Laboratori bilong Rulet · Laboratori · Statistik · 8 min

Statistik i Makim Samting Tru

“I makim samting” i min olsem stretpela wil i no save mekim olsem planti taim. I no min “i tru”, na i no min “bai yu winim mani”.

Sotpela bekim

P-velyu 0.05 i min olsem stretpela wil bai givim pasin i no stret olsem wanpela taim long tupela ten. Sapos yu lukim tupela ten wil, wanpela bai pas long laki nating.

pasin ol i save bihainim
p < 0.05
hamas taim giaman ansa i kamap
5% long wan wan tes
stretim
mekim gen
em i no
p(wil i gat bias)

P-velyu em i wanem

Assuming the stated null model, the p-valuelaboratorip-velyuSans long lukim senis i bikpela olsem o i winim dispela yu lukim, sapos wil i stret.Ritim olgeta → is the probabilitymatematikSansSans bilong wanpela resal, long 0 i go 1 o 0% i go 100%. Wanpela namba: 1/37 = 2.70%.Ritim olgeta → of a test statistic at least as extreme as the observed one. It is not the probability that the wheel is fair or that the result arose by chance. Interpretation also depends on the design, assumptions and tests attempted.

toktoki stret?
p = 0.03 i min olsem stretpela wil i givim ol namba olsem 3% bilong taimyes
p = 0.03 i min olsem i gat 97% sans olsem wil i gat biasnogat
p = 0.03 i min olsem hat hul bai i stap hatnogat
p = 0.03 bihain long yu lukim 30 wil em i strongpela maknogat; yu ken wetim wanpela o tupela ansa olsem long laki nating
p = 0.03 tupela taim, long tupela samting yu kaunim i no hangamap long narapela, long wanpela wilnau em i gutpela long lukluk gut

Planti tes

Olgeta tes yu mekim long stretpela wil i gat 5% sans long pas long mak 5%. Testim tupela ten wil na wanpela i pas. Testim wanpela wil long tupela ten hap taim na wanpela i pas. Testim wanpela samting yu kaunim wantaim 37 pasin bilong bungim hultulHulWanpela long ol 37 hul i gat namba long rota. Narapela nem: slot, cup.Ritim olgeta →, sikis pasin bilong bungim seksenwilSeksenWanpela hap bilong wil we ol hul i stap wan bihain long narapela, i narakain long wanpela grup long laiaut.Ritim olgeta → na tripela pasin bilong bungim kala, na wanpela samting bai pas. Marasin em: makim tes pastaim long yu lukim ol namba, mekim wanpela taim tasol, na strongim long nupela namba. Planti taim, painim nabaut long planti wil i mekim ansa olsem wil i “gat bias”.

Sans bilong wanpela giaman ansa o moa long p < 0.05%
1 tes
5%
5 tes
22.6%
10 tes
40.1%
20 tes
64.2%
37 tes (wanpela long wan wan hul)
85.0%
%
042.5085.0%
1 − 0.95ⁿ. Sapos yu testim wan wan hul long em yet, klostu oltaim yu bai painim wanpela i “makim samting”.

Makim samting na bikpela bilong en

A large sample can detect a small departure from fairness that still does not make a bet profitable. At a true pocket probability of 1/36.5, a 35:1 bet has negative expectation. At 1/36 it breaks even. A p-value does not measure the size or practical value of an effect.

Strong bilong tes

Power is the chance of rejecting the null under a specified alternative. For one pocket chosen in advance, an exact one-sided 5% binomial test on 1,000 spins rejects at 37 or more hits. If the true rate is 1/30, its power is about 28%, not 50%.

Mak bilong wok tru

Bilong toktok long bias i gat as: tes yu makim pastaim, p < 0.01 i no 0.05, mekim gen long narapela samting yu kaunim i no hangamap na i wankain bikpela, na bihain long yu mekim gen, ol hul i kamap planti i mas kamap 1/34 o moa. Long histori bilong pilai, ol i kamap long dispela mak sampela taim tasol. Ol lista ol i putim ples klia i no soim wanpela taim long nupela wil ol i lukautim gut.

Ol askim

Bilong wanem 0.05?

Pasin i kam long ol yia 1920. I no gat wanpela samting i narakain long en. Taim yu lukim planti wil, 0.01 em i mak i gat as moa.

Sapos p i liklik tru, wil i gat bias tru?

Ating em i no givim ol namba long wankain sans. Em i ken kam long dila, long rong long raitim, o long rong long putim ol namba. Mekim gen.

Ansa i makim samting i tokim mi long wanem hul bilong putim bet?

No. That requires a separate estimate of hit probability, its uncertainty and the actual payoutmaniPeMani bet i win i givim yu, ol i tok olsem “X long 1”, antap long mani bilong bet i kam bek.Ritim olgeta →.

Amamas wantaim Matematik

Traim ol dispela yu yet pastaim. Bekim bilong wanwan i hait; opim taim yu redi.

  1. Suppose a dozenbetDosen1–12, 13–24 o 25–36, pe em i 2:1.Ritim olgeta → chosen before data collection appears 742 times in 3,000 independent double-zerowilSiroWanpela grinpela hul we i no gat bet wan-long-wan, dosen o kolam i karamapim; em i as bilong edj bilong kasino. Narapela nem: wanpela siro, 0.Ritim olgeta → spins. Under the uniform model, how many standard deviationsmatematikStandet diviesenSkwea rut bilong varians; hamas longwe resal i save stap long ekspektet velyu.Ritim olgeta → below the mean is that? What is the probabilitymatematikSansSans bilong wanpela resal, long 0 i go 1 o 0% i go 100%. Wanpela namba: 1/37 = 2.70%.Ritim olgeta → of 742 or fewer hits?

    Soim bekim

    Bekim Klostu 8.1 standet diviesen daunbilo. Stretpela wil bai mekim klostu wanpela taim long 3 × 10¹⁵ traim, em i no save kamap. Bekim ol i prinim wantaim askim i tok 11.75; mipela i kisim 8.07.

    Under the stated model, X is binomial with n = 3,000 and p = 12/38. Its mean is np ≈ 947.37 and its standard deviation is √(np(1 − p)) ≈ 25.46.

    The standardized difference is (742 − 947.37)/25.46 ≈ −8.07. Summing the exact binomial probabilities from 0 through 742 gives P(X ≤ 742) ≈ 9.77 × 10⁻¹⁷. A normal approximation with continuity correction gives about 4.25 × 10⁻¹⁶; the relative error is large this far into the tail.

    This is the probability of the specified data event under the model, not the probability that the wheel is fair. Such a result calls for checking the observations, independence, uniformity and analysis plan. Selecting the worst-looking dozen after inspecting data, or searching many wheels or time periods, requires accounting for that selection.

Liklik lap

Wanpela long tupela ten stretpela wil i “makim samting”. Em i mining bilong dispela namba.

Wanpela long tupela ten stretpela wil i “makim samting”. Em i mining bilong dispela namba.

Ol As bilong Save

  1. Pearson, K. “On the criterion that a given system of deviations from the probable…” https://www.amstat.org/asa/files/pdfs/P-ValueStatement.pdf
  2. Feller, W. An Introduction to Probability Theory and Its Applications, Vol. 1. https://www.itl.nist.gov/div898/software/dataplot/refman1/auxillar/chsqgood.htm
model: Yurop, wanpela sirometodoloji ↗ol as bilong save ↗