On the first point, Kurt, who now does consider himself professional, concedes it is "not a lucrative business." On the second, well, just you wait.

Over the following years, Kurt began to make a name as one of the best stone skippers in the world. Between 2002 and 2006 he held the record for the most bounces. Then in 2007 his worldcame crashing down. Someone wrested the t.i.tle from him with a stone that skipped more than 50 times.

"That triggered it. I realized I had to unpack everything I knew and start like I did not know anything. I had to try every nuance and concept." That was when he explained to his wife that he was going to quit work to become the finest stone skipper the world had ever seen.

The five years that followed are best understood by viewing them as an eighties movie montage, from Rocky, say-but with a wife occasionally in the background wondering why "Eye of the Tiger"

is playing and her husband is reading books about stones.

Kurt began training his body and honing his mind. He studied the scientific literature and a.n.a.lyzed his opponents' techniques. "I went deep into physics a.n.a.lysis-into geometry and force considerations. I started doing things with the stone that I have never seen explained in the literature."

There is a significant body of work on stone skipping. Most physics papers advise the stone is best angled to hit the water at a shallow angle. They offer advice on its tilt, its internal inclination, and its spin. But, says Kurt, they are all flawed.

"The models a.s.sume an idealized disc, in an idealized situation, with no air resistance. But the stone doesn't behave in that way when the disc is not symmetrical, the water surface is not flat, and there is air resistance." All of this means in the real world, stones do not behave in what he calls the "cla.s.sical" style.

"In the cla.s.sical style, you want the first bounce high because a cla.s.sically skipping stone is very much like a ball bouncing on a surface. Every bounce is lower than the last, so if the first is high the others will be, too." This a.n.a.lysis would have you tilt the stone at an angle so that it bounces up after the first hit.

The first clue that this might be incorrect came from listening. "You can tell when you throw a good combination of spin and speed because you can hear it-the stone sounds like a bird fluttering." Air resistance is not negligible at all.

"When I throw, I want the first bounce to come out almost horizontal. I aim to go in at thirty degrees and come out at five. Instead of, say, going in at fifteen and coming out at ten. In this way more energy is captured in the forward movement of the stone. As soon as you add vertical oscillation you also lose a tremendous amount of energy to air resistance."

Air resistance can also have its uses. In conventional models, the stone is a.s.sumed to be thrown in a parabola, pulled down by gravity and curving to its first contact with the water. He has found that his stones sometimes curve almost the other way. Such is the speed and spin that, "in practice, wingand propeller forces can lengthen the travel path as the stone glides into contact. It depends on the shape and weight of the stone, but even an aggressively steep throw can impact at angles closer to the horizontal." The stone is quite literally flying.

"By embracing and harnessing complexities, I've been able to make a stone skip in ways that existing descriptions do not even consider. At least, that is my feeling; I have not had my specific concepts reviewed or tested by a professional. I've asked, of course, but it's like they have better things to do!"

There are five variables you can control in stone skipping, of which the angle of entry is just one.

The others are simpler, though, at least in theory. There is the stone's own tilt-how level it is- which in Kurt's model should be kept close to flat, but never pointing down. There is its "twist"-its leftright inclination. This also needs to be flat. Then there is spin and velocity, which both need to be as high as you can make them.

It is achieving the latter that is the hard part. "Making it impart the particular forces you need is not simple." This should not be surprising. "If you get someone off the street to throw the hammer or discus, they're not going to be able to do it. It's about training your body to a very specific action."

He says most of his colleagues have had sports injuries, including some necessitating back surgery.

After his five years of investigation, Kurt realized there were two styles that work: whip and drive.

He is a rarity in that he can do both. "Drive is more like a baseball throw. A lot of strength comes from the back legs, lunging forward and low. In whip you are contorting more, throwing the stone down at the ground almost at your feet." It is with whip that he uses a thirty-degree angle rather than the shallower approach that the literature recommends. "The shoulders are rotating in such a way that they make a whip out of the arm-at the last second your arm forces that stone forward." He demonstrates both throws in a YouTube video t.i.tled "Whip vs Drive."

Both approaches have their merits. "Whip amplifies spin. Drive amplifies velocity. With drive you can throw a heavier rock faster, but with not so much spin." His main piece of advice to novice skippers is to just experiment, to find a style that works for them, as he did. "Don't be afraid to do something that feels uncomfortable. If you only do what feels natural you may not hit on the best actual method for you."

He had learned to throw by using drive, but on the day Kurt felt ready to try for the world record, he chose whip. It is worth watching what happened next on YouTube.

This throw was Kurt's gift to the world, his Sistine Chapel. The rock seems less to bounce than to float. Lifting off the water after the first impact, it temporarily takes flight, then returns to break the surface of the lake in batches-rat-a-tat-tat, rat-a-tat-tat.

On that day, everything went right-forty bounces, fifty, and still it continued. It was for throws like this that Kurt got into stone skipping. "The great thing is it's so primal," he says. "In this world, where everything is so connected and everybody is so focused on some device in their hand, I find it essential to replace that with a piece of the earth, a rock."There's nothing more basic or satisfying than that. Basically, you're making a stone float. You're in control of that magic. It's like watching a fire or a sunset; it's putting a person in physical contact with something millions of years old." In all those millions of years, though, no one had done this.

Sixty bounces, then seventy. A slight corrugation of ripples on the lake helps it far past the previous record-the undulations seemingly coming up to meet each bounce. It doesn't so much rebound from a flat surface as slice off the tops of successive peaks. Eighty bounces, then rat-a-tat-tat, eighty-eight.

The stone reaches skipping apotheosis, seemingly giving up on bouncing at all and just sliding along the water instead until, victorious, it at last sinks beneath the surface.

It was as close to an unbreakable record as most stone skippers could imagine. All Kurt's work had been worthwhile-for him, at least.

Did his wife agree? Well, he says, these days he is in semiretirement, and when he talks about skipping, "she says, 'Hey, I thought we were done with that now?'?" She is, he says, "inscrutable . . .??

I.

think she likes to see me successful in the end. On the way there she has been a little bit long- suffering. She is . . ." he thinks of a better word, "tolerant."TWENTY QUESTIONSWHAT'S IT ALL ABOUT?

Animal, vegetable, mineral, or game? Game.

Does it involve a board? No.

Does it lead to intense and persistent feuds? Not normally.

Can you do it outside? Yes.

HOW DO YOU PLAY?.

Does it require a stick? No.

Is it stone skipping? No.

HOW DOES IT END?.

I give up. It's twenty questions.

But you don't do that outside. You don't, but you can if you want.

You're an idiot.

a.n.a.lYSIS.

Robin Burgener has a pretty good idea about what you are thinking of when you play twenty questions. "You can almost guarantee it," he says. "People new to the game will go for a cat, or a dog. Then they will think of a vegetable-usually a carrot." Amazed that Robin's online computer program manages to guess both, they will then resolve to stop going easy on it. "They'll make up something really difficult, like a duck-billed platypus. Then they're blown away when it gets it in ten goes. But, really," he says, "how many egg-laying mammals are there, covered in fur?"

In 1988 Robin designed a computer program to play twenty questions. People using it think of an animal, vegetable, or mineral and, just as in the parlor game, it asks questions in an attempt to guess it. These days, it normally succeeds.

When it started, though, it was only in extremely specific circ.u.mstances that it would defeat the human. Namely, if the human was thinking of a cat. "All it knew was that," says Robin. Slowly, over time, it learned from its losses-starting with the most common subject. "Next came a dog, then a carrot." Then, presumably, a duck-billed platypus.These days, twenty-five years later and after being made into a website and an app and used for millions of runs, how many things does it know? Well, a bit over 5,000. Out of all of the world's billions of animals, vegetables, and minerals, it seems, people who play twenty questions are only ever choosing one of 5,000 of them. Even when people are trying to be obscure, they are obscure in a predictable way.

In fact, it's even more limited than that. "Ninety percent of games draw from fewer than 100 objects. The top 1,000 objects gets you one or two percent more."

In theory, this means 90 percent of games should end in seven guesses: the most efficient search strategies, binary searches, eliminate 50 percent of options each time. If you start with 100 options, after one guess you have 50, then 25, and so on. So, to play twenty questions successfully, when looking at all the likely objects remaining, each guess should try to get rid of half of them. This is not the way a human works. "When guessing, a person narrows it down quickly to one thing in their head and tries to ask questions about that. It's very much human nature to fixate on one thing and try to prove and disprove it. A machine's ability to question and remain objective is so much better-it can hold 1,000 options in its head."

While a human, having established they are looking for an animal, might try to find out if it is a reptile, amphibian, mammal, fish, insect, or bird, say, a more efficient question might be to find out if it lives in the sea or above the sea. That way, whatever the answer, you have gotten rid of about half the options. If you ask if it is an insect and the answer is no, you have only eliminated around a seventh, which is hugely less efficient.

Using a binary search, if people are drawing from 1,000 objects, in theory 10 questions should find the answer. With 5,000, it is 13. With twenty questions, you can guarantee correctly guessing anything from over one million objects.

In the real world this does not always work. a.s.suming you are not allowed to ask about its spelling ("Does it come between Aardvark and Mantis in the dictionary?") there is not always an obvious way to divide a group in two. A few simple tactics can help, though.

The first, which often trips up beginners, is to use comparators efficiently. If you ask, "Is it large?"

unless it is either an atom (definitively small) or the universe (definitively big), almost any answer is correct. A stag beetle, for instance, is big for a beetle, but small for an animal. So instead ask if it is larger, say, than your head. Then there is no ambiguity, no wiggle room for a pedantic guessee, and you are able to create an arbitrary cutoff point-much like specifying Mantis in the above alphabetical example-through which you can divide the remaining objects in two and get closer to a binary search.

The second is to use words like usually. If you ask someone who is thinking of a sheep, "Wouldlions eat them?" the answer has to be Yes-lions, especially in zoos, have probably eaten sheep and, even if they haven't, given half a chance they certainly would. This will lead you down a cul-de-sac of asking about African herbivores. Ask instead, "Would lions usually eat them?" and the answer has to be No.

All of this a.s.sumes your guessee actually knows what they are talking about, mind you. This particular problem could be called "the dolphin constraint." Robin, whose computer learns from guesses rather than encyclopedias, is familiar with the issue. "I can ask our program, is a dolphin a fish? It's pretty sure it is a fish. Most of the people playing the game think that a dolphin is a fish." So even if you were able in theory to eliminate half of all options at each stage, you might find at the end that when you had carefully removed all the mammals, the dolphin had persisted: a Linnaean taunt, in parlor game form.

There are two ways to respond to this. One, fifteen questions on when you are trying to work out how a turbot could have starred in a hit 1960s drama series called Flipper, is to declare yourself the victor and the guessee an idiot. This is how many games, and friendships, end in the real world.

Robin's website has a more liberal approach. "The program is not an authority on knowledge, it is an authority on human knowledge. We call it folk taxonomy." It accepts that there is a 60 percent chance of a dolphin's being a fish in the mind of the guessee and never quite eliminates it. Which seems like a far more sensible approach. After all, what could be more satisfying than defeating your friend at twenty questions and then adding to the triumph by explaining to them that a dolphin is a mammal?EPILOGUE I learned the true meaning of a family Christmas when I was eleven or twelve. Gathered with my relatives in Ireland, we were playing a board game, I forget which. What I do remember is what happened when I lost, and my uncle Terry won.

Sitting in front of a blazing fire, gla.s.ses of port for the adults and comforting hot chocolates for the children, Terry looked at me-a nephew he had known since the day I was born, who had played with his own children and spent holidays running in his garden.

Then, his performance began.

"Ha ha, I won, you lost," he sang (yes, he actually sang). He leapt on the sofa and bounced up and down, with each bounce pointing at me and singing "you lost" for emphasis. Finally, he sat down, brushed himself off, regained his composure, and asked for another game. No one pa.s.sed comment.

Terry was far from the only uncle who a.s.sisted in the creation of this book. There was Uncle Ken, who structured his year around his annual Christmas quiz. In his case the pleasure was not in winning -he was the quizmaster-but in devising questions so devious that we were all losers. Then there was Uncle Simon, who introduced me to Diplomacy, a game few people have ever won, but whose sine qua non is forming friendships and alliances with more trusting nieces and nephews, then brutally double-crossing them after twelve hours of game play.

It is always invidious to single out people in an acknowledgments section, but equally it would be wrong not to mention the important role my cousins and my own sister had in forming the philosophy behind this book. It was my cousin Olivia who first introduced me to the delights of daylong games of Monopoly. It was her brother, Christopher, often in alliance with my sister, Clare, who showed me how such games ordinarily end: with someone tipping up the board.

But my final lesson in gamesmanship, one that I save for another book, came not from relatives but from my neighbors. For much of my childhood we lived next door to a gang of unruly children named the Salisburys. They taught me that, ultimately, the real victory comes not on the game board at all; that is merely a prelude. The real victory comes to whoever can afterward put his opponent in a headlock, wrestle him to the ground, and then sit on his face.

Uncle Terry: watch your back.ACKNOWLEDGMENTS It is always the case that a lot of people are involved in the creation of a book. That is especially true here. Each chapter required the time, effort, and experience of experts in the field. Not only were they kind enough to talk to me about their jobs and lives, they were also kind enough to do so in the cause of a book about games.

One of the privileges of my day job as a science writer is that I get to chat to people about their life's work and distill it into 600 words. I long ago realized that someone who is pa.s.sionate about quantum mechanics, or snails, or stone skipping, is a hugely more interesting interviewee than an A- list actor or a celebrity chef. That theory was proved here.

My only regret is that two experts were left out: Daniel Sheppard on Rubik's cube and Michael Bowling on poker. With both the fault was mine. They were engaging and knowledgeable-it is difficult to think, in Sheppard's case, of how I could have found someone more qualified than a man who can solve a Rubik's cube blindfolded, in under a minute-but my own talents were lacking: I was unable to transfer their expertise into something comprehensible on the page.

Jamie Joseph at Ebury was all I could have hoped for in an editor: just enough praise to keep my fragile ego happy, and enough gentle suggestions for structural improvements hopefully to keep readers happy, too. He is responsible, along with designers at Clarkevanmeurs Design and ill.u.s.trator Tiffany Beucher, for overseeing the beautiful final look of the book. Helena Caldon helped get rid of some of the sillier mistakes. Both she and Jamie should be thankful to my wife, who saved a lot of editing and probably a lot of ego ma.s.saging, too, by eliminating the more self-indulgent pa.s.sages and clunkier phrases. Without Sarah Williams this book simply would not be here. She is the perfect agent: efficient, kind, wise-and she doesn't wear red trousers or drink Nespressos.

Finally, I would like to thank The Times of London. Early in my career one of the editors gave me a piece of advice. He said, "Whenever important people agree to interviews or invite you to parties, remember they are not talking to you at all. They are talking to The Times."

I owe everything to the paper. Far too many people at Times Towers have helped me get to the stage where I could pitch a book and have a publisher notice for me to name them all, but it would be wrong not to mention several: Ben Preston, Martin Fletcher, and Bronwen Maddox, who gave me my first job; Roland Watson and Emma Tucker, who helped me get my second; and Nicola Jeal, who for a while kept me occupied every Friday finding trends in male beauty products, Great White Sharks in Cornwall, and even self-awareness in Jedward.

It was James Harding who moved me to science, and John Witherow, Fay Schlesinger, and Jeremy Griffin who decided to promote someone whose milieu is research papers about seals having s.e.x with penguins and why dogs defecate in a north-south direction. I am hugely grateful. It is the homenews desk-Dan Parkinson, Mark Sellman, James Burleigh, Andrew Ellson, Devika Bhat, Claire Bishop, Dee Howey, and Robin Stacey-who have to live with the daily consequences of that decision. Lastly, ma.s.sive thanks to Mike Smith and Michael Moran.

Any mistakes are mine and mine alone.

APPENDICES.

APPENDIX 1: QUIZ ANSWERS.

1. Aldrin 2. Saint Barthelemy 3. Arachne APPENDIX 2: TWO-LETTER SCRABBLE WORDS.

All words are in Merriam-Webster's Official SCRABBLE Players Dictionary (4th Edition). Lists compiled by the North American SCRABBLE Players a.s.sociation (www.scrabbleplayers.org).

aa (Hawaiian) a type of lava > AAS.

ab An abdominal muscle > ABS.

ad (Coll.) advertis.e.m.e.nt > ADS.

ae (Scots) one. No -S.

ag (Short for) agricultural; (noun) agriculture > AGS.

ah Interjection expressing surprise, joy, etc. > AHS, AHING, AHED.

ai (Tupi) the three-toed sloth > AIS.

al An E. Indian shrub > ALS.

am Present tense of be.

an Indefinite article; (noun) something that might have happened but did not, as in ifs and ans > ANS.

ar The letter r > ARS.

as In whatever way; (noun) a Norse G.o.d > AESIR; a gravel ridge or KAME > ASAR; a Roman coin > a.s.sES.

at Preposition denoting position in s.p.a.ce or time; (noun) a monetary unit of Laos > ATS.

aw Interjection expressing disappointment, sympathy, etc.

ax (US) axe > AXES.

ay (Noun) an affirmative vote > AYS. Also AYE.

ba In Ancient Egyptian religion, the soul > BAS.

be To exist.bi (Short for) bis.e.xual > BIS.

bo Fellow; pal, buddy > BOS.

by Beside, near; (noun) same as BYE > BYS.

de From (as used in names).

do A musical note: DOS; (verb) to perform > DOES, DOING, DID, DONE.

ed (Short for) education > EDS.

ef The letter f.

eh Interjection expressing enquiry; (verb) to say "eh" > EHS, EHING, EHED.

el The letter l > ELS.

em The letter M; a unit of measurement in printing > EMS.