I stared in awe at one of my favourite (coal) stocks. It was up 15% yesterday:      
What will it do today?

>Drop like a rock!

I made up a spreadsheet that downloaded a bunch of daily prices.
I calculated how often a 15% increase happened.
I checked to see what happened next.

>It dropped like a rock!

Pay attention! The spreadsheet looks like this (where you can click on the picture to download it):

You can ask for a particular daily gain ... like 20% (as in the above picture).

See Nov 26, 2008? It gained 20% (or more) and continued up the following day!      
>See Sep 19, 2008? It dropped like a rock!
So what'll it do tomorrow?
>Your guess is worser than mine.
I'm betting on something like this

I pray for this

I got this!

After playing this game I realize that I'd like to see the open-to-close variation, not the close-to-close.

>I coulda told you that!
Okay, so here's another spreadsheet with candlestick charts:

The six charts have the up-big jump as the middle candle ... and it's green ... of course!
In the above example, you'd have made money 40% of the time by buying at the Open and selling at the Close (following an up-big jump).

>Just 40%? That's pretty lousy!
You can, if you wish, borrow this.

Note, however, that you can get an up-big gain one day
... then open higher the following day but drop during the day.
If you bought at the Open and sold at the Close (the following day), you'd lose.

On the other hand, it might start up the following day and end even higher (after the up-big jump).    
If you bought at the Open and sold at the Close (the following day), you'd win.

>Why only six charts? Why not ...?
There ain't enough room for more, but if you wanted more recent charts just change the Start and End Dates.

You can also move from one chart to the next with that slider. See it?
>What happens when you try to go past the last chart?
The world comes to an end.