This system has been proposed a couple of times before in this thread. I actually thought it was clear from the beginning that they would use this method before other people suggested option 1 and 2.elkjiaer is back wrote: ↑Tue Jun 11, 2024 09:30I am glad you like my idea and actually surprised havent´seen it mentioned anywhere so far. But this is the only realistic way to perform the draw, limiting the number of constrains to be checked and still keeping the excitment for the audience (i mean scenario 3 of course)greenbay wrote: ↑Tue Jun 11, 2024 08:51 While playing around a little with your option 2, I'd say this is off the table. I'm just a hobby coder, not a software engineer, but I wonder if someone will really have an idea how to prevent deadlocks by fail safe code, not just but heuristic try and error.
Option 1 is basically just the same as option 2. You need to prevent deadlocks by code, not just by heuristic try and error.
Therefore option 3. And the longer I think about it, the more attractive this option looks to me. Why not publishing a fixture list before the draw? MD1: A1-A2, B1-C3, D4-A2 and so on. MD2: A2-D3, B2-B1 etc... UEFA can make one up to their liking. Making sure that i.e. C3 play away on MD1 and MD 7, while playing home on MD2 and MD8.
Then we could have an actual draw. Not some computer generated stuff leading to conspiracy theories by the dozen. Pick a ball from the 36 teams, say some time during the draw it's now Bayer Leverkusen's turn, from pot B, formerly known as pot 2. A quick computer check to which spots they can go, like before there was a computer check to which groups they can go, computer says it's B1, B2, B3, B5, B6 and B9. Then place the corresponding balls into a pot, and let an actual person do the draw, grabbing B1 from the pot, so we know that Bayer Leverkusen play C3 on MD1 at home, with C3 earlier been allocated to i.e. AC Milan. While playing B2, not allocated yet, away on MD2.
For sure this will not take 4 hours. It will definitely take far less than a minute per team to just grab some ball. So the whole draw can be done is half an hour or less.
They could just show on a screen the 8 matchdays and every time a team is set to specific slot (A1, C3, D9 ETC) then it is highlited so it is easy to see all its opponents in the 8 matchdays
Draw System for League Phases of European Cups 2024 onwards
-
elkjiaer is back
- Posts: 279
- Joined: Fri Jul 28, 2017 09:08
After a lot of trials I finally managed to generate the 144 fixtures split among the 8 matchdays . In general it doesn’t seem possibile to strictly fullfill all requirements regarding calendar order . This is why Uefa in the regulations also says “ in principle “ . It looks like we can’t guarantee a maximum of 2 home or away days in a row for all teams . Beside that I now have a template where team names are replaced by a code (A1 to A9 for pot 1 teams , B1 to B9 for pot 2 etc )
Whenever I get the time I will try to play around with the draw to see how team allocation in the different “code” system can work . If anyone of you has good programming skills I can of course share the calendar template
Whenever I get the time I will try to play around with the draw to see how team allocation in the different “code” system can work . If anyone of you has good programming skills I can of course share the calendar template
Please share it in a Google Sheet here!elkjiaer is back wrote: ↑Wed Jun 12, 2024 16:18 After a lot of trials I finally managed to generate the 144 fixtures split among the 8 matchdays . In general it doesn’t seem possibile to strictly fullfill all requirements regarding calendar order . This is why Uefa in the regulations also says “ in principle “ . It looks like we can’t guarantee a maximum of 2 home or away days in a row for all teams . Beside that I now have a template where team names are replaced by a code (A1 to A9 for pot 1 teams , B1 to B9 for pot 2 etc )
Whenever I get the time I will try to play around with the draw to see how team allocation in the different “code” system can work . If anyone of you has good programming skills I can of course share the calendar template
If anyone is interested to create a simulator for the draw and has good programming skills may contact (either he or at twitter) with me to assist him create one. I have found one was but my computer is old enough and requires much more time to run and test it during the development and I don't have such time.
The following "random thoughts" are based on the assumption that all higher-seeds will win the qualifiers.
Step 1:
Pair two teams each from the same country but from different pots as long as possible (by random draw)
Internazionale ITA 1 + AC Milan ITA 3
Borussia Dortmund GER 1 + VfB Stuttgart GER 4
Atalanta ITA 2 + Bologna ITA 4
Paris Saint-Germain FRA 1 + AS Monaco FRA 4
Manchester City ENG 1 + Arsenal ENG 2
Glasgow Rangers SCO 2 + Celtic SCO 4
Bayern München GER 1 + Bayer Leverkusen GER 2
Atletico Madrid ESP 2 + FC Barcelona ESP 1
Girona FC ESP 4 + Real Madrid ESP 1
Liverpool ENG 1 + Aston Villa ENG 4
Benfica POR 2 + Sporting CP Lisbon POR 3
Step 2:
Pair two remaining teams each from different pots (by random draw).
Juventus ITA 2 + Feyenoord NED 3
Sturm Graz AUT 4 + Dinamo Zagreb CRO 3
Maccabi Tel-Aviv ISR 3 + RB Leipzig GER 1
Ferencvaros HUN 4 + PAOK Thessaloniki GRE 3
Stade Brestois FRA 4 + PSV Eindhoven NED 3
Club Brugge BEL 2 + Slavia Praha CZE 3
Shaktar Donetsk UKR 2 + Red Star Belgrade SRB 3
Step 3:
Form 9 pools by again pairing 2 pairs each from above, steps 1+2, containing one team from every pot, max two teams from the same country in one pool (by random draw).
Atalanta ITA 2 + Bologna ITA 4 + Maccabi Tel-Aviv ISR 3 + RB Leipzig GER 1
Manchester City ENG 1 + Arsenal ENG 2 + Sturm Graz AUT 4 + Dinamo Zagreb CRO 3
Juventus ITA 2 + Feyenoord NED 3 + Paris Saint-Germain FRA 1 + AS Monaco FRA 4
Benfica POR 2 + Sporting CP Lisbon POR 3 + Girona FC ESP 4 + Real Madrid ESP 1
Club Brugge BEL 2 + Slavia Praha CZE 3 + Borussia Dortmund GER 1 + VfB Stuttgart GER 4
Shaktar Donetsk UKR 2 + Red Star Belgrade SRB 3 + Liverpool ENG 1 + Aston Villa ENG 4
Bayern München GER 1 + Bayer Leverkusen GER 2 + Stade Brestois FRA 4 + PSV Eindhoven NED 3
Atletico Madrid ESP 2 + FC Barcelona ESP 1 + Ferencvaros HUN 4 + PAOK Thessaloniki GRE 3
Internazionale ITA 1 + AC Milan ITA 3 + Glasgow Rangers SCO 2 + Celtic SCO 4
Step 4:
For every team draw 8 opponents, one each from every other pool except for its own pool. Of course making sure, 2 from each pot. This automatically makes sure that all* teams can face max two opponents from the same country.
*Only one exception (for the 24-25 season): Germany with three teams in pot 1 causing trouble, being allocated to three pools in this scenario. So it's the only deadlock to prevent, that some Non-German happens to draw a German team from all of those three pools.
Edit: Even this can be circumvented: Just make an extra ruleset (for the 24-25 season):
a) First team to draw a German side from pot 1, automatically gets a Non-German team allocated from the "Leverkusen pool" or the "Stuttgart pool".
or (whatever comes first)
b) First team to draw a non-German side from pot 1, automatically gets Leverkusen or Stuttgart on their dance card.
So it will be like this:
UEFA will draw a ball from the pot containing all 36 teams. UEFA will then compute their 8 opponents, my guess at least. First opponent to be assigned must be a pot 1 opponent. Extra rule a) or b) will automatically be carried out. No more "max 2 per country" to check for all of the remaining draw. Computing time for valid draws is now only a marginal fraction of what it used to be without "pairing and pooling".
Step 1:
Pair two teams each from the same country but from different pots as long as possible (by random draw)
Internazionale ITA 1 + AC Milan ITA 3
Borussia Dortmund GER 1 + VfB Stuttgart GER 4
Atalanta ITA 2 + Bologna ITA 4
Paris Saint-Germain FRA 1 + AS Monaco FRA 4
Manchester City ENG 1 + Arsenal ENG 2
Glasgow Rangers SCO 2 + Celtic SCO 4
Bayern München GER 1 + Bayer Leverkusen GER 2
Atletico Madrid ESP 2 + FC Barcelona ESP 1
Girona FC ESP 4 + Real Madrid ESP 1
Liverpool ENG 1 + Aston Villa ENG 4
Benfica POR 2 + Sporting CP Lisbon POR 3
Step 2:
Pair two remaining teams each from different pots (by random draw).
Juventus ITA 2 + Feyenoord NED 3
Sturm Graz AUT 4 + Dinamo Zagreb CRO 3
Maccabi Tel-Aviv ISR 3 + RB Leipzig GER 1
Ferencvaros HUN 4 + PAOK Thessaloniki GRE 3
Stade Brestois FRA 4 + PSV Eindhoven NED 3
Club Brugge BEL 2 + Slavia Praha CZE 3
Shaktar Donetsk UKR 2 + Red Star Belgrade SRB 3
Step 3:
Form 9 pools by again pairing 2 pairs each from above, steps 1+2, containing one team from every pot, max two teams from the same country in one pool (by random draw).
Atalanta ITA 2 + Bologna ITA 4 + Maccabi Tel-Aviv ISR 3 + RB Leipzig GER 1
Manchester City ENG 1 + Arsenal ENG 2 + Sturm Graz AUT 4 + Dinamo Zagreb CRO 3
Juventus ITA 2 + Feyenoord NED 3 + Paris Saint-Germain FRA 1 + AS Monaco FRA 4
Benfica POR 2 + Sporting CP Lisbon POR 3 + Girona FC ESP 4 + Real Madrid ESP 1
Club Brugge BEL 2 + Slavia Praha CZE 3 + Borussia Dortmund GER 1 + VfB Stuttgart GER 4
Shaktar Donetsk UKR 2 + Red Star Belgrade SRB 3 + Liverpool ENG 1 + Aston Villa ENG 4
Bayern München GER 1 + Bayer Leverkusen GER 2 + Stade Brestois FRA 4 + PSV Eindhoven NED 3
Atletico Madrid ESP 2 + FC Barcelona ESP 1 + Ferencvaros HUN 4 + PAOK Thessaloniki GRE 3
Internazionale ITA 1 + AC Milan ITA 3 + Glasgow Rangers SCO 2 + Celtic SCO 4
Step 4:
For every team draw 8 opponents, one each from every other pool except for its own pool. Of course making sure, 2 from each pot. This automatically makes sure that all* teams can face max two opponents from the same country.
*Only one exception (for the 24-25 season): Germany with three teams in pot 1 causing trouble, being allocated to three pools in this scenario. So it's the only deadlock to prevent, that some Non-German happens to draw a German team from all of those three pools.
Edit: Even this can be circumvented: Just make an extra ruleset (for the 24-25 season):
a) First team to draw a German side from pot 1, automatically gets a Non-German team allocated from the "Leverkusen pool" or the "Stuttgart pool".
or (whatever comes first)
b) First team to draw a non-German side from pot 1, automatically gets Leverkusen or Stuttgart on their dance card.
So it will be like this:
UEFA will draw a ball from the pot containing all 36 teams. UEFA will then compute their 8 opponents, my guess at least. First opponent to be assigned must be a pot 1 opponent. Extra rule a) or b) will automatically be carried out. No more "max 2 per country" to check for all of the remaining draw. Computing time for valid draws is now only a marginal fraction of what it used to be without "pairing and pooling".
"Put it in your signature to save you the trouble of writing it over and over again."
And another random thought, part 427:
Deadlocks in the later stages of the draw, and therefore computing time, can be reduced by implementing a more severe "max 2 opponents from the same country rule". The idea behind all this: Why waste a "second country slot" too early?
Say Liverpool have already drawn Inter and Bayern from pot 1. Now they need opposition from pot 2. If they are assigned either Atalanta or Juventus and then Leverkusen on top, there are a bunch of invalid draws for Liverpool now in the way. So let's make the "max two" rule more severe. Kind of "max one". Teams are allowed a second opponent from the same country only and only if there is no other valid draw.
So in the Liverpool sceario as before, Liverpool cannot draw Atalanta or Juventus or Leverkusen from pot 2. So they get maybe Benfica. Way less deadlocks now in the way for Liverpool further down the road. So once they get down to pot 4, Bologna and Stuttgart are very likely still in the mix, increasing the number of possible opponent, therefore preventing a deadlock.
Deadlocks in the later stages of the draw, and therefore computing time, can be reduced by implementing a more severe "max 2 opponents from the same country rule". The idea behind all this: Why waste a "second country slot" too early?
Say Liverpool have already drawn Inter and Bayern from pot 1. Now they need opposition from pot 2. If they are assigned either Atalanta or Juventus and then Leverkusen on top, there are a bunch of invalid draws for Liverpool now in the way. So let's make the "max two" rule more severe. Kind of "max one". Teams are allowed a second opponent from the same country only and only if there is no other valid draw.
So in the Liverpool sceario as before, Liverpool cannot draw Atalanta or Juventus or Leverkusen from pot 2. So they get maybe Benfica. Way less deadlocks now in the way for Liverpool further down the road. So once they get down to pot 4, Bologna and Stuttgart are very likely still in the mix, increasing the number of possible opponent, therefore preventing a deadlock.
"Put it in your signature to save you the trouble of writing it over and over again."
This seems like something that will be solved by AI by next year tbh...
There is no need for AI here as the (average) human brain might be to weak for coming up with a solution. It's about reducing the computing power necessary on one hand, as UEFA have created rules for a draw that has a gazillion possible outcomes of which via a completely random draw the majority are in violation of such rules. And on the other hand it's about having a real draw, so balls manually picked by actual people, not some computations in the back, leading to conspiracy theories, if say Real get a draw against a bunch of foes from non top-coutries, while Barca have a schedule packed with nearly every opponent coming from a top-5 country. As mentioned before, all this can be done manually. It's just about reducing computing power for not having to check a gazillion possible outcomes but only a million or so before each ball is drawn. And about reducing TV time needed for a manual draw as picking a couple of hundred balls manually would take way too long, so we need to get it down a few dozen as it has always been before. That's why (at least me) prefers the so called "option 3". No need for a computer draw and it's only 36 balls to pick from. Computations in the back are reduced to a minimum, only for to check, if i.e. Real can go to A2 if Barca were already drawn in A1, stuff like what we had before with "red" groups and "blue" groups. But as in option 3 the schedule is being publishing before the draw, everything is transparent. Only remaining question is about the Tuesday/Wednesday schedule and alternating home/away games. Or to be more precise, reducing the computing time need therefore to prevent deadlocks.
"Put it in your signature to save you the trouble of writing it over and over again."
Which is of course a priority.
I guess that the majority agrees with you here, at least I do, for what I understood from this thread.
-
elkjiaer is back
- Posts: 279
- Joined: Fri Jul 28, 2017 09:08
Here you are (just a bit in a hurry so writing them here instead).Stadion wrote: ↑Thu Jun 13, 2024 11:44Please share it in a Google Sheet here!elkjiaer is back wrote: ↑Wed Jun 12, 2024 16:18 After a lot of trials I finally managed to generate the 144 fixtures split among the 8 matchdays . In general it doesn’t seem possibile to strictly fullfill all requirements regarding calendar order . This is why Uefa in the regulations also says “ in principle “ . It looks like we can’t guarantee a maximum of 2 home or away days in a row for all teams . Beside that I now have a template where team names are replaced by a code (A1 to A9 for pot 1 teams , B1 to B9 for pot 2 etc )
Whenever I get the time I will try to play around with the draw to see how team allocation in the different “code” system can work . If anyone of you has good programming skills I can of course share the calendar template
The following criteria were applied (if anyone has the type you can indeed check if they were fulfilled):
a )The home and away matches for each team alternate correctly between matchdays 1 and 2, as well as between matchdays 7 and 8.
b) For each team, there is at least one and at most two home matches in the matchday range 1-3 (inclusive).
c) For each team, there is at least one and at most two home matches in the matchday range 4-6 (inclusive).
d) Matchtypes (AA, AB, BC etc, meaning for example 2 teams from pot A, or one from pot A vs one from pot B etc) maxiumum 2 fixtures per matchday involving 2 teams from the same pots (AA; BB;CC;DD) . This is not a requirement by UEFA, but i suppose they want to spread those out evenly to maximize the interest and avoiding too many big matches on the same matchday.
Suggestions are welcome!
Matchday 1:
A2-D29
A4-A5
A8-A9
B10-C19
B11-B12
B14-A6
B15-A7
B16-B17
B18-A1
C20-A3
C21-C22
C24-C25
D30-B13
D31-C23
D32-D33
D34-C26
D35-C27
D36-D28
Matchday 2:
A1-A2
A3-D30
A5-D32
A6-C24
A7-B16
A9-D36
B12-A4
B13-B14
B17-D35
C19-C20
C22-D31
C23-B15
C25-A8
C26-B18
C27-B10
D28-B11
D29-C21
D33-D34
Matchday 3:
A3-B12
A5-C23
A6-A7
B10-A2
B11-D29
B13-D31
B17-B18
C19-D28
C20-C21
C22-B14
C24-B16
C26-A9
C27-A1
D30-A4
D32-B15
D33-C25
D34-A8
D35-D36
Matchday 4:
A2-B11
A3-C21
A4-D31
A6-D33
A8-C26
A9-A1
B13-A5
B14-B15
B16-D34
C20-B12
C23-D32
C24-A7
C25-B17
C27-C19
D28-D29
D30-C22
D35-B18
D36-B10
Matchday 5:
A1-B10
A5-A6
A7-D34
A8-B17
A9-C27
B11-A3
B12-B13
B15-D33
B16-C25
B18-D36
C19-A2
C21-A4
C22-C23
C26-D35
D28-C20
D29-D30
D31-B14
D32-C24
Matchday 6:
A1-D28
A2-C20
A4-C22
A7-A8
B10-B11
B12-D30
B14-D32
B15-C24
B17-A9
B18-C27
C21-B13
C23-A6
C25-C26
D29-A3
D31-A5
D33-B16
D34-D35
D36-C19
Matchday 7:
A2-A3
A4-B13
A7-C25
A8-D35
A9-B18
B10-D28
B12-C21
B14-C23
B15-B16
C19-B11
C20-D29
C22-A5
C24-D33
C26-C27
D30-D31
D32-A6
D34-B17
D36-A1
Matchday 8:
A1-C19
A3-A4
A5-B14
A6-B15
B11-C20
B13-C22
B16-A8
B17-C26
B18-B10
C21-D30
C23-C24
C25-D34
C27-D36
D28-A2
D29-B12
D31-D32
D33-A7
D35-A9
Random thoughts, volume 653:
I was just looking up the tiebreakers in case two teams are tied after group stage but haven‘t had a direct encounter or the H2H ended in a draw. Amongst other tiebreakers, UEFA will introduce the concept of common games. A tiebreaker well known (at least to me) from the NFL.
Common games explained for non NFL fans: if two teams are tied after the regular season and if there is no H2H winner, they will look for common opponents both tied teams have played. Tiebreaker then goes to the team with the better record against these opponents only.
So I started to wonder: will the rules care for a maximum number of common opponents? Or is it theoretically possible that two teams can face the exact same eight opponents?
I was just looking up the tiebreakers in case two teams are tied after group stage but haven‘t had a direct encounter or the H2H ended in a draw. Amongst other tiebreakers, UEFA will introduce the concept of common games. A tiebreaker well known (at least to me) from the NFL.
Common games explained for non NFL fans: if two teams are tied after the regular season and if there is no H2H winner, they will look for common opponents both tied teams have played. Tiebreaker then goes to the team with the better record against these opponents only.
So I started to wonder: will the rules care for a maximum number of common opponents? Or is it theoretically possible that two teams can face the exact same eight opponents?
"Put it in your signature to save you the trouble of writing it over and over again."
Couldn't the common games tiebreaker be a little unfair though?greenbay wrote: ↑Fri Jun 14, 2024 15:06 Random thoughts, volume 653:
I was just looking up the tiebreakers in case two teams are tied after group stage but haven‘t had a direct encounter or the H2H ended in a draw. Amongst other tiebreakers, UEFA will introduce the concept of common games. A tiebreaker well known (at least to me) from the NFL.
Common games explained for non NFL fans: if two teams are tied after the regular season and if there is no H2H winner, they will look for common opponents both tied teams have played. Tiebreaker then goes to the team with the better record against these opponents only.
So I started to wonder: will the rules care for a maximum number of common opponents? Or is it theoretically possible that two teams can face the exact same eight opponents?
It might be possible for example that the teams in question have 3 common opponents and that one team has played them all at home and the other has played them all away. I don't know how far down the tiebreaker list this condition is, and therefore how unlikely that it will be that this is used to separate the teams in question.
Good point. But UEFA obviously thinks otherwise.
Which imho presents us with a bigger unfairness: Assume that Bayer Leverkusen and FC Barcelona are tied for #8 in the final standings, therefore one goes automatically to R16, the other must qualify via the extra playoffs. Say, Leverkusen even won the H2H by 1-0 in Nou Camp. Doesn't matter. Barcelona grabs the #8 spot because there was one common game: Barca beat Graz 4-0 at home, while Leverkusen only won 3-1 away. So imho UEFA got it wrong here. No more H2H because it's considered unfair with only one side having had a home game. But then going for common games, a far worse tiebreaker than H2H, but then not caring at all for the home/away situation.
Interestingly there is no more H2H. Maybe with regards to H2H they thought it would be unfair as there is only a single game between them on home soil for a tied team.If two or more teams are equal on points on completion of the league phase matches, the following criteria are applied, in this order, to determine their rankings:
Superior goal difference in the league phase
Higher number of goals scored in the league phase
Higher number of away goals scored in the league phase
Higher number of wins in the league phase
Higher number of away wins in the league phase
Higher number of points obtained collectively by league phase opponents
Superior collective goal difference of league phase opponents
Higher number of goals scored collectively by league phase opponents
Lower disciplinary points total based only on yellow and red cards received by players and team officials in all league phase matches (red card = 3 points, yellow card = 1 point, expulsion for two yellow cards in one match = 3 points)
Higher club coefficient
Which imho presents us with a bigger unfairness: Assume that Bayer Leverkusen and FC Barcelona are tied for #8 in the final standings, therefore one goes automatically to R16, the other must qualify via the extra playoffs. Say, Leverkusen even won the H2H by 1-0 in Nou Camp. Doesn't matter. Barcelona grabs the #8 spot because there was one common game: Barca beat Graz 4-0 at home, while Leverkusen only won 3-1 away. So imho UEFA got it wrong here. No more H2H because it's considered unfair with only one side having had a home game. But then going for common games, a far worse tiebreaker than H2H, but then not caring at all for the home/away situation.
"Put it in your signature to save you the trouble of writing it over and over again."
I think that the issue is also what if 3 clubs (A, B, C) ends up with the same number of points, and two of them played eachother (A-B) but the third one (C) did not play an head to head; or if A-B happened, A-C happened, but B-C did not happened.
Generically speaking, I like the head to head rule as a tiebreaker the best, but In this new format it is not easy to apply an head to head rule, and this besides what you already pointed out, that one club has the advantage of playing the head to head match at home while the other plays it away.
Unless I am missing something, of course.
Where is the rule regarding the common game ?greenbay wrote: ↑Fri Jun 14, 2024 15:59 Good point. But UEFA obviously thinks otherwise.Interestingly there is no more H2H. Maybe with regards to H2H they thought it would be unfair as there is only a single game between them on home soil for a tied team.If two or more teams are equal on points on completion of the league phase matches, the following criteria are applied, in this order, to determine their rankings:
Superior goal difference in the league phase
Higher number of goals scored in the league phase
Higher number of away goals scored in the league phase
Higher number of wins in the league phase
Higher number of away wins in the league phase
Higher number of points obtained collectively by league phase opponents
Superior collective goal difference of league phase opponents
Higher number of goals scored collectively by league phase opponents
Lower disciplinary points total based only on yellow and red cards received by players and team officials in all league phase matches (red card = 3 points, yellow card = 1 point, expulsion for two yellow cards in one match = 3 points)
Higher club coefficient
Barcelona grabs the #8 spot because there was one common game: Barca beat Graz 4-0 at home, while Leverkusen only won 3-1 away.
The way I read the rules they sum the points of all 8 opponents (not just the common teams they played against):
"Higher number of points obtained collectively by league phase opponents" so each team has 8 opponents, so the points obtained by those 8 opponents determine which team played against the hardest opponents, and that team will be higher in ranking.