One quick nod that I don’t usually do in an article: If you’ve been enjoying what we’re up to on The SWU.Report, please consider dropping by the ko-fi page. I’d love to get the hosting covered at least – and if we end up over that, folks that know me from AGOT know I’ll find a way to turn it back into some cool stuff for the community. Thanks, all.
Nine weeks in. One hundred and nine events. 8,206 registered decklists. One hundred and nine winners.
That last number includes Admiral Ackbar.
Prague was billed as the season’s final exam, the moment when nine weeks of cumulative APR data would meet the most prepared, most concentrated competitive field the LAW season had to offer. In some ways the results delivered exactly that. In others, they served as a useful reminder of what the numbers can and can’t actually tell us. We’ll get to both, including the part where I was wrong.
The Prediction: A Miss With Caveats
The W8 article predicted Dedra Meero / Colossus to win Prague. That was wrong. I was wrong. Admiral Ackbar / Data Vault won Prague. (But seriously, who’d have guessed that?)
It’s worth taking a look at that Dedra call, though, because the reasoning wasn’t bad even if the outcome was. Through nine weeks of data, Dedra posted the third-best F→T8 APR in the tracked field at +31.2%, with a matchup profile that beats Lando (56.4%), Aurra (76.4%), and Vader (67.3%): the three archetypes most likely to dominate any given top 8 bracket. Her Stage 2 had been improving week over week. The prediction was grounded in the data available, and the data was pointing at a deck that had a solid, beefy case for being the right call.
There’s also an interesting wrinkle here that the data couldn’t capture. Dedra entered Prague as a widely discussed threat. It’s entirely possible that writing publicly about Dedra as the Prague frontrunner scared some pilots off registering her and convinced others to prepare harder for the matchup, and I am, of course, deeply honored to have potentially single-handedly suppressed Dedra’s field presence at the season’s largest event. The result was she came in underregistered relative to her power level, with only 28 players bringing her to Day 1 out of 663 total. A deck that represents 4% of the field is harder to count on winning a large event regardless of its matchup spread, simply because there’s less probability mass in the bracket. The prediction may have been right about Dedra being good. It just couldn’t account for how few pilots trusted her on the day.
The problem, of course, is that Prague had a rather different answer entirely…
The Prague Top 8: What Actually Happened
The full Prague top 8 tells a story that’s worth documenting precisely before we draw any conclusions from it:
- Admiral Ackbar / Data Vault (NXS_BigONotation)
- Mother Talzin / Yellow Force (Alfie Crust)
- Chewbacca / Blue (Jonas Skjold Frederiksen)
- Kazuda Xiono / Data Vault (FatherComstock_TDM)
- Colonel Yularen / Red (PL_kaiser)
- Kazuda Xiono / Data Vault (bartomanix_TDM)
- Mother Talzin / Green Force (Ambroise Labelle)
- Luke Skywalker / Data Vault (EC_DarkRider)
Read that list and then look at the Day 1 field breakdown. Boba Fett ran 113 players (16.5% of the Day 1 field, the single largest archetype in the room). Lando / Lake Country ran 76 (11%). Obi-Wan Kenobi / Blue Force ran 56 (8%). Those three archetypes collectively represented roughly 35% of the Day 1 field.
Not a single one made the top 8.
The Day 2 field tells you exactly why. Every archetype that survived to the second day is at one, two, or three copies out of the full Day 2 field. The diversity is extraordinary: Luke, Maul, Thrawn, Dedra (Blue), The Client, Ackbar, Qui-Gon, Sly Moore, Sabé: a full spread of archetypes that have been living outside the top tier all season. This wasn’t a meta that converged on its dominant decks in elimination rounds. It was a meta that was actively built to beat them, and it worked completely. Prague played out as the most pronounced single-event example of the hub-and-spoke dynamic we’ve been tracking all season: the decks in the center of the wheel got preyed upon so thoroughly that none of them had a pilot still standing in the top 8.
Kazuda Xiono / Data Vault putting two players into the top 4 is one of the more interesting late-season signals. He’s been appearing in PQ top 8s with increasing regularity in the back half of the season, and Prague suggests the deck has honest-to-goodness legs rather than being a series of local spikes. Worth watching when the next season kicks off.
Chewbacca / Blue finishing 3rd is worth a closer look too. This isn’t the fast Chewie Blue variant that occasionally popped up in PQ data earlier in the season. Jonas Skjold Frederiksen’s list was built around incremental board advantage, using healing and extended-value units to grind opponents down rather than racing. The result suggests there’s a beefy midrange version of this archetype that the aggregate season data never fully captured, in part because it was, of course, rare enough to fly under the tracking radar.
Two Talzin variants in the top 8 (Yellow Force in 2nd, Green Force in 7th) is also a useful data point. Both builds share the core Talzin engine but skew into different aspect packages, and both survived a Day 2 field that eliminated Boba, Lando, and OB entirely. The matchup profile bears this out: Talzin runs reasonably well into the three most common Day 1 archetypes, which made her a natural survivor in a bracket designed to punish the season’s dominant decks.
What Prague Actually Told Us: The Player Behind the Deck
One of the foundational limitations of APR as a predictive methodology is that it measures deck performance in aggregate, the collective result of every registered copy of an archetype across every eligible event. What it cannot measure is the individual pilot. And in a 663-player event, the winner was one of exactly one Admiral Ackbar / Data Vault players in the entire Day 2 field.
This is, in a real sense, the sport being the sport. A player who has dedicated significant time to a single off-meta choice, honing their lines against the expected field, learning the matchup intricacies that most opponents haven’t prepared for, can absolutely rise to the top of a large event through sheer craft even when the aggregate signal on their deck is quiet. That’s a beefy skill expression, and it’s one the numbers simply can’t capture. NXS_BigONotation, playing for the Nexus team, did exactly that. The APR doesn’t know that this particular pilot has been running Ackbar / Data Vault for weeks and knows every interaction. It only knows that 28 players registered the archetype all season, and it treats each of them identically.
To that end, the log normalization in the APR formula does try to account for low player count. The adjustment compresses extreme values at small sample sizes, which is why Ackbar’s F→T8 APR in the spreadsheet shows -24.8% rather than something astronomically negative. The formula is working as designed in that sense. But even with the normalization, a single-entry archetype that goes 1-for-1 on T8 appearances and 1-for-1 on wins remains an outlier, and the numbers should be read accordingly.
The conversion rate metric is worth flagging here too. Ackbar / Data Vault shows a 100% conversion rate: one T8 appearance, one win. That looks extraordinary, and in a sense it is. But it’s exactly the kind of case that illustrates why conversion rate as a standalone metric is one of the weakest signals in the dataset. A single player navigating one elimination bracket is not a sample size. APR is better than raw conversion rate precisely because the log normalization dampens these extremes, but even APR has limits when the field count approaches one.
The honest read on Prague is this: Dedra was the right prediction given what the data could tell us. Ackbar was the right result given what the data couldn’t. My predictive record at the season’s largest event sits at a solid 0-for-1. There you have it.
The Season-Long Signals: What Held Up
Set aside Prague for a moment (and my predictions for Prague, which I’m happy to do) and look at what nine weeks of data actually produced as stable signals.
Lando Calrissian / Lake Country ends the season as the dominant entry archetype. Final numbers: 445 decks, 107 T8 appearances, 16 Hub wins, +85.5% F→T8 APR, +13.6% T8→W APR. The entry efficiency signal was present from Week 3 and never wavered. Prague’s result (zero top 8 appearances despite 76 Day 1 registrations) is the clearest single-event illustration of what it means to be the deck everyone has a plan for. The F→T8 APR measures how efficiently Lando converts field presence into T8 appearances across the full season. It does not, of course, measure how well a room full of prepared pilots targets the most popular deck at the season’s most prestigious event.
| Archetype | Field | T8 | Wins | F→T8 APR | T8→W APR |
| Lando Calrissian / Lake Country | 445 | 107 | 16 | +85.5% | +13.6% |
Aurra Sing / Data Vault had a dark W9 (zero T8 appearances, zero wins at Prague) but finishes the season with the second-best entry APR in the tracked field at +40.0%. Her W7 surge was real and sustained. The structural story (strong into Lando, Boba, and Vader; weak to Dedra) played out exactly as the matchup matrix suggested it would. Prague specifically may have exposed a secondary problem: Yularen ran 50 players at Day 1 (7% of the field, third highest), and PL_kaiser’s 5th place finish with Yularen / Red is direct confirmation that the archetype had real teeth in this specific field. Yularen runs meaningfully higher in European events than US events, which means Aurra’s aggregate matchup numbers across the full season may have understated the matchup problem in the context of a Prague-style event. That’s worth keeping in mind as a regional artifact rather than a general verdict on the Aurra/Yularen pairing. In a similar vein, for those of us who played A Game of Thrones for years, this isn’t a surprise: the European meta always ran a bit askew from the US numbers, with different decks rising and falling based on regional testing groups and local favorites. It appears SWU is no exception on that front.
Boba Fett / Lake Country did exactly what Boba always does across the full season. 837 decks, 130 T8 appearances, 18 wins. Steady, consistent, unspectacular F→T8 APR (+34.5%), healthy Stage 2 (+7.8%), positive win O/U throughout. The workhorse of the season, of course, from beginning to end. Prague’s zero top 8 appearances from 113 players has already generated some talk about whether Boba is in decline. The nine-week APR record says no. What Prague actually shows is that Boba was the most-targeted deck in the room, with specific tech appearing in multiple lists designed to answer him. That’s not a deck falling off. That’s a deck that was so good for so long that the final event of the season built around beating him. There’s a difference, and the cumulative numbers make it clear.
| Archetype | Field | T8 | Wins | F→T8 APR | T8→W APR |
| Boba Fett / Lake Country | 837 | 130 | 18 | +34.5% | +7.8% |
The Season’s Most Interesting Closing Story: Sabé
Sabé / Data Vault ends the season as the most efficient closer in the tracked field with a meaningful sample size. Final: 117 decks, 12 T8 appearances, 4 wins, +63.1% T8→W APR.
| Archetype | Field | T8 | Wins | F→T8 APR | T8→W APR |
| Sabé / Data Vault | 117 | 12 | 4 | -1.8% | +63.1% |
The F→T8 at -1.8% means she never found her entry efficiency, cutting to top 8 at a rate barely below expectation, not dominating Swiss. But once she made the cut, her conversion was elite. The matchup matrix bore this out: she goes even or better into Lando (46.3%) and Boba (58.6%), with a roughly even Vader matchup at 46.2%, a solid spread against the most common elimination round opponents. She’s a solid closer who needed better pilots in Swiss or a larger registration base to really show what the deck could do. Her Day 2 appearance at Prague (one copy) is a small confirmation that she can get there at a large event.
The Season’s Persistent Frustration: Obi-Wan / Blue Force
Obi-Wan Kenobi / Blue Force ends the season exactly where he’s been since Week 4. 589 decks, 78 T8 appearances, 5 wins, +17.4% F→T8 APR, -31.4% T8→W APR. Win O/U of -4.75, the most negative in the tracked field.
| Archetype | Field | T8 | Wins | F→T8 APR | T8→W APR |
| Obi-Wan Kenobi / Blue Force | 589 | 78 | 5 | +17.4% | -31.4% |
This is the clearest structural verdict in the dataset. OB/BF converts field presence into T8 appearances reliably. He simply cannot close. His most common T8 opponents are, of course, also his worst matchups, and that doesn’t change because the players bringing those decks know it and prepare for it. Prague was just the final punctuation mark: 56 players, zero top 8 appearances, zero impact on the bracket. Nine weeks. Five wins from 78 T8 appearances. The signal is the verdict.
Other Notable Season-End Numbers
Luke Skywalker / Data Vault ends at -3.2% F→T8, -30.5% T8→W, 2 Hub wins from 37 T8 appearances. The prediction that he’d end the season without winning was wrong. He picked up wins in both W7 and W9. His Prague top 8 appearance (EC_DarkRider, 8th place) is the best single-event result of his season. The Stage 2 numbers remain some of the weakest in the dataset across the full season, but Prague suggests there are pilots who know how to make this deck work when the bracket lines up…
Grand Admiral Thrawn / Yellow Force finishes the season as the most extreme outlier in the closing-efficiency band at +47.9% T8→W with a -27.4% F→T8. Two wins from six T8 appearances all season, a remarkable closing rate structurally limited by its own niche appeal. His Day 2 presence at Prague (two copies) and the top 8 bracket suggests the European meta may value this deck more than the aggregate numbers reflect.
Dedra Meero / Colossus ends the season with a curious split: excellent F→T8 APR (+31.2%), persistent negative Stage 2 (-5.9%). She also, of course, produced exactly zero Prague top 8 appearances despite being my confident pick to win the whole thing. The matchup data explains the ceiling. OB/Vergence at 80% against her remains the structural problem in elimination play, and Boba and OB appearing regularly in top 8s means she’s fighting against two difficult opponents more often than her matchup spread would prefer. The Prague top 8 contained zero Dedra / Colossus, though two Dedra / Blue players made Day 2, suggesting the base pairing matters considerably at the highest level of play.
The Final Table
| Archetype | Field | T8 | Wins | F→T8 APR | T8→W APR |
| Lando Calrissian / Lake Country | 445 | 107 | 16 | +85.5% | +13.6% |
| Dedra Meero / Blue | 21 | 6 | 2 | +57.9% | +47.9% |
| Boba Fett / Data Vault | 65 | 14 | 0 | +47.7% | -28.6% |
| Lando Calrissian / Data Vault | 123 | 23 | 1 | +40.6% | -30.6% |
| Aurra Sing / Data Vault | 220 | 39 | 5 | +40.0% | +1.4% |
| Boba Fett / Lake Country | 837 | 130 | 18 | +34.5% | +7.8% |
| Vel Sartha / Red 27HP | 55 | 10 | 1 | +31.8% | -7.1% |
| Dedra Meero / Colossus | 223 | 36 | 4 | +31.2% | -5.9% |
| Admiral Piett / Blue | 157 | 24 | 0 | +24.6% | -39.6% |
| Chewbacca / Yellow | 154 | 23 | 4 | +22.7% | +18.4% |
| Obi-Wan Kenobi / Vergence Temple | 214 | 31 | 2 | +21.6% | -24.8% |
| Mother Talzin / Green Force | 100 | 15 | 1 | +21.1% | -19.1% |
| The Client / Blue | 75 | 11 | 0 | +18.3% | -23.4% |
| Kazuda Xiono / Data Vault | 83 | 12 | 2 | +17.7% | +12.6% |
| Obi-Wan Kenobi / Blue Force | 589 | 78 | 5 | +17.4% | -31.4% |
| Chewbacca / Alliance Outpost | 108 | 15 | 1 | +16.0% | -19.1% |
| Leia Organa / Data Vault | 80 | 11 | 1 | +14.3% | -10.0% |
| Boba Fett / Blue | 90 | 12 | 2 | +12.8% | +12.6% |
| The Client / Red | 75 | 10 | 0 | +12.2% | -21.3% |
| Admiral Piett / Red | 111 | 14 | 4 | +9.8% | +51.4% |
| Chewbacca / Blue | 47 | 6 | 0 | +8.7% | -9.6% |
| Colonel Yularen / Red | 293 | 35 | 5 | +7.8% | +7.6% |
| Tobias Beckett / Red | 176 | 21 | 3 | +7.1% | +6.5% |
| Mother Talzin / Yellow Force | 267 | 31 | 2 | +5.8% | -24.8% |
| Darth Vader / Yellow | 347 | 40 | 5 | +5.5% | 0.0% |
| Aurra Sing / Red | 99 | 11 | 1 | +2.3% | -10.0% |
| Mother Talzin / Red Force | 81 | 9 | 3 | +2.2% | +56.7% |
| Sabé / Data Vault | 117 | 12 | 4 | -1.8% | +63.1% |
| Lando Calrissian / Blue | 206 | 21 | 1 | -2.4% | -28.3% |
| Luke Skywalker / Data Vault | 366 | 37 | 2 | -3.2% | -30.5% |
| Darth Maul / Blue Force | 148 | 14 | 3 | -6.1% | +28.6% |
| Qui-Gon Jinn / Green Force | 158 | 13 | 2 | -12.7% | +9.0% |
| Avar Kriss / Yellow Force | 70 | 5 | 1 | -15.5% | +15.9% |
| Lando Calrissian / Colossus | 43 | 3 | 1 | -14.4% | +34.1% |
| Colonel Yularen / Red 27HP | 68 | 4 | 1 | -21.0% | +23.8% |
| Admiral Ackbar / Data Vault | 28 | 1 | 1 | -24.8% | +71.7% |
| Boba Fett / Yellow | 110 | 6 | 1 | -25.4% | +9.6% |
| Grand Admiral Thrawn / Yellow Force | 117 | 6 | 2 | -27.4% | +47.9% |
Ackbar/DV APR values reflect n=1 field entry (Prague Regional, 663 players). Values are mathematically accurate but not meaningful as predictive signals. Hub win counts used throughout. Meta Stats win counts are higher as they include sub-32 player events not tracked by the Hub.
Looking Back
Nine weeks, 109 events, and one of the most analytically interesting competitive seasons the game has seen (my Prague predictions notwithstanding). The format produced a dominant deck (Lando), a coherent counter-meta (Aurra, Dedra), a clear structural loser despite high volume (OB/Blue Force), a surprising closer (Sabé), and a Prague top 8 where the three most-played Day 1 archetypes produced zero top 8 appearances between them.
To that end, the APR methodology held up where it was supposed to. It correctly identified Lando’s dominance early and sustained it. It correctly flagged OB/BF’s Stage 2 problem before it was community consensus. It correctly predicted Aurra’s surge in Week 7. What it couldn’t account for, and isn’t designed to account for, is a room full of pilots who collectively decided to build for the metagame rather than play the metagame, and a single pilot who brought true mastery to an unexpected deck at exactly the right moment on exactly the right weekend.
There you have it. The LAW season, in the books.
Ending here with a bit of a teaser – if all goes well, I should have something brand new for the site ready to premiere next week week. Keep your fingers crossed that I can finish getting it set up correctly.
Archetype counts from SWU Competitive Hub (109 Hub-eligible events, >32 players, Premier format). Field estimates from Meta Stats deck count page (May 26 2026). Prague top 8 and Day 1/Day 2 field breakdowns from the official Regional Championship standings and Fanfinity event graphics. APR methodology from “We’re Talking About It Wrong” (swu.report, August 2025). Total field: 8,206 registered decklists across the season.
