Two Format Changes

[aermet69]

If SOS was a perfect measure for the trials you went through, I'd start to argue that we need to let some 5-3s in, over the 6-2s.

6. Casper J. 12 65
7. Koen M. 12 64
8. Quirin F. 10 82
9. Jerome N. 10 79

Looking at the standings from Europeans (since I didn't see SOS posted for Worlds yet), it would seem fair that Jerome should replace me or Koen in the top 8. Since I had a fair bit of easier opponents, me getting 6-2 seems like less of a feat than Quirin and Jerome who went 5-3.

Now, I don't think anyone (not even myself) would take such an argument seriously, but what I'm trying to illustrate, is that SOS is not a perfect measure and I think it's absolutely reasonable to set a target for wins, where you at least have a shot at top 8 rather than being off'ed on your SOS.

Especially when we have pair-ups and downs and the randomness of the first 1-2 rounds, we need to take SOS with a grain of salt.

[quickdraw3457]

My team was talking on saturday about this but the tournament software currently randomizes pairings each round based on record right? What if it calculated sos in real time and paired based on that? Would that help or hurt things?

[imrahil327]

My team was talking on saturday about this but the tournament software currently randomizes pairings each round based on record right? What if it calculated sos in real time and paired based on that? Would that help or hurt things?

Hurt, I'd think- then TRGR when it comes to SOS.

[WiseMarsellus]

probably someone like mht or jimli could answer this better than me. but my instinct is it would polarize the sos more. like the 2-2s who lost their first two games would face each other, while the 2-2s who won their first two would also face each other. and the 3-2s, and the 4-2s, etc. to the point where if you're 0-2 you're going to have an even more miserable sos than you already do.

however you're going to have the top players facing each other more regularly. so with the top players playing other top players and the bottom players facing more bottom players, maybe you have a more centralized final standings? or maybe i'm overestimating the impact. it's an interesting question though, i'm not sure if anyone uses such a system but i'd be curious to see what results they've found if there is one

[gogolen]

probably someone like mht or jimli could answer this better than me. but my instinct is it would polarize the sos more. like the 2-2s who lost their first two games would face each other, while the 2-2s who won their first two would also face each other. and the 3-2s, and the 4-2s, etc. to the point where if you're 0-2 you're going to have an even more miserable sos than you already do.

however you're going to have the top players facing each other more regularly. so with the top players playing other top players and the bottom players facing more bottom players, maybe you have a more centralized final standings? or maybe i'm overestimating the impact. it's an interesting question though, i'm not sure if anyone uses such a system but i'd be curious to see what results they've found if there is one

It seems like that is how it would work out, and would just make a bad start even harder to overcome. Its bad enough that you started 0-2, but now you are guaranteed to be playing the worst 0-2 player, making it that much harder to bounce back.

it would also be near-impossible to duplicate if there was an issue and the software crashed and pairings had to be done by hand.

[quickdraw3457]

probably someone like mht or jimli could answer this better than me. but my instinct is it would polarize the sos more. like the 2-2s who lost their first two games would face each other, while the 2-2s who won their first two would also face each other. and the 3-2s, and the 4-2s, etc. to the point where if you're 0-2 you're going to have an even more miserable sos than you already do.

however you're going to have the top players facing each other more regularly. so with the top players playing other top players and the bottom players facing more bottom players, maybe you have a more centralized final standings? or maybe i'm overestimating the impact. it's an interesting question though, i'm not sure if anyone uses such a system but i'd be curious to see what results they've found if there is one

I thought this too but when I thought about it more I think it would just hurt your opponent's SOS (2nd tiebreaker) more than your own SOS. Cause if you beat a bottom end 2-2 player, they're going to play a bottom end 2-3 player next round and theoretically have a better chance of winning and helping your SOS.

Or so I think.

[WiseMarsellus]

probably someone like mht or jimli could answer this better than me. but my instinct is it would polarize the sos more. like the 2-2s who lost their first two games would face each other, while the 2-2s who won their first two would also face each other. and the 3-2s, and the 4-2s, etc. to the point where if you're 0-2 you're going to have an even more miserable sos than you already do.

however you're going to have the top players facing each other more regularly. so with the top players playing other top players and the bottom players facing more bottom players, maybe you have a more centralized final standings? or maybe i'm overestimating the impact. it's an interesting question though, i'm not sure if anyone uses such a system but i'd be curious to see what results they've found if there is one

I thought this too but when I thought about it more I think it would just hurt your opponent's SOS (2nd tiebreaker) more than your own SOS. Cause if you beat a bottom end 2-2 player, they're going to play a bottom end 2-3 player next round and theoretically have a better chance of winning and helping your SOS.

Or so I think.

oh jeez this is a rabbithole. you're right. i'm going to amend my post to "i have absolutely no clue"

[vhstapes]

I see the merits of the play in game but having to play a 9th game of super competitive star wars sounds like a nightmare, and I'd 100% concede to the my opponent if they had the better SOS. If I was the higher SOS player, I'd be really annoyed at having to further prove myself and might just scoop to go have dinner. But I am a fat kid so there's that

[JediJer]

In what sense would you re-evaluate? IMO the goals you laid out were met, and the system worked as intended.

Simply based on the feedback from the community on the play in game, I think it would be wise to review and debrief to see if the correct decision was made. If the committee thinks that it wasn't a net positive for the community, then they should reevaluate the decision and either try something different or go back to a straight top 8 cut as before.

I think the random top/bottom pairing was overall a positive thing. I wouldn't want to change that.

[imrahil327]

In what sense would you re-evaluate? IMO the goals you laid out were met, and the system worked as intended.

Simply based on the feedback from the community on the play in game, I think it would be wise to review and debrief to see if the correct decision was made. If the committee thinks that it wasn't a net positive for the community, then they should reevaluate the decision and either try something different or go back to a straight top 8 cut as before.

I think the random top/bottom pairing was overall a positive thing. I wouldn't want to change that.

Oh sure, and that's what we are doing (including the survey) I thought you meant "It was definitely wrong and should be fixed"

[itcouldbewirfs]

probably someone like mht or jimli could answer this better than me. but my instinct is it would polarize the sos more. like the 2-2s who lost their first two games would face each other, while the 2-2s who won their first two would also face each other. and the 3-2s, and the 4-2s, etc. to the point where if you're 0-2 you're going to have an even more miserable sos than you already do.

however you're going to have the top players facing each other more regularly. so with the top players playing other top players and the bottom players facing more bottom players, maybe you have a more centralized final standings? or maybe i'm overestimating the impact. it's an interesting question though, i'm not sure if anyone uses such a system but i'd be curious to see what results they've found if there is one

I thought this too but when I thought about it more I think it would just hurt your opponent's SOS (2nd tiebreaker) more than your own SOS. Cause if you beat a bottom end 2-2 player, they're going to play a bottom end 2-3 player next round and theoretically have a better chance of winning and helping your SOS.

Or so I think.

If the software used SOS in real time for pairings how about using the swiss-system where top half plays bottom half. So if 8 players have the same pts then #1 plays #5, #2 plays #6 etc. Top players gets a somewhat easy match but not so easy or SOS crushing. Bottom players don't get stuck playing each other. Ofcourse the swiss system relies on external rankings so you can do this from the first round. We would still have random pairing for much of the first several games, which seems fine to me.

[karrde225]

I mentioned this to team Pgh at dinner on Saturday, but I really felt like I would have rather played the play in game on Sunday morning than Saturday night. But there were some voiced concerns about being able to prep for your opponent, etc.

But what if instead, we have either an 8 or 12 person Day 3 (depending on if there are any 6-2's outside of the top 8), and start it a round earlier in the latter case, giving the top 4 players a bye. Then you reseed when you have 8 players. So, no teching for specific opponents, no playing a 9th game on Day 2. Sure some would play 8 games max each day, but with only 8 people to corral, you can start Day 3 much earlier for those that don't get the top 4 seeds.

[mryellow]

I mentioned this to team Pgh at dinner on Saturday, but I really felt like I would have rather played the play in game on Sunday morning than Saturday night. But there were some voiced concerns about being able to prep for your opponent, etc.

But what if instead, we have either an 8 or 12 person Day 3 (depending on if there are any 6-2's outside of the top 8), and start it a round earlier in the latter case, giving the top 4 players a bye. Then you reseed when you have 8 players. So, no teching for specific opponents, no playing a 9th game on Day 2. Sure some would play 8 games max each day, but with only 8 people to corral, you can start Day 3 much earlier for those that don't get the top 4 seeds.

I would definitely be open to re-jiggering Day 3 start times. There is more wiggle room there than there is at the end of Day 2.

[Advocate]

I agree with Jonny. With our 13 combined Day 3 qualifications, that should carry some weight.

[Karrdeshark]

I mentioned this to team Pgh at dinner on Saturday, but I really felt like I would have rather played the play in game on Sunday morning than Saturday night. But there were some voiced concerns about being able to prep for your opponent, etc.

But what if instead, we have either an 8 or 12 person Day 3 (depending on if there are any 6-2's outside of the top 8), and start it a round earlier in the latter case, giving the top 4 players a bye. Then you reseed when you have 8 players. So, no teching for specific opponents, no playing a 9th game on Day 2. Sure some would play 8 games max each day, but with only 8 people to corral, you can start Day 3 much earlier for those that don't get the top 4 seeds.

I would definitely be open to re-jiggering Day 3 start times. There is more wiggle room there than there is at the end of Day 2.

If this was combined with dropping at least the quarter and semi-final time limits to 1:15, that could be plausible. As long as it's not too early (subjectively, 9) because that can really mess with jet lagged folks like myself.

Other considerations, though, are that the play in game should probably still be only one game, otherwise 4 full matches is a looooong day. Also, there needs to be some manner of radio silence to prevent the players with byes from getting a significant scouting advantage. I'm not anti-scouting, but when one player is having his games watched and the other isn't revealing his decks, that does grant a potentially significant advantage in terms of game planning.

[mryellow]

Other considerations, though, are that the play in game should probably still be only one game, otherwise 4 full matches is a looooong day. Also, there needs to be some manner of radio silence to prevent the players with byes from getting a significant scouting advantage. I'm not anti-scouting, but when one player is having his games watched and the other isn't revealing his decks, that does grant a potentially significant advantage in terms of game planning.

Also would these Day 3 play-in games be done with Day 2 decks? I'm assuming yes. If that's the case there's less of an issue with scouting, but either way that can be somewhat mitigated by having the games done in a separate room.

[Karrdeshark]

Also would these Day 3 play-in games be done with Day 2 decks? I'm assuming yes. If that's the case there's less of an issue with scouting, but either way that can be somewhat mitigated by having the games done in a separate room.

This is true (re: separate room) and would definitely be ideal. One possible problem, however, is that using yesterday's decks could force players to have to physically change their decks in between the play in round and quarterfinals if they can't build them both separately. Then they need a deck check for their own tired sake to ensure they didn't leave out something important. And would they still be able to change their list for the top 8 at that point?

[mryellow]

This is true (re: separate room) and would definitely be ideal. One possible problem, however, is that using yesterday's decks could force players to have to physically change their decks in between the play in round and quarterfinals if they can't build them both separately. Then they need a deck check for their own tired sake to ensure they didn't leave out something important. And would they still be able to change their list for the top 8 at that point?

All valid points. They should be able to change lists of course but the logistics would be more complicated. I guess it would just be another variable to consider for them the night before/the morning of. The time needs to be spent regardless - it either comes out of dinner with friends or pm/am sleep.

[Hayes]

Other considerations, though, are that the play in game should probably still be only one game, otherwise 4 full matches is a looooong day. Also, there needs to be some manner of radio silence to prevent the players with byes from getting a significant scouting advantage. I'm not anti-scouting, but when one player is having his games watched and the other isn't revealing his decks, that does grant a potentially significant advantage in terms of game planning.

Also would these Day 3 play-in games be done with Day 2 decks? I'm assuming yes. If that's the case there's less of an issue with scouting, but either way that can be somewhat mitigated by having the games done in a separate room.

For the record, I am a fan of play in games (and really, really liked the random pairings 1-4 vs 5-8). We all recognize that there is no perfect system, so I am in favor of whatever methodology diminishes the impact of factors that are outside of a players control.

(I.e. it's not my fault if I am randomly assigned the bye round 1 and my opponents in rounds 2 and 3 each finish 2-6. Similarly if I went 0-2 then won 6 games, I should not be upset that I might end up in a playoff game (or games) because I likely faced "easier" competition as a direct result of my poor start.)

I like this idea of expanding Day 3 (Sunday). Anyone that is involved in a playoff game (or games) has the same opportunity as everyone else to tweak or change their decks over night. Those game(s) just have to start at 9am Sunday. Effectively, this just grants a bye for anyone not involved (because they earned a higher placing after 8 games on Day 2).

[Death Star]

I'm very curious to hear from the four people who played in the play-in games about their opinions.

Me too They (along with a few others) will be receiving a survey shortly, although they can of course discuss it here as well.

I did not receive a survey (going 6-2 but fortunately not needing the play-in game). Did this ever take place?

To add my opinions cause I did not voice them immediately after worlds:

Loved the random pairings. Please keep those exactly like that for all further majors (at least worlds and europeans).

Not a fan of the play-in games and in favor of ditching those again. It’s a zero-sum game and takes away from time to decompress and get ready for day 3.