%. But from a rank taken more than 12 hours before, 8th advanced 40% and 9th 31% — basically a coin toss. Even 6th and 7th were only 67–77%. Ranks 1–4 advanced 97%+ no matter when you looked.
Short version: top 4 at the midpoint is all but locked, and 5th–12th gets decided on the final day. If someone in your alliance is hovering around 8th, tell them to save their remaining attempts for the end.
Seeds take more than half the seats
The game marks some qualifier players as seeds. Groups usually had 8 seeds (median), and of the 8 advancing spots seeds took a median of 5. Seeds advanced 63.5% of the time; non-seeds 3.4%.
Put differently, a non-seed is really fighting for about three spots per group, and those only open up when three of the eight seeds fall short. When the group list comes out, count the seeds first. If you’re not one of them, your real target is the weakest three or four seeds.
Three servers take half the seats
In the 15 events with a full 128-player semifinal roster, we looked at how the 16 servers split the 128 spots. The fair share would be 8 each. It’s nowhere near even:
- The single best server: median 18% (about 23 spots)
- Top three servers combined: median 48%
- Bottom eight servers combined: median 12%
- Events with at least one server sending nobody: 14 of 15 (up to 9 servers in one event)
The top 32 looks the same — across 27 events, the top three servers took a median of 18 of 32 knockout spots. Server age within an event didn’t matter: sort an event’s 16 servers by number into quarters and the average semifinalists per server were 8.4, 6.9, 8.6 and 8.0. Servers in the same event are similar in age anyway; what separates them is how many high-power accounts each one has.
Semifinals belong to the qualifier group winners
We took everyone who reached the semifinals and split them by their qualifier group rank.
Qualifier group winners made the top 32 100% of the time. 2nd place made it 74% of the time, 3rd 22%, and 4th and below 2% or less. With 2 of 8 advancing from each semifinal group the naive odds are 25%, but in practice the order set in qualifiers survived almost intact: a group winner takes one spot and the 2nd-place finishers fight over the other.
In knockouts, the higher qualifier score wins
In 122 knockout matches we could compare both players’ qualifier scores. The higher qualifier score won 89%. With a 5–30% score gap it was 94–98%; only when the scores were within 5% did it drop to 76%. In semifinal group matches (303) the favourite won 70% — lower, probably because some of those matches don’t matter once the group’s advancers are already clear.
The matches themselves are lopsided too: 79% of the 1,621 decided matches were 3–0 sweeps. Comebacks are rare.
The one Season 6 event where we captured the final ranking makes it plain. All eight of the final top 8 were qualifier group winners, and their qualifier scores were almost in final-ranking order (1st 48.7M, 2nd 48.0M, 3rd 47.5M…).
So how do you prepare?
The whole qualifier window is the fight. There’s little room to luck your way up in the knockouts. Qualifier score is your seed and your strength rating in one, so finishing hero upgrades, gear and skills in the week before the event is what makes the biggest difference.
Set a realistic goal. If you’re not seeded, making the semifinals (top 8 in your group) is goal one, and that means beating the weaker seeds. The top 32 is effectively reserved for group 1st and 2nd; from 3rd down, reaching the semifinals is the achievement.
If you’re around 8th, go all-in on the last day. Twelve hours before close, more than half of 8th-place standings still flip. Inside the last half-day, a top-8 spot almost always holds.
Caveats
These records only exist when a collector user opened the Champion Duel screens, so coverage varies by event and half of the 30 events with qualifier scores are Season 6. Season 1 had no near-close qualifier capture and is missing from the table. Qualifier ranks are the rank at capture time, not necessarily the final one, and the game doesn’t say what makes someone a seed, so we couldn’t check that. Match winners were taken from the points in the match record. We’ll update the numbers as more events come in.
Unofficial fan-site content · no partnership or sponsorship with the developer · observational analysis from collected data.
By: lastwarhub.com
