A Couple of New Toys

[uDevil]

I've added a couple of new programs to the software section of my poker page.

This one is a bankroll calculator



This one will let you see the effects of variance on your win rate and bankroll



These are both Java applets, so you need to have the Java Runtime Environment (JRE) installed to use them. I've only tried them on Firefox, but they should be fine in IE. They are not quite in final form, but you're welcome to play with them.

If you have any trouble, let me know by responding to this post or send me a pm.

[Gonso]

Both very neat. Only thing I noticed was that you have what looks like dark blue text and a blue background. Had a hard time reading it.

Thanks though, very cool.

[uDevil]

[ QUOTE ]
Only thing I noticed was that you have what looks like dark blue text and a blue background. Had a hard time reading it.

[/ QUOTE ]

Sorry about that. It's black on blue. I have no sense of aesthetics, so maybe I'll have to consult someone who does. How about red on orange? [img]/images/graemlins/smile.gif[/img]

[Ludanto]

Those are very nice applets.

Some questions:
1) win rate is callculated per (the usual) 100 hands?
2) what is the meaning of "confidence" here?
3) is it possible to explain standard deviation in very simple terms) (such as for pot equity: if your pot equity is 35% it means that on average you will win 35 games from one hundred and lose the other 65 if you play the hand until showndown for all 100 hands. Something that is easy to imagine.)
If would be nice to have an easy explanation of standard deviation.

[uDevil]

[ QUOTE ]
1) win rate is callculated per (the usual) 100 hands?

[/ QUOTE ]

Yes. This is explained in the notes below the applet. Maybe I need to put it on the graph.

[ QUOTE ]
2) what is the meaning of "confidence" here?

[/ QUOTE ]

Let's say your win rate is 1 BB/100. You play 100 hands. Do you expect to win exactly 1 BB? On average yes, but usually you won't. Sometimes you win more, sometimes less. You win
1 BB +/- "some amount".

If we assume results are normally distributed, then we know that 95% of the time, you will win 1 +/- 1.96*SD*sqrt(N/100) BB.

This is the 95% confidence interval. If you play many sets of 100 hands, in 95% of those sets, you can expect to win an amount that falls within the range defined by the confidence interval. This range is between the red lines on the graph.

[ QUOTE ]
3) is it possible to explain standard deviation in very simple terms)

[/ QUOTE ]

I can try. It is a measure of the spread in your results-- the swings between the amounts you win in 100 hands, and the amounts you lose. Roughly speaking, it is the "average" size of those swings.

You can find a technically correct definition at

http://en.wikipedia.org/wiki/Standard_deviation

[AlienBoy]

[ QUOTE ]
[ QUOTE ]
Only thing I noticed was that you have what looks like dark blue text and a blue background. Had a hard time reading it.

[/ QUOTE ]

Sorry about that. It's black on blue. I have no sense of aesthetics, so maybe I'll have to consult someone who does. How about red on orange? [img]/images/graemlins/smile.gif[/img]

[/ QUOTE ]


A substantially lighter blue background would be fine.

Aesthetically, you never want a BG element to be more than 20% of the foreground (text) element - I.e. if the TEXT is BLACK (100%) then the BG should not be darker than 20%. Alternately, if the text is WHITE (100%) then the BG should not be lighter than 20% of that level.

For black text (#000000) you don't want a bg darker than (#66CCFF) as an example.



AB

[Gonso]

Or, you could just make the text white

[uDevil]

[ QUOTE ]
For black text (#000000) you don't want a bg darker than (#66CCFF) as an example.

[/ QUOTE ]

Thanks, AB. I changed it.

[Ludanto]

Thanks for the explanation uDevil. Good job with the applets [img]/images/graemlins/smile.gif[/img]