One of my favorite Melee upsets of all-time is when Armada beat Mew2King at the first Genesis. While beating Mang0 definitively remains the greatest overall win from Armada’s run, I think that beating Mew2King was the turning point in which the stakes of the event suddenly jumped up a notch. After beating Mew2King, beating Mang0 was on the table, and so was winning the tournament – all for a guy who played Peach and had come all the way across the Atlantic Ocean for his first major.
Magic tends to happen at Genesis, and seemingly more often than other Melee tournaments. It’s practically the birthplace of our game’s greatest players, rivalries, and breakouts. Decades after Armada came onto the scene, and then retired, we witnessed the rise of Jmook, a current generational great, whose rise to prominence wouldn’t have been possible without the existence of Genesis.
These are the most prominent examples of amazing upsets at Genesis. Many more happened off-stream, on crowded setups in the corner of the venue. In today’s column, I’m going to explore the numbers behind the biggest upsets in Genesis history.
Methodology
When it came to assessing the biggest upsets, naturally, I had much to investigate across ten iterations of the world’s premier Melee tournament. I began from the latest one, perusing through the start.gg pages of Genesis X, looking through top eight, top 64, round two pools, and even Reddit threads of round one pools to ensure that I didn’t miss anything particularly crazy. I collected data related to not only the players involved in upsets, but their seeds as well. Using their seeds, I then created normalized scores for “Expected Placement by Seed (EPS) and measured the difference between both players in this category to evaluate their respective “Upset Factor.” This was inspired by PGStats’ metric of the same name, though my methodology is slightly different; see the appendix for more details.
This was a long and tedious process to do for every Genesis tournament. In hindsight, I could have likely found a way to automate the process of checking individual player pages to determine their seeds. It could still be possible that I’m missing data here. But I find it unlikely. I cross-referenced my list with my dear friend and Last Night In Melee reporter self-flagellate to ensure that I was not missing anything substantial, at least up until Genesis 5.
Why not Genesis 4? The long story short: the bracket software tournament organizers use today was not always as powerful as it is in 2025. Back in the days of start.gg’s launch (when it was smash.gg), it was not uncommon for major organizers to manually reshuffle bracket seeding after round two pools. The idea was that by doing this, they could avoid repeats in bracket matches (the double jeopardy problem), and overcome logistical constraints of not having software that could execute these tasks. Unfortunately, for my purposes, this made the “seed” listed for certain players on the Genesis 4 and Genesis 3 pages extremely dubious. This type of issue was similarly present for Genesis 2 and the original Genesis, but with another hassle in the presence of round-robin pools before the final bracket of each event.
Given the constraints of me writing this (one week), I had to make a judgment call: exclude old results or do my best to ‘recreate’ seeding through checking the bracket structure and determining ‘projected seeds’ for each player? It was initially difficult to make a decision, but then I realized something that all people recognize at some point or another in their life – it’s okay to ask for help.
I reached out to both CurlyW and Dr. Z, co-head TO of Genesis, for the original seeding sheets for the first four Genesis events. Lo and behold, I finally had something concrete to work with (at least in terms of the third and fourth Genesis). Keep this in mind as you notice that upsets don’t always cleanly add up in terms of seeding for when they happen in the tournament (for example, SFAT, as the 12 seed, played Mew2King, the 4 seed, at Genesis 3 in winners round of 16, rather than PewPewU, the 13 seed).
While working on compiling data I also made an executive decision: I am not going to count round-robin pools results toward my methodology. It’s just not really as important or scalable to my method of evaluating upsets as looking at results that happened in the final bracket. So in the end, I included everything, made ‘educated’ guesses for the first two Genesis tournaments, and then excluded pools results.
Why You’re Not Allowed To Get Mad
I initially positioned this article as me overviewing the biggest upsets in Genesis history. But that wasn’t painting the whole picture. More than being my personal list of greatest upsets in Genesis history, this article is my ‘attempt’ to create a method to evaluating their impact. As you read through this list, be warned: you’re going to see sets here that probably “shouldn’t” be on a greatest upsets list. Some of you will likely see red and immediately go “why is Edwin counting this as an upset when it clearly isn’t one?” In fact, you might see results that aren’t there and then go “wow, I can’t believe Edwin didn’t count this.”
Remember: I’m intentionally sticking to an imperfect methodology. Why? Out of morbid curiosity. Obviously, a more proper way to determine upsets would involve measuring head-to-head history, weighing character matchups accordingly, or other factors beyond seeding. But that’s not what this column is about.
Upset Factor of 4
| Event | Underdog | Favorite |
| Genesis | Armada (6) | Mang0 (1) |
| Genesis | Eddy Mexico (52) | Colbol (13) |
| Genesis | Scar (21) | Lambchops (5) |
| Genesis 2 | MacD (29) | Fly Amanita (7) |
| Genesis 3 | n0ne (50) | PewPewU (13) |
| Genesis 3 | Milkman (97) | Wobbles (28) |
| Genesis 5 | Kira (126) | Colbol (30) |
| Genesis 5 | Lucky (15) | Mew2King (4) |
| Genesis 5 | Santiago (34) | ChuDat (11) |
| Genesis 5 | Plup (6) | Hungrybox (1) |
| Genesis 6 | Snowy (99) | Kalamazhu (30) |
| Genesis 6 | Darkatma (51) | Lucky (14) |
| Genesis 6 | Chango (162) | ChuDat (34) |
| Genesis 7 | Forrest (50) | moky (15) |
| Genesis 7 | SluG (85) | Spark (21) |
| Genesis 7 | FliiNcHy (163) | Magi (35) |
| Genesis 7 | Fishbait (171) | Kalamazhu (43) |
| Genesis 7 | Sufy (141) | Kalvar (45) |
| Genesis 7 | Sock (314) | Franz (58) |
| Genesis 8 | Jmook (21) | Plup (5) |
| Genesis 8 | Krudo (42) | KoDoRiN (10) |
| Genesis 8 | Noire (163) | Zuppy (35) |
| Genesis 8 | Gooms (100) | ChuDat (29) |
| Genesis 8 | John Wick (87) | Joshman (23) |
| Genesis 9 | Salt (25) | Axe (8) |
| Genesis 9 | Quang (101) | Chickenman400 (28) |
| Genesis X | aMSa (7) | Zain (2) |
| Genesis X | Raz (43) | Salt (11) |
| Genesis X | Medz (50) | Magi (15) |
Against – not all of these are results most Melee people would reasonably consider the greatest Genesis upsets ever. Nobody is really going to bat an eyelash at aMSa defeating Zain, Krudo beating KoDoRiN, or Plup beating Hungrybox. However, you’ll still find some of the more notable upsets in genesis history here. These ones might not sniff anywhere close to the Top 10, but you have hidden gems like Sock beating Franz in there, as well as the crown jewel of Armada’s Genesis run in his win over Mang0.
To address the elephant in the room here: I obviously do not think that Upset Factor is an adequate way to represent the thunderous nature of Armada defeating Mang0 at the first Genesis. With that said, for some of the other surprising results here, it’s not too far off, and it’s pretty funny to see it grouped into the same category as Raz defeating Salt.
Upset Factor of 5
| Event | Underdog | Favorite |
| Genesis | Hungrybox (16) | DaShizWiz (3) |
| Genesis 3 | Nintendude (22) | Mew2King (4) |
| Genesis 3 | Alex19 (53) | Silent Wolf (10) |
| Genesis 5 | Kalvar (159) | Nintendude (31) |
| Genesis 5 | Bunk (154) | dizzkidboogie (26) |
| Genesis 6 | Trif (28) | Mew2King (5) |
| Genesis 6 | Bimbo (80) | Bananas (16) |
| Genesis 7 | Darrell (119) | Lucky (23) |
| Genesis 9 | skaht (426) | Kalamazhu (42) |
| Genesis X | Juicebox (78) | Spark (14) |
| Genesis X | Medz (50) | Aklo (10) |
| Genesis X | Buckets (1482) | Justus (85) |
Here, we get into recognizable upsets, results that hardcore fans might reminisce about, and a few shenanigan-packed sets. Which one shall we start with? Let’s go with the shenanigans. The two most jarring upsets on this list in terms of raw differential between involved players are Buckets over Justus and and skaht (formerly known as EdwinBuddingFan) over Kalamazhu. In the case of Buckets, a power-ranked NorCal Falco, the seeders just missed him. It happens. Meanwhile, I would love to treat skaht over Kalamazhu as exactly what it appears on the surface: the largest raw seeding differential between a Top 48 supermajor seed and an underdog in Genesis history. Sadly, I have to tell the truth: Kalamazhu played Fox dittos, not a single game of Peach. Is it an all-time upset? No, probably not. Is it great for my friend though? Absolutely.
My favorite of the lesser-known upsets is Alex19. This guy is often retroactively remembered as a streamer who got voted into Summit as Mang0’s friend, but nobody ever talks about how good he actually was. His win over Silent Wolf at Genesis 3, along with beating Wizzrobe, is proof that this guy was actually insane at the game. This, mind you, came in a timespan when he was also high up on SoCal’s power rankings and beating people like n0ne, S2J, and Westballz. Alex was a monstrous player and had a huge presence in the scene. It’s absolutely insane that nobody seems to ever talk about him any more.
Mostly everything else makes sense. Some of the character-matchup specific upsets are inflated by Upset Factor (like Texas Peach Bunk defeating Ice Climbers dizzkidboogie), but then you get into your classic event-defining upsets, like Hungrybox beating DaShizWiz and then crying afterwards at the original Genesis, or Trif beating Mew2King and having a monster pop off.
Upset Factor of 6
| Event | Underdog | Favorite |
| Genesis 5 | Moky (75) | ChuDat (11) |
| Genesis 6 | Joyboy (75) | Swedish Delight (11) |
| Genesis 6 | JBash (233) | Shroomed (24) |
| Genesis 8 | Jmook (21) | Cody Schwab (3) |
| Genesis | Zhu (14) | Mew2King (2) |
We aren’t even in the Top 10, and we’re getting into the good shit. You have a legacy-defining set in Zhu defeating Mew2King at the original Genesis and the grand finale (at least the one we like to remember) of Jmook’s performance at Genesis 8. There’s also moky’s first win on the big stage with his victory over ChuDat.
However, the other two of these are practically forgotten about in Melee history. It can’t be overstated how wild it was that JBash, a nationally unranked Marth player from MDVA, beat Shroomed, not only one of the all-time great NorCal legends, but someone who consistently was a thorn in the side of PewPewU, one of the greatest Marths ever. The last result on this list is Joyboy over Swedish Delight. It’s tempting to ret-con this as a “Fox-Sheik” upset, but it’s also the same Sheik that had just previously beaten Armada’s Fox six months previously. Conversely, Swedish dropped out of the event due to dealing with an arm injury, so maybe this result should be taken with a grain of salt.
Upset Factor of 7
| Event | Underdog | Favorite |
| Genesis 2 | Taj (13) | Mang0 (1) |
| Genesis 6 | Mvlvchi (146) | ARMY (15) |
| Genesis 7 | Ryobeat (59) | Wizzrobe (6) |
| Genesis 8 | Eggy (117) | Moky (12) |
I have a story related to Eggy defeating moky. Before Genesis 8, I asked s-f a very silly prompt: “what’s the most fucked upset that could conceivably happen at Genesis?” His answer: Eggy defeating moky. All this is to say that if you’re looking for someone to blame, it’s s-f, who willed this result into existence. Remember this story, as I’ll get to my own answer to this prompt later.
In all seriousness though, this mostly adds up. mvlvchi over ARMY is obviously inflated by Upset Factor given their character matchup involved, but by seed, it undeniably had a ripple effect throughout the event. Meanwhile, Ryobeat over Wizzrobe was another takedown of a top seed that also played a large role in getting Ryobeat to Smash Summit. I briefly wrote about Taj beating Mang0 and have done so in even more depth, so I won’t repeat myself there.
Upset Factor of 8
| Event | Underdog | Favorite |
| Genesis | Axe (61) | Jman (4) |
| Genesis 6 | Magi (61) | Mang0 (4) |
Axe defeating Jman has a whole documentary, but it also changed Axe’s life and convinced him to take competing even more seriously beyond being known as a good Pikachu player. Meanwhile, Magi over Mang0 speaks for itself, with Magi becoming the first woman to defeat one of the gods. Having witnessed it in person too, it’s one of only two times I’ve ever been in a venue where the vast majority of people rooted for someone to beat Mang0. The other time was when aMSa won The Big House 10.
The main stipulation I’d have to these sets is that while they’re extremely important, they’re also all best-of-three. I still think they’re all-time caliber results, but at the same time, by virtue of coming in a higher variance format, it does make them a little less crazy.
Upset Factor of 9
| Event | Underdog | Favorite |
| Genesis 7 | Panda (68) | Axe (4) |
| Genesis 9 | 2saint (32) | aMSa (1) |
And now we get into two results that are definitely two of the biggest upsets in the history of Genesis. I think there’s a strange irony to Panda over Axe here – with Axe essentially turning from the person who had arguably the biggest upset in Genesis history to now becoming the victim of an arguably even bigger upset (though one that also happened in best-of-three).
I think that the scale of 2saint defeating aMSa is honestly really underappreciated among most Melee fans. aMSa headed into this tournament as a contender for No. 1, and one that looked impervious to any Jigglypuff that wasn’t Hungrybox, and in best-of-five as well. I’ll go out on a limb and say that this result is honestly crazier than when aMSa beat Mew2King for the first time.
That’s the Mang0! (Upset Factor of 10+)
| Event | Underdog | Favorite | Upset Factor |
| Genesis 9 | Taj (127) | Mang0 (2) | 12 |
| Genesis 8 | Fizzwiggle (64) | Mang0 (1) | 11 |
| Genesis 9 | Matteo (63) | Mang0 (2) | 10 |
Obviously, the sets with Matteo and Taj (the second one at least) are not serious contenders for being the biggest upsets of all-time, given that Mang0 played secondaries. However, I do think that Fizzwiggle beating Mang0 is an undeniably historic result, though the fact that it happened in a best-of-three does slightly work in favor of the underdog. Sadly, for my previous prompt to s-f (“what is the most fucked upset that could conceivably happen at Genesis 8”), this was my pick. Keeping that in mind, maybe this wasn’t so shocking. Actually, no; that’s BS. It is shocking. This is literally a 64-seed defeating a one seed. What am I talking about?
Although Mang0 won the first Genesis, the truth is he’s underperformed or sandbagged at the last three events. Maybe it’s corny, but as a fan of the game, I’d love nothing more than one last run from the GOAT where he gives it his all at Genesis and wins.
Conclusion
There’s a really great Bill James quote I learned about this week while talking to Chroma about this column. “When the numbers acquire the significance of language, they acquire the power to do all of the things which language can do: to become fiction and drama and poetry.”
To an extent, I think the ‘numbers’ behind some of the biggest upsets aren’t quite accurate to reality. At the same time though, what is reality if it isn’t our perception of what happened; what is our perception of things around us if they aren’t always changing? This is the fun of applying numbers to history; you challenge your preconceived notions. Sometimes they don’t change, and other times, maybe you missed something, and that’s okay.
More than creating a strict, ordered, undeniable list of the biggest upsets in Genesis history, I hope this column was a fun trip down memory lane, as well as a neat little way of exploring some cool stories that you never knew that you wanted to learn about.
Special thanks to Dr. Z and CurlyW for directly sharing original sheets for seeding for Genesis 3 and 4.
