I started playing HU last month and got off to a bad start but things started to pick up towards the end of the month. I thought this month was going pretty well but then in the last 3 days I've gone through a 20+BI downswing and basically lost what I've worked towards so far this month. I don't really think its a standard downswing, but its possible, right? Feels like I'm constantly being coolered.
Are graphs even supposed to be this swingy?
Yes, not unusual - 600 games is a relatively small sample.
Run some simulations with the variance calculator to improve your objectivity in interpreting your results.
http://www.husng.com/content/husng-variance-calculator
Yes, not unusual - 600 games is a relatively small sample.
Run some simulations with the variance calculator to improve your objectivity in interpreting your results.
http://www.husng.com/content/husng-variance-calculator
I ran the simulations but what exactly does it represent? How do I read it?
Poker is a game of skill & luck.
You can control the skill element, but not the luck - it is random.
Random events can be modelled using math - the variance calculator is an experimental application of this math.
Imagine a random number generator which generates a result somewhere on a continuous number line between 0 & 1, represented as [0,1].
If you have a win rate, w, then [0,w] may represent a win and [w,1] represents a loss to emulate the binomial distribution of wins vs losses in HUSNGs.
On a number line this would look like:
[0 ---------------------- w ------------- 1]
The result of a single HUSNG is binary: you win or you lose.
The random outcome , [X] sits somewhere on this contiunous distribution.
An example of a win would be:
[0 ----------- X ----------- w ------------- 1]
An example of a loss would be:
[0 ---------------------- w ------- X ------ 1]
SO WHAT DOES THIS MEAN FOR INTERPRETING THE GRAPH?
The graph models experimentally a series of consecutive indepent random events with a random probability distribution function defined by the inputs you gave.
This emulates what happens when you play: your expected win rate is a function of your skill relative to your opponent and the extent to which this translates to a return on your investment depends on the size of your edge relative to the rake in the games you are playing.
Your expected return E(R) on each game can be calculated by taking your chance of winning multiplied by the payoff when you win (U) and adding your chance of losing multiplied by the payoff when you lose (D).
E(R) = w * U + (1-w) * D
For a given rake (r) as a percentage of each buyin (B) we have:
U = B * (1-2r)
D = - B
so then
E(R) = [w * B * (1-2r)] - [(1-w) * B]
or
E(R) = 2 * w * B * (1-r) - 1
HOWEVER, your winrate w, is also function of the random variable [X] above.
Playing with the variance calculator allows you visualise possible experimental results and how they can deviate from your theoretical E(R) due to the luck element of poker.
Excellent post Cdon.
The post may be great, but is no good as a response to OP. OP has no sample size in order to figure out trough math if it is a downswing or if he just sucks.
Hi.
I answered the poster's question about interpretting the variance calculator simulator by offering an explantion of the fundamentals underlying the simulator.
In my experience, understanding a model from first principles allows you to better understand the outputs.
The key take home message was: poker results are underpinned by a probability distribution with a small (relative to the amount of money in play) expected return and large variance. The variance calculator is an excellent tool for accepting this reality of poker.
If the poster was, as you suggest, in fact interested in statistical confidence regarding win rates this is covered comprehensively here:
http://www.husng.com/content/given-my-results-what-should-my-confidence-level-be-about-my-true-winrate-0
Regardless, I think we can both agree that players will get a lot more value out of working on their game to improve their expected return than tilting themselves worrying about the underlying randomness of poker.
probably we could get all downswing posts here and use the explanation above (that I like very much) together with these couple bis below graphs for couple hundred of hands and make special what is NOT a downswing check first thread :)
(no offence to the original poster I was the same when I started playing husngs)
just be aware that if this is hypers you have to understand that for people who play full time or are really dedicated 600 games means like 3days of grind. So to you it feels like a much bigger deal because you have spent a month working on this.