Pokemon Showdown is the number one Pokemon simulator in the world. Every month, we log several million battles.
Primarily, we use these logs for determining usage statistics, but there's so much more information in these logs just asking to be distilled.
Data mining has taken on a negative connotation in recent months, but data mining is just a set of tools, and data scientists are simply simply scientist, intrepid explorers delving into the deep murky waters of complex systems to try to find golden gems of insights.
As Smogon's resident data miner, I attempt to produce these gems each month in terms of metagame and moveset breakdowns, but that's really just the tip of the iceberg.
With that in mind, I would like to provide people with the opportunity to ask me specific questions that can be answered through data mining of the logs but aren't suited for scripting into something I would want to post every month and for every tier. The idea would be that I'd pick a few of these questions and answer them in an occasional (every issue?) feature of The Smog.
Examples might be: "Who is the most common scarfer in OU?" "What is the most used move in NU?" "In Ubers, what percentage of Earthquakes end up doing no damage due to immunities?"
Let me know if this idea sounds like something people might be interested in.
Primarily, we use these logs for determining usage statistics, but there's so much more information in these logs just asking to be distilled.
Data mining has taken on a negative connotation in recent months, but data mining is just a set of tools, and data scientists are simply simply scientist, intrepid explorers delving into the deep murky waters of complex systems to try to find golden gems of insights.
As Smogon's resident data miner, I attempt to produce these gems each month in terms of metagame and moveset breakdowns, but that's really just the tip of the iceberg.
With that in mind, I would like to provide people with the opportunity to ask me specific questions that can be answered through data mining of the logs but aren't suited for scripting into something I would want to post every month and for every tier. The idea would be that I'd pick a few of these questions and answer them in an occasional (every issue?) feature of The Smog.
Examples might be: "Who is the most common scarfer in OU?" "What is the most used move in NU?" "In Ubers, what percentage of Earthquakes end up doing no damage due to immunities?"
Let me know if this idea sounds like something people might be interested in.
