Playoff and match probabilities

Discussion in 'LA Galaxy' started by skydog, Aug 6, 2019.

  1. cleschke

    cleschke Member+

    Aug 16, 2004
    Fullerton, CA
    Club:
    Los Angeles Galaxy
    Nat'l Team:
    United States
    Honestly though, if the Galaxy has shown us anything this year it's that outcomes haven't conformed to what the odds predicted they should be. 20/1 seems worth plopping down a small, innocuous amount of money just to see what happens.

    But yes, I supposed the houses will always set the line so that the math, when borne out, will show the odds in their favor. There's never really a "good" bet. Just smarter bets than others.
     
    skydog repped this.
  2. MPNumber9

    MPNumber9 Member+

    Oct 10, 2010
    Club:
    Los Angeles Galaxy
    Nat'l Team:
    United States
    Someone would've made a buttload off the Vancouver game. I think they were like 8% to win or something ridiculous.
     
  3. skydog

    skydog Member+

    Aug 1, 1999
    Durham, NC
    Club:
    Los Angeles Galaxy
    Pretty much - 538 had them at 9%.
    And two weeks ago we were 99% to finish 2nd->4th. But somehow we failed to accomplish that modest goal.
     
  4. skydog

    skydog Member+

    Aug 1, 1999
    Durham, NC
    Club:
    Los Angeles Galaxy
    To counter the pessimism of my earlier post I want to note a couple of arguments in our favor:

    One - We can win these playoff games by either -
    1) Leading at the end of regulation.
    2) Being tied at the end of regulation and then winning in overtime or in pks. The last time we played at both MIN and LAFC we were tied at end of regulation, so good precedents.

    Two - Because of the high intensity levels usually seen in the playoffs there is a tendency for "special" players who have athletic traits that others can't match to be the difference makers. Stars matter more in playoffs than they do in the regular season games. See the 2018 NBA playoffs (or any other playoffs) for examples of how "really good regular season players" often get exposed by the Durants and Kawahi Leonards of the world. Ibra and Pavon are special athletes who can overcome the best efforts of MLS regulars. If our team can somehow manage to play even average defense these two can take us to victory.

    tldr vers: I'm saying we have a chance! Go Galaxy!
     
    MPNumber9 and TrickHog repped this.
  5. TrickHog

    TrickHog Moderator
    Staff Member

    Oct 14, 2002
    Los Angeles, CA
    Club:
    Los Angeles Galaxy
    Nat'l Team:
    United States
    [​IMG]
     
    galvanator and The Cadaver repped this.
  6. Thanasi Dalkos

    Thanasi Dalkos New Member

    LA Galaxy
    Greece
    Sep 23, 2019
    MPNumber9 repped this.
  7. skydog

    skydog Member+

    Aug 1, 1999
    Durham, NC
    Club:
    Los Angeles Galaxy
    We will win our first playoff game because higher seeds are 4-0 and an upset is due? Ok that's classic "gambler's fallacy" thinking, so I'll try again. We will break the lower seed losing streak because our 38% win probability is the best of any lower seed. (The others were/are 22% -> 34%* to win.)

    *All 538.com probabilities.
     
    kaburu, FlapJack and TrickHog repped this.
  8. FlapJack

    FlapJack Member+

    Mar 3, 2006
    Los Angeles
    Update that probably to 1. :)
     
    hav77, skydog, TrickHog and 1 other person repped this.
  9. skydog

    skydog Member+

    Aug 1, 1999
    Durham, NC
    Club:
    Los Angeles Galaxy
    Fantastic win! The score flattered us but we executed when we needed to, they didn't. We move on, they go home.

    Speaking of moving on -- our task now gets much tougher according to most objective measures. While 538 gave us the best chance of upsetting a higher seed in the 1st round that same objective, algorithm driven metric has us the least likely to win in the 2nd round. In fact their model gives us only a 15% ( :eek: ) of victory at LAFC. Ooomph....

    I know 99% of us LAG fans will disagree with this. Why so low from 538? Because these models are based solely on team strength and game location. And 538's model is really high on LAFC - they have an SPI of 60.3 which would put them mid-table in Serie A. No other MLS team even reaches the 50's. We come in at 42.4 - only RSL is rated weaker among the remaining playoff teams. For some weird reason they think our defense is a problem. Pssshhaw...

    Second reason why they give us such a low win prob is because -- unlike most human fans (who tend to have pretty strong recency biases) -- they don't give extra points for having a good history against a particular team. They see that as more likely small sample noise than representing anything real.

    Who's correct - the fans or the models? Who knows. In the end it doesn't matter unless you are trying to set up a fair bet. Because even teams with a 15% chance of winning do win sometimes, by definition. Hopefully this is our time!
     
    The Cadaver and TrickHog repped this.

Share This Page