Thescienceofstatisticsis a waytotrytosortthroughit.
それを整理してみるのが統計学という学問です。
Soifyoudon't haveanystatisticalknowledgeorunderstandingorperspectivethenyou'relikelytosaywellthismustbetruebecausemyfriendsaiditorthismustbetruebecause I hearditonthenewsor I justkindofthinkitmustbetrue.
WehavequestionfromLawrence I tvsaysquestionforstatisticians, whydidthepollsgetitsowrongexplanationsplease.
統計学者への質問ですが、なぜ世論調査はこんなに間違ったのか、説明してください。
Yeah.
そうですね。
Sopublicopinionpolling, especiallywhenit's predictingelectionsis a veryhighprofilething, butalso a hardthingtodoandusuallypeoplenoticethemistakesmorethanthecorrections.
Butwhatmakesitespeciallyhardforthepollstersisthatit's hardtoget a randomsample.
しかし、世論調査会社にとって特に難しいのは、ランダムなサンプルを得ることが難しいということです。
Andthemainreasonisbecausemostpeopledon't wanttotalktopollsters, pollingcompaniesdon't necessarilyliketotalkaboutit, buttheirresponseratesareusuallylessthan 10% andthatcanleadto a lotofbiasesbecausemaybepeoplewhosupport a certaincandidateare a littlebitmorelikelytoagreetotalktothepollstersthanpeoplewhosupportanothercandidate.
ButthenifyousayoutofallthemillionsoffamiliesintheUnitedKingdomorinthewholeworldwas a chancethatsomewhere, there's a familywheretwokidsbothdiedofSIDSextremelylikelyanditseemslikethatwasthecasewithher.
Theshortansweriseverythingthatistosayifyou'retalkingaboutwinning a lotteryjackpotforoneofthebiglotterieslikemegamillionsorpowerball, thenthechanceofwinningthatjackpotwith a singleticketisonechancein a coupleof 100 milliondependingonwhichlottery.
Socomparedtothat, almostanythingyoucanthinkofbeingkilledby a boltoflightningorthenextpersonyoumeetwillonedaybethepresidentoftheUnitedStatesoranycrazythingyoucancomeupwith.
Andinfactonethat I liketouseasanexampleisifyoudrivetothestoretobuyyourlotteryticket, you'rewaymorelikelytobekilledin a carcrashonyourwaytothestorethanyouaretowinthejackpot.
Nextwehave a questionfrom s malimall, I'm justpatientlywaitingforpeopletorealizethatallstatisticsareskewedbecausethedataisskewedinsomanywaysthat I can't evenlistthemall.
A questionfromSixlatinsixloversix, whomakesbettingodds.
ベッティングオッズを作るラテン系6人組の恋人シックスさんからの質問です。
Isitanalgorithm?
アルゴリズムなのでしょうか?
Soit's a reallyinterestingproblemforthebookiesorthepeoplewhoaremakingtheseodds.
だから、ブッキーやオッズをつける人たちにとっては、実に興味深い問題なのです。
Nowthegoalisprettyeasytounderstandbecauseifyou're a bookie, whatyouwantisprettymuchtohavethesameamountofbeddingonbothsides.
このゴールは非常にわかりやすいものです。
Sothatintheendyoudon't reallycareifthehorsewinsornotoryoudon't reallycareiftheteamwinsornotbecauseeitherwayyou'regonnamakemoneybecauseyou'regonnagetyourcut, whereasifeverybodybetononesideandthentheyallonethenyoucouldlose a lotofmoney.
最終的には、馬が勝とうが勝つまいが、チームが勝とうが勝つまいが、どちらでも構わないということです。
Butontheotherhandhowtheydothatiskindof a challengeandusuallythey'reupdatingtheiroddsastheygoandiftheysayeverybody's bettingonthisoneteam G webetterchangetheoddssothatthenextbettorsaremorelikelytobetontheothersideand I'm not a bookie.
Whereasnowthere's somuchonlinegamblingthat a lotofitisautomatedandtheyhavealgorithmswhich I thinkarenotsimplebasedonhoweverybody's bettingandtryingtoadjustthings.
So I'lldothathereandweimaginewehave a frogwhicheverysecondrandomlydecideseithertomoveonestepthiswayortomoveonestepthiswayandonceitdoesthenthenextseconditagaindecidesrandomlytomoveonestepthiswayoronestepthisway.
Soeveninthisonefamilyprobablythere's a lotofotherpeopleineachofthosegenerationsandifanythreeofthemhadmatcheduptheirbirthdaysthenthisthesametweetcouldhavebeenwritten.
Buthowmuchmoreandthe p valuequestionwouldbewhat's theprobabilityifwehadn't givenanytreatmentThatthatsamenumberormoreofthepeoplewouldstillhavegottenbetter.
Okay, sonextquestionfromKing M Boussosaysstatisticallyoneofthechancesandright, andthisis a displayof a drawresultsand I believethiswasfromtheSouthAfricapowerballlotterybackindecemberof 2020.
So a questionfrom a Tetraformsays, hey, whatisthemoststatisticallyimprobablethingtohappentoyou?
テトラフォームからの質問で、「ねえ、統計的に最もありえないことは何?
Well, when I wasinmyearlyteens, myfamilywenton a triptoDisneyworldfloridaandinthemiddleofitall, welookedupandwesawmyfather's cousinPhilandhelivedinConnecticutatthetimeandwelivedinTorontoCanadaandwehadnoideahewasgoingtobethere.
Whataboutmydad's othercousinsormymom's cousinsormycousinsormypianoteacherormyfriendfromschoolorthere's probably a few 100 peoplethatwewouldhavebeenreallysurprisedtosee.
And I endupcomputing, there's aboutonechancein 200 orso, abouthalfof 1% thatifyougoon a triptoDisneyworldandspend a coupleofdaystherealltherisethatyourunintosomebodythat, youknow, soit's notsoincredible, eventhoughitsurewas a surpriseatthetime.