Return Home

The Rules Can Set You Free

Back to Episode

Chess board Chess board (b3d_/flickr)

Play is something we all do--it seems so natural, it feels a little bit ridiculous to ask why we need it. But psychology professor Alison Gopnik explains some profound benefits, and tells us about a noticeable shift that happens somewhere between age 3 and 6--a shift that causes a kind of tug-of-war between total imaginative freedom...and the rules.

Brian Christian, who we talked to for our Talking to Machines show, weighs in on this same philosophical battle with an example from the 1863 World Championship of Checkers...where every single game in the 40-game series was a draw. According to Brian, checkers reached a point where all the best moves were known, and top-notch players realized it was possible to play an ideal game by using a set of plays known as "The Book." And this pretty much killed the sucking out all chance for improvisation.

Then, Frederic Friedel, the cofounder of ChessBase, tells us about creating a database that's so widely used...his critics call him the "man who ruined chess." But according to Frederic, and Bobby Fischer biographer Frank Brady, the sinister studiousness of The Book in chess is counter-balanced by a totally unpredictable realm of possibility known as The Novelty.

Further reading:

Alison Gopnik: The Philosophical Baby

Brian Christian: The Most Human Human

Frank Brady: Endgame: Bobby Fischer's Remarkable Rise and Fall


Brian Christian and Alison Gopnik

Comments [16]

Rud from USA

Play free funny Panda Checkers for android

Mar. 03 2016 02:33 AM
dave from seattle.

Great story on 'Book' in chess, ruined by the horrible attempt at notation. Seriously, the pawns don't start where you are describing and castling is not pieces switching places.
Oh, and for 'Dirtypants', the answer to your question of how there can be more moves stemming from one position than from an earlier one is transposition. A position develops that is the same as a different, more popular opening or variation of an opening. This means that the pieces got to the same squares, but in a different move order.

Dec. 30 2013 12:20 PM
Scott from Austin, TX

You should check out the game of Go some time: much more popular than chess (at least in Asia), much older, vastly more move and game possibilities, a much larger "Book", and yet much simpler rules. Students and pros relish studying the vast collection of recorded games because there is no chance of repeating even part of one. Best of all, there are no draws.

Dec. 03 2012 01:45 PM

I don't understand how the calculation of the moves / board layouts in the "live" chess match could be correct. They should be in a consistent decline (or static), given that you can't make the incidence of a particular countermove happen MORE often than the moves that came before it, right? Yet between move 7 and 8, the previous occurrences were 2428 times and then 2613 times, respectively. Does this mean that the counts is of the actual MOVE being made independent of the previous moves? Or was there an editing / calculating error? I don't see how that dataset would be interesting if it was just any time that a particular move of a particular piece had ever been moved unless it was in the context of the board layout.

I'm guessing editing error, calculating error, or an error on the part of the interviewee (misspeaking, etc.).

Can someone explain?

May. 19 2012 05:04 PM

Heya Nikolas! Sounds interesting, can I read it too?

odo AT duke DOT edu

Mar. 15 2012 12:10 AM
juniorloks7 from gilroy


Jan. 20 2012 01:24 AM
Daniel from USA

Hey Nikolas from the NL - I am very interested in reading your undergrad thesis on this .. is it available? I am daniel.bilar at gmail

Thanks a bundle


Dec. 30 2011 02:12 PM
Nikolas from Netherlands

Hey, that was part of my (undergraduate) thesis! I took things one step further though: The way you present it makes is seem that "The Novelty" arises if the state-space of a game is large. That's not really true though. If you take checkers you'll also probably get an astronomical figure (probably not quite as big as chess, but still).

The point is that for the Novelty to arise you need more than a large state-space: you need an incentive for investigating it. In my research, I represented this as an information problem: the harder it is to compress the state-space, the more information there is to it, the more interesting it is to play the game.

Dec. 07 2011 05:53 AM
Ge Yu from Washington DC

I was recently in NYC for a workshop
at the New York Film Academy. One of the tasks of the workshop is that
every participant must make a short film with extremely short
turnaround time, demonstrating visual story telling (i.e. no sync
sound dialogues). I thought about the Games episode of RadioLab I
recently heard and made my movie based on that. I want to share it
with you guys and tell you I really appreciate the works you guys do.
It has been a source of inspiration for me for many years. Here is the

Oct. 08 2011 11:13 PM
Jose from San Diego

Since they were talking about "the book" and Bobby Fischer, they should have mentioned that Fischer felt that book openings were destroying the creativity of chess. To remedy this, he invented "Fischer Random" chess, where the pieces on each side are randomized at the start, so that openings can't simply be memorized. Near the end of his life, he refused to play regular chess and would only agree to play his "fischer chess".

Sep. 27 2011 10:59 PM
William Moore from Pittsburgh, PA, USA

Xochipilli - I found a reference to the 1863 World Championship at

Starting in 1863 the openings used in competitive checkers were restircted to 43 possibilites and randomly selected. Two games would be played, with each player using each side of the random opening once.

Sep. 23 2011 01:34 PM
Xochipilli from Paris, France

Great podcast! About checkers I'm very surprised to hear that the game was "solved" back in 1863. Actually, checkers have been solved only in 2007, and only "weakly solved", meaning that although every possibility has been explored (500 billion billion possible positions), there isn't any algorithm yet, able to know what the perfect move should be, at any point of the game. I couldn't find any reference on the web, about this famous 1863 World Championship of Checkers. Anyone knows more about it?

Sep. 03 2011 12:54 PM
Derek Bruff

There are terms for the two different kinds of gameplay Dr. Gopnik described, ludus (structured gameplay) and paidia (freeform gameplay). I'm kind of surprised you discussed these two forms and the tension between them so much without referencing the terms themselves!

See for a bit more on these two forms of gameplay.

Sep. 01 2011 04:26 PM

Andrew Bird "Tenuousness"

Aug. 31 2011 02:17 PM
Brandon from Boston

Does anyone know the music that was playing under the girl's story about mice in Antarctica?

Aug. 30 2011 10:12 AM
Amber Bobnar from Boston

The little girl with the imaginary friend is adorable, but did anyone notice that she has a friend who lived in antarctica, moved to the moon, and dines on house mice? My favorite album of all time is The Moon and Antarctica by Modest Mouse. Maybe her daddy listens to this too?

Aug. 28 2011 08:09 PM

Leave a Comment

Email addresses are required but never displayed.