Jump to content
advertisement_alt

Underdog

Members
  • Content Count

    20,063
  • Joined

  • Last visited

Community Reputation

1 Neutral

About Underdog

  • Rank
    MVP of OT (self-proclaimed)
  • Birthday 04/14/1969

Profile

  • Real name
    Jon Wiesman
  • Your gender
    Male

Personal

  • About Yourself
    The P5 "scouting report" spelled my name wrong. Nice scouting, guys.
  • Favorite poker hand
    ASAH

Screen names

Rankings

  • Worldwide

    N/A

  • All-time high

    13,974 (2010)

Cashes

  • Lifetime total

    $11,610

  • Biggest Cash All Time

    $2,508

  • Number of cashes

    135

  • Average cash

    $86

Latest post

  1. The 180-person Sit-n-Go is one of the most popular formats in online poker. It was only a little over three years ago that they were introduced at the $20+2 level, and some of you may remember my 180x180 SNG Challenge article back then. In a thread about that article, I mentioned possibly doing a follow-up about the newly-introduced $4+.40 buy-in tournaments. Well, three years have gone by and I have now played 1,800 $4.40 180-person SNG tournaments on PokerStars. What have I learned after 1,800 tournaments? Well, first of all, I have learned I never want to play another $4.40 180-person SNG tournament ever again. I'm half-joking; they can be fun, but are a grind and the stakes are low, but there are some other lessons to be learned too, and I hope you'll find them interesting. Before I get into the numbers, let me give two disclaimers. First, I am not a professional poker player, but I am a player with solid numbers (in terms of ROI) and I continually strive to make my game better. Secondly, in my original article, I talked about how 180 tournaments probably wasn't enough data to draw any solid conclusions, and of course that turns out to be an understatement. The shocking truth: even 1,800 tournaments may not be enough data. I'll show you what I mean later. The Numbers If we look back at that original article, here were my numbers for 180 $22 SNGs: Played: 180 In the money: 31 (17.22%) Final tables: 15 (8.33%) Wins: 4 (2.22%) Buy-ins: $3,960 Paid: $7,574.40 Profit: $3,614.40 ROI: 91.27% At the time, this is what I wrote: "Frankly, I'm pleased." Well, what I know now is that I should have been very pleased indeed. While obviously a lot of players can and do exceed 90% over any given 180 tournaments, I doubt there are many players who could maintain a 90% ROI consistently. It turns out that I was the beneficiary of some positive variance in my first experiment. Here are the numbers for my 1,800 $4.40 SNGs: Played: 1,800 In the money: = 324 (18.00%) Final Tables: 162 (9.00%) Wins: 20 (1.11%) Buy-ins: $7,920 Paid: $13,296.56 Profit = $5,376.56 ROI = 67.89% At first glance, the numbers look remarkably similar. My percentages of being in the money and getting to the final table were both about the same. The big differences are the percentage of wins and the overall ROI, but remember this: in the original experiment of 180 $22s, I never finished in 2nd place. I won all four times that I got heads-up. I said at the time I knew this didn't really prove much, and I was right. I finished in 2nd place 18 times, and if I had won those 18 tournaments, the win percentage and ROI would match the original experiment almost exactly. Or, looking at the other more realistic way, if I had lost two of those four wins in the original experiment, my ROI would have been in line with this recent experiment. Oh, Variance Given that these numbers are so similar to the ones I posted three years ago, it is tempting to use the word consistent, but a closer look will reveal some wild swings within the data. Here is the chart of my profit over games played for $4.40 tournaments, generated by SharkScope. At a glance, it appears to be a pretty consistent graph of about a 70% profit rate over 1,800 tournaments. Look closely, however, at the red line beginning at game 1,218. In 374 games, I posted a loss of $449.68, for an ROI of -27.33%. In fact, game 1,217 marks the high-water mark of my experiment. From game 1,218 through game 1,800, I posted a loss of $73.28 for an ROI of -2.86% over 583 tournaments, and that's including a first place in tournament 1,799. If I had stopped after game 1,217, I would have made $5,449.84 for an ROI of 101.77%. This chart also reflects how fortunate I was to have had good results in my original experiment three years ago. Within this run of 1,800 tournaments, my worst run of 180 consecutive tournaments started with game 1,216 where I went on to lose $330.48 for an ROI of -41.73%. (If I had done that three years ago, I certainly wouldn't have written an article about it, and even if I had, pocketfives probably wouldn't have been too interested in publishing it.) However, my best run started at tournament 934 where I went on to win $1,609.92 for an ROI of 203.27%! You can see that run on the chart; what looks like a rather large vertical line was when I actually won two consecutive tournaments on March 3rd, 2008. At that time, I had climbed to 2nd place on the SharkScope.com leaderboard for $5 multi-table games. At this time, I'm not on the leaderboard at all. The point of all this is that we never really get away from our old friend variance. As poker players, we all know this on an intellectual level, but sometimes knowing that the big dip on a rollercoaster is perfectly safe doesn't remove the emotional impact of riding it. When I am dealing with negative swings, I find it is sometimes useful to "zoom out" and look at my results over a larger period of time. I know I can consistently win; I have numbers to show it. Variance happens to everybody, and my experience is not unique. The universe is not out to get me. Here are the results over games played chart of a successful grinder at 180-person SNGs (profit stats omitted): It looks pretty consistent, doesn't it? And it is: that's a steady upward trend over a large sample size. But okay, let's zoom in just a bit, looking at data just over the last month. Yikes. That's a short-term loss over nearly 700 games. Losing a relatively small amount might not seem like a big deal compared to the long-term, but it can be difficult to maintain the right mindset during downswings of this nature. Having said all that about learning to accept variance, I have to admit: I was suspicious. While it's true that accepting variance is something we all need to learn, that doesn't mean we shouldn't take a long, hard look at ourselves when we do have a long negative streak. Each of us needs to take responsibility for giving ourselves the best chance to succeed. It's not always appropriate to chalk up a bad run to variance, nor is it necessarily correct to pat ourselves on the back for a good run. Being able to discern the difference between simple (but possibly extreme) variance and the expected consequences of bad (or good) decisions is one of the hardest parts about this game. So being a programmer I wrote a Monte Carlo simulation to test how likely some of these swings are. All these simulations test the likelihood of these scenarios using the actual data. Results between 1 and 1217: Profit = $5449.84, ROI = 101.77% Likelihood of playing 1217 tournaments with ROI of 101.77% = 4.11% Results between 1218 and 1591: Profit = $-449.68, ROI = -27.33% Likelihood of playing 374 tournaments with ROI of -27.33% = 0.05% Results between 1218 and 1800: Profit = $-73.28, ROI = -2.86% Likelihood of playing 583 tournaments with ROI of -2.86% = 0.18% Worst run of 180 started at 1216. Profit = $-330.48, ROI = -41.73% Likelihood of playing 180 tournaments with ROI of -41.73% = 0.27% Best run of 180 started at 934. Profit = $1609.92, ROI = 203.27% Likelihood of playing 180 tournaments with ROI of 203.27% = 0.75% As you can see, all of these scenarios are relatively unlikely. The run of 374 tournaments with an ROI of -27.33% has only a 0.05% or 5-in-10,000 chance of happening given the data set. And yet it did happen. Now I could just accept it as variance, but I think it's probably a little more likely that there might be something to pick up in review. Sometimes we pick up leaks in our game and good poker players (even just casual ones like me) are the ones who find those leaks and plug them. Comparing Levels Three years ago one of the questions people asked me was whether I saw a huge difference in the skills of players at different buy-in levels. Obviously there is a difference, although I don't think it is as large as some people may think. After my initial 180 $22 tournaments at 91% ROI I have come down to a 55.8% ROI at that level. Clearly there are some better players at that level. However, one of the misconceptions on pocketfives that I have encountered is the idea that the $4.40 tournament player is terrible, and that it's like playing bingo. Just last week someone posted a hand from a $12 180 in Poker Discussion and said that he was sure that the small blind villain who pushed on the flop must have had a good hand because "donks at this level don't know about the stop-and-go." I can assure you that the stop-and-go move is alive and well at the $4.40 level. Generally, it's not a good idea to make too many assumptions about the players in a game. While having a general idea as to the level of a random player is appropriate, don't get sucked into complacency and fail to notice that some players are playing much better (or worse) than the buy-in level would seem to indicate. The truth is that you often have no idea why a player is playing at a particular level. A good player might have terrible bankroll management and has been forced to go back to a lower level after some bad variance has wiped her out. Conversely, she might be a bankroll nit who is determined to stay within very tight parameters for game selection. You just don't know. One of the worst players I ever encountered was in the Sunday Million on PokerStars. This person would absolutely not fold under any circumstances. I was amazed to see players who should have known better try to bluff this person with air and get called with bottom pair. Eventually he busted out but not before he knocked out a lot of players. Having said that, I do play slightly different at the $4.40 level than I would at higher levels. As you would expect, I don't bluff unless I have some evidence that the villain is capable of folding a hand. I have to have seen a big fold before I attempt a big bluff. Likewise, I bet stronger for value because I am more likely to get paid off. It's pretty basic common sense stuff, nothing that most good players haven't already figured out. Lessons from the data While these numbers have been helpful in trying to determine my skill level, there is additional information that we can glean from the data. Take a look at these numbers, keeping in mind that a 180-person tournament pays 18 spots: Finishes at: 20th: 14 19th: 9 18th: 19 17th: 20 What do these numbers tell you? Go ahead, you can say it; I know. That's me being weak on the bubble. That's what it looks like when I am trying to back into the money instead of waffle-crushing on the bubble to go for a top 3 payout. Now look at this: 11th: 21 10th: 14 9th: 21 8th: 21 Again, that's me being a scared little donk on the Final Table bubble. I don't like it, but ignoring it isn't going to help me get better. What are the results of this? Let's look: 3rd: 15 2nd: 18 1st: 20 I need to be turning some of those 18ths and 17ths into chances at Final Tables, and I need to be turning those 9ths and 8ths into top 3 finishes. Imagine that in those 19 tournaments where I finished 18th, I decided to play very aggressively on the bubble, forcing midstacks to fold or look me up. Let's say that it resulted in 10 additional bubbles, but that in the other 9 tournaments, I signficantly increased my stack. Perhaps in five of those tournaments, I still finished between 10th and 18th, but maybe in four I managed to get to 9th or better. Four additional 9th-place finishes, plus the original twenty-one, would now give me twenty-five. Let's do the exercise again, and covert thirteen of those into 10ths and the other twelve into top-three finishes. Of course it isn't that easy, and my conversion rate of 50/50 is too optimistic, but you get the idea. A solid player should have more first place finishes than I have. Yes being aggressive on the bubble will cause me to have more 19ths and more 10ths, but I will more than make up for it with more top 3 finishes where all the money is. The question is, now that I have identified the problem, can I fix it? I intend to find out. Conclusions Just as I said three years ago at the end of my first article, once again I'm not sure if I have enough data to draw any solid conclusions. Instead, I hope I've given you some things to think about as you look at your own data and try to evaluate your game. One thing to consider of course is that poker is not played in a vacuum. Other than the variance of the cards, every game is a unique mix of players and circumstances. It is best to be very deliberate when analyzing our results, constantly collecting data and reevaluating based on new information. --- More Articles by Underdog34 Underdog's 180x180 SNG Challenge Feb 25, 2006 A Year in the Life of a PocketFiver Jan 29, 2006 -----
×

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.