Official Smogon Doubles Tournament V - Round 11

Status
Not open for further replies.
TPP

TPP

is a Tournament Directoris a Top Social Media Contributoris a Forum Moderatoris a Top Community Contributoris a Top Metagame Resource Contributoris a Tiering Contributoris a Top Dedicated Tournament Hostis a Top Tutor Alumnusis a Top Team Rater Alumnusis a Senior Staff Member Alumnusis a Past SPL Championis a Past WCoP Champion
Moderator
0TxlDTW.png

Artwork by Blazenix
​
Welcome to the fifth edition of the Official Smogon Doubles Tournament! This year's format will be a Bo3 SV Doubles OU swiss tournament with a top cut, in which the winner walks away with the OSDT trophy
vgc.png.m.1
!

Rules & Tournament Format

General Tournament Rules
  • This is a standard SV DOU tournament.
  • OSDT V will be a swiss tournament with a top cut of 16.
    • Players will be automatically dropped when they hit three losses as they are no longer in contention for top cut.
    • Players will be assigned a random opponent out of the pool of players with the same record as you each round.
  • Rounds will be Best of Three. You may switch teams in between battles of the same set.
  • Matches are to be played on Pokemon Showdown!. If you can't agree on which server to play on, the Smogtours server is the default option.
  • Replays are required and must be posted every round.
  • There are no extensions, so please get your games done on time.
  • Your hosts are TPP, a fairy, FayaWizard and Dead by Daylight
Activity Decisions / Coin Flips: In the event that there is an uncompleted match at the deadline, it will be coin flipped unless one or both of the players involved in the match specifically request an activity win and explain sufficiently why they deserve to win over their opponent. Attempting to schedule a battle is required in order to receive the win by activity. To make it clear that you are requesting an activity win when you post, you must open your post with "Activity Win Request" in bold.

For example, if you believe your opponent is making an unfair request and that they don't deserve the activity win over you, you can either make your own request or quote their post and explain why they don't deserve it. You should try to make your activity win request posts at least 24 hours before the deadline to give your opponent a reasonable opportunity to contest it with their own activity win request.

In case of disconnections, if you played in the SmogTours server, call a Tournament Director as soon as possible and we'll see about the possibility of having a rematch.

Prize

Courtesy of the Pokemon Showdown! Upper Staff, in addition to the trophy, the winner of OSDT V will win an additional prize. The winner will have a choice of receiving a Custom Avatar or reclaiming an inactive username on Pokemon Showdown!, with the runner-up receiving the option not chosen by the winner.

Resources

Smogon Tournament Discord
Doubles OU Discord
Doubles OU Overview
Doubles OU Strategy Dex
Doubles OU Sample Teams
Doubles OU Viability Rankings

9-1 Group

Xrn  vs  Fangame10
SeaLife  vs  MichaelderBeste2
sir jelloton  vs  Chris32156

8-2 Group

Sapphire  vs  Grandmas Cookin
Fragments  vs  breakfast enjoyer
bagel  vs  Yoda2798
DaAwesomeDude1  vs  Charlotte
entrocefalo  vs  MADARAAAA
Voltix  vs  zee
Tenebricite  vs  Nido-Rus
Yellow Paint  vs  Actuarily
SoulWind  vs  xqiht
Nakano Nino  vs  Penter
SMB  vs  Always!

There are no extensions!
Replays are mandatory!
The deadline for this round is Sunday, May 4th at 11:59 PM GMT -5.
Resistance Sheet​
 
Last edited:
9-1 Group

Xrn  vs  Fangame10 Fangame10 is favored by Furret and will win.
sir jelloton  vs  Chris32156 I really like Jello's team structure, these teams may have unexpected effects.

8-2 Group

Sapphire  vs  Grandmas Cookin high light
Fragments  vs  breakfast enjoyer enjoy the game.
bagel  vs  Yoda2798 Highlight, actually I don't want any of you to be eliminated, but that's the truth.
DaAwesomeDude1  vs  Charlotte dad1 :)
entrocefalo  vs  MADARAAAA MADARAAAA Go!
Voltix  vs  zee Another decisive game, zee also has Furret's blessing, good luck.
Tenebricite  vs  Nido-Rus 50/50 sry, I don't know much about you.
Yellow Paint  vs  Actuarily Actuarial always wins.
SoulWind  vs  xqiht can't go wrong with chinese prodigy.
Nakano Nino  vs  Penter Nakano Nino Go!
SMB  vs  Always! SMB is always reliable like this.
 
EDIT: This is outdated, please see the 2nd post.

For the Top Cut, obviously all of the 9-1 players will automatically qualify, and the two 8-2 players with the lowest Resistance will do a Play-in round.
The current order on the 8-2 Resistance sheet will be accurate to the order after Week 11; only Sapphire and Nakano Nino could swap positions.
Nido-Rus and all players above them will skip the Play-ins (if they go 9-2) because two players below them are playing each other by the random draw.
 
Last edited:
Magic Mayhem Maiden said:
For the Top Cut, obviously all of the 9-1 players will automatically qualify, and the two 8-2 players with the lowest Resistance will do a Play-in round.
The current order on the 8-2 Resistance sheet will be accurate to the order after Week 11; only Sapphire and Nakano Nino could swap positions.
Nido-Rus and all players above them will skip the Play-ins (if they go 9-2) because two players below them are playing each other by the random draw.
Click to expand...
my res is only 60% if I were to lose to Xrn I may have to play for 16th no?
 
Fangame10 said:
my res is only 60% if I were to lose to Xrn I may have to play for 16th no?
Click to expand...
My bad, I forgot to account for the 9-1 players, and also forgot that other players can affect Resistances.

But I investigated the Resistance Spreadsheet, and it turns out that it has not yet been updated for Round 10, so I manually edited the sheet.
Screenshot 2025-04-29 011005.png


It will take too much time to determine who will skip the Play-ins because all results affects everyone's Resistance, but I think this is more accurate to the Playoffs.
 
dropping from the tour, was super fun playing DOU again and seeing all the cool teambuilding ideas, glhf to the remaining players!


The past 2 iterations of OSDT has had horrible hosting decisions made regarding the swiss format, and is frankly unacceptable for what is supposed to be the premier Smogon Doubles tournament. Honestly I think most of these bad decisions can be avoided if you simply read the wikipedia article on swiss tournament structure https://en.wikipedia.org/wiki/Swiss-system_tournament

tldr: the current bye-filling system for swiss tournaments is counter-productive, the number of rounds and expected outcomes actually need to be calculated properly, up/downpairs are not a bad thing, hosting tools for official Smogon tournaments needs to be polished.

1. Added Byes
From what I've gathered, the purpose of adding a crap ton of byes in early rounds is to create a 2^n bracket size, which in theory guarantees 0 byes throughout the duration of the tournament excluding drops and bans. This bye idea is in host consideration because OSDT autodrops players that cannot qualify for top cut (x-3), which means that an odd number of remaining players would give a bye to a player still eligible for top cut. Official Pokemon tournaments don't have this issue as they don't autodrop players, meaning that they can assign 0-x players the potential bye every round with minimal consequence. I've documented below how many byes were added this year, also counting free wins due to players wanting to drop and not being able to due to this system:
85 byes
85 byes
85 byes [2 drop]
21 byes
0 byes
1 bye [8 drops]
0 byes [2 drops]
1 bye [1 drop]
0 bye [1 drop]
0 bye [0 drop]
0 bye [1 drop]
As we can see from these numbers, a total of 293 byes were handed out in the swiss portion, with at least one bye happening in every round except Round 5 and 10. In the absolute worst case scenario in a regular swiss hosting, we'd have a maximum 11 byes, one in each round as having an even amount of byes cancels each other out through the pairing process. The added early-round byes barely did anything to help the later rounds anyways, and then has the massive downside of adding over 200 early-round byes. Swiss is designed to be adaptive; it always provides 1 bye maximum in the event of weird numbers, drops, and bans. Trying to manipulate numbers to create an ideal scenario doesn't actually work because humans don't act ideally, nor did the hosts actually execute ideally. As we'll talk about in the next part, the hosts added an extra round, which explicitly creates up/down pair scenarios, something the 2^n bracket size was supposed to prevent.

2. Tournament numbers and Top Cut
In theory with a 2^n bracket size, you're supposed to get a perfect binomial distribution of players at each rating, with no up/down pairs or byes. (I assume this is what the host team was hoping for?) Here's what we would get if nobody dropped or got banned:
1746034428196.png
Projected Swiss calculations found with: https://limitlesstcg.com/tools/swisscalc [it doesn't actually simulate brackets, but the generated numbers are usually very close to reality, and specifically for 2^n works perfectly.]
Numbers and Swiss Rounds found here: https://www.pokemon.com/static-asse...play-pokemon-tournament-rules-handbook-en.pdf [pdf page 27, VG Variant 1]
2^9 = 512, which is the math behind 512 players being 9 rounds. With zee's proposal to add an extra round to reduce top cut matches, we'd arrive at 10 rounds, not 11.

You may notice that this option for 9 rounds yields a large top cut size, which is expected for a tournament with numbers manipulated to the maximum cut size via byes (which is probably why they decided another swiss round was the play). Let's see what numbers we'd expect to get with 10 and 11 rounds:
1746062015227.png

1746062058902.png

1746062068956.png

1746062075391.png

(yes I wrote my own code to simulate actual brackets, these numbers are obtained over 32000 trials])
We obtain cumulative values of ~23.305, ~28, ~13.96, ~16.84, or most probably, cut sizes of 23, 28, 14, 17.
The number of Top Cut Round 1 Byes are as follows: Top 9, Top 4, Top 2, Top 15.
None of these numbers are amazing, but 11 rounds 512 players is the definitely the worst one: punishing the unlucky 16/17 seeds. Players much prefer rewarding a few lucky top players rather than punish a few unlucky x-2 players. Furthermore, 11 rounds for 512 players fails to accomplish the goal of removing the extra top cut round in exchange for the swiss round.
While this decision to add an extra round did give a chance to have a clean 16 player top cut (a 25% chance), we did not roll that 25% chance (shoutouts to sir jelloton for beating Xrn in R10 so we could see this blunder in action LOL) and thus rolled a 17 player cut, which is not only a waste of a week, but absolutely brutal for the unlucky 16/17 seeds.
I don't necessarily disagree to run 2 extra swiss rounds to eliminate 2 cut rounds, but the risks and numbers must be calculated before you follow through with them.

3. Up/Down pairs
I'm genuinely unsure why people think these are an issue. Every single swiss tournament has these (Pokemon, Chess, other TCGs, etc.) and besides resistance for top cut seeding, nobody complains about them. Implementing a forced 2^n bracket size, round robin, or whatever to avoid it comes with such little upside, but often with major downside (not that anti-up/down pairing was implemented properly anyways)

4. Misc.
Some minor things about swiss tours that were also swept under the rug:
a) Byes are not supposed to be awarded to players more than once. There are probably many who received multiple explicit byes, but an example is zee receiving a bye in Round 1 and Round 6. I don't know the specifics, but you should be rerolling seeds to ensure this doesn't happen, if your bracket generator doesn't already prevent players from receiving multiple byes. (Top Cut byes are separate)
b) Repeat pairings are not supposed to happen. Keanu vs Opelucid happened in Round 3, and in Round 9. We all know this, but yet it happened in the most important Smogon Doubles tour. OSDT needs a reliable, accurate bracketmaker that can be used long-term. I know Princess Autumn and Dark Pulse have both made swiss tour bracket managers that have worked to high degrees of success, and certainly worth a consideration for OSDT
c) Inactive players continue to clog up the tournament. We spend so much time worrying about the singular maximum bye created by the swiss tournament, yet somehow neglect the tens of byes created by inactive players not playing the game. Maybe we can consider bringing this inactivity proposal back? I have 8 wins in this tournament, yet I've only created 5 winposts, which really shouldn't happen. Works just fine too if you don't force 2^n bracket size.
d) Listening to the Doubles community. It feels like the Doubles community is always unhappy with many of the hosting decisions made for OSDT; maybe we should actually consider listening to the community that knows swiss best, and makes up a majority of the OSDT playerbase.

Some ideas to fix the current problems:
x-2 bye issue: Either you embrace the potential 1 bye, or you change x-3 autodrop to x-4 autodrop to give x-2 players no room to receive a late-round bye
Inactive players: remove them if they show 2-3 weeks of inactivity or willingness to play.
Up/down pairs: it's not a problem that needs fixing, up/down pairs are fine and don't hurt anyone.
Bracketmaker: Upgrade the bracketmaker, current one doesn't work as intended enough.
Tournament entries: Calculate the numbers ahead of time and plan the number of rounds accordingly. You can also do what Pokemon does and pre-determines formats based on number of players.
Sorry for the sloppy writing, I've been travelling non-stop the past week and barely have any spare time and energy to spend writing this x_x
 
Last edited:
Sapphire said:
dropping from the tour, was super fun playing DOU again and seeing all the cool teambuilding ideas, glhf to the remaining players!


The past 2 iterations of OSDT has had horrible hosting decisions made regarding the swiss format, and is frankly unacceptable for what is supposed to be the premier Smogon Doubles tournament. Honestly I think most of these bad decisions can be avoided if you simply read the wikipedia article on swiss tournament structure https://en.wikipedia.org/wiki/Swiss-system_tournament

tldr: the current bye-filling system for swiss tournaments is counter-productive, the number of rounds and expected outcomes actually need to be calculated properly, up/downpairs are not a bad thing, hosting tools for official Smogon tournaments needs to be polished.

1. Added Byes
From what I've gathered, the purpose of adding a crap ton of byes in early rounds is to create a 2^n bracket size, which in theory guarantees 0 byes throughout the duration of the tournament excluding drops and bans. This bye idea is in host consideration because OSDT autodrops players that cannot qualify for top cut (x-3), which means that an odd number of remaining players would give a bye to a player still eligible for top cut. Official Pokemon tournaments don't have this issue as they don't autodrop players, meaning that they can assign 0-x players the potential bye every round with minimal consequence. I've documented below how many byes were added this year, also counting free wins due to players wanting to drop and not being able to due to this system:
85 byes
85 byes
85 byes [2 drop]
21 byes
0 byes
1 bye [8 drops]
0 byes [2 drops]
1 bye [1 drop]
0 bye [1 drop]
0 bye [0 drop]
0 bye [1 drop]
As we can see from these numbers, a total of 293 byes were handed out in the swiss portion, with at least one bye happening in every round except Round 5 and 10. In the absolute worst case scenario in a regular swiss hosting, we'd have a maximum 11 byes, one in each round as having an even amount of byes cancels each other out through the pairing process. The added early-round byes barely did anything to help the later rounds anyways, and then has the massive downside of adding over 200 early-round byes. Swiss is designed to be adaptive; it always provides 1 bye maximum in the event of weird numbers, drops, and bans. Trying to manipulate numbers to create an ideal scenario doesn't actually work because humans don't act ideally, nor did the hosts actually execute ideally. As we'll talk about in the next part, the hosts added an extra round, which explicitly creates up/down pair scenarios, something the 2^n bracket size was supposed to prevent.

2. Tournament numbers and Top Cut
In theory with a 2^n bracket size, you're supposed to get a perfect binomial distribution of players at each rating, with no up/down pairs or byes. (I assume this is what the host team was hoping for?) Here's what we would get if nobody dropped or got banned:
View attachment 736416
Projected Swiss calculations found with: https://limitlesstcg.com/tools/swisscalc [it doesn't actually simulate brackets, but the generated numbers are usually very close to reality, and specifically for 2^n works perfectly.]
Numbers and Swiss Rounds found here: https://www.pokemon.com/static-asse...play-pokemon-tournament-rules-handbook-en.pdf [pdf page 27, VG Variant 1]
2^9 = 512, which is the math behind 512 players being 9 rounds. With zee's proposal to add an extra round to reduce top cut matches, we'd arrive at 10 rounds, not 11.

You may notice that this option for 9 rounds yields a large top cut size, which is expected for a tournament with numbers manipulated to the maximum cut size via byes (which is probably why they decided another swiss round was the play). Let's see what numbers we'd expect to get with 10 and 11 rounds:
View attachment 736483
View attachment 736485
View attachment 736486
View attachment 736487
(yes I wrote my own code to simulate actual brackets, these numbers are obtained over 32000 trials])
We obtain cumulative values of ~23.305, ~28, ~13.96, ~16.84, or most probably, cut sizes of 23, 28, 14, 17.
The number of Top Cut Round 1 Byes are as follows: Top 9, Top 4, Top 2, Top 15.
None of these numbers are amazing, but 11 rounds 512 players is the definitely the worst one: punishing the unlucky 16/17 seeds. Players much prefer rewarding a few lucky top players rather than punish a few unlucky x-2 players. Furthermore, 11 rounds for 512 players fails to accomplish the goal of removing the extra top cut round in exchange for the swiss round.
While this decision to add an extra round did give a chance to have a clean 16 player top cut (a 25% chance), we did not roll that 25% chance (shoutouts to sir jelloton for beating Xrn in R10 so we could see this blunder in action LOL) and thus rolled a 17 player cut, which is not only a waste of a week, but absolutely brutal for the unlucky 16/17 seeds.
I don't necessarily disagree to run 2 extra swiss rounds to eliminate 2 cut rounds, but the risks and numbers must be calculated before you follow through with them.

3. Up/Down pairs
I'm genuinely unsure why people think these are an issue. Every single swiss tournament has these (Pokemon, Chess, other TCGs, etc.) and besides resistance for top cut seeding, nobody complains about them. Implementing a forced 2^n bracket size, round robin, or whatever to avoid it comes with such little upside, but often with major downside (not that anti-up/down pairing was implemented properly anyways)

4. Misc.
Some minor things about swiss tours that were also swept under the rug:
a) Byes are not supposed to be awarded to players more than once. There are probably many who received multiple explicit byes, but an example is zee receiving a bye in Round 1 and Round 6. I don't know the specifics, but you should be rerolling seeds to ensure this doesn't happen, if your bracket generator doesn't already prevent players from receiving multiple byes. (Top Cut byes are separate)
b) Repeat pairings are not supposed to happen. Keanu vs Opelucid happened in Round 3, and in Round 9. We all know this, but yet it happened in the most important Smogon Doubles tour. OSDT needs a reliable, accurate bracketmaker that can be used long-term. I know Princess Autumn and Dark Pulse have both made swiss tour bracket managers that have worked to high degrees of success, and certainly worth a consideration for OSDT
c) Inactive players continue to clog up the tournament. We spend so much time worrying about the singular maximum bye created by the swiss tournament, yet somehow neglect the tens of byes created by inactive players not playing the game. Maybe we can consider bringing this inactivity proposal back? I have 8 wins in this tournament, yet I've only created 5 winposts, which really shouldn't happen. Works just fine too if you don't force 2^n bracket size.
d) Listening to the Doubles community. It feels like the Doubles community is always unhappy with many of the hosting decisions made for OSDT; maybe we should actually consider listening to the community that knows swiss best, and makes up a majority of the OSDT playerbase.

Some ideas to fix the current problems:
x-2 bye issue: Either you embrace the potential 1 bye, or you change x-3 autodrop to x-4 autodrop to give x-2 players no room to receive a late-round bye
Inactive players: remove them if they show 2-3 weeks of inactivity or willingness to play.
Up/down pairs: it's not a problem that needs fixing, up/down pairs are fine and don't hurt anyone.
Bracketmaker: Upgrade the bracketmaker, current one doesn't work as intended enough.
Tournament entries: Calculate the numbers ahead of time and plan the number of rounds accordingly. You can also do what Pokemon does and pre-determines formats based on number of players.
Sorry for the sloppy writing, I've been travelling non-stop the past week and barely have any spare time and energy to spend writing this x_x
Click to expand...
You writing this much shows that you care way too much for me to interpret your decision to drop from the tour at this stage as one made in good faith.

I find it intriguing how you complain about all this stuff while in the same post creating more of the same problem by dropping. Please have some integrity and play your game.

I'm a huge fan of gma, rooting for him to win the whole thing and wouldn't want people to discredit what hopefully will be his grand victory because of stuff like this.
 
Unfortunately calling act, my oppo could only play on sunday night and we agreed 9.30-10 pm +2 (also on his smogon wall), I waited more than 30-40 min online but never got response (conversation is in Italian but you can translate) and I cannot be available later
 

Attachments

  • Screenshot_20250504-223004.png
    Screenshot_20250504-223004.png
    389.9 KB · Views: 77
  • Screenshot_20250504-223013.png
    Screenshot_20250504-223013.png
    397.6 KB · Views: 72
Last edited:
Activity Wins:

Sapphire  vs  Grandmas Cookin
entrocefalo  vs  MADARAAAA
Nakano Nino  vs  Penter
 
Status
Not open for further replies.
Back
Top