Monday, 26 September 2011

Explaining and understanding a monster downswing

So if anyone has been following my results since I have been living in Costa Rica they will have noticed a pretty unusual sight in my sharkscope graph, I like to call it the 'monster downswing'. I guess to a lot of regs this swing isn't so bad and for me mentally I feel it's ok to handle too, but just the uniqueness of this swing due to moving up and down stakes is really quite interesting and felt like explaining why a swing like this can and will happen when playing high stakes super turbos over hundreds of thousands of hands.

First of all I would like to share my all in ev since recorded from mid March, the graph is missing around 10k hands from the first week when I made the switch to Pokerstars.
For people who do not understand these graphs basically the orange line is your all in expected value (ev) which basically gives you a good idea of how well you are playing only using maths and no other factors. The blue line shows you what is actually happening with the luck element which is involved in poker. In the short term you can see major differences in these two lines but in the long run you will see a trend that these two lines will overlap each other on many occasions and the difference over the very long run will be almost impossible to see, which also can explain there isn't luck involved in long term poker.

This downswing doesn't look to crazy when looking at this graph, in fact it looks pretty similar to the others I have had in the past. So why does this one seem to drag on for so long and how comes in actual results I am down $30,000 when my blue line says I haven't lost anything, we need to look into this small time period in a bit more depth to see what is going on.


Here we can see the same graph but only the time whilst I have been in Costa Rica. The orange line here indicates I should be up around 40+ buyins (1*) and the blue line indicates I should be up around 1buy in and I actually am up one buy in since I been here. So if I am up one buy in, why am I down $30,000?
These two graphs will help explain:

The first graph shows the actual amount of money through the time period and the second graph shows the 'ROI' (Return on investment) at each different stake as you can see I have run way worse in high stake games to lower stake games. Now this really helps explain where things have gone wrong. The fact that I am playing multiple stakes is really the key to this swing. Running bad in high stakes and running good in low stakes can really make a downswing last a lot longer and a lot more drastic than it would have if I was only playing one stake. If I had been playing one stake the entire way through the downswing I would actually be in profit through this period.
I obviously cannot complain about this because I have had many experiences of running better at high stakes than low stakes, and it's something that will Iron out over a longer time period these things just kind of have to happen. I also can never complain because as some of you know I had some killer months before coming here where I really crushed the Hyper turbo's on stars well the opposite thing was happening there, I was running above ev but I was also running better in the higher stake games and worse in the lower stake games so I have to have a neutral feeling towards the downswing as I do the upswing just to keep my head cool and concentrated to the fullest.

(1*)Calculating how many buy ins I should be up is pretty hard because the program I use 'Poker Copilot' calculates the hands in 'bb' (big blinds not big bets) and not in actual buy ins but I give 25bbs for one buy in seen as 25bbs = 1 buy in when you buy into the Hyper turbo Sng. I am not sure if this is mathematically correct due to the constant change in blinds in each Sng but for arguments sake I will keep it like this to keep it simple.

Ok so that's the maths behind the swing, what about the psychology.

Using all of these graphs we can actually see some interesting stuff that can help explain how my mind set may have been effected at the early points of the swing although the entire way through I have been pretty chilled out and am happy with how I have been playing there is some points you can probably see a change in my game with result orientated adaptions my brain made unconsciously through confidence and trying to reduce the variance. (2*) Obviously we cannot use graphs like these to predict how someone is thinking but I am going to break down the swing and explain some of what has been going through my brain through out the downswing.



(2*)A note to people: normally the things our brain tell us to do in poker to reduce variance will increase it more ie, reducing our 3betting range and waiting out a better spot when we have a clear ev+ spot in front of us trust in your decision making if it's been correct in the past.

Well I hope this can explain a better view to anyone who is questioning my downswing and also to the people who might be worried, will help them to understand the maths and the psychology behind swings like this. In the end this is a business with high variance at points, all we can do is grind and play our best at all times, we can only change the out come of the future, try not to hold on to the past. I also hope this helps others who happen to have a downswing like this in the future. Don't give up, don't make irrational decisions, don't be result orientated and understand these things happen and most importantly don't let poker control your happiness levels, let other things in life and the things that poker has given you decide on how happy you are.

9 comments:

  1. Interesting read, particularly how our instinctive attempts to reduce variance are flawed. Think I've been guilty of tightening my 3betting range when taking shots at higher stakes. Could you give any other examples of it?

    ReplyDelete
  2. Very nice blogpost, I got the same issues that you have countered the last few weeks or whatever it is, ''glad'' to see that I am not the only one, who has these specific kind of things.

    ReplyDelete
  3. Thanks for the comments guys. W1thecr0w there are many different examples of things people do as short term adjustments, not sure of examples to give you as could be anything from how fast you bet when your on tilt to how fast you bet when your in the zone, if you play a draw passive of fast might depend on your mood etc.

    ReplyDelete
  4. Cheers man. Not thought about how my timings might alter, tbh I've always been pretty sceptical of the worth of timing tells online. What sort of things do you look out for? Good luck coming through the other side of this patch.

    ReplyDelete
  5. i'm looking at graph 1 where u are mostly winning and wondering shouldn't your winning line be higher than the equity line not the other way around since u are actually winning. this what i understand all in equity to be about

    http://www.pokertracker.com/forums/viewtopic.php?f=45&t=32184

    ReplyDelete
  6. Hey I was just wondering, when I an downswing you feel like you are getting owned all over. How do you adjust to that, meaning:

    For example you have Q6, flop K56 or whatever, normally you would cbet I guess, but do you ever tend to check back those kinda hands when in a downswing?

    If you get the feeling they always have it, so you just rather check it behind?

    ReplyDelete
  7. Chipolino, you can be either above or under ev in a life time, 50% of people will be above and 50% of people will be under.

    To the last question. I would always expect my cbet to be just as profitable on a downswing as it is on an upswing, the play or cbetting 2nd pair if it is optimal on a particular board when I am on a heater then it is also profitable when I am not, nothing changes, only in your head. I will seek the most profitable line and take that one always. Don't let external factors change what you are doing. Math is Math

    ReplyDelete
  8. This is a really great post. Curious about your statement "50% of people will be above [EV] and 50% of people will be under". I think that is a misconception because all players do not deviate from expected value in equal amounts. For example, one person can run extremely far above expectation throughout the course of his career resulting in many of his opponents running under expectation. The opposite holds true as well, if one person runs extremely under EV that will lead to many of his opponents running above EV.

    ReplyDelete
  9. Yeah I agree with this but in a theory in would have to be 50/50 and if everyone in the world was to play poker it would average out at 50/50 as the math would add up over the extreme sample size.

    ReplyDelete