What's the purpose of an advantaged final?

[the return of AHAN]

This is a bit of leading question because it was motivated by a specific situation today where the pre-determined tournament format of prelims, playoffs, and super-playoffs, followed by an advantaged final, yielded the following final standings in 30-team field;

1.) Anytown A 12-2 (5-0 in prelims; 3-2 in playoffs)
2.) Bedford Falls A 12-2 (5-0 in prelims; 5-0 in playoffs)
3.) Pottersville A 10-2 (5-0 in prelims; 5-0 in playoffs)

So, my question REALLY is, what's the purpose of an advantaged final if this set of final standings is possibly on the table?

[Cheynem]

Well, the *purpose* of an advantaged final is to correctly ascertain the best team.

However, in most quizbowl contexts, if a team "clears" the field (2 games or more), there is no need for an advantaged final, so while I'm not 100% sure what happened in your tournament, this seems like a tournament where the rules/format are the issue more than the concept of an advantaged final in general.

[jonpin]

The purpose of an advantage final is to balance the benefit of a final that provides public display of the best teams and the crowning moment against the detriment of a team going X-0 all day, then losing a one-off final to a team that went (X-1)-1 who they had beaten earlier in the day.

Well, the *purpose* of an advantaged final is to correctly ascertain the best team.

However, in most quizbowl contexts, if a team "clears" the field (2 games or more), there is no need for an advantaged final, so while I'm not 100% sure what happened in your tournament, this seems like a tournament where the rules/format are the issue more than the concept of an advantaged final in general.

The best I can ascertain: In one of two parallel playoff groups, Bedford was 5-0, Anytown was second at 3-2. In the other group, Pottersville was 5-0. Then in the superplayoffs, Bedford and Anytown both beat Pottersville and the fourth team, and since Anytown dropped their second loss (as it came to a team that wasn't in the top four), Bedford was 3-0 in the super, Anytown was 2-1, Pottersville was 1-2, then Anytown won both games of the advantage final.
If the above is accurate (and of course with no allowance for the margins of these games), I think that's a fair result. Anytown won 2-of-3 from Bedford, and Pottersville lost their only game to the two teams above them. Is it "unfair" for some definition of that word that Anytown is not punished for suffering an upset loss? That is an argument QB can have and has had for the entirety of the 21st century, if not longer.

[the return of AHAN]

The purpose of an advantage final is to balance the benefit of a final that provides public display of the best teams and the crowning moment against the detriment of a team going X-0 all day, then losing a one-off final to a team that went (X-1)-1 who they had beaten earlier in the day.

Well, the *purpose* of an advantaged final is to correctly ascertain the best team.

However, in most quizbowl contexts, if a team "clears" the field (2 games or more), there is no need for an advantaged final, so while I'm not 100% sure what happened in your tournament, this seems like a tournament where the rules/format are the issue more than the concept of an advantaged final in general.

The best I can ascertain: In one of two parallel playoff groups, Bedford was 5-0, Anytown was second at 3-2. In the other group, Pottersville was 5-0. Then in the superplayoffs, Bedford and Anytown both beat Pottersville and the fourth team, and since Anytown dropped their second loss (as it came to a team that wasn't in the top four), Bedford was 3-0 in the super, Anytown was 2-1, Pottersville was 1-2, then Anytown won both games of the advantage final.
If the above is accurate (and of course with no allowance for the margins of these games), I think that's a fair result. Anytown won 2-of-3 from Bedford, and Pottersville lost their only game to the two teams above them. Is it "unfair" for some definition of that word that Anytown is not punished for suffering an upset loss? That is an argument QB can have and has had for the entirety of the 21st century, if not longer.

Right, this is what happened. It hit differently because, a year ago, with a smaller field initial field, playoffs consisted of one eight-team pool, where Anytown went 7-0, Bedford Falls went 5-2 (suffering losses to Anytown and another team that finished last in the pool), and Pottersville was 4-3. Anytown was awarded the championship, with no further matches necessary. A year ago, Bedford Falls was punished for losing to a team not in the top 4, but in their pool. This year, the roles were switched, but Anytown was offered a path back to the trophy via a disadvantaged final that wasn't there a year ago.

[the return of AHAN]

The purpose of an advantage final is to balance the benefit of a final that provides public display of the best teams and the crowning moment against the detriment of a team going X-0 all day, then losing a one-off final to a team that went (X-1)-1 who they had beaten earlier in the day.

And if the X-0 team is paired with a team that went (X-2)-2 in the playoffs, and team (X-2)-2 capitalizes on that free shot at improving their placement to now win the tournament with an identical X-2 record? How is that fair to the team that had gone X-0?
Moreover, I've been at this a while and it feels like I see the disadvantaged team win a lot more often than the 25% you'd expect to see if the teams were evenly matched and the outcome was a coin flip. I wonder if such records exist or could be easily collected.

[jonpin]

The purpose of an advantage final is to balance the benefit of a final that provides public display of the best teams and the crowning moment against the detriment of a team going X-0 all day, then losing a one-off final to a team that went (X-1)-1 who they had beaten earlier in the day.

And if the X-0 team is paired with a team that went (X-2)-2 in the playoffs, and team (X-2)-2 capitalizes on that free shot at improving their placement to now win the tournament with an identical X-2 record? How is that fair to the team that had gone X-0?
Moreover, I've been at this a while and it feels like I see the disadvantaged team win a lot more often than the 25% you'd expect to see if the teams were evenly matched and the outcome was a coin flip. I wonder if such records exist or could be easily collected.

I think that might be caused by your personal history and the 2022 nationals. Until and unless someone makes a concerted effort to collect that data, we have a pretty limited sample size.
In the modern NSC history (2010-present), there have been two no-finals, five one-games, and five advantage finals. Three were won on the first game (2015, 2017, 2021), one was won by the leader on the second game (2019), one was won by the trailer via both games (2022).
In the past 20 years of the HSNCT, there have been five one-game finals and 14 advantage finals. Six were won on the first game (03, 04, 05, 07, 10, 18), four won by the leader in the second game (06, 08, 09, 12), four won by the trailer via both games (11, 17, 21, 22).
This means at major HS nationals, 5 of 19 advantage finals were won by the second-place team, which is not appreciably different from 25%.

Does someone have this data organized for ICT, MSNCT, etc, or want to be the one who collects it?

[Cody]

The purpose of an advantage final is to balance the benefit of a final that provides public display of the best teams and the crowning moment against the detriment of a team going X-0 all day, then losing a one-off final to a team that went (X-1)-1 who they had beaten earlier in the day.

And if the X-0 team is paired with a team that went (X-2)-2 in the playoffs, and team (X-2)-2 capitalizes on that free shot at improving their placement to now win the tournament with an identical X-2 record? How is that fair to the team that had gone X-0?

It wouldn't be fair assuming the win-loss record was based on the same schedule. An advantaged final requires the team coming from behind to finish with a better record to fairly win the tournament, so in your scenario the team behind by 2 games would have to win 3 in a row. Otherwise, it's not an advantaged final.

The basic point of an advantaged final (and playoffs, etc.) is to avoid a single elimination schedule – one loss should not eliminate a team from winning the tournament. It seems like your main problem from the OP is not with an advantaged final but rather with the structure of a prelims → playoffs → superplayoffs schedule that reduced the number of games teams were compared on, which is largely unavoidable at large field, regular-season tournaments given the limited number of matches.

[the return of AHAN]

The purpose of an advantage final is to balance the benefit of a final that provides public display of the best teams and the crowning moment against the detriment of a team going X-0 all day, then losing a one-off final to a team that went (X-1)-1 who they had beaten earlier in the day.

And if the X-0 team is paired with a team that went (X-2)-2 in the playoffs, and team (X-2)-2 capitalizes on that free shot at improving their placement to now win the tournament with an identical X-2 record? How is that fair to the team that had gone X-0?

It wouldn't be fair assuming the win-loss record was based on the same schedule. An advantaged final requires the team coming from behind to finish with a better record to fairly win the tournament, so in your scenario the team behind by 2 games would have to win 3 in a row. Otherwise, it's not an advantaged final.

The basic point of an advantaged final (and playoffs, etc.) is to avoid a single elimination schedule – one loss should not eliminate a team from winning the tournament. It seems like your main problem from the OP is not with an advantaged final but rather with the structure of a prelims → playoffs → superplayoffs schedule that reduced the number of games teams were compared on, which is largely unavoidable at large field, regular-season tournaments given the limited number of matches.

Right, and through the playoffs and superplayoffs, Anytown and Bedford Falls played the exact same set of seven opponents.

[the return of AHAN]

The purpose of an advantage final is to balance the benefit of a final that provides public display of the best teams and the crowning moment against the detriment of a team going X-0 all day, then losing a one-off final to a team that went (X-1)-1 who they had beaten earlier in the day.

And if the X-0 team is paired with a team that went (X-2)-2 in the playoffs, and team (X-2)-2 capitalizes on that free shot at improving their placement to now win the tournament with an identical X-2 record? How is that fair to the team that had gone X-0?
Moreover, I've been at this a while and it feels like I see the disadvantaged team win a lot more often than the 25% you'd expect to see if the teams were evenly matched and the outcome was a coin flip. I wonder if such records exist or could be easily collected.

I think that might be caused by your personal history and the 2022 nationals. Until and unless someone makes a concerted effort to collect that data, we have a pretty limited sample size.
In the modern NSC history (2010-present), there have been two no-finals, five one-games, and five advantage finals. Three were won on the first game (2015, 2017, 2021), one was won by the leader on the second game (2019), one was won by the trailer via both games (2022).
In the past 20 years of the HSNCT, there have been five one-game finals and 14 advantage finals. Six were won on the first game (03, 04, 05, 07, 10, 18), four won by the leader in the second game (06, 08, 09, 12), four won by the trailer via both games (11, 17, 21, 22).
This means at major HS nationals, 5 of 19 advantage finals were won by the second-place team, which is not appreciably different from 25%.

Does someone have this data organized for ICT, MSNCT, etc, or want to be the one who collects it?

Thanks, and that's what I'd expect to see. When I review the stats for Illinois NAQT State since 2016, when the tourney began being held on SCT D2, there have been 5 advantaged finals, with the disadvantaged team winning all 5. One year the final two teams were both 12-0 and played a best of three, with Hinsdale Central A winning game 1, and Auburn A winning games 2 and 3. I don't know if I'd necessarily call that an advantaged final though, even if the same thing functionally occurred.

[Cody]

The purpose of an advantage final is to balance the benefit of a final that provides public display of the best teams and the crowning moment against the detriment of a team going X-0 all day, then losing a one-off final to a team that went (X-1)-1 who they had beaten earlier in the day.

And if the X-0 team is paired with a team that went (X-2)-2 in the playoffs, and team (X-2)-2 capitalizes on that free shot at improving their placement to now win the tournament with an identical X-2 record? How is that fair to the team that had gone X-0?

It wouldn't be fair assuming the win-loss record was based on the same schedule. An advantaged final requires the team coming from behind to finish with a better record to fairly win the tournament, so in your scenario the team behind by 2 games would have to win 3 in a row. Otherwise, it's not an advantaged final.

The basic point of an advantaged final (and playoffs, etc.) is to avoid a single elimination schedule – one loss should not eliminate a team from winning the tournament. It seems like your main problem from the OP is not with an advantaged final but rather with the structure of a prelims → playoffs → superplayoffs schedule that reduced the number of games teams were compared on, which is largely unavoidable at large field, regular-season tournaments given the limited number of matches.

Right, and through the playoffs and superplayoffs, Anytown and Bedford Falls played the exact same set of seven opponents.

But you can't isolate two teams out of four and compare them because you are ordering all the teams in the bracket at once. Anytown can't simultaneously be X-2 compared to Bedford and X-1 compared to Pottersville because then they would have two different records and it would be impossible to compare any of the teams. This is a fundamental limitation of a prelims → playoffs → superplayoffs schedule that reduces the comparable schedule to 3 games, not an advantaged final. The only way to avoid this situation is to devise a schedule with more comparable games (e.g. have a superplayoffs of top 6) which is only possible if you have more packets and time.

[the return of AHAN]

But you can't isolate two teams out of four and compare them because you are ordering all the teams in the bracket at once.

Says who? It's four teams. It's not an inscrutable number of permutations.

[Cody]

But you can't isolate two teams out of four and compare them because you are ordering all the teams in the bracket at once.

Says who? It's four teams. It's not an inscrutable number of permutations.

then do it standings first, then we can discuss generalizing.

[CPiGuy]

The purpose of an advantage final is to balance the benefit of a final that provides public display of the best teams and the crowning moment against the detriment of a team going X-0 all day, then losing a one-off final to a team that went (X-1)-1 who they had beaten earlier in the day.

And if the X-0 team is paired with a team that went (X-2)-2 in the playoffs, and team (X-2)-2 capitalizes on that free shot at improving their placement to now win the tournament with an identical X-2 record? How is that fair to the team that had gone X-0?
Moreover, I've been at this a while and it feels like I see the disadvantaged team win a lot more often than the 25% you'd expect to see if the teams were evenly matched and the outcome was a coin flip. I wonder if such records exist or could be easily collected.

I think that might be caused by your personal history and the 2022 nationals. Until and unless someone makes a concerted effort to collect that data, we have a pretty limited sample size.
In the modern NSC history (2010-present), there have been two no-finals, five one-games, and five advantage finals. Three were won on the first game (2015, 2017, 2021), one was won by the leader on the second game (2019), one was won by the trailer via both games (2022).
In the past 20 years of the HSNCT, there have been five one-game finals and 14 advantage finals. Six were won on the first game (03, 04, 05, 07, 10, 18), four won by the leader in the second game (06, 08, 09, 12), four won by the trailer via both games (11, 17, 21, 22).
This means at major HS nationals, 5 of 19 advantage finals were won by the second-place team, which is not appreciably different from 25%.

Does someone have this data organized for ICT, MSNCT, etc, or want to be the one who collects it?

Here's the data for ICT since 2004, because I can't really tell what the hell happened in finals before then.

D1:
8: Advantaged team wins in one game (2004, 2005, 2006, 2007*, 2008, 2011***, 2012, 2019)
3: Advantaged team wins in two games (2009*, 2018, 2021)
2: Disadvantaged team wins (2015, 2022)

D2:
9: Advantaged team wins in one game (2004, 2006, 2009, 2011*, 2012*, 2013*, 2014, 2015, 2019)
2: Advantaged team wins in two games (2008, 2021)
2: Disadvantaged team wins (2005**, 2017)

*: this team won an advantaged final despite clearing the field by 2 games or more because of old NAQT rules
**: this team won a disadvantaged final despite being down by 2 games or more because of old NAQT rules
***: Andy Watkins cheated but the final was still played so I'm counting it.

So disadvantaged teams won 4 out of 26 finals. However several of these finals wouldn't have happened under modern rules (and were probably significantly skewed towards the team with advantage because they were more than 2 games better than the next team and therefore almost certainly strongly favored to win even one game); including only finals which would have happened under current advantaged final rules, disadvantaged teams won 3 out of 20 finals (15%).

This suggests to me that the true expected win rate for disadvantaged teams is less than 25%. This makes sense to me because in general one would expect the team with advantage to be the better team (which is usually why they earned an advantage), although this is not always true (2022 Stanford was, statistically, better than Harvard despite being disadvantaged in the final. That they're also one of two D1 teams ever to overturn a disadvantage is probably related to this.)

[the return of AHAN]

The purpose of an advantage final is to balance the benefit of a final that provides public display of the best teams and the crowning moment against the detriment of a team going X-0 all day, then losing a one-off final to a team that went (X-1)-1 who they had beaten earlier in the day.

And if the X-0 team is paired with a team that went (X-2)-2 in the playoffs, and team (X-2)-2 capitalizes on that free shot at improving their placement to now win the tournament with an identical X-2 record? How is that fair to the team that had gone X-0?
Moreover, I've been at this a while and it feels like I see the disadvantaged team win a lot more often than the 25% you'd expect to see if the teams were evenly matched and the outcome was a coin flip. I wonder if such records exist or could be easily collected.

I think that might be caused by your personal history and the 2022 nationals. Until and unless someone makes a concerted effort to collect that data, we have a pretty limited sample size.
In the modern NSC history (2010-present), there have been two no-finals, five one-games, and five advantage finals. Three were won on the first game (2015, 2017, 2021), one was won by the leader on the second game (2019), one was won by the trailer via both games (2022).
In the past 20 years of the HSNCT, there have been five one-game finals and 14 advantage finals. Six were won on the first game (03, 04, 05, 07, 10, 18), four won by the leader in the second game (06, 08, 09, 12), four won by the trailer via both games (11, 17, 21, 22).
This means at major HS nationals, 5 of 19 advantage finals were won by the second-place team, which is not appreciably different from 25%.

Does someone have this data organized for ICT, MSNCT, etc, or want to be the one who collects it?

Here's the data for ICT since 2004, because I can't really tell what the hell happened in finals before then.

D1:
8: Advantaged team wins in one game (2004, 2005, 2006, 2007*, 2008, 2011***, 2012, 2019)
3: Advantaged team wins in two games (2009*, 2018, 2021)
2: Disadvantaged team wins (2015, 2022)

D2:
9: Advantaged team wins in one game (2004, 2006, 2009, 2011*, 2012*, 2013*, 2014, 2015, 2019)
2: Advantaged team wins in two games (2008, 2021)
2: Disadvantaged team wins (2005**, 2017)

*: this team won an advantaged final despite clearing the field by 2 games or more because of old NAQT rules
**: this team won a disadvantaged final despite being down by 2 games or more because of old NAQT rules
***: Andy Watkins cheated but the final was still played so I'm counting it.

So disadvantaged teams won 4 out of 26 finals. However several of these finals wouldn't have happened under modern rules (and were probably significantly skewed towards the team with advantage because they were more than 2 games better than the next team and therefore almost certainly strongly favored to win even one game); including only finals which would have happened under current advantaged final rules, disadvantaged teams won 3 out of 20 finals (15%).

This suggests to me that the true expected win rate for disadvantaged teams is less than 25%. This makes sense to me because in general one would expect the team with advantage to be the better team (which is usually why they earned an advantage), although this is not always true (2022 Stanford was, statistically, better than Harvard despite being disadvantaged in the final. That they're also one of two D1 teams ever to overturn a disadvantage is probably related to this.)

I wonder if HS quizbowl national or state levels championships are more likely to have multiple, high level teams that are closer to a coin flip than college championships. Or if there's something about high schoolers' psyches that they're more likely to fumble the ball when the pressure is on. Given that line of thought, here is the MSNCT results since its 2011 inception;

1: Advantaged team wins in one game (2018)**
3: Advantaged team wins in two games (2012, 2013, 2017)
0: Disadvantaged team wins
7: one game final; (2011, 2014, 2015, 2016, 2019, 2021, 2022)
** - Can you call it a team when it was a one-man team, Shiva Oswal?!? I'll never stop being amazed with that outcome.

WOW! I didn't realize advantaged finals were so rare at MSNCT! But it's just as interesting there has yet to be someone who has overcome the disadvantage.

[CPiGuy]

I wonder if HS quizbowl national or state levels championships are more likely to have multiple, high level teams that are closer to a coin flip than college championships. Or if there's something about high schoolers' psyches that they're more likely to fumble the ball when the pressure is on. Given that line of thought, here is the MSNCT results since its 2011 inception;

I suspect it's just more likely that at nationals (or state tournaments in strong states like Illinois) the teams are more likely to be evenly matched, whereas at local tournaments there is frequently one team which is vastly stronger (I think most advantaged finals I ran during my time at Michigan were won in one game, because it was usually DCC vs whoever the second best team was, which DCC usually won handily).