Artikel über Varianz in SNGs

    • fummi
      fummi
      Bronze
      Dabei seit: 13.02.2006 Beiträge: 7.317
      Hi :) Hat einer mal nen Artikel (engl. oder dt.) zur Hand, der die mathematische Seite von Varianz in SNGs behandelt? Interessant wären Sachen wie, in welchem Umfang sich der ROI verändert, wenn man z.B. von 40 Coinflips oder 3:2 Favoriten mal 30 am Bubble verliert, was ja durchaus realistisch ist. Ich denke das kann einen bei einer Samplesize von 300 doch schon recht stark runterziehen. Also ein (vorallem mathematischer) Artikel darüber wäre ganz interessant. fummi
  • 8 Antworten
    • Feldhase7
      Feldhase7
      Bronze
      Dabei seit: 04.05.2005 Beiträge: 1.631
      #2
    • fummi
      fummi
      Bronze
      Dabei seit: 13.02.2006 Beiträge: 7.317
      Hab was gefunden auf 2+2. Nur leider blick ich das Ding noch nich ganze ;) http://archiveserver.twoplustwo.com/showflat.php?Cat=0&Board=&Number=1941324&page=3&view=&sb=5&o=14&fpart=1
    • gr8ermoon
      gr8ermoon
      Bronze
      Dabei seit: 05.09.2006 Beiträge: 37
      ich zitiere hier mal einen artikel von curtains zum thema varianz: This month I’m going to talk a bit about variance. A lot of readers know all about it but I know for sure that some don’t. I have a funny story about a player who loved to play poker and decided to try playing online. He especially loved tournaments and started playing the sit and go’s on Pokerstars. Needless to say, he started out incredibly well, finishing in the money in maybe about 18 of his first 24 low buyin tournaments. Since some of these tournaments were of the 2-table variety, it was clear to this player that there was no chance that this could simply be a stroke of good fortune. No, instead this player was quite certain that he was probably one of the better poker tournament players in the world. He went to Daniel Negreanu’s website and posted about his results and asked honestly whether it was possible for someone to have such results and not be an amazing player. However instead of hearing “Wow I’ve never heard of such incredible results, you seem destined to be one of the all time greats” he was greeted with “Variance variance, blah blah, etc etc”. Of course Negreanu wasn’t clearly reading what this poster said! 18 OUT OF 24, clearly this is the next Phil Ivey we are talknig about!!! Anyway if it’s not obvious by now, this poster was me. I have gone from about the most clueless of all sit and go players, whom after a 150 tournament sample would proudly tell everyone I knew about how I could come in the top 3 about 50% of the time to one who can make a good living at the game. Fortunately most of the people I bragged to don't know enough about poker to come back at me now to mock me. In this article I’m going to share with you the likelihood of certain swings and what they mean. All of the below results are determined using a program called “ROI Predictor” which instantly simulates thousands of tournaments for you. Let’s start by examining someone with a nice healthy 13.6 ROI (return on investment), which is achieved by getting 1st place 13% of the time, 2nd place 12% of the time and 3rd place 12% of the time. Someone who acheives these results will make a healthy living 8 tabling almost any limit of sit and go poker. However even with results of this nature, there is about a 2.5% chance that you will be down money after 500 sit and gos. To people who play 8-12 games at once, this tmie can come and go in a flash, however if you are playing 1-2 tables at a time, this losing streak can drag on forever, and leave you feeling like you are a hopeless player. If we lower this stretch of tournaments to 200, then there is about a 9% chance that you will be down money at the end of this stretch, and there is about a 20% chance of being down money after 100 tournaments. Now 13.6% is a very healthy ROI, and is a bit higher than most anyone would expect at the highest levels. Let’s take an ROI of 9% (Assuming you get1st, 2nd and 3rd place all 12%) and see what kind of swings you should expect… 2000 touranments: only about .1% of the time will you be down after 2000 events. 1000 tournaments: about 2% of the time you will be down after 1000 events. 500 tournaments: about 7.5% of the time you will be down after 500 events 200 tournaments: about 18% of the time you will be down after 200 events 100 tournaments: about 27% of the time you will be down after 100 events Looking at the numbers above, you see that even for a very healthy winning player, someone who will make about $20 per tournament at the $215 level, 2% of the time they will be down money after 1000 tournaments. When you have lots of pro players playing many games, you can be sure that some of these players have gone through these 1000 tournament downswings. When it’s happening it’s almost impossible to believe that you can actually be a winning player. I’ve talked to people who have been in the midst of 500-1000 tournament downswings, and they are just so worried about what they are doing wrong, when in reality they could be playing quite reasonably and still be having such bad fortune. Most of the time such bad streaks will actually be due to bad player. I personally have never had a losing streak longer than 400 tournaments, but I consider myself one of the lucky ones, and I know that if I keep playing it will happen eventually. If you haven’t had any extended downswings, you almost surely haven’t played enough. Also please note that the larger your ROI is, the much less likely it is for you to go on one of these extended losing streaks. A few more interesting notes: Let’s take a look at someone who is a pure breakeven player, with a 0% ROI: This player will have an ROI of above 10%: After 1000 tournaments: 2.25% of the time After 500 tournaments: 8.4% of the time After 200 tournaments: 17% of the time (above 20% ROI - 4% of the time) So with the above figures you can also see that a lot of break even/mediocre players can experience extended streaks of good fortune. This especially holds true for those who play maybe one table a time, and may take months to play 500-1000 tournaments. They can go a long time thinking they are very good at poker, without being very good at all. If you are a good player you will probably meet a lot of people who fit this bill, and will be more than happy to tell you how good they are and how good their results are. It’s hard to count just how many people I’ve heard mention that they finish in the money 75-80% of the time in the online sit and go’s. The main point of this article was to demonstrate just how long a bad streak can go on for in sit and go play. If you play 20 tournaments per day, and are a solid winning player of around 9% ROI, you will have a breakeven/losing month about 1 in every 50 months. This may not seem like a lot, but you probably have seen enough one outers in your day to make you reconsider. Also the toughest part is fighting through this bad streak, and continuning to play your best game. A lot of people stop playing after such streaks, or even move to cash games (Which may or not be a bad thing, but it’s often done for the wrong reason). Let me stress that the most likely reason that one is losing, is because they aren’t a good player. However if you have had good results for thousands of tournaments, and suddenly hit a rough patch of 500-1000, it’s a lot more likely you are experiencing the negative side of variance. It is possible for almost any player in the world to lose money over this many events.
    • Ratskin
      Ratskin
      Bronze
      Dabei seit: 19.04.2006 Beiträge: 411
      Original von gr8ermoon 2000 touranments: only about .1% of the time will you be down after 2000 events.
      bedeutet "to be down" hier break even oder was soll das heißen?
    • FrtZ
      FrtZ
      Black
      Dabei seit: 25.08.2006 Beiträge: 3.412
      gewagte these: down = unten = verlust :rolleyes:
    • fummi
      fummi
      Bronze
      Dabei seit: 13.02.2006 Beiträge: 7.317
      Also das Verlust würde ich weg machen und mit "unten" könnte dann durchaus BE gemeint sein.
    • FrtZ
      FrtZ
      Black
      Dabei seit: 25.08.2006 Beiträge: 3.412
      hallo? to be down heißt verluste gemacht haben.
    • fummi
      fummi
      Bronze
      Dabei seit: 13.02.2006 Beiträge: 7.317
      Es geht doch darum, dass man eigentlich Winning Player ist .. und ich deute es so, dass man nach xxxx SNGs mit einer WSK von xx% bestenfalls BE ist. Also Verluste gemacht haben, wenn man es so sieht wie du.