Probably most of you already understand what data mining is and actively use if for your everyday grinding sessions, but I will just go over the term for those who are less than familiar with it.
Data mining in online poker is the process of collecting hands that have been played by other players and collected by specialist data mining centers, for example HHMailer.com .
Bundles of this mined data is then available for purchase. How this is done is not revealed and strictly is not relevant – what is important is the amount of information, usually millions of hands, that is quickly sold to serious poker players who mostly play on the low and middle limits.
Why does this happen?
Within present-day poker there are normally thousands of players all hanging around on the same limit (let’s say, for example, NL50 PokerStars). You can gain an advantage by knowing the common tendencies of the field and adjusting your play accordingly. Here mining is a true aid, as it not only helps you to identify at once the kind of player who is sitting with you at the table, but it also allows you to scrutinize his play outside the table effectively, in order to exploit his mistakes more efficiently.
Especially timely is the following aspect of mining:
While you are multi-tabling it is extremely difficult to follow play closely enough, or deal with rare situations effectively. This information often determines the outcome of a deal. Thanks to data mining, the player has not only the basic stats at his disposal (such as VPIP/PFT/3Bet), but also everything else there often is to know about somebody’s style of play. This includes the rate of three-barrel bluffing in 3-bet pots or the check-raise percentage on a dry pair flop.
All this data can be gathered and analyzed after uploading the obtained data mined on your favorite tracker and filtering on the situations that interest you. In general, it concerns the game analysis of most of the better regular players on your own limit in order to improve the quality of your play.
Apart from anything else it becomes significantly easier to move up to a new limit if you don’t feel uncomfortable because of the unknown quantities of the players on your new level. All the information you need is there before you. How you use this information is a different matter, but that is for another post.
All in all, data mining in present-day online poker became, as they say, a «must have» and the standard some time ago (if your most-clicked button in your poker client is not «Deposit» anyhow). If you play on some limits without preparing yourself first with mining, it’s akin to sailing a ship in a fog bank without any radio positioning equipment.
In conclusion, it would be worth a few words about poker rooms’ attitudes to stats collecting and the usage of data mining in general. The official position of many of the largest poker rooms towards mining is that they do not support (and indeed, in some case, prohibit) data capture and its use while playing at their tables, but mining does have a direct relation to the amount of tables played by a particular grinder, and – importantly – the amount of rake that is generated. In reality, you are not likely to be penalized by purchasing mining and using it, and you certainly won’t be penalized by using it for game analysis outside of the poker client. You can draw your own conclusions.
In the next article we’ll discuss certain mining types, and analyze the typical rate of average regular players on NL25-NL100 tables. Stay tuned!
