I will start off by saying I'm not suggesting anyone should actually do this. I'll also say I have never and have no plans to do such.

However, I've been curious as to what actually stops it, especially if it is in pgn format and not some custom format. If you write a book, you can copyright it because it is your created work. If you annotate a game, it is your annotated work that would be protected.

But, with chess games they are simply notated moves of an actual game. Anyone can notate those same moves, so how is some giant 10 million game database protected from someone just buying it and either putting it out there for free, or worse reselling it themselves?

Something I've wondered for a while. I think I first wondered it when I noticed that chess.com's game database seemed to be a lot smaller than other sites or ones that are sold and I thought "Why don't they just buy one of those 10 million chessbase databases and shove it in their game database?"

A specific database represents a specific unique collection and thus is covered by copyright protections.

Even though someone else could notate those exact same games since the games themselves freely exist? Wouldn't taking two different databases and combining them together create a specific unique collection?

No, it would not.

You have to actually DO something different and unique to be able to claim a copyright.  Combining the copyrighted works of two others is just stealing from both of them.

There are plenty of sources of free games and databases already, and some of them are rather well groomed.  Why resort to theft?

Assuming it is a hypothetical and not a real question, simply because why would everyone keep repeating the same work of notating 10 million games? It seems kind of silly.

If it is creating something new, like a story or a work of art, of course people should create their own. This would just be redundant as the games have already been accurately notated.

It also seems like there'd be some big loop holes in there that couldn't be covered. Like notating 10 new games and adding it into an existing group, there by creating a new and different group of games.

I have no stake or interest in this as I don't search through databases that much and when I do chess.com's is big enough for what I'm looking for. Just seemed like an interesting situation.

I suspect the database makers have inserted a fake game or two. If they show up in a database where they don't belong, they have evidence that their property has been misappropriated.

edit: best if annotations have been lifted, as bare game scores are not copyrightable. Collections? Maybe they are, in the same way as dictionary word-collections are protectable.

All the data bases you could ever want (at least for otb games) are out there for free.  What you need is software to search and filter them.  The search engines are patented just like any other software.  With chessbase software there is a serial number with each copy that you have to enter in order to install it.

Copyright laws have come up for discussion before. 

You cannot copyright a game or even series of opening moves in theory, because it's not unique and impossible to determine the origin. 

Simplified example but who invented 1. e4 ? Nobody knows.

As for collections of games, same problem - but the dog is probably right that fake games are inserted here and there, just as mapmakers like the Thomas Book makes intentional errors or omissions to detect copyright infringement. 

Of course if you happen to be looking for something on the map page that has the "error" you are screwed.

I didn't realize there were multi-million game databases out there for free, that would be interesting.

I fully agree and accept with the software that searches through those games. That is something that is created and therefore definetly owned.

 Maybe that is all that it is. Chessbase (and similar programes) creates a good search engine for game databases and then sells new and updated databases. They probably aren't worried about the databases themselves as the people who bought their software will likely buy the database too.

I own chess base 10.  I update it weekly with TWIC (the week in chess).  The biggest problem I have is that TWIC gives me more games than I want.  I am not that interested in games played for the under 10 years of age championship.  I prune the TWIC file down so that it just includes games where at least one of the players has a FIDE rating of over 2500, before I merge the file with my big data base.  The latest version of chess base contains a lot of correspondence games in its data base which would be nice to have, but so far I haven't been able to justify the expense.    I wait about 2 months before pruning and merging a TWIC files.  This is mainly so I can find games mentioned in Chess publishing.com.  The free portion of the updates on chess publishing.com usually tell at what point a novelty was played and some useful information on critical variations.

I make my own database of CC games by downloading them directly from ICCF. Their archives are very messy, so I spent several hours one day and combined every package into one giant database. Then I used chessbase to clean the entries and remove all doubled games. The final count was over 310,000 games, with 200,000 having been played in the last 5 years.

Thank you, I have started downloading.  Do you have any suggestions at what ratings are worth keeping.  I am thinking of eliminating anything where the ICC of either rating is under 2000.

Actually you want to keep everything from 2007 on, no matter what the rating. When I started my ICCF account back then, I found that often the new provisional 1800 players were playing at the same strength as the 2400+ players I faced. This is because people were using the same engines (Rybka, Houdini, etc.) and so the only difference was in opening preparation. I remember a 2100 I drew with quickly rose to nearly 2600 in rating, yet I knew all he was doing was using Houdini and just entering massive numbers of tournaments.

At what point does a collection of games become a database? If all "collections" (taking it to mean more than 1) were allowed to be copyrighted, couldn't you copyright, say, all of Paul Morphy's games? I really doubt that that's possible, so perhaps there's a proper definition of collection.

No collection of raw game data can be copyrighted. They are considered free public domain. As far as annotated collections go, I'm not sure what the law provides for those. They might be copyrightable, but I don't know that for sure.

Game scores can not be copyrighted that has been settled in court many years ago. Some IM and GM want games to be but the simple fact is they can not be, who can say if a game was produced by an amateur first or by a master
Annotations can be (how book authors protect their work) and are.

games are just data so depends on what you want them information for.

I would also keep games from say 2200+ rated up if your looking for opening refutations unless you want to computer analyze everything. at this point the games have some educational value especially if you want to look for how to beat silly mistakes.

also going strictly on rating can be a bad idea since sometimes databases dont have the right rating or no rating entered and you can prune games from older players like Lasker, Alekhine, etc. even some modern game databases the players 'forget' to enter a rating

Annotations can be copyrighted but the variations in the annotations are in the public domain.   I recall that Keene and Levy plagiarized wholesale from Erstin's book on the Two Knights Defense in their book An Opening Repertoire for the Atacking Player.  Their variations followed the exact same order and length as Erstin's.   I have left the original chessbase data base in tact so I have the historical games, it's only the TWIC updates that I prune for ratings.  The reason I only require one player to be rated above 2500 is so I can see how strong players deal with non-theoretical moves.

In terms of US law, it seems that games in themselves do not qualify (a little short, this analysis is probably a bit shallow, but its all I could find):

Melekhina, A. & Orkin, N., “Intellectual property issues in chess games” Journal of Intellectual Property Law & Practice (2011) 6 (10): 723-727.

Annotations are IP protected, raw games, scores, tables, and list of openings are not as suggested by FrbrdX. Similar to sports stats & scores for fantasy sports which the US SC ruled were public domain ~2 years ago.

I figured as much, which makes my question all the more interesting. What stops people from just distributing those databases that they bought from a company freely as all of the games, without annotations, wouldn't be copyrighted.

Or even better, what stops chess.com from getting one of those 10 million game databases and sticking it the chess.com one?