Using sharkscope
Sharkscope.com is a powerful tool you should be aware of if you play online poker seriously. Sharkscope essentially tracks every user’s performance in sit’n'go tournaments (sngs) for several sites, including the Ongame network (which includes several popular sites such as pokerroom, bwin etc), full tilt, etcetera.
There are two principal uses for this nice tool. The thing I like the most is to be able to track my own performance. I do not need to write the results down after any sng or to use any pesky spreadsheet to calculate my global wins (or, god forbid, losses), average winnings and global return over investment. They also provide nice graphs where you can see how you’ve been doing across time, your current trends, and your performance in each level of stakes (e.g. you may notice that you win at $5 sngs but you lose at $10).
Here is a sample of the information I extracted from a shark, whose nickname I won’t disclose:
The second use, and probably the most spectacular is that you can check the stats for your opponents with the same amount of detail I showed above. This can lead to a great advantage. For example, if your opponent is strong, you know you must be more careful. If your opponent is a fish, you can avoid risking too much in every hand since you want to reduce the variance - you will show a profit sooner or later - or you may concentrate on finding an obvious weakness.
As usually happens, things are not so good and easy as they seem. Sharkscope place a limit of 5 free daily searches. Therefore you can not search for all the opponents in a tournament (let alone all the tournaments you play during the day). That is, for free. If you pay, you get more. The other major complaint about this site is that they do not have a 100% accuracy. That is, sometimes they miss some data and the graphs are not accurate.
However I do not agree with this complaint, because the important thing is not to have precise data, but a general idea about the strength of a player, his current form, trends in your results, etcetera. And you can achieve this by having incomplete data.