<?xml version="1.0" encoding="utf-8"?>
<feed xmlns="http://www.w3.org/2005/Atom">
  <title>Decision Design</title>
  <subtitle>Trasformare l&#39;evidenza in decisioni</subtitle>
  <link href="https://decisiondesign.org/feed.xml" rel="self"/>
  <link href="https://decisiondesign.org/"/>
  <updated>2026-07-28T00:00:00.000Z</updated>
  <id>https://decisiondesign.org/</id>
  <author>
    <name>Alessandro Saccoia</name>
    <email>ciao@decisiondesign.org</email>
  </author>
  <entry>
    <title>L&#39;ospedale piccolo e la vostra campagna di ieri</title>
    <link href="https://decisiondesign.org/articoli/ospedale-piccolo-campagna/"/>
    <updated>2026-07-28T00:00:00.000Z</updated>
    <id>https://decisiondesign.org/articoli/ospedale-piccolo-campagna/</id><category term="legge-dei-piccoli-numeri"/><category term="regressione-verso-la-media"/>
    <summary>Un esperimento del 1974 spiega perché il vostro miglior canale del mese scorso è probabilmente il vostro canale più piccolo. E perché lo scoprirete di nuovo il mese prossimo.</summary>
    <content type="html">&lt;p&gt;Nel 1974 Kahneman e Tversky sottoposero a un gruppo di persone una domanda che
sembra banale. In una città ci sono due ospedali. In quello grande nascono circa
45 bambini al giorno, in quello piccolo circa 15. Sappiamo che il 50% dei neonati
sono maschi, ma la percentuale esatta varia da un giorno all’altro. Per un anno,
ciascun ospedale ha registrato i giorni in cui più del 60% dei nati era maschio.
Quale ospedale ne ha registrati di più?&lt;/p&gt;
&lt;p&gt;La maggioranza rispose “circa lo stesso numero”. La risposta corretta è
l’ospedale piccolo, e non di poco: circa il triplo dei giorni. La percentuale
attesa è identica nei due ospedali — è il 50% per costruzione — ma la &lt;em&gt;varianza&lt;/em&gt;
non lo è affatto. Con quindici nascite al giorno bastano dieci maschi per
superare la soglia; con quarantacinque ne servono ventotto, e la legge dei grandi
numeri lavora contro gli estremi.&lt;/p&gt;
&lt;div class=&quot;riquadro&quot;&gt;
&lt;span class=&quot;etichetta&quot;&gt;Legge dei piccoli numeri&lt;/span&gt;
&lt;p&gt;La tendenza a trattare un campione piccolo come se avesse le stesse proprietà
statistiche di un campione grande — in particolare, ad attendersi da esso la
stessa stabilità. Non è ignoranza della statistica: è una intuizione che
sopravvive intatta anche in chi la statistica la conosce. Con &lt;span class=&quot;katex&quot;&gt;&lt;span class=&quot;katex-mathml&quot;&gt;&lt;math xmlns=&quot;http://www.w3.org/1998/Math/MathML&quot;&gt;&lt;semantics&gt;&lt;mrow&gt;&lt;mi&gt;n&lt;/mi&gt;&lt;/mrow&gt;&lt;annotation encoding=&quot;application/x-tex&quot;&gt;n&lt;/annotation&gt;&lt;/semantics&gt;&lt;/math&gt;&lt;/span&gt;&lt;span class=&quot;katex-html&quot; aria-hidden=&quot;true&quot;&gt;&lt;span class=&quot;base&quot;&gt;&lt;span class=&quot;strut&quot; style=&quot;height:0.4306em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord mathnormal&quot;&gt;n&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt; piccolo,
&lt;span class=&quot;katex&quot;&gt;&lt;span class=&quot;katex-mathml&quot;&gt;&lt;math xmlns=&quot;http://www.w3.org/1998/Math/MathML&quot;&gt;&lt;semantics&gt;&lt;mrow&gt;&lt;mi mathvariant=&quot;normal&quot;&gt;S&lt;/mi&gt;&lt;mi mathvariant=&quot;normal&quot;&gt;E&lt;/mi&gt;&lt;/mrow&gt;&lt;annotation encoding=&quot;application/x-tex&quot;&gt;&#92;mathrm{SE}&lt;/annotation&gt;&lt;/semantics&gt;&lt;/math&gt;&lt;/span&gt;&lt;span class=&quot;katex-html&quot; aria-hidden=&quot;true&quot;&gt;&lt;span class=&quot;base&quot;&gt;&lt;span class=&quot;strut&quot; style=&quot;height:0.6833em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord&quot;&gt;&lt;span class=&quot;mord mathrm&quot;&gt;SE&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt; cresce come &lt;span class=&quot;katex&quot;&gt;&lt;span class=&quot;katex-mathml&quot;&gt;&lt;math xmlns=&quot;http://www.w3.org/1998/Math/MathML&quot;&gt;&lt;semantics&gt;&lt;mrow&gt;&lt;mn&gt;1&lt;/mn&gt;&lt;mi mathvariant=&quot;normal&quot;&gt;/&lt;/mi&gt;&lt;msqrt&gt;&lt;mi&gt;n&lt;/mi&gt;&lt;/msqrt&gt;&lt;/mrow&gt;&lt;annotation encoding=&quot;application/x-tex&quot;&gt;1/&#92;sqrt{n}&lt;/annotation&gt;&lt;/semantics&gt;&lt;/math&gt;&lt;/span&gt;&lt;span class=&quot;katex-html&quot; aria-hidden=&quot;true&quot;&gt;&lt;span class=&quot;base&quot;&gt;&lt;span class=&quot;strut&quot; style=&quot;height:1.0503em;vertical-align:-0.25em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord&quot;&gt;1/&lt;/span&gt;&lt;span class=&quot;mord sqrt&quot;&gt;&lt;span class=&quot;vlist-t vlist-t2&quot;&gt;&lt;span class=&quot;vlist-r&quot;&gt;&lt;span class=&quot;vlist&quot; style=&quot;height:0.8003em;&quot;&gt;&lt;span class=&quot;svg-align&quot; style=&quot;top:-3em;&quot;&gt;&lt;span class=&quot;pstrut&quot; style=&quot;height:3em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord&quot; style=&quot;padding-left:0.833em;&quot;&gt;&lt;span class=&quot;mord mathnormal&quot;&gt;n&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span style=&quot;top:-2.7603em;&quot;&gt;&lt;span class=&quot;pstrut&quot; style=&quot;height:3em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;hide-tail&quot; style=&quot;min-width:0.853em;height:1.08em;&quot;&gt;&lt;svg xmlns=&quot;http://www.w3.org/2000/svg&quot; width=&quot;400em&quot; height=&quot;1.08em&quot; viewBox=&quot;0 0 400000 1080&quot; preserveAspectRatio=&quot;xMinYMin slice&quot;&gt;&lt;path d=&quot;M95,702
c-2.7,0,-7.17,-2.7,-13.5,-8c-5.8,-5.3,-9.5,-10,-9.5,-14
c0,-2,0.3,-3.3,1,-4c1.3,-2.7,23.83,-20.7,67.5,-54
c44.2,-33.3,65.8,-50.3,66.5,-51c1.3,-1.3,3,-2,5,-2c4.7,0,8.7,3.3,12,10
s173,378,173,378c0.7,0,35.3,-71,104,-213c68.7,-142,137.5,-285,206.5,-429
c69,-144,104.5,-217.7,106.5,-221
l0 -0
c5.3,-9.3,12,-14,20,-14
H400000v40H845.2724
s-225.272,467,-225.272,467s-235,486,-235,486c-2.7,4.7,-9,7,-19,7
c-6,0,-10,-1,-12,-3s-194,-422,-194,-422s-65,47,-65,47z
M834 80h400000v40h-400000z&quot;/&gt;&lt;/svg&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;vlist-s&quot;&gt;​&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;vlist-r&quot;&gt;&lt;span class=&quot;vlist&quot; style=&quot;height:0.2397em;&quot;&gt;&lt;span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;: la media resta al suo posto, gli
estremi si moltiplicano.&lt;/p&gt;
&lt;/div&gt;
&lt;p&gt;Fin qui è un aneddoto da manuale, il tipo di cosa che si legge annuendo e si
archivia come capita. Il problema è che l’ospedale piccolo è nella
vostra dashboard, adesso, e ha un nome più rassicurante: si chiama “segmento
emergente”, oppure “canale in crescita”, oppure “la creatività che sta andando
forte”.&lt;/p&gt;
&lt;h2&gt;Il campione più piccolo vince sempre la classifica&lt;/h2&gt;
&lt;p&gt;Prendete una lettura di performance qualunque, per canale. Sette canali, un mese
di dati, ordinati per tasso di conversione. Chi arriva primo?&lt;/p&gt;
&lt;div class=&quot;tabella-scorrevole&quot;&gt;
&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Canale&lt;/th&gt;
&lt;th&gt;Utenti&lt;/th&gt;
&lt;th&gt;Conversioni&lt;/th&gt;
&lt;th&gt;Tasso&lt;/th&gt;
&lt;th&gt;Scarto atteso&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;Newsletter partner&lt;/td&gt;
&lt;td&gt;240&lt;/td&gt;
&lt;td&gt;14&lt;/td&gt;
&lt;td&gt;5,8%&lt;/td&gt;
&lt;td&gt;±1,5 pt&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Affiliazione B&lt;/td&gt;
&lt;td&gt;610&lt;/td&gt;
&lt;td&gt;27&lt;/td&gt;
&lt;td&gt;4,4%&lt;/td&gt;
&lt;td&gt;±0,8 pt&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Ricerca a pagamento&lt;/td&gt;
&lt;td&gt;18.400&lt;/td&gt;
&lt;td&gt;662&lt;/td&gt;
&lt;td&gt;3,6%&lt;/td&gt;
&lt;td&gt;±0,14 pt&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Social a pagamento&lt;/td&gt;
&lt;td&gt;22.100&lt;/td&gt;
&lt;td&gt;707&lt;/td&gt;
&lt;td&gt;3,2%&lt;/td&gt;
&lt;td&gt;±0,12 pt&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Display&lt;/td&gt;
&lt;td&gt;41.500&lt;/td&gt;
&lt;td&gt;1.079&lt;/td&gt;
&lt;td&gt;2,6%&lt;/td&gt;
&lt;td&gt;±0,08 pt&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;
&lt;/div&gt;
&lt;aside class=&quot;margine&quot;&gt;
&lt;p&gt;Lo scarto atteso è l’errore standard della proporzione,
&lt;span class=&quot;katex&quot;&gt;&lt;span class=&quot;katex-mathml&quot;&gt;&lt;math xmlns=&quot;http://www.w3.org/1998/Math/MathML&quot;&gt;&lt;semantics&gt;&lt;mrow&gt;&lt;mrow&gt;&lt;mi mathvariant=&quot;normal&quot;&gt;S&lt;/mi&gt;&lt;mi mathvariant=&quot;normal&quot;&gt;E&lt;/mi&gt;&lt;/mrow&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;msqrt&gt;&lt;mrow&gt;&lt;mi&gt;p&lt;/mi&gt;&lt;mo stretchy=&quot;false&quot;&gt;(&lt;/mo&gt;&lt;mn&gt;1&lt;/mn&gt;&lt;mo&gt;−&lt;/mo&gt;&lt;mi&gt;p&lt;/mi&gt;&lt;mo stretchy=&quot;false&quot;&gt;)&lt;/mo&gt;&lt;mi mathvariant=&quot;normal&quot;&gt;/&lt;/mi&gt;&lt;mi&gt;n&lt;/mi&gt;&lt;/mrow&gt;&lt;/msqrt&gt;&lt;/mrow&gt;&lt;annotation encoding=&quot;application/x-tex&quot;&gt;&#92;mathrm{SE}=&#92;sqrt{p(1-p)/n}&lt;/annotation&gt;&lt;/semantics&gt;&lt;/math&gt;&lt;/span&gt;&lt;span class=&quot;katex-html&quot; aria-hidden=&quot;true&quot;&gt;&lt;span class=&quot;base&quot;&gt;&lt;span class=&quot;strut&quot; style=&quot;height:0.6833em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord&quot;&gt;&lt;span class=&quot;mord mathrm&quot;&gt;SE&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.2778em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mrel&quot;&gt;=&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.2778em;&quot;&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;base&quot;&gt;&lt;span class=&quot;strut&quot; style=&quot;height:1.24em;vertical-align:-0.305em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord sqrt&quot;&gt;&lt;span class=&quot;vlist-t vlist-t2&quot;&gt;&lt;span class=&quot;vlist-r&quot;&gt;&lt;span class=&quot;vlist&quot; style=&quot;height:0.935em;&quot;&gt;&lt;span class=&quot;svg-align&quot; style=&quot;top:-3.2em;&quot;&gt;&lt;span class=&quot;pstrut&quot; style=&quot;height:3.2em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord&quot; style=&quot;padding-left:1em;&quot;&gt;&lt;span class=&quot;mord mathnormal&quot;&gt;p&lt;/span&gt;&lt;span class=&quot;mopen&quot;&gt;(&lt;/span&gt;&lt;span class=&quot;mord&quot;&gt;1&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.2222em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mbin&quot;&gt;−&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.2222em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord mathnormal&quot;&gt;p&lt;/span&gt;&lt;span class=&quot;mclose&quot;&gt;)&lt;/span&gt;&lt;span class=&quot;mord&quot;&gt;/&lt;/span&gt;&lt;span class=&quot;mord mathnormal&quot;&gt;n&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span style=&quot;top:-2.895em;&quot;&gt;&lt;span class=&quot;pstrut&quot; style=&quot;height:3.2em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;hide-tail&quot; style=&quot;min-width:1.02em;height:1.28em;&quot;&gt;&lt;svg xmlns=&quot;http://www.w3.org/2000/svg&quot; width=&quot;400em&quot; height=&quot;1.28em&quot; viewBox=&quot;0 0 400000 1296&quot; preserveAspectRatio=&quot;xMinYMin slice&quot;&gt;&lt;path d=&quot;M263,681c0.7,0,18,39.7,52,119
c34,79.3,68.167,158.7,102.5,238c34.3,79.3,51.8,119.3,52.5,120
c340,-704.7,510.7,-1060.3,512,-1067
l0 -0
c4.7,-7.3,11,-11,19,-11
H40000v40H1012.3
s-271.3,567,-271.3,567c-38.7,80.7,-84,175,-136,283c-52,108,-89.167,185.3,-111.5,232
c-22.3,46.7,-33.8,70.3,-34.5,71c-4.7,4.7,-12.3,7,-23,7s-12,-1,-12,-1
s-109,-253,-109,-253c-72.7,-168,-109.3,-252,-110,-252c-10.7,8,-22,16.7,-34,26
c-22,17.3,-33.3,26,-34,26s-26,-26,-26,-26s76,-59,76,-59s76,-60,76,-60z
M1001 80h400000v40h-400000z&quot;/&gt;&lt;/svg&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;vlist-s&quot;&gt;​&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;vlist-r&quot;&gt;&lt;span class=&quot;vlist&quot; style=&quot;height:0.305em;&quot;&gt;&lt;span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;. Non è una raffinatezza: è la scala minima entro
cui i numeri della colonna precedente non significano niente.&lt;/p&gt;
&lt;/aside&gt;
&lt;p&gt;La newsletter partner è prima con un margine che sembra schiacciante: 5,8% contro
il 3,6% della ricerca a pagamento, cioè il 61% in più. In una riunione questo
diventa “spostiamo budget sulla newsletter partner”. Ma quattordici conversioni
su duecentoquaranta utenti hanno un errore standard di un punto e mezzo: due
conversioni in più o in meno — due persone — muovono il tasso di quasi un punto
percentuale. L’intervallo di confidenza al 95% va da circa il 2,8% all’8,8%.
Sovrappone quello della ricerca a pagamento per intero.&lt;/p&gt;
&lt;p&gt;Il canale non è il migliore. È il più piccolo. Sono due proprietà diverse che nel
formato “classifica ordinata per tasso” diventano indistinguibili, perché la
tabella ordina per stima puntuale e la stima puntuale di un campione piccolo è
rumore con un decimale.&lt;sup class=&quot;footnote-ref&quot;&gt;&lt;a href=&quot;#fn1&quot; id=&quot;fnref1&quot;&gt;[1]&lt;/a&gt;&lt;/sup&gt;&lt;/p&gt;
&lt;h2&gt;Perché si ripresenta ogni mese&lt;/h2&gt;
&lt;p&gt;C’è un meccanismo che rende questo errore non solo frequente ma &lt;em&gt;ciclico&lt;/em&gt;. Il
canale piccolo che questo mese sta in cima, il mese prossimo tornerà verso la
media — non perché qualcosa si sia rotto, ma perché la regressione verso la media
è ciò che fa il rumore. A quel punto succede una delle due cose peggiori.&lt;/p&gt;
&lt;p&gt;Se nel frattempo avete spostato budget su quel canale, la sua discesa verrà
attribuita alla saturazione, all’audience esaurita, al deterioramento creativo:
tutte spiegazioni causali plausibili per un fenomeno che non richiede nessuna
causa. Se invece non avete fatto niente, un altro canale piccolo sarà salito in
cima, e la classifica avrà un nuovo vincitore da spiegare. In entrambi i casi
l’organizzazione impara qualcosa di falso e lo impara con convinzione crescente,
perché ogni mese arriva una nuova conferma.&lt;/p&gt;
&lt;blockquote&gt;
&lt;p&gt;Le persone hanno un’intuizione statistica pessima e una capacità narrativa
eccellente: la seconda copre i danni della prima.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;aside class=&quot;margine&quot;&gt;
&lt;p&gt;Questa è la ragione per cui i “test creativi” a basso volume producono learning
più velocemente di quanto la statistica consenta: non stanno misurando le
creatività, stanno campionando la propria varianza.&lt;/p&gt;
&lt;/aside&gt;
&lt;p&gt;Il costo non è il budget spostato su un canale mediocre — quello è recuperabile.
Il costo è che il team ha ora una teoria del proprio mercato costruita su
fluttuazioni casuali, e le teorie sono appiccicose. Nella mia esperienza da questo
lato del tavolo, disinstallare una convinzione fondata su tre mesi di dati
rumorosi richiede molto più di tre mesi di dati puliti.&lt;/p&gt;
&lt;h2&gt;Tre difese che costano poco&lt;/h2&gt;
&lt;p&gt;Non servono strumenti nuovi. Servono tre abitudini, tutte fastidiose, tutte
economiche.&lt;/p&gt;
&lt;p&gt;&lt;strong&gt;Primo: nessun tasso senza denominatore, nessuna classifica senza intervallo.&lt;/strong&gt;
Se una tabella mostra una percentuale, deve mostrare accanto la numerosità e una
misura di incertezza. È una modifica di venti minuti al vostro foglio di calcolo
e cambia quello che le persone &lt;em&gt;vedono&lt;/em&gt; prima ancora di quello che pensano. Chi
guarda un 5,8% ± 3,0 non propone di spostare il budget.&lt;/p&gt;
&lt;p&gt;&lt;strong&gt;Secondo: soglia minima di volume per entrare in classifica.&lt;/strong&gt; Decisa in
anticipo, non caso per caso. Sotto quella soglia il canale esiste, si finanzia,
si osserva — ma non compete nella graduatoria e non genera decisioni. La soglia
si calcola: fissate la differenza &lt;span class=&quot;katex&quot;&gt;&lt;span class=&quot;katex-mathml&quot;&gt;&lt;math xmlns=&quot;http://www.w3.org/1998/Math/MathML&quot;&gt;&lt;semantics&gt;&lt;mrow&gt;&lt;mi&gt;δ&lt;/mi&gt;&lt;/mrow&gt;&lt;annotation encoding=&quot;application/x-tex&quot;&gt;&#92;delta&lt;/annotation&gt;&lt;/semantics&gt;&lt;/math&gt;&lt;/span&gt;&lt;span class=&quot;katex-html&quot; aria-hidden=&quot;true&quot;&gt;&lt;span class=&quot;base&quot;&gt;&lt;span class=&quot;strut&quot; style=&quot;height:0.6944em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord mathnormal&quot; style=&quot;margin-right:0.0379em;&quot;&gt;δ&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt; che vi interesserebbe davvero
rilevare e ricavate la numerosità che serve per distinguerla dal rumore,&lt;/p&gt;
&lt;p class=&quot;katex-block&quot;&gt;&lt;span class=&quot;katex-display&quot;&gt;&lt;span class=&quot;katex&quot;&gt;&lt;span class=&quot;katex-mathml&quot;&gt;&lt;math xmlns=&quot;http://www.w3.org/1998/Math/MathML&quot; display=&quot;block&quot;&gt;&lt;semantics&gt;&lt;mrow&gt;&lt;mi&gt;n&lt;/mi&gt;&lt;mtext&gt;  &lt;/mtext&gt;&lt;mo&gt;≈&lt;/mo&gt;&lt;mtext&gt;  &lt;/mtext&gt;&lt;mfrac&gt;&lt;mrow&gt;&lt;mn&gt;2&lt;/mn&gt;&lt;mtext&gt; &lt;/mtext&gt;&lt;mo fence=&quot;true&quot; stretchy=&quot;true&quot; minsize=&quot;1.2em&quot; maxsize=&quot;1.2em&quot;&gt;(&lt;/mo&gt;&lt;msub&gt;&lt;mi&gt;z&lt;/mi&gt;&lt;mrow&gt;&lt;mn&gt;1&lt;/mn&gt;&lt;mo&gt;−&lt;/mo&gt;&lt;mi&gt;α&lt;/mi&gt;&lt;mi mathvariant=&quot;normal&quot;&gt;/&lt;/mi&gt;&lt;mn&gt;2&lt;/mn&gt;&lt;/mrow&gt;&lt;/msub&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;msub&gt;&lt;mi&gt;z&lt;/mi&gt;&lt;mrow&gt;&lt;mn&gt;1&lt;/mn&gt;&lt;mo&gt;−&lt;/mo&gt;&lt;mi&gt;β&lt;/mi&gt;&lt;/mrow&gt;&lt;/msub&gt;&lt;msup&gt;&lt;mo fence=&quot;true&quot; stretchy=&quot;true&quot; minsize=&quot;1.2em&quot; maxsize=&quot;1.2em&quot;&gt;)&lt;/mo&gt;&lt;mn&gt;2&lt;/mn&gt;&lt;/msup&gt;&lt;mtext&gt;  &lt;/mtext&gt;&lt;mi&gt;p&lt;/mi&gt;&lt;mtext&gt; &lt;/mtext&gt;&lt;mo stretchy=&quot;false&quot;&gt;(&lt;/mo&gt;&lt;mn&gt;1&lt;/mn&gt;&lt;mo&gt;−&lt;/mo&gt;&lt;mi&gt;p&lt;/mi&gt;&lt;mo stretchy=&quot;false&quot;&gt;)&lt;/mo&gt;&lt;/mrow&gt;&lt;msup&gt;&lt;mi&gt;δ&lt;/mi&gt;&lt;mn&gt;2&lt;/mn&gt;&lt;/msup&gt;&lt;/mfrac&gt;&lt;/mrow&gt;&lt;annotation encoding=&quot;application/x-tex&quot;&gt;n &#92;;&#92;approx&#92;; &#92;frac{2&#92;,&#92;bigl(z_{1-&#92;alpha/2} + z_{1-&#92;beta}&#92;bigr)^{2}&#92;; p&#92;,(1-p)}{&#92;delta^{2}}
&lt;/annotation&gt;&lt;/semantics&gt;&lt;/math&gt;&lt;/span&gt;&lt;span class=&quot;katex-html&quot; aria-hidden=&quot;true&quot;&gt;&lt;span class=&quot;base&quot;&gt;&lt;span class=&quot;strut&quot; style=&quot;height:0.4831em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord mathnormal&quot;&gt;n&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.2778em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.2778em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mrel&quot;&gt;≈&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.2778em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.2778em;&quot;&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;base&quot;&gt;&lt;span class=&quot;strut&quot; style=&quot;height:2.4852em;vertical-align:-0.686em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord&quot;&gt;&lt;span class=&quot;mopen nulldelimiter&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mfrac&quot;&gt;&lt;span class=&quot;vlist-t vlist-t2&quot;&gt;&lt;span class=&quot;vlist-r&quot;&gt;&lt;span class=&quot;vlist&quot; style=&quot;height:1.7992em;&quot;&gt;&lt;span style=&quot;top:-2.368em;&quot;&gt;&lt;span class=&quot;pstrut&quot; style=&quot;height:3.054em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord&quot;&gt;&lt;span class=&quot;mord&quot;&gt;&lt;span class=&quot;mord mathnormal&quot; style=&quot;margin-right:0.0379em;&quot;&gt;δ&lt;/span&gt;&lt;span class=&quot;msupsub&quot;&gt;&lt;span class=&quot;vlist-t&quot;&gt;&lt;span class=&quot;vlist-r&quot;&gt;&lt;span class=&quot;vlist&quot; style=&quot;height:0.7401em;&quot;&gt;&lt;span style=&quot;top:-2.989em;margin-right:0.05em;&quot;&gt;&lt;span class=&quot;pstrut&quot; style=&quot;height:2.7em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;sizing reset-size6 size3 mtight&quot;&gt;&lt;span class=&quot;mord mtight&quot;&gt;&lt;span class=&quot;mord mtight&quot;&gt;2&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span style=&quot;top:-3.284em;&quot;&gt;&lt;span class=&quot;pstrut&quot; style=&quot;height:3.054em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;frac-line&quot; style=&quot;border-bottom-width:0.04em;&quot;&gt;&lt;/span&gt;&lt;/span&gt;&lt;span style=&quot;top:-3.7992em;&quot;&gt;&lt;span class=&quot;pstrut&quot; style=&quot;height:3.054em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord&quot;&gt;&lt;span class=&quot;mord&quot;&gt;2&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.1667em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mopen&quot;&gt;&lt;span class=&quot;delimsizing size1&quot;&gt;(&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;mord&quot;&gt;&lt;span class=&quot;mord mathnormal&quot; style=&quot;margin-right:0.044em;&quot;&gt;z&lt;/span&gt;&lt;span class=&quot;msupsub&quot;&gt;&lt;span class=&quot;vlist-t vlist-t2&quot;&gt;&lt;span class=&quot;vlist-r&quot;&gt;&lt;span class=&quot;vlist&quot; style=&quot;height:0.3448em;&quot;&gt;&lt;span style=&quot;top:-2.5198em;margin-left:-0.044em;margin-right:0.05em;&quot;&gt;&lt;span class=&quot;pstrut&quot; style=&quot;height:2.7em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;sizing reset-size6 size3 mtight&quot;&gt;&lt;span class=&quot;mord mtight&quot;&gt;&lt;span class=&quot;mord mtight&quot;&gt;1&lt;/span&gt;&lt;span class=&quot;mbin mtight&quot;&gt;−&lt;/span&gt;&lt;span class=&quot;mord mathnormal mtight&quot; style=&quot;margin-right:0.0037em;&quot;&gt;α&lt;/span&gt;&lt;span class=&quot;mord mtight&quot;&gt;/2&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;vlist-s&quot;&gt;​&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;vlist-r&quot;&gt;&lt;span class=&quot;vlist&quot; style=&quot;height:0.3552em;&quot;&gt;&lt;span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.2222em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mbin&quot;&gt;+&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.2222em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord&quot;&gt;&lt;span class=&quot;mord mathnormal&quot; style=&quot;margin-right:0.044em;&quot;&gt;z&lt;/span&gt;&lt;span class=&quot;msupsub&quot;&gt;&lt;span class=&quot;vlist-t vlist-t2&quot;&gt;&lt;span class=&quot;vlist-r&quot;&gt;&lt;span class=&quot;vlist&quot; style=&quot;height:0.3361em;&quot;&gt;&lt;span style=&quot;top:-2.55em;margin-left:-0.044em;margin-right:0.05em;&quot;&gt;&lt;span class=&quot;pstrut&quot; style=&quot;height:2.7em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;sizing reset-size6 size3 mtight&quot;&gt;&lt;span class=&quot;mord mtight&quot;&gt;&lt;span class=&quot;mord mtight&quot;&gt;1&lt;/span&gt;&lt;span class=&quot;mbin mtight&quot;&gt;−&lt;/span&gt;&lt;span class=&quot;mord mathnormal mtight&quot; style=&quot;margin-right:0.0528em;&quot;&gt;β&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;vlist-s&quot;&gt;​&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;vlist-r&quot;&gt;&lt;span class=&quot;vlist&quot; style=&quot;height:0.2861em;&quot;&gt;&lt;span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;mclose&quot;&gt;&lt;span class=&quot;mclose&quot;&gt;&lt;span class=&quot;delimsizing size1&quot;&gt;)&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;msupsub&quot;&gt;&lt;span class=&quot;vlist-t&quot;&gt;&lt;span class=&quot;vlist-r&quot;&gt;&lt;span class=&quot;vlist&quot; style=&quot;height:1.054em;&quot;&gt;&lt;span style=&quot;top:-3.3029em;margin-right:0.05em;&quot;&gt;&lt;span class=&quot;pstrut&quot; style=&quot;height:2.7em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;sizing reset-size6 size3 mtight&quot;&gt;&lt;span class=&quot;mord mtight&quot;&gt;&lt;span class=&quot;mord mtight&quot;&gt;2&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.2778em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord mathnormal&quot;&gt;p&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.1667em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mopen&quot;&gt;(&lt;/span&gt;&lt;span class=&quot;mord&quot;&gt;1&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.2222em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mbin&quot;&gt;−&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.2222em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord mathnormal&quot;&gt;p&lt;/span&gt;&lt;span class=&quot;mclose&quot;&gt;)&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;vlist-s&quot;&gt;​&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;vlist-r&quot;&gt;&lt;span class=&quot;vlist&quot; style=&quot;height:0.686em;&quot;&gt;&lt;span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;mclose nulldelimiter&quot;&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/p&gt;
&lt;p&gt;che con &lt;span class=&quot;katex&quot;&gt;&lt;span class=&quot;katex-mathml&quot;&gt;&lt;math xmlns=&quot;http://www.w3.org/1998/Math/MathML&quot;&gt;&lt;semantics&gt;&lt;mrow&gt;&lt;mi&gt;p&lt;/mi&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mn&gt;0,035&lt;/mn&gt;&lt;/mrow&gt;&lt;annotation encoding=&quot;application/x-tex&quot;&gt;p = 0{,}035&lt;/annotation&gt;&lt;/semantics&gt;&lt;/math&gt;&lt;/span&gt;&lt;span class=&quot;katex-html&quot; aria-hidden=&quot;true&quot;&gt;&lt;span class=&quot;base&quot;&gt;&lt;span class=&quot;strut&quot; style=&quot;height:0.625em;vertical-align:-0.1944em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord mathnormal&quot;&gt;p&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.2778em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mrel&quot;&gt;=&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.2778em;&quot;&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;base&quot;&gt;&lt;span class=&quot;strut&quot; style=&quot;height:0.8389em;vertical-align:-0.1944em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord&quot;&gt;0&lt;/span&gt;&lt;span class=&quot;mord&quot;&gt;&lt;span class=&quot;mpunct&quot;&gt;,&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;mord&quot;&gt;035&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;, &lt;span class=&quot;katex&quot;&gt;&lt;span class=&quot;katex-mathml&quot;&gt;&lt;math xmlns=&quot;http://www.w3.org/1998/Math/MathML&quot;&gt;&lt;semantics&gt;&lt;mrow&gt;&lt;mi&gt;δ&lt;/mi&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mn&gt;0,01&lt;/mn&gt;&lt;/mrow&gt;&lt;annotation encoding=&quot;application/x-tex&quot;&gt;&#92;delta = 0{,}01&lt;/annotation&gt;&lt;/semantics&gt;&lt;/math&gt;&lt;/span&gt;&lt;span class=&quot;katex-html&quot; aria-hidden=&quot;true&quot;&gt;&lt;span class=&quot;base&quot;&gt;&lt;span class=&quot;strut&quot; style=&quot;height:0.6944em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord mathnormal&quot; style=&quot;margin-right:0.0379em;&quot;&gt;δ&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.2778em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mrel&quot;&gt;=&lt;/span&gt;&lt;span class=&quot;mspace&quot; style=&quot;margin-right:0.2778em;&quot;&gt;&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;base&quot;&gt;&lt;span class=&quot;strut&quot; style=&quot;height:0.8389em;vertical-align:-0.1944em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord&quot;&gt;0&lt;/span&gt;&lt;span class=&quot;mord&quot;&gt;&lt;span class=&quot;mpunct&quot;&gt;,&lt;/span&gt;&lt;/span&gt;&lt;span class=&quot;mord&quot;&gt;01&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt; e i valori convenzionali di &lt;span class=&quot;katex&quot;&gt;&lt;span class=&quot;katex-mathml&quot;&gt;&lt;math xmlns=&quot;http://www.w3.org/1998/Math/MathML&quot;&gt;&lt;semantics&gt;&lt;mrow&gt;&lt;mi&gt;α&lt;/mi&gt;&lt;/mrow&gt;&lt;annotation encoding=&quot;application/x-tex&quot;&gt;&#92;alpha&lt;/annotation&gt;&lt;/semantics&gt;&lt;/math&gt;&lt;/span&gt;&lt;span class=&quot;katex-html&quot; aria-hidden=&quot;true&quot;&gt;&lt;span class=&quot;base&quot;&gt;&lt;span class=&quot;strut&quot; style=&quot;height:0.4306em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord mathnormal&quot; style=&quot;margin-right:0.0037em;&quot;&gt;α&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt; e
&lt;span class=&quot;katex&quot;&gt;&lt;span class=&quot;katex-mathml&quot;&gt;&lt;math xmlns=&quot;http://www.w3.org/1998/Math/MathML&quot;&gt;&lt;semantics&gt;&lt;mrow&gt;&lt;mi&gt;β&lt;/mi&gt;&lt;/mrow&gt;&lt;annotation encoding=&quot;application/x-tex&quot;&gt;&#92;beta&lt;/annotation&gt;&lt;/semantics&gt;&lt;/math&gt;&lt;/span&gt;&lt;span class=&quot;katex-html&quot; aria-hidden=&quot;true&quot;&gt;&lt;span class=&quot;base&quot;&gt;&lt;span class=&quot;strut&quot; style=&quot;height:0.8889em;vertical-align:-0.1944em;&quot;&gt;&lt;/span&gt;&lt;span class=&quot;mord mathnormal&quot; style=&quot;margin-right:0.0528em;&quot;&gt;β&lt;/span&gt;&lt;/span&gt;&lt;/span&gt;&lt;/span&gt; dà qualche migliaio di utenti per gruppo. Non duecentoquaranta.&lt;/p&gt;
&lt;p&gt;&lt;strong&gt;Terzo: la domanda “quanto sarebbe estremo per caso?”&lt;/strong&gt; prima della domanda
“perché è estremo?”. Una simulazione banale — mille ripetizioni con lo stesso
tasso base e le stesse numerosità — vi dice quanto spesso il canale più piccolo
finisce in cima per pura fortuna. La risposta, nella maggior parte dei mix
reali, è: quasi sempre.&lt;/p&gt;
&lt;p&gt;Nessuna delle tre richiede di saper fare inferenza causale. Sono igiene, non
metodo. Ma hanno una proprietà utile: spostano il momento del dubbio &lt;em&gt;prima&lt;/em&gt;
della riunione, quando l’unica cosa in gioco è un numero, e non dopo, quando in
gioco c’è la faccia di qualcuno.&lt;/p&gt;
&lt;h2&gt;L’osservazione scomoda&lt;/h2&gt;
&lt;p&gt;Kahneman e Tversky notarono che i loro soggetti sbagliavano anche quando erano
psicologi con formazione statistica. Il bias non risiede nella mancanza di
conoscenze: risiede nel fatto che l’intuizione arriva prima del calcolo e si
presenta con la stessa sensazione di ovvietà. Questo esperimento lo conosco da
quindici anni, e ogni volta che guardo una classifica per tasso di conversione
sento, per un istante, che il primo canale sia il migliore.&lt;/p&gt;
&lt;p&gt;La differenza non la fa sapere. La fa aver messo la colonna dell’incertezza
accanto a quella del tasso, mesi prima, quando nessuno stava decidendo niente.&lt;/p&gt;
&lt;hr class=&quot;footnotes-sep&quot;&gt;
&lt;section class=&quot;footnotes&quot;&gt;
&lt;ol class=&quot;footnotes-list&quot;&gt;
&lt;li id=&quot;fn1&quot; class=&quot;footnote-item&quot;&gt;&lt;p&gt;È la stessa struttura del celebre risultato sulle scuole più piccole:
negli anni Novanta diverse fondazioni americane finanziarono la riduzione
delle dimensioni delle scuole perché le scuole con i risultati migliori erano
piccole. Erano piccole anche quelle con i risultati peggiori. &lt;a href=&quot;#fnref1&quot; class=&quot;footnote-backref&quot;&gt;↩︎&lt;/a&gt;&lt;/p&gt;
&lt;/li&gt;
&lt;/ol&gt;
&lt;/section&gt;
</content>
  </entry>
</feed>
