I was just wondering how the knockout rounds would look if the seeding was done like the Conmebol Copa Libertadores (and to the smaller extent, the Concacaf Champions League), where the group winners are seeded 1-8, the group runners-up are seeded 9-16, and the fixed-bracket is created from the seeds. Each set (winners, and runners-up) are seeded via: (1) points (2) goal difference (3) goals for (4) fewer red cards (5) fewer yellow cards So the seeding tables come out as: SeedWinnersTeamsPtsGFGAGDYCRC 1Group BNetherlands91037 2Group CColombia9927 3Group FArgentina9633 4Group HBelgium9413 5Group EFrance7826 6Group GGermany772510 7Group ABrazil772540 8Group DCosta Rica7413 SeedRunners-UpTeamsPtsGFGAGD 9Group AMexico7413 10Group BChile6532 11Group ESwitzerland6761 12Group DUruguay6440 13Group HAlgeria4651 14Group GUSA4440 15Group FNigeria4330 16Group CGreece424-2 ...creating the knockout rounds as: Round-16 #1 Netherlands [v] #16 Greece #2 Colombia [v] #15 Nigeria #3 Argentina [v] #14 USA #4 Belgium [v] #13 Algeria (note: teams are from the same group; this has to be allowed unless you put in complex rules to bypass this case) #5 France [v] #12 Uruguay #6 Germany [v] #11 Switzerland #7 Brazil [v] #10 Chile #8 Costa Rica [v] #9 Mexico One more note on how the R-16: obviously, the actual matchups would not be known until all of the groups finish play, but each group winner would be know on each group's final matchday. So to put some clarity in the sites used in the R-16, you say that group A winner plays at city 1, group B winner plays at city 2, and so on, so that the group winners' travels would at least be known. Quarterfinals (#1 Netherlands -v- #16 Greece) [v] (#8 Costa Rica -v- #9 Mexico) (#2 Colombia -v- #15 Nigeria) [v] (#7 Brazil -v- #10 Chile) (#3 Argentina -v- #14 USA) [v] (#6 Germany -v- #11 Switzerland) (#4 Belgium -v- #13 Algeria) [v] (#5 France -v- #12 Uruguay) Semifinals (#1 Netherlands -v- #16 Greece -v- #8 Costa Rica -v- #9 Mexico) [v] (#4 Belgium -v- #13 Algeria -v- #5 France -v- #12 Uruguay) (#2 Colombia -v- #15 Nigeria -v- #7 Brazil -v- #10 Chile) [v] (#3 Argentina -v- #14 USA -v- #6 Germany -v- #11 Switzerland) third / Final (#1 Netherlands -v- #16 Greece -v- #8 Costa Rica -v- #9 Mexico -v- #4 Belgium -v- #13 Algeria -v- #5 France -v- #12 Uruguay) [v] (#2 Colombia -v- #15 Nigeria -v- #7 Brazil -v- #10 Chile -v- #3 Argentina -v- #14 USA -v- #6 Germany -v- #11 Switzerland)
I like the idea, but the concern you mentioned regarding teams not knowing where they would go until the last match would probably make it impossible. Today teams, and most importantly fans, only have a couple of routes they could possibly follow to the final, if this branches into 8 different possibilities, it will create a planning havoc in such a short competition. In this case, I believe the cons outweight the pros.
Instead of creating new threads in each tournament, might as well append to this one. Here are what the knockout rounds would look like for the Copa American Centenario: SeedGrpTeamPtsGDGFGA 1DArgentina9 2CMexico7+43+2+1=61+0+1=2 3BPeru7+21+2+1=40+2+0=2 4AUSA6 5CVenezuela7 6DChile6+21+2+4=72+1+2=5 7AColombia6+22+2+2=60+1+3=4 8BEcuador5 Theoretical knockout rounds: QF: #1 Argentina -v- #8 Ecuador #2 Mexico -v- #7 Colombia #3 Peru -v- #6 Chile #4 USA -v- #5 Venezuela SF: [#1 Argentina -v- #8 Ecuador] -v- [#4 USA -v- #5 Venezuela] [#2 Mexico -v- #7 Colombia] -v- [#3 Peru -v- #6 Chile] F: [#1 Argentina -v- #8 Ecuador -v- #4 USA -v- #5 Venezuela] -v- [#2 Mexico -v- #7 Colombia -v- #3 Peru -v- #6 Chile] Interesting that both SF rounds would have had all four groups. Actual knockout rounds (my seedings listed) QF: #4 USA -v- #8 Ecuador #1 Argentina -v- #5 Venezuela #3 Peru -v- #7 Colombia #2 Mexico -v- #6 Chile The interesting fact here is that all four matches have seeds (n) -v- (n+4): 4-v-8, 1-v-5, ... SF: [#4 USA -v- #8 Ecuador] -v- [#1 Argentina -v- #5 Venezuela] [#3 Peru -v- #7 Colombia] -v- [#2 Mexico -v- #6 Chile] Both SF's have the traditional seeds of 1-4-5-8 and 2-3-6-7, albeit in different orderings. (So if all four group winners would have won the QF's, it would have been the standard 1-v-4 and 2-v-3 in the SF's.) F: [#4 USA -v- #8 Ecuador -v- #1 Argentina -v- #5 Venezuela] -v- [#3 Peru -v- #7 Colombia -v- #2 Mexico -v- #6 Chile] * Coming on Wednesday: the hypothetical final-16 bracket of UEFA.
The obvious problem is scheduling. Even if you follow what you have suggested for the Group winners the runners-up will have wildly different amounts of rest. Assuming for simplicity that the Group A winner is ranked 1, Group B 2 etc, if the Group H runner-up finished ranked 16th (ie, worst runner-up) they have to play their round of 16 match the day after their third group match, while (if the Group A runner-up finished ranked 9th) another might have 6 rest days (3 3rd-match days plus the first 3 R16 days) You could have a break in the tournament to prevent the 0 rest days thing, but no-one wants the thing extended for another few days just on the chance that the draw might be weird or you could have some sorts of limits, but no matter how you do it you are going to run the risk of this problem. J
True that the format is not really perfect for a one-month tournament (as opposed to a Champions League where rounds have plenty of time between them), but the main point is to give some incentive to have teams play out the full 90 minutes of each game. (So, if you win group C, in the actual tournament your place in the R-16 is fixed, where in my form you have to try to keep scoring to increase your seed.) So the table (do not have to go past the GF tie-breaker in any case): sdgpteamtmfinptgdgf 1AFranceFRA1734 2CGermanyGER1733 3DCroatiaCRO1725 4BWalesWAL1636 5EItalyITA1623 6FHungaryHUN1526 7CPolandPOL2722 8DSpainESP2635 9EBelgiumBEL2624 10FIcelandISL2514 11BEnglandENG2513 12ASwitzerlandSUI2512 13BSlovakiaSVK3403 14ERepublic of IrelandIRL34-22 15FPortugalPOR3304 16CNorthern IrelandNIR3302 DTurkeyTUR33-22 AAlbaniaALB33-21 ARomaniaROU41-22 ESwedenSWE41-21 DCzech RepublicCZE41-32 FAustriaAUT41-31 BRussiaRUS41-42 CUkraineUKR40-50 R-16: #1 France -v- #16 Northern Ireland #2 Germany -v- #15 Portugal #3 Croatia -v- #14 Republic of Ireland #4 Wales -v- #13 Slovakia #5 Italy -v- #12 Switzerland #6 Hungary -v- #11 England #7 Poland -v- #10 Iceland #8 Spain -v- #9 Belgium The 4-v-13 matchup is from the same group, but it is a little difficult to completely remove the case in a best-third-place-advance tournament, so you would either accept this or put in complex rules to counter it. QF: [#1 FRA -v- #16 NIR] -v- [#8 ESP -v- #9 BEL] [#2 GER -v- #15 POR] -v- [#7 POL -v- #10 ISL] [#3 CRO -v- #14 IRL] -v- [#6 HUN -v- #11 ENG] [#4 WAL -v- #13 SVK] -v- [#5 ITA -v- #12 SUI] Notice how the top countries are well-distributed throughout the four QF matches, as opposed to the actual tournament which has one "heavy" bracket and one "lighter" bracket SF: [#1 FRA -v- #16 NIR -v- #8 ESP -v- #9 BEL] -v- [#4 WAL -v- #13 SVK -v- #5 ITA -v- #12 SUI] [#2 GER -v- #15 POR -v- #7 POL -v- #10 ISL] -v- [#3 CRO -v- #14 IRL -v- #6 HUN -v- #11 ENG] F: [#1 FRA -v- #16 NIR -v- #8 ESP -v- #9 BEL -v- #4 WAL -v- #13 SVK -v- #5 ITA -v- #12 SUI] -v- [#2 GER -v- #15 POR -v- #7 POL -v- #10 ISL -v- #3 CRO -v- #14 IRL -v- #6 HUN -v- #11 ENG]
True. It's not something that bothers the Libertadores (I think, I could be wrong) but rules could presumbly be introduced if this was a big problem. You're sort of deluding yourself of your own cleverness with this remark. Because the teams didn't play to maximise their placement (because 1st in the group was all that mattered - and even then the way that Spain played in their third match led many to claim it was only advancement that mattered) it's a bit rich to then claim that using these results gives you a "better" outcome. If they had played to maximise results it's completely conceivable that the same "lopsided" draw would have resulted - there's nothing in the formulation that really addresses this, particularly given that a Group of Death situation is much more likely to affect the teams' final rankings than their actual effort levels. J
I'm not doing any kind of boasting in this respect. I am simply using the hard numbers, as they occurred, and see that it (in this case, at least) it generates a much better bracket than the pre-determined bracket. (And of course, I'm not the only one who pointed out the immense disparity in the two bracket halves.) This would be similar to saying, "if a win was worth four points, then the standings would look like...." You couldn't speculate whether it would have changed teams' strategies during the season; you could only state how things shook out using the raw data. I've seen other comments (in individual ESPN articles, during the knockouts) saying that the top 4 group winners should have played the four third-place teams, so it just isn't me thinking this.