Start with https://likesport.biz/articles/ice-dance-vs-figure-skating-olympic-differences-explained.html if you need a break from frame data–then queue training mode and study these ten names, because they will decide who lifts the Evo trophy next year.
Reynald once again tops Japanese power rankings after resetting grand finals against GO1 depleted crew, proving The King of Fighters XV still belongs to him when the stakes hit $50 k. Watch his 2-bar Iori kill combo that clocks 412 damage off any jump-in; he lands it nine times out of ten in tournament sets, according to DashFight stats.
16-year-old Fenrich cracked the Street Fighter 6 CPT Europe with a 72-win streak using only JP. His frame-perfect corner wall-bounce setup leaves opponents frozen for 27 frames–enough to confirm into level-three super for 40 % life. If Capcom keeps the current patch through December, he will enter Evo seeded second only to MenaRD.
Chanel swapped Leo for A.K.I. at Evo Japan and still finished top three. His poison fog oki locks 75 % of the cast in block-stun for 42 frames, letting him dash through for a left-right mix that even Daigo couldn’t fuzzy-guard on LAN. Expect every Street Fighter lab monster to copy his fog-dash shimmy by April.
Spinkop grinded Mortal Kombat 1 for 1,300 hours since launch and shows zero rust. He punished NinjaKilla breakaway with Scorpion spear > teleport > KB for 48 % life, closing out the set 3-0 at Combo Breaker. NetherRealm patched the route within 48 hours, but Spinkop already posted three new optimized punishes on his YouTube–subscribe now before they go private.
Gamera returns to Tekken 8 after a two-year military break and hits a 900-rating peak on the Korean server using only Victor. His heat-dash cancel into 4,3 creates a 12-frame gap that looks interruptible but is actually plus-three; opponents who press eat a full launcher. Harada hinted no balance patch until after Evo 2026, so Gamera setup stays legal for the big stage.
Bookmark these ten players, enable Twitch notifications, and keep your phone charged–ticket alerts drop in three weeks and their pools will sell out faster than a PS5 Pro restock.
Data-Driven Tier List: Who Leaked the Hidden Metrics?
Grab the JSON dump from the EVO 2025 warm-up PCs–those files still sit in the public lobby cache under /tmp/evo_stats. Run grep -E "player_id|dmg_per_opening|framedata_adv" and pipe the output into a small Python script (pandas + scikit-learn) to reproduce the 0.87 correlation between dmg_per_opening and tournament placement. The leak came from a TO who forgot to scrub the rehearsal build before handing the machines to the university volunteers; the hash of the fileset matches the one that briefly appeared on X at 03:14 GMT the night before pools.
Three clusters jump out once you plot the data. Cluster A: 11 players who convert every 7.3 seconds of neutral into 182 damage on average–Tokyo Akira "Rei" Hayashi, Seoul Lim "RoRo" Ji-hu, and NYC Maya "M10" Singh sit here. Cluster B: 38 grinders who need 11.4 seconds for the same damage. Cluster C: everyone else. If you scrim against Cluster A, set Discord to record frame-by-frame; their neutral gap is only 4–6 frames, so you need a 120 fps capture to spot the shimmy trigger. Cluster B still punishes sloppy jumps with 80 % consistency, so warm up your anti-air OS in the PS5 version instead of PC–its 2.14 extra frames of input lag mirror the offline cabinets they rent for East-Asia side events.
- Rei: 0.92 throw-tech rate vs. left-side approach–force right-side by switching your controller port after the coin-flip.
- RoRo: whiffs jab into parry on frame 9 every third reset–buffer a 7-frame low to score a full combo.
- M10: drops anti-air after 19 minutes of play–stretch sets to best-of-five in top-8 when possible.
Zip the parsed CSV together with the raw .brec files, drop them into a private GitHub repo, and share the link only through Fightcord DMs–Reddit auto-flags the domain and Twitch mutes on-screen URLs. If you want to stay off radar, rename the extension to .b64 and include a one-line decoder in the README. And next year, factory-reset every station before you let randos queue up; the metrics you leak today become the tier list tomorrow.
Frame-Data Deep Dive: 5-Frame Buffer Tech That Catapulted a Rookie to Top 8
Map your fastest normal to L1, toggle the PlayStation 5 "Accurate" controller latency preset, and grind the timing in training mode until the green 5-frame flash feels like a heartbeat–this alone shaved 0.18 s off rookie Seo "Rush" Jun punish window and turned every blocked shoryu into a full-combo starter. Rush recorded 1,200 attempts per night, logged the hit-confirm accuracy spreadsheet-side, and noticed that micro-releasing the trigger 2 ms before the buffer window removed the 3 % input-drop rate; three weeks later he was whiff-punishing Tokido 13-frame sweep with a 6-frame jab into 42 % life drain.
The payoff peaked at Evo 2026 pools: down 0-1 against a seasoned Karin, Rush buffered st.LP during the last 4 frames of her dash, stuffed it clean, confirmed with st.MP xx run, carried to the corner, and closed the round with 871 damage–one more set of those disciplined buffers and he punched a ticket to top 8 while commentators still pronounced his tag wrong.
Win-Rate Curves: 67% Corner Carry Stat That Separates Contenders from Pretenders
Filter every 2025 CPT replay for the moment a player corners the opponent at 40 % life or less; if the attacker converts at least two further knockdowns without leaving the corner, they win 67 % of those games within 17 seconds. Load that filter into your training-mode script tonight–set the dummy to random wake-up, record your optimal wall-splat route, and loop it 50 times per side. Players who hit the 67 % benchmark average 4.2 Evo points higher than those stuck at 55 %, and they survive loser's bracket 1.8 rounds deeper. The gap isn't magic; it's muscle memory built on three data-verified habits: always spend one bar to keep the wall, never auto-pilot after the third hit (opponent's mash-out rate spikes to 38 %), and delay your fourth normal by 6 frames–whiff data shows 41 % of reversals come out one frame too early, letting you crush-counter for 320 damage and seal the round.
| Player | Corner Carry % | Evo 2025 Finish | Avg. Seconds to Kill |
|---|---|---|---|
| Rei | 71 % | 2nd | 15.3 |
| Garo | 69 % | 3rd | 16.1 |
| Kyo | 68 % | 5th | 16.8 |
| Lian | 57 % | 17th | 21.4 |
Track your own percentage with the free CPT-replay parser–drop your CFN ID, export the CSV, and divide successful corner kills by total corner openings. If you sit below 63 % after 200 matches, swap to a character with a built-in wall-splat special; the parser shows Kyosuke and Meryl gain a 9 % boost in this metric without extra execution hours. Above 63 %, stop grinding combos and start studying opponent patterns: players who reach 70 % spend 42 % of their lab time reviewing defensive tendencies on wake-up rather than chasing max-damage routes. Hit that 67 % threshold before Evo 2026 registration closes and you'll land directly in the top-32 bracket skip pool, saving four hours of early Swiss sets and cutting your fatigue index by 12 %–the same edge Rei used to reset the grand finals last year.
Bracket Heatmaps: Why Pool D2 is the 2026 Death Slot No One Wants
Skip the warm-up stations and book a 6 a.m. flight–Pool D2 starts at 8:17 a.m. Friday and every seeded player drops into losers before lunch.
Last year data scrape shows 42 % of D2 openers end in a 0-2 exit, the worst survival rate across 48 pools. The culprit is a three-way collision: Korean Tekken prodigy ULSAN anchors the winner side, Smash phenom Sparg0 lands at 17th seed after a late registration, and Street Fighter Celeb-side rush-down demon Tokido slips in at 33rd. All three share the same 64-man quarter-pool, so two of them face elimination before the swiss round even starts.
EVO 2026 seeding logic uses a 70 % weight on 2025 CPT, Tekken World Tour and Smash major points, then fills gaps alphabetically by gamer tag. Because Sparg0 registered after the deadline, the algorithm treated his tag as "ZZ-sparg0" and parked him next to early birds ULSAN and Tokido. TOs won’t reseed; they only swap identical regions to avoid passport clashes, so the trio stays locked.
If you’re a low-profile qualifier, scout the stream schedule instead of the seed number. D2 matches hit broadcast on evo3 at 9 a.m., meaning every set is archived within 30 minutes. Download the replay immediately, because coaches on Discord already clip counter-pick stages and upload them to private servers before pools close. Run those VODs at 1.25×, tag every neutral reset, and you’ll spot that ULSAN breaks 75 % of throws on the seventh frame–input your break on the sixth and you’ll push him to the wall.
Hotel Wi-Fi buckles under 2.4 GHz when 3 000 fighters post clips, so tether through 5 GHz or you’ll drop frames in Discord callouts. Bring a 20 000 mAh power bank; last year D2 ran 53 minutes over because the Granblue cabinet lost sync and forced a restart. You’ll need the extra juice to review notes between matches.
Book a room at the Luxor, not the expo venue–D2 players get shuffled to Hall H, a 12-minute walk from the main stage. The hallway carpet slows your stride more than you think; testers clocked 9:42 versus 6:15 on tile. That 3.5-minute gap decides whether you hit the bathroom, find a sponsor for a quick handshake, or reach the setup early enough to counter-pick the Tri-Line stage ban.
Bottom line: if your badge reads D2, treat Thursday night like finals eve. Squeeze in one last death-match against a local who mirrors your main, then sleep at 11 p.m. with blackout curtains drawn. The bracket won’t wait for your coffee, and by 10 a.m. half the pool is already texting their flights home.
Scouting Checklist: What Coaches Track Before Day-1 Pools
Pull the last 90 days of match VODs and tag every neutral-to-hit-confirm conversion with a time-stamp; if a player lands the same 2-touch sequence more than 35 % of the time, drill your counter-hit setup in that exact spacing window until muscle memory overrides autopilot.
Watch for controller choice: a sudden shift from Hitbox to Pad between CEO and Frosty Faustings correlates with a 0.4 s slower aerial drift reaction–drop a safejump three frames earlier and you’ll trade into full combo.
Track warm-up length. Players who spend less than six minutes in training mode before their first pool match whiff 22 % more anti-airs under stadium lights; queue up a two-game casual set against them while they’re still chatting at the merch booth.
Scrape Discord logs for patch-day complaints–if they moan about "useless" pushback on their main 5MK, expect an immediate switch to an alternate pocket that has lost the same match-up in only two recorded sets; prep your counter-pick before bracket upload closes.
Log sleep data from public tweets: red-eye flights cut reaction time by 11 ms per lost hour; schedule your meeting for 8 a.m. pools and run a staggered pressure string they literally can’t fuzzy-guard on 5.5 h of sleep.
Finally, save their button config from the last streamed local; if the macro row shows L+M+H set to the top shoulder, they buffer burst on the first stagger every single round–bait it with a micro-drop combo and cash out with a 70 % rage-triggered punish.
Button-to-Super Lag Logs: How 0.17 s Reaction Gap Became the New Benchmark
Record your next ten ranked matches with OBS at 240 fps, then run the .mp4 through the open-source tool fgc-laglog.py; it spits out a CSV that flags every super that triggered ≥6 frames after the input light turned green. If your average gap is above 170 ms, swap your leverless PCB to an RP2040-based board (GamerFinger G4-2025 or Haute42 T16) and set the polling rate to 8 kHz–this alone shaved 11 ms off Kusanagi Evo 2025 bracket run and moved him from 17th to 3rd.
Top players now treat 0.17 s as the hard ceiling because the most common kill combo in 2026 meta–drive-rush, st. MP, st. HP, super–leaves only a 9-frame window to confirm. Whiff one frame and you eat 40 % life; whiff two and you’re stunned. The leaderboard proves it: every Evo 2026 finalist averaged ≤165 ms on 200 samples, while everyone who placed 33rd or worse sat between 175–190 ms. Buy a €25 Raspberry Pi Pico, flash the free lag-hunter firmware, and wire it inline with your USB-C cable; the onboard OLED flashes red the instant your hardware adds more than 1.4 ms. Pair that with a weekly 15-minute drill in FrameAdvantage 2.3 random-block mode–set the CPU to mash jab on wake-up and refuse to press super until you see the white hit-spark–and you’ll drop 8 ms in two weeks, guaranteed.
Side-Switch Denial Count: The Micro-Stat That Predicts Upsets
Track how many times a player refuses to switch sides after losing a round; if the count climbs above three, bet against them. Evo 2025 showed a 71 % upset rate when the higher seed clung to the same corner, and the stat already repeats in 2026 qualifiers.
Side-switch denial exposes three weaknesses: flawed counter-pick logic, shaky adaptation, and tilt. Stubborn players lock themselves into sub-optimal ranges, miss wall-corner set-ups, and hand the initiative to opponents who happily rotate. The micro-stat surfaces before the bracket shows it.
Japanese prodigy "Raito" denied six straight swaps at Kagaribi X, lost four neutral rounds, then got perfected by a 197th-seeded Blanka. His denial count spiked from 0.8 to 4.2 per set, and his Evo odds lengthened from +400 to +1800 overnight. Sportsbooks still lag; you can beat them by watching the player-cam instead of the scoreboard.
Grab the free "EvoObserver" overlay; it parses the broadcast HUD, logs every side-switch refusal, and pings your phone when the tally hits the danger zone. Combine it with five-round windows: if the denier drops two of the last five rounds while staying put, slam the live bet before odds adjust. The edge lasts about 18 seconds on average.
Coaching staffs already game this. Team KRO analyst sits behind the stage wall with a whiteboard tally; between games he flashes "SWITCH" to his player if the count reaches three. Since adopting the protocol, KRO reverse-sweep rate jumped from 11 % to 34 %. Copy the tactic in your local; a simple notebook column works just as well.
Watch for the tell: a player who glances at the crowd instead of the screen before declining the side swap. Eye-tracking data from 120 pools matches links that glance to a 64 % chance of losing the next round. Mash the bet button the instant you spot it, then enjoy the upset before the casters catch up.
Q&A:
Who the most surprising name on the 2026 list, and why did the writers bump them ahead of older gods like Daigo or SonicFox?
The writers gave the final slot to 18-year-old Vietnamese Akira one-trick "bún_chả". In the last twelve months he taken seven straight CPT events using only VT-4 Akira, bodying Tokido Chun-Li in grand finals twice. The panel said pure win-rate against top-eight opposition outweighed legacy points this year; Daigo missed two Super Premiers and SonicFox new main still has a 3–7 spread against Akira, so the math pushed the rookie in.
How did the article decide who counts as a "rising star" versus a "legend"? Some of these players have been around since 2019.
They drew the line at Evo top-eight finishes. Anyone with two or more Evo grand-final Sundays before 2024 is labeled a legend; everyone else is a riser. That why "rising" includes people like Amari, who actually competed for five years but only cracked Evo top eight for the first time in 2025.
Is there any data on how many of these pros are playing Street Fighter 6 versus other titles? I mainly follow Tekken.
Seven of the ten are SF6 specialists. The other three are Tekken 8: Ulsan, Chanel, and the article lone wild-card Pakistani pick, 16-year-old Atif "Butter" Khan, who just won EVO Japan with an undefeated loser-bracket run. If you care about Tekken only, skim straight to sections 4, 6, and 9 those are the T8 players.
Did the piece mention any training routines or setups I could copy? I’m stuck in Diamond and want to hit Master.
Yeah, they shadowed three players for a week. The most repeatable routine came from Amari: two 45-minute aim-lab sessions on a 144 Hz monitor, then three straight hours of ranked with a 30-second break every three games to jot notes in a paper notebook no phone allowed. He claims the handwriting step forces the brain to tag mistakes more deeply than a spreadsheet. Since he started it last winter he climbed from Diamond 3 to Master 2800 LP.
Are there any dark-horse names the article almost included but cut at the last minute?
The writers kept an "on the bubble" sidebar. Three names got nixed: British Cammy player F-Word, who missed the list because he placed 9th at both Evo and CPT Finals; American zombie-main "Rotten"; and Korean Tekken phenom CuddleCore, who took a six-month hiatus. The editors said if any of the final ten withdraws from the first 2026 major, CuddleCore is the first alternate they’ll slot in.
Which of the newcomers on the 2026 list has the best shot at knocking an Evo champion off the throne, and what makes their style so annoying to deal with?
The name everybody whispers right now is "Kite." He mains a zoner in the new SNK fighter who can cancel fireball recovery into a teleport, so he never stays in the same place longer than eight frames. Veteran champions are used to bullying younger players by walking them to the corner; against Kite the corner keeps disappearing. Daigo lost a five-game money set to him at a Kawasaki arcade last month and, after the last round, simply stared at the screen for thirty seconds without pressing rematch. That tells you how obnoxious the matchup feels when your footsies get answered by a teleport cross-up that leaves you blocking the wrong way. If he survives pools without running into a strong counter-pick specialist, top-eight is almost guaranteed.
Reviews
Sophia Williams
Omg these queens are slaying my life! Momochi wife luck finally ran out when that 15yo Korean girl reverse-perfected him with *Kuma* I screeched so loud my cat levitated. Buying her merch faster than I swipe on Bumble, periodt.
Emily
So Evo 2026 fresh meat list dropped guess I’ll cancel my "get good" plans and just marry a pro for free flights, duh
Evelyn
so none of you noticed the kid who beat Daigo at locals last month didn’t even make the list? I was folding laundry and still caught the stream, but whoever typed this thing must’ve been microwaving dinner. are we really handing out "rising star" badges like Girl-scout cookies now, or did the rest of you fall asleep on the sofa with the chat open again?
BlazeRider
Mate, who slipped you the 2026 crystal ball? My cat predicts better after sniffing cold pizza. Where my name? I once beat a Wi-Fi Ryu with a Guitar Hero controller counts, yeah?
Julian Mercer
Why omit ReptilianZombie? His Baiken reset loops shredded Evo 2025 pools stats show 78% perfect rate.
Olivia Brown
my heart thumping like a broken arcade stick half these names are boys i kissed on twitchcon bleachers, the rest are ghosts who beat my love out of brackets. i still keep their wristbands under my pillow; they smell like soda and promises. if any of you read this, please pick up. i’ll trade my last heart token for a rematch, or just a "hi" between rounds.