"I don't think it's really possible to create a good database with member uploads"
Chessgames.com is probably the finest database around, not because of quantity, although it's not small, but because of a) the general quality of the games included and b) because each game, or each player, has a page where the game or player can be discussed. Most of the games at chessgames.com were uploaded by members.
What i dont understand is some people feel the need to oppose the call for a database thats atleast on the level of already free DBs like 365...if you dont plan on using this chess.com DB or cant because of being nonpremium, why do you care so much that you oppose it?
Who opposes to what? If the paying member happen to receive a brand new great DB, i'm glad for them :) the thing certainly doesn't bother me! I was just opposing the view that such a database is easy to create.
By the way, is chess.com DB that bad? I've never used it, but i know that chess365 is already pretty bad!
bresando The Game Explorer DB here is a lot better than nothing, but it's about 500,000 games the last time I counted, and not by 1-2-3-4 etc :=). By way of comparison, my CB BIG DB 2011 DB was about 3.5 million games "out of the box" and is now upgraded to 5.3 million games.
CB BIG DB 2012 is about 5.1 million games "out of the box"
http://www.amazon.com/gp/product/B006H2N4IA/ref=olp_product_details?ie=UTF8&me=&seller=
"I don't think it's really possible to create a good database with member uploads"
Chessgames.com is probably the finest database around, not because of quantity, although it's not small, but because of a) the general quality of the games included and b) because each game, or each player, has a page where the game or player can be discussed. Most of the games at chessgames.com were uploaded by members.
Yes, but the people at chess.com would surely upload unreal games. There are so many trolls here.
Even though the games we play are fishy games, I think they do add good information to the database. Its just another chess situation that gets added with some kind of result. If a crazy guy plays 1. a4 and loses, then its another statistic. If a crazy guy plays 1. a4 and wins, the its also another statistic.
I don't think the problem(s) being referred to here regarding member submitted games is (exclusively) the fact that they're fishy, but that the DB will end up with incorrect: game scores, player data, duplicate games, triplicate games etc. And yes someone playing 1.a4 is "also another statistic" but most of us are more interested in what Kramnik and Anand are playing, IMHO.
The default GAME EXPLORER DB seems to be comprised of "Master Games" when you go to look something up in it BTW.
Even though the games we play are fishy games, I think they do add good information to the database. Its just another chess situation that gets added with some kind of result. If a crazy guy plays 1. a4 and loses, then its another statistic. If a crazy guy plays 1. a4 and wins, the its also another statistic.
this is NOT statistics. contrary to what people think statistics are not just random gathering of information. this is infact is counter to what statistical research is about. the point is to gather important quality information from a large quanity of information to make assessments about the expected outcome. having poor players win with 1 a4 is crap information if the level of play is poor. garbage-in garbage-out.
You should create a group to discuss this matter. It'd be interesting to see just how many people really want a huge database from chess.com. I know I am certainly awaiting the feature anxiously.