How well can we, at Less Wrong, tell the difference between truth and fiction? Let's play a little game, which I once saw in a movie

In this game, we each say five things about ourselves, four of which are true. We then try to guess which ones are true and which ones are lies. (Go ahead and use Google, if you like.) I'll start.

My five facts:

1) I have another LessWrong commenter's autograph.

2) I once received the first place prize in an (unsanctioned, online) Magic tournament that lasted a total of ten rounds (seven rounds of Swiss pairings, plus three single elimination rounds) but only beat an opponent at Magic during four of them.

3) I've beaten Battletoads, on an actual NES, without using a Game Genie or other cheat device.

4) Not too long ago, I made a $500 donation to Population Services International, using my debit card. Unfortunately, this overdrew my checking account. My parents were not pleased, not only because I overdrew my checking account, but also because they disapproved of my donating such a large amount of money.

5) My favorite Star Wars movie is "Attack of the Clones."

(Note: I used a random number generator to determine which position would contain the lie.)

New to LessWrong?

New Comment


69 comments, sorted by Click to highlight new comments since:

I like this game, but you've got to make it interesting. In that vein, here's mine:

1) Eliezer once wrote an Overcoming Bias post where he said something like "Make one little mistake in a whole long chain of inferences, and you end up in Outer Mongolia." The first time I read this post was from a webcafe in Outer Mongolia.

2) I have pet several full-grown tigers, and one of them almost fell asleep on my lap.

3) I once broke up with a girl partly because she was a deontologist.

4) One of my school teachers once accused me, in all seriousness, of being a witch. Not witch as in "Wiccan", but witch as in "has magic powers and consorts with the devil".

5) I was given a blessing by a Tibetan lama in the mountains of Nepal.

Gur snyfrubbq vf guerr.

Ahzore bar vf gehr; V jnf va Nfvn ng gur gvzr, nyjnlf jnagrq gb frr Zbatbyvn, naq fcrag gur jrrx va Hyna Ongne naq fheebhaqvatf.

Ahzore gjb vf nyfb gehr: frr uggc://ra.jvxvcrqvn.bet/jvxv/Gvtre_Grzcyr.

Ahzore guerr vf snyfr: V qvq tb ba BXPhcvq, frr fbzrbar jub fnvq va ure cebsvyr gung fur jnf n qrbagbybtvfg, naq guvax "Vg'q arire jbex", ohg gung'f nf sne nf V'ir tbggra.

Ahzore sbhe vf gehr; vg jnf n penml byq ernyyl eryvtvbhf fhofgvghgr grnpure jub fgnegrq bss trarenyyl pbaqrzavat jvgpupensg, V qrpvqrq gb nethr jvgu uvz orpnhfr V nz trarenyyl nethzragngvir, naq uvf penmvarff qvq gur erfg.

Ahzore svir vf gehr: frr uggc://ra.jvxvcrqvn.bet/jvxv/Gratobpur

When you're reasoning out new knowledge in the absence of crushingly overwhelming guidance, you can miss one point and wake up in Outer Mongolia fifty steps later.

-- Do Scientists Already Know This Stuff?, Eliezer Yudkowsky, May 16, 2008

If we're being strict then #1. Though the name's still informally used to refer to Mongolia, there's really been no such place as Outer Mongolia during any period of history in which it's likely to have contained a webcafe.

If we're not then I'd guess #2, as your favorite David Berman thought experiment involves clearly and distinctly imagining tigers.

Well I read 2 at first as "I have several pet tigers" and I definitely would've guessed that was the lie! I still guess 2.

[-][anonymous]00

del

I'm going with number 3, battletoads was hard! :(

My five:

1) I have the complete Feynman Lecture Series

2) I have beaten professional Counter-Strike players while never having played in any tournaments or leagues.

3) I am a smoker (cigarettes).

4) I never finished Half-Life.

5) My watch has Mickey Mouse on it (his hands point to the time!).

I assume that #2 refers to beating them at Counter-Strike...

My dad used to wear a watch like that. I guess 4.

I guess that statement 2 is the lie.

Ahzore bar vf gur snyfrubbq, V bayl jvfu V unq vg. V arire svavfurq unys-yvsr orpnhfr V qvfpbirerq V cersrerq pbhagre-fgevxr qhr gb univat (zbfgyl) vagryyvtrag bccbaragf. V jbhaq hc orvat tbbq sevraqf jvgu n srj cebsrffvbanyf naq cynlrq jvgu gurz frzv-erthyneyl. Jr jrer nobhg rdhny, jubrire jnf zber njnxr jbhyq hfhnyyl jva. V qvqa'g cynl gbheanzragf orpnhfr V fnj ubj zhpu jbex gurl chg vagb cresrpgvat grnz-cynl naq svtherq V unq orggre guvatf gb qb (abg fb fher vg jnf gur evtug qrpvfvba abj, ohg gur cnfg vf gur cnfg). Cvpxrq hc fzbxvat ng 19, sevraqf naq V tbg fbzr pvtnef sbe Znex Gjnva'f oveguqnl naq jryy, fyvccrel fybcr. Nf sbe zl jngpu, vg jnf erpragyl er-qvfpbirerq sebz n gevc gb Qvfarl Jbeyq jura V jnf n xvq.

Gung znxrf frafr, npghnyyl. Fubhyq V or unccl gung gur npghny yvr jnf gur bar gung V gubhtug jnf gur frpbaq zbfg yvxryl gb or gur yvr?

V fubhyq fnl fb, bhg bs svir fgngrzragf gung lbh unir 0 vasbezngvba nobhg bgure guna gur fgngrzragf gurzfryirf rira trggvat pybfr vf dhvgr na nppbzcyvfuzrag (gung be yhpxl, juvpurire lbh cersre).

I guess #3 since the cigarettes is superfluous. Perhaps you do smoke, but do not smoke cigarettes?

[-][anonymous]00

del

[This comment is no longer endorsed by its author]Reply

I think it's (2). To win a 7/SS+top 8 tournament, you would definitely need at least 4 and probably 5 wins in the main tournament, plus three wins in elimination. The rules of the Swiss system don't allow any player to receive more than one bye, so you would need to have had 4 wins, 1 bye, and at least 2 but probably 3 forfeits. Online tournaments may be different, but of all the Swiss tournaments I've played in, I have only ever seen one forfeit, so the same player getting multiple forfeits is highly unlikely.

A bit of relevant clarification: I didn't say I that I won. I said I received the prize for first place.

Only piece of information we're missing is how many people started in the tournament which would allow us to find out how many points he would need to get top8.

You sound like you know what a Magic tournament is about way more than I do (don't know what counts as a draw, or if there even is such a thing) and have revised my estimates accordingly.

(1) .1 (2) .35 (3) .35 (4).18 (5).02

5 is low due to pjeby's comment, the .02 is my probability that he would slash my tires or is wrong.

Draws are possible in Magic (kill both players at the same time), but they're quite rare. Unless for some reason you wanted to draw instead of win, in which case you could do so intentionally.

In an online tournament, in addition to anything the tournament format itself allows, there's always another way to get eliminated: disconnect / failure to show up. I don't remember the disconnect rate in Magic online being particularly high, but a tournament does involve a significant amount of time for people to misplan.

Drawn games are rare, but drawn matches are fairly common. A match is drawn if it runs over the time limit without either player winning, or if it's best 2 of 3 and each player has one win when the time limit is reached.

I didn't consider disconnects. Those would tend to greatly increase the forfeit rate, over the forfeit rate I observed from in-person tournaments.

We don't know how many people were in the tournament, but we do have a very good proxy: the number of rounds in the Swiss portion. Swiss tournaments are supposed to have between log_2(P) and log_2(P)+1 rounds, where P is the number of players, so the tournament had between 64 and 128 players.

My (crude) probability assignments for which one is the lie: 0.15, 0.1, 0.15, 0.1, 0.5. I know nothing about Magic and have never even heard of Battletoads, but I'll assume that #3 is impressive because otherwise you wouldn't have chosen it.

(My immediate reaction to your list: The lie must be #5 because the other statements are merely somewhat improbable whereas that one is an actual logical impossibility.)

Well, here's my reveal, encoded in ROT-13 for those who don't want to be spoiled.

Gur nzbhag V qbangrq jnf fvk uhaqerq naq svsgl qbyynef, abg svir uhaqerq qbyynef. Rirelguvat ryfr npghnyyl unccrarq. Gur nhgbtencu vf sebz Miv Zbjfubjvgm. Gur bayvar Zntvp gbheanzrag qvq vaqrrq unccra. (Qrgnvyf urer.) V unir, vaqrrq, orngra Onggyrgbnqf, naq Rcvfbqr 2 npghnyyl vf zl snibevgr Fgne Jnef svyz. Guvf vf abg orpnhfr V gubhtug vg jnf hahfhnyyl tbbq, ohg engure orpnhfr V gubhtug gur bgure svyzf, vapyhqvat gur bevtvany gevybtl, jrer rira jbefr. Gurer'f cyragl bs tbbq zngrevny va gur Fgne Jnef Rkcnaqrq Havirefr, ohg nf sne nf V'z pbaprearq, Trbetr Yhpnf unf nyjnlf orra n onq jevgre.

The game works better (at least as an inference game) if you make the wrong statement very wrong and not just a little bit wrong. And I've never seen the appeal of Fgne Jnef either.

[-][anonymous]00

del

Yeah, but I'm not good at lying...

Ha, the link's relevant details are awesome.

For people who need one, here is a ROT-13 converter.

Rcvfbqr 2 npghnyyl vf zl snibevgr Fgne Jnef svyz. Guvf vf abg orpnhfr V gubhtug vg jnf hahfhnyyl tbbq, ohg engure orpnhfr V gubhtug gur bgure svyzf, vapyhqvat gur bevtvany gevybtl, jrer rira jbefr. Gurer'f cyragl bs tbbq zngrevny va gur Fgne Jnef Rkcnaqrq Havirefr, ohg nf sne nf V'z pbaprearq, Trbetr Yhpnf unf nyjnlf orra n onq jevgre.

Guvf pbashfrf zr orpnhfr bar bs gur guvatf gung znxrf gur bevtvany gevybtl orggre jnf gung crbcyr jrer noyr gb gnxr gur ervaf njnl sebz Trbetr, naq zhpu bs gur jevgvat jnf n pbyynobengvir rssbeg. Rzcver jnf pyrneyl shyy bs jva. Gur cevapvcyr bs punevgl unq zr ershfvat gb npprcg gung #5 jnf gehr, fvapr V pbhyq fvzcyl abg guvax gung nobhg n crefba.

Gurer jrer gjb guvatf va Nggnpx bs gur Pybarf gung znqr vzcerffrq zr zber guna gur bgure svyzf. Bar jnf gung, jura Pbhag Qbbxh nfxrq Bov-Jna naq Nanxva gb wbva uvf eroryyvba, rirel fgngrzrag ur znqr va fhccbeg bs uvf eroryyvba jnf npghnyyl gehr. Gur Frangr ernyyl jnf orvat pbagebyyrq naq znavchyngrq ol n Fvgu Ybeq! Gur yrnqref bs gur Genqr Srqrengvba xarj gung, rira gubhtu gur Wrqv qvqa'g, orpnhfr gurl hfrq gb jbex sbe uvz hagvy, qhevat Rcvfbqr 1, ur oebxr uvf cebzvfr gb fhccbeg gurve vainfvba bs Anobb - yrnivat gurz gb ghea gb Pbhag Qbbxh sbe uryc. Naq erzrzore, rira gur Wrqv unq pbafvqrerq Pbhag Qbbxh bireyl vqrnyvfgvp orsber uvf nyyrtrq snyy gb gur Qnex Fvqr. (Fher, ur ernyyl jnf Cnycngvar'f arj ncceragvpr naq ercynprzrag sbe Qnegu Znhy, ohg ubj jnf gur Genqr Srqrengvba fhccbfrq gb xabj gung?) Hayvxr gur bgure Fgne Jnef zbivrf, Nggnpx bs gur Pybarf pbagnvarq zbeny nzovthvgl - zbfg bs gur crbcyr svtugvat ba gur fvqr bs gur ivyynvaf unq ernfbaf sbe qbvat fb gung n tbbq crefba pbhyq fhccbeg, vafgrnq bs fvzcyl orpnhfr gurl jrer Onq Thlf svtugvat ba gur fvqr bs bgure Onq Thlf. Gur bgure cneg bs gur svyz gung V yvxrq jnf gur zhpu-znyvtarq "ebznapr" orgjrra Nanxva naq Cnqzr. Vg'f abg fhccbfrq gb or n snvel gnyr ebznapr. Nanxva vfa'g "va ybir" jvgu Cnqzr. Ur'f bofrffrq jvgu Cnqzr, va gur znaare bs n fgnyxre. Gur fprarf bs gur gjb bs gurz, nybar, gnyxvat jvgu rnpu bgure, nera'g fhccbfrq gb or ebznagvp. Gurl'er fhccbfrq gb or perrcl, naq gb zr, gurl jrer.

Gung'f jul V guvax Nggnpx bs gur Pybarf vf gur yrnfg onq Fgne Jnef zbivr.

[-][anonymous]00

del

Well, the villain does ask the hero to say five things about himself, four of them true. I don't remember what happened next, though.

[-][anonymous]00

del

Yeah, I think that's about right.

This game is a relatively common ice-breaker. I think I also remember seeing it in the movie Cry_Wolf. (No link because I am lazy.)

[-][anonymous]00

deleted

[-][anonymous]00

del

[This comment is no longer endorsed by its author]Reply

Maybe we should modify the game so that the lie has to be the approximate opposite of something that's true? (Actual opposite when used in a logic class.)

[-][anonymous]00

del

[This comment is no longer endorsed by its author]Reply

My rule is a restriction to prevent an underwhelming reveal, not a heuristic for choosing good lies.

If you're interested in the subject of choosing good lies: All of your statements (whether they are truth or false) should be statements that are unlikely to describe a randomly chosen person. That's why the witchcraft one would not make a good truth or lie.

My probabilities are: .1, .3, .2, .2, .2

In case people don't know that, hotornot scores are so heavily biased, so getting 9.6 is quite simple

[-][anonymous]00

This game is a relatively common ice-breaker. I think I also remember seeing it in the movie Cry_Wolf. (No link because I am lazy.)

I guess #1. Here are my five:

1) I've climbed Pão de Açucar

2) Every man I've kissed has been dressed like a woman

3) I've never had a single tooth cavity

4) I once hung off the wing of a plane (that was flying)

5) I'm going bald

[-][anonymous]10

deleted

Well, by devoting so much explanation to 2, you make it pretty clear that's not it (or you're just a very good faker.) I guess (.1, .1, .1, .5, .2)

My analysis:

1) Pão de Açucar is in Rio de Janeiro, Brazil, and, according to Wikipedia, is climbed by a lot of people. I have no reason to consider this any more unlikely than any other random statement someone might claim.

2): This seems plausible, if we assume that Jordan is male. I wouldn't be surprised to find at least some men dressed like women during an event like Carnival. Another not too weird circumstance in which this might happen is if both men are actors, and the kiss is during a performance.

3): Unusual, but possible. Again, this doesn't throw up any more red flags.

4): This is just plain weird, and seems to be the least plausible. Indeed, is that even physically possible?

5) This is very common.

I guess that #4 is the lie.

Nice analysis. One comment though...

Ahzore sbhe vf vaqrrq cbffvoyr. V'ir qbar vg =Q.

Care to offer more details?

Also, which was the lie?

Well, the plane was too small to jump out of directly without hitting the tail, so first I had to grab the wing and work my way to the tip before letting go. I've got to say, however much of a rush skydiving is, hanging from a wing is way more exhilarating.

The lie was (1), it's definitely something I'd like to do, just haven't got around to it.

I would guess #2 due to ambiguous clothing definitions. What does "dressed like a woman" mean?

With regard to #2, it occurred to me that he could have never kissed a man.

I suppose I could clarify:

Every man I've kissed has been dressed in clothing traditionally accepted by Western culture to be feminine and moreover this group is nonempty.

Likewise by (3) I don't mean that every time I've had a cavity I've had at least two.

3 seems likely to be true, given this Google search for "cronoDAS Battletoads."

When should I reveal which statement is a lie?

[-][anonymous]00

deleted

If you had actually donated $400 to PSI, but the rest of the statement were true, would that count as the lie?

yes, which means 4 contains a lot of burdonsome detail, which means you might want to suspect 4, which means Crono should put the lie elsewhere so...

it probably evens out in the end

[-][anonymous]00

Yes.

Now this is creepy -- don't tell me your second commenter's autograph is the eunuch troll!

[-][anonymous]00

I had nothing to do with that post; I'm not Lojban and I don't think I've ever met him.

5) My favorite Star Wars movie is "Attack of the Clones."

Please, please tell me this is the lie.

Please, please tell me this is the lie.

Sadly, it is not. I would say which statement is, but I believe we're not supposed to post spoilers on LW. However, I will note that only one of his five statements actually contains enough information for it to be trivially falsifiable by anyone with internet access...

What do you mean by spoilers, in this case?

What do you mean by spoilers, in this case?

In this case, a spoiler would be pointing people to the place on the internet where they can find out why one of your statements is factually impossible. (Unless I'm very confused, which of course is also possible.)