Playoff Tiebreakers: Stop the Madness

Discussion in 'MLS: News & Analysis' started by The Perfesser, Oct 18, 2009.

  1. The Perfesser

    The Perfesser New Member

    May 23, 1999
    AthensGA/NewburyptMA
    Club:
    New England Revolution
    Nat'l Team:
    United States
    OK, can any of you explain within the next 30 minutes who will be in the playoffs? I didn't think so.

    If Chicago loses to CUSA on Thursday, and FCD, DCU, and TFC win on Saturday, and the Revs win on Sunday, there will be 5 teams on 42 points.

    What's the problem? MLS chooses to use the head-to-head method as the first tiebreaker. Here are the gory details (see rule 8 below, and then add two more teams). But before you do that, figure out who finishes 2nd in the east, and THEN run the results for the remaining 4 teams. I'll wait:

    ==================

    TEAM-STANDINGS TIE-BREAKING PROCEDURES
    The team awarded the highest position in the MLS standings will be the team with the greatest number of points (3 points for a win, 1 point for a tie, 0 points for a loss). In the event that two teams finish the regular season with an equal number of points, the following system will be used to break the tie:

    1.The highest position shall be awarded to the team with the better win/loss record in current regular season games against all other teams equal in points. (head-to-head competition)

    2.If the teams are still equal in the standings, the highest position shall be awarded to the team with the greater goal difference against all other teams during the regular season. (goal differential)

    3.If the teams are still equal in the standings, the highest position shall be awarded to the team scoring the greatest number of total goals against all other teams during the regular season. (total goals)

    4.If the teams are still equal in the standings, the procedures described in this section shall be applied only to games played on the road by each team against all other teams during the regular season. (road 1-3)

    5.If the teams are still equal in the standings, the procedures described in this section shall be applied only to games played at home by each team against all other teams during the regular season. (home 1-3)

    6.If the teams are still equal in the standings, the highest position shall be awarded to the team with the fewest team disciplinary points in the League Fair Play table during the regular season.

    7.If the teams are still equal in the standings, the highest position in the standings shall be determined by the toss of a coin.

    8.The first tiebreaker in a three-way tie is also head-to-head, but it is determined via points-per-game versus the other two teams. If two teams are tied in points-per-game head-to-head, the next tie breaker is goal difference.

    NOTE: If two clubs remain tied after another club with the same number of points advances during any step, the tie breaker reverts to step 1 of the two-club format.

    =======================

    What's wrong with this picture? Try explaining this to someone whom you're trying to recruit as an MLS fan, or to a statistician, or a lawyer. :eek:

    There are three compelling arguments for an alternative method:

    Total Goal Difference and Total Goals Scored,

    the method used in all FIFA competitions, the Bundesliga and several other leagues.

    • Philosophical. What are you trying to measure? Just two (or more) teams on the same number of points vs each other, or how all of the tied teams are doing vs the entire league. What's the point of having a league table then if you don't measure how the tied teams have done against the whole league? The playoff teams should be about who is the better team AGAINST THE ENTIRE LEAGUE FOR THE WHOLE SEASON because that's what Goal Difference and Goals Scored actually measure. By measuring the two (or more) teams against each other, you don't know on those few dates if players were injured, on international duty, etc. By measuring total performance for the entire season, you've got a better sense of which is actually the better team and deserves to be in the league playoffs. Just because other US sports use head-to-head as a tiebreaker, doesn't mean MLS has to. MLS used to have a countdown clock and (crap)shootouts too but then wisely adopted practices used for soccer in other countries.
    • Simplicity. Just look at the standings, it's as plain as day and you don't need a legal document to figure it out. If you are trying to get fans to understand who's in and who's out, it takes 2 seconds to see who's got the better goal diff and goals scored. Why laboriously consult the "MLS Monks" to figure out how many head-to-head matchups can dance on the head of a pin?
    • Fairness. Many of these teams tied on points have played 3 games vs each other. Clearly the team with 2 home games has a much greater chance of coming out ahead on points if you measure those three games. Doing against the entire league solves that problem. Until we get a balanced schedule next year (but apparently ONLY for next year because with 18 teams in 2011 we will be unbalanced again, unless there is a 34 game season) there will ALWAYS be instances when teams will have played each other an odd number of times. With more teams it's likely that you will be only measuring one head-to-head game which is truly unfair.

    * note. El Jefe to enter stage right to disagree in 5...4...3...2...1.... since we had a "preview" of this debate on the Week 31 multiple results thread. ;)
     
  2. Knave

    Knave Member+

    May 25, 1999
    I believe I have something to say about this later, but I have question first.

    I understand La Liga uses head-to-head for tie-breakers. Does anyone have a link for their exact tie-breaking procedure? I can't seem to find anything for some reason. But my hunch is that it is actually different from the head-to-head procedure used by MLS.

    Anyone got that info?
     
  3. Eric Shinn

    Eric Shinn Member

    Dec 28, 2007
    I would assume they would just take the results of all 5 teams vis a vis each other and whoever has the best record of the 5 wins and if any teams are tied with best record they will go to goal difference.

    But i agree, conceivably the team with the best overall GD of the 5 tied teams could have the worst head to head record of the 5 and thus not have their good GD counted for them later.
     
  4. ThreeApples

    ThreeApples Member+

    Jul 28, 1999
    Smurf Village
    Club:
    San Jose Earthquakes
    Nat'l Team:
    United States
    I don't have a link, but I believe it is the same for the first step, but different if they have to go beyond that. Head-to-head points is the first tiebreaker, which is the same as MLS as long as the opponents all played twice, but if there's still a tie they go to head-to-head goal difference and goals scored, whereas MLS goes to full season goal difference and goals scored.
     
  5. Mucky

    Mucky Member+

    Mar 30, 2009
    Manchester England
    Club:
    Manchester United FC
    All well reasoned points though I would think an alternative philosophical argument could be made by supporters of the current system because even though you are comparing against all teams in the league those teams were strong and weak at different times in the season too - injuries, streaks, etc all come into play so head to heads could be seen as a more relevant yardstick for who should advance in that respect.

    I cant think of a reasonable argument against your points on fairness and simplicity so on balance I think the goal differential system is better - at least ,as you say, until the schedule becomes balanced.

    I made a post a while ago about another problem which is all the final fixtures should kick off at the same time so as not to allow an advantage to those playing later or last.(This could be a very significant advantage and in my view far outweighs any unfairness in the tie breaker system)

    Of course the US stretches across time zones so it is a difficult problem to overcome.
     
  6. jass

    jass Member

    Oct 12, 2006
    Club:
    Parana Curitiba
    Nat'l Team:
    Mexico
    I agree. Even worse is when these dumb rules were used for Superliga. Remember when some teams thought they qualified past the group stage but actually hadnt because MLS tiebreakers were used instead of international ones?
     
  7. JasonMa

    JasonMa Member+

    Mar 20, 2000
    Arvada, CO
    Club:
    Colorado Rapids
    Nat'l Team:
    United States
    Because the other 3 threads where this discussion is happening wasn't enough?

    It looks like Greece, Italy, Poland, Portugal, Russia, and Spain all use H2H first. That's enough for a difference for me to say that both methods have validity and MLS isn't doing anything obviously wrong, as much as some people want to claim that's the case.
     
  8. TrueCrew

    TrueCrew Member+

    Dec 22, 2003
    Columbus, OH
    Club:
    Columbus Crew
    Nat'l Team:
    United States
    I agree wholeheartedly.

    One thing I'll say about La Liga, they all play each other twice, home and away. That isn't the case in MLS (not all the time anyway).

    I'm in the GD camp all the way. I can maybe see for a 2 team (but only if they played twice, not three times), but not 3 team TB's or more.

    And throw in MLS' stupid interpretation of the 3 team (if three teams are tied on points, and 2/3 tied on on PPG then:

    1) If one team advances (they are along in 1st), then they advance and one uses the 2 team tie-breakers for the remaining two teams (makes sense to me).

    2) If two teams are tied, and a third team is behind them, then don't eliminate the third team and proceed to a two team TB for the tied 2 teams, but proceed to step 2 of the three way tiebreaker (GD), but only consider the goal difference of the top 2 teams. In other words, eliminate the third team, but still use the 3-team tiebreakers to break a two team tie.

    How dumb (and inconsistent) can one get?
     
  9. BSGuy321

    BSGuy321 Member

    Sep 2, 2008
    Agree. Too complicated.
     
  10. Knave

    Knave Member+

    May 25, 1999
    To me there is something wrong with the MLS tie-breaking procedures. But it's likely not what you think it is.

    The Perfesser asks the right question above: "What are you trying to measure?" But I don't think the answer to that question is unambiguous.

    You're trying to define the better team. Do you measure that via head-to-head or via some league-wide standing (like GD)? Is the better team determined by how teams do against each other or by how team did against the entire league? There is no right answer for these questions and that ambiguity is why some leagues and competitions use head-to-head and others use things like GD.

    In and of itself, using head-to-head or things like GD is neither wrong nor right, nor better nor worse. So what is wrong with the MLS tie-breaking procedure if neither head-to-head nor GD etc. can be right or wrong in themselves? What's wrong is that the MLS tie-breaking system is internally inconsistent.

    Many, many leagues and competitions (like La Liga and the UEFA Champions League) use head-to-head as the first tie-breaker. And with two exceptions those leagues and competitions continue with head-to-head measures for the second and third tie-breakers: head-to-head GD, head-to-head total goals and so forth.

    The exceptions are Serie A and MLS, both of which use overall GD as the second tie-breaker. This seems wrong to me. If you opt for head-to-head as the first tie-breakers, then you're saying the better team is best measured by how they did against each other. In that case, to be consistent the second tie-breaker should be head-to-head GD.

    The question is not which is the better first tie-breaker, but if you choose head-to-head (or, alternatively, GD) as the first tie-breaker, then what makes consistent sense for the second tie-breaker. If you choose head-to-head as the primary tie-breaker, then you should be reasonably exhaustive about it and not switch to an overall measure like GD for the second tie-breaker. By the same token, if you choose GD as your first tie-breaker, then your second tie-breaker should not be head-to-head.

    MLS chooses head-to-head on the first tie-breaker, and then gives up on it in favor of overall measures. The result is that MLS has this strange hybrid system that given its internal inconsistency actually defies rational explanation.

    All that said, The Perfesser is right about this:
    That, to me, is the best argument that can be made in favor of GD and other overall, league-wide measures breaking ties. If you have a table with GF, GA and maybe even a helpful GD column, then you ought to be able to figure out the standings with minimal effort.

    GD is clean and straightforward. And those are things, all other arguments aside, that head-to-head simply cannot offer.
     
    1 person likes this.
  11. BSGuy321

    BSGuy321 Member

    Sep 2, 2008
    Simplicity = Beauty
     
  12. ElJefe

    ElJefe Moderator
    Staff Member

    Feb 16, 1999
    Colorful Colorado
    Club:
    FC Dallas
    Nat'l Team:
    United States
    To me, where it looks like MLS is screwing up is NOT in the use of head-to-head as the first tiebreaker (which in my mind is the right way, but I'll admit that it's personal preference), but in "resetting" the tiebreakers in multiteam situations after the top team is chosen.

    Take the four-way Dallas/DC/New England/Toronto tiebreaker:

    Toronto: 2-1-3 (1.50 PPG)
    Dallas: 2-2-2 (1.33 PPG), +4 GD
    New England: 2-2-2 (1.33 PPG), -5 GD
    DC: 1-2-3 (1.00 PPG)

    Now, that would seem to provide to some sort of order on its own (save for the need to go to a second tiebreaker for Dallas and New England), but instead, MLS resets the tiebreakers for Dallas, New England, DC, which New England wins, then against for Dallas and DC, which DC wins.

    So you instead get the following order:

    1. Toronto
    2. New England
    3. DC
    4. Dallas

    And to get the result, you have to go through three different head-to-head computations, instead of one.
     
  13. cthomer5000

    cthomer5000 Member+

    Apr 23, 2007
    Raleigh NC
    Club:
    New York Red Bulls
    I like it as-is. There are equally valid philosophical arguments to be made on both sides.
     
  14. deron

    deron New Member

    Jul 25, 2006
    Centennial, CO
    Club:
    Colorado Rapids
    Nat'l Team:
    United States
    I wish it were simply GD, followed by head-to-head w/away goal bonus. I know some teams meet three times, but I reckon the two home games is balanced by the away goal bonus.
     
  15. Q*bert Jones III

    Q*bert Jones III The People's Poet

    Kingston Stockade
    Feb 12, 2005
    Woodstock, NY
    Club:
    DC United
    Muddle League Soccer.
     
  16. ChefJim27

    ChefJim27 Member

    Feb 9, 2008
    I don't think the problem is how MLS chooses to resolve ties. I think it has to do with there being so many clubs tied. It seems almost silly that a 30 game regular season could solve so little. Trying to resolve a 4 or 5 team tiebreak scenario is enough to make one's head spin. Before we all get too psychotic, let's let week 32 happen, and see where we are. Right now, it seems eerily like trying to tinker with the BCS formula, which seems to change every year attempting to solve last year's injustice. If MLS decided not to switch the format for this season, 3 Eastern clubs would be guaranteed, as opposed to 2. It's an imperfect world, but it corresponds to the ancient Chinese curse...

    "May you live in interesting times."
     
  17. Emile

    Emile Member

    Oct 24, 2001
    dead in a ditch
    A question that needs to be asked is what the league should be encouraging - in what manner do they want to motivate teams?

    Goal differential is not that great a way to break ties. It does have simplicity in it's favor, but it also violates a tenant of sport, that a win is a win, the simple objective. In making a 2-0 win more valuable than a 1-0 win, you are encouraging running up the score on lesser teams and playing defensively when behind. This may be necessary in a short tourney like the World Cup, because head-to-head will quite often be a tie, but it has questionable outcomes. If you wanted to go the simple route, I would suggest goals scored is the better marker since it encourages teams to try and get goals all the time, although that is also not the purpose of the game. Even better might be first ranking teams by total wins, since you are then adding a small encouragement for not playing for ties.

    I think the method they are using is quite fair, and ameliorates the issues with the unbalanced schedule. Resetting after each team is removed is probably fair too. Complexity is an issue, but how much of one? Are people tearing their hair out over this? Or is it captivating people's interest? The NFL has pretty complex tiebreaker rules which always cause a giant piece of the newspaper agate being devoted to all the permutations of so many 9-7 teams. Nobody jumps off a bridge.

    The point just made is pretty key. The likelihood of a four or five-team tiebreaker in an MLS season is not terribly likely, and may not even happen this year. MLS has parity, but this is a pretty unique finish to my memory. More teams will mean a larger middle and perhaps make this a more common event, but it is exciting to have the last playoff spots feel like a crazy knifefight. But most of all, this is yet another example of how MLS can never win - because if there were a two-team tie with a clear head-to-head winner, and MLS decided the spot with a lame and arbitrary breaker like GD, they would get all kinds of shit about it. How unfair, this is not how they do it in Europe, etc.

    Me, I'm just going to enjoy the madness.
     
    2 people repped this.
  18. The Artist

    The Artist Member+

    Mar 22, 1999
    Illinois
    Club:
    DC United
    Nat'l Team:
    United States
    Totally agree. I admit that simplicity and fairness are powerful arguments in favor of GD, but philosophically I think there is something wrong with saying a 2-0 win is clearly better than a 4-3 win. Game situations change all the time. I believe Nowak once said it's easy to win a game 5-0, but it's hard to win 1-0, meaning that in a 1-0 game you have to be at your best for 90 minutes while in a laugher the other team will usually give up after falling behind 3-0 and the last two goals (see the Gold Cup Final) are merely the result of a disheartened and overstretched defense.

    On the other hand I'm willing to admit that entertainment trumps all and there's something very entertaining about watching a team with a 3-0 lead still attacking all out because they know they need to win 4-0 to advance and seeing a team losing 3-0 playing bunker ball (see the Confederations Cup). Plus, if we ever are able to have all the games played simultaneously you would have those great moments of flipping back and forth between games, knowing that one goal in either game could change the standings.
     
  19. profiled

    profiled Moderator
    Staff Member

    Feb 7, 2000
    slightly north of a mile high
    Club:
    Los Angeles Galaxy
    My only big complaint is the mixing and matching of conferences and seedings.

    If you're going to allow teams to swap sides, then there should be no automatic qualification and the rankings should be given based on 1-8 overall not 1-4 west and 1-4 east. The East for example is getting easier games do to this, they'll end up with teams 1 and then likely 6-8 on their side, will the west be battled out between teams 2-5, which hardly makes any sense.
     
    2 people repped this.
  20. Autogolazo

    Autogolazo BigSoccer Supporter

    Feb 19, 2000
    Bombay Beach, CA
    They have to have the final week's games begin at the same time next year.

    If the chips should fall the Revs' way, then they have a big advantage knowing from the outset that they only need a draw on Sunday to advance.

    Ditto for Chivas USA knowing the final Galaxy and Sounders point totals in pursuit of #1 in the West.
     
  21. Clint Eastwood

    Clint Eastwood Member+

    Dec 23, 2003
    Somerville, MA
    Club:
    FC Dallas
    Nothing about it makes sense.

    I've learned to just let it be..........and see what happens.

    MLS is trying to have it "so many ways" by having conferences plus what amounts to a single table for the bottom half of the league trying to get in the playoffs.........that it's mindnumbing.

    I just don't care anymore. Tell me who the playoff teams are........then I'll see if I can make myself interested.

    I don't feel horrible deciding which of the 7th, 8th, 9th, or 10th best team in the league end up getting "screwed". Perhaps if all the teams weren't so overwhelmingly mediocre, they wouldn't have this problem. Did anybody sit through that Revs-Fire abomination this weekend. The Revs could have played for another 3 hours and not scored. I'm sure if I was a Revs fan I'd be pissed if my team didn't make the playoffs based on a convoluted formula I didn't understand.....but as a neutral I can't say I'd feel any sympathy for a team playing that poorly right now.

    The best team (and most entertaining) in the league right now could easily be FCD. I wish there was some way for MLS to jigger it's formula to get those guys in. Hell...........in the event of multi-team ties...........order the teams by best second-half record. That way you'll at least get the "hot" teams in the playoffs. (I realize that wouldn't happen. I'm just brainstorming.)
     
  22. JasonMa

    JasonMa Member+

    Mar 20, 2000
    Arvada, CO
    Club:
    Colorado Rapids
    Nat'l Team:
    United States
    I find it interesting that using H2H favors taking risks and getting a result. For example if DC, NE, and Colorado end up tied at 40 Colorado gains more from its 1 win and 1 loss (1.5 pts/game) to DC than its 2 draws to NE (1 pt/game), even though you would say that they were .500 against both teams.
     
  23. Emile

    Emile Member

    Oct 24, 2001
    dead in a ditch
    I couldn't agree more. If people want to be bothered by the playoff structure, this is the problem that should be addressed.

    Also, if you want an effed-up tiebreaker, I think that if Egypt beats Algeria by two goals in the final qualifier, they will flip a coin to see who goes to the World Cup.
     
  24. revsrock

    revsrock Member+

    Jul 24, 1999
    Boston Ma
    Club:
    New England Revolution
    Nat'l Team:
    United States
    This is from:
    http://wbx.me/l/?p=1&u=http://blog.revolutionsoccer.net/?p=5353


    This does not take in account 2 teams tied for a spot

    Basically, if multiple teams are tied, they go into a mini-pool and the team with the most points per game in matches involving those teams advances on. So, in those cases, the following scenarios exist:

    Mini Pool (winner)

    Chicago-Dallas-New England: Dallas

    Colorado-Salt Lake-New England: Salt Lake

    Dallas-D.C. -New England: New England

    Dallas-New England-Toronto: Dallas

    D.C.-New England-Toronto: Toronto

    Chicago-Dallas-D.C.-New England: D.C.

    Colorado-Salt Lake-Dallas-New England: Goal Difference between RSL and FCD

    Colorado-Salt Lake-D.C.-New England: Salt Lake

    Colorado-Salt Lake-New England-Toronto: Toronto

    FC Dallas-D.C.-New England-Toronto: Toronto

    Colorado-Salt Lake-Dallas-D.C.-New England: Colorado

    Colorado-Salt Lake-Dallas-D.C.-New England-Toronto: Toronto
     

Share This Page