ᱡᱩᱣᱟᱲᱤᱭᱟᱜ ᱵᱷᱩᱞ
ᱢᱤᱫ ᱨᱮᱡᱟᱞᱴ ᱫᱚ ᱡᱮᱞᱮᱧ ᱚᱠᱛᱚ ᱵᱟᱝ ᱩᱫᱩᱜ ᱞᱮᱱ ᱛᱮ “ᱦᱮᱡ ᱦᱚᱨ” ᱠᱟᱱᱟ, ᱱᱚᱣᱟ ᱵᱤᱥᱣᱟᱥ᱾

ᱜᱮᱞ ᱟᱨᱟᱜ ᱛᱟᱭᱚᱢ ᱦᱚᱸ ᱦᱮᱸᱫᱮ ᱨᱮᱭᱟᱜ ᱥᱚᱢᱵᱷᱟᱵᱱᱟ ᱫᱚ 18/37 ᱜᱮ ᱛᱟᱦᱮᱸᱱᱟ᱾ ᱪᱟᱠᱟ ᱫᱚ ᱡᱟᱦᱟᱸᱱ ᱦᱤᱥᱟᱹᱵ ᱵᱟᱝ ᱫᱚᱦᱚᱭ ᱟ ᱟᱨ ᱡᱟᱦᱟᱸᱱ ᱨᱤᱱ ᱵᱟᱝ ᱦᱟᱛᱟᱣ ᱟᱠᱟᱫᱟ᱾
ᱵᱷᱩᱞ
If redᱵᱟᱡᱤᱟᱨᱟᱜ18 ᱟᱨᱟᱜ ᱮᱞ ᱨᱮ ᱥᱚᱢᱟᱱ ᱴᱟᱠᱟ ᱵᱟᱡᱤ᱾ ᱯᱷᱨᱮᱧᱪ ᱨᱮ: ᱨᱩᱡᱽ᱾ᱯᱩᱨᱟᱹ ᱯᱟᱲᱦᱟᱣ → has come up ten times, many players feel that blackᱵᱟᱡᱤᱦᱮᱸᱫᱮ18 ᱦᱮᱸᱫᱮ ᱮᱞ ᱨᱮ ᱥᱚᱢᱟᱱ ᱴᱟᱠᱟ ᱵᱟᱡᱤ᱾ ᱯᱷᱨᱮᱧᱪ ᱨᱮ: ᱱᱣᱟᱨ᱾ᱯᱩᱨᱟᱹ ᱯᱟᱲᱦᱟᱣ → is “due” and bet accordingly. Others feel red is “hot” and bet that. Both are reasoning from a sequence that is already over to a spin that has not happened. In the fair single-zeroᱪᱟᱠᱟᱡᱤᱨᱚᱢᱤᱫᱜᱮ ᱦᱟᱹᱨᱭᱟᱹᱲ ᱠᱷᱟᱱᱟ ᱡᱟᱦᱟᱸ ᱥᱚᱢᱟᱱ ᱴᱟᱠᱟ, ᱰᱚᱡᱮᱱ ᱟᱨᱵᱟᱝ ᱠᱚᱞᱚᱢ ᱵᱟᱡᱤ ᱵᱟᱭ ᱠᱚᱵᱷᱟᱨᱟ; ᱠᱮᱥᱤᱱᱳ ᱵᱟᱲᱛᱤ ᱨᱮᱭᱟᱜ ᱢᱩᱞ᱾ ᱮᱴᱟᱜ ᱧᱩᱛᱩᱢ: ᱥᱤᱝᱜᱚᱞ ᱡᱤᱨᱚ, 0᱾ᱯᱩᱨᱟᱹ ᱯᱟᱲᱦᱟᱣ → model with independent spinsᱜᱚᱱᱤᱛᱟᱡᱟᱫ ᱜᱷᱩᱨᱟᱹᱣᱢᱟᱨᱮ ᱪᱮᱫ ᱦᱚᱸ ᱦᱩᱭ ᱠᱷᱟᱱ ᱦᱚᱸ ᱡᱚᱛᱚ ᱜᱷᱩᱨᱟᱹᱣ ᱨᱮ ᱥᱚᱢᱵᱷᱟᱵᱱᱟ ᱢᱤᱫ ᱞᱮᱠᱟ ᱜᱮ᱾ ᱱᱚᱶᱟ ᱥᱟᱭᱤᱴ ᱨᱮᱭᱟᱜ ᱡᱚᱛᱚ ᱦᱤᱥᱟᱹᱵ ᱨᱮᱭᱟᱜ ᱵᱮᱥ᱾ᱯᱩᱨᱟᱹ ᱯᱟᱲᱦᱟᱣ →, the next-spin probabilityᱜᱚᱱᱤᱛᱥᱚᱢᱵᱷᱟᱵᱱᱟᱢᱤᱫ ᱨᱮᱡᱟᱞᱴ ᱨᱮᱭᱟᱜ ᱥᱚᱢᱵᱷᱟᱵᱱᱟ, 0 ᱠᱷᱚᱱ 1 ᱟᱨᱵᱟᱝ 0% ᱠᱷᱚᱱ 100%᱾ ᱢᱤᱫ ᱮᱞ: 1/37 = 2.70%᱾ᱯᱩᱨᱟᱹ ᱯᱟᱲᱦᱟᱣ → of black remains 18/37.
ᱪᱮᱫᱟᱜ ᱱᱚᱣᱟ ᱥᱟᱹᱨᱤ ᱞᱮᱠᱟ ᱵᱩᱡᱷᱟᱹᱣᱚᱜ-ᱟ
Under the independent fair-wheel model, the proportions of red and black each tend toward 18/37 as the sample grows. This does not require a compensating run of black. After ten initial reds and N further spins, expected red and black counts differ by ten, while the contribution of those initial reds to the total proportion becomes smaller. Actual counts still fluctuate.
ᱢᱚᱱᱴᱮ ᱠᱟᱨᱞᱚ, 18 ᱟᱜᱚᱥᱴ 1913
ᱠᱮᱥᱤᱱᱳ ᱫᱮ ᱢᱚᱱᱴᱮ-ᱠᱟᱨᱞᱚ ᱨᱮᱭᱟᱜ ᱢᱤᱫ ᱴᱮᱵᱩᱞ ᱨᱮ ᱦᱮᱸᱫᱮ ᱞᱟᱜᱟᱛᱟᱨ 26 ᱫᱷᱟᱣ ᱩᱫᱩᱜ ᱞᱮᱱᱟ᱾ ᱠᱚᱢ ᱵᱮᱥᱤ ᱜᱮᱞ ᱢᱚᱬᱮ ᱥᱟᱭ ᱦᱮᱸᱫᱮ ᱠᱷᱚᱱ, ᱠᱷᱮᱞᱚᱸᱰ ᱠᱚ ᱟᱨᱟᱜ ᱨᱮ ᱫᱷᱮᱨ ᱵᱟᱡᱤ ᱠᱚ ᱞᱟᱜᱟᱣ ᱞᱮᱫᱟ, ᱵᱟᱨ ᱜᱩᱱ ᱟᱨ ᱫᱚᱦᱲᱟ ᱵᱟᱨ ᱜᱩᱱ ᱠᱟᱛᱮ, ᱡᱚᱛᱚ ᱜᱷᱩᱨᱟᱹᱣ ᱢᱟᱨᱮ ᱠᱷᱚᱱ “ᱵᱮᱥᱤ ᱥᱟᱹᱨᱤ”᱾ ᱠᱮᱥᱤᱱᱳ ᱫᱚ ᱢᱤᱞᱤᱭᱚᱱ ᱢᱤᱞᱤᱭᱚᱱ ᱯᱷᱨᱟᱸ ᱮ ᱡᱤᱛᱟᱹᱣ ᱞᱮᱫᱟ᱾ ᱯᱟᱦᱤᱞ ᱦᱮᱸᱫᱮ ᱠᱷᱚᱱ ᱚᱱᱟ ᱞᱟᱜᱟᱛᱟᱨ ᱨᱮᱭᱟᱜ ᱥᱚᱢᱵᱷᱟᱵᱱᱟ ᱫᱚ 137 ᱢᱤᱞᱤᱭᱚᱱ ᱠᱷᱚᱱ 1 ᱛᱟᱦᱮᱸᱱ ᱠᱟᱱᱟ, ᱟᱨ ᱫᱤᱥᱚᱢ ᱨᱮᱭᱟᱜ ᱠᱮᱥᱤᱱᱳ ᱨᱮ ᱡᱟᱦᱟᱸ ᱴᱷᱮᱱ ᱦᱚᱸ ᱢᱤᱫ ᱫᱤᱱ ᱱᱚᱣᱟ ᱦᱩᱭᱩᱜ ᱨᱮᱭᱟᱜ ᱥᱚᱢᱵᱷᱟᱵᱱᱟ ᱫᱚ ᱡᱷᱚᱛᱚ ᱞᱮᱠᱟ ᱢᱤᱫ ᱛᱟᱦᱮᱸᱱ ᱠᱟᱱᱟ᱾ ᱱᱚᱣᱟ ᱞᱟᱹᱜᱤᱫ ᱜᱮ ᱡᱩᱣᱟᱲᱤᱭᱟᱜ ᱵᱷᱩᱞ ᱫᱚ ᱢᱚᱱᱴᱮ ᱠᱟᱨᱞᱚ ᱵᱷᱩᱞ ᱦᱚᱸ ᱢᱮᱱᱟ ᱠᱟᱱᱟ᱾
ᱵᱤᱥᱮᱥ ᱵᱟᱝ᱾ 44 ᱫᱷᱟᱣ ᱠᱷᱚᱱ ᱢᱤᱫ ᱫᱷᱟᱣ ᱦᱩᱭᱩᱜ-ᱟ᱾
ᱠᱷᱮᱞᱚᱸᱰ ᱠᱚ ᱧᱮᱞ ᱮᱦᱚᱵ ᱮᱫᱟ᱾ ᱟᱨᱟᱜ ᱫᱚ “ᱦᱮᱡ ᱦᱚᱨ”᱾
ᱴᱮᱵᱩᱞ ᱨᱮ ᱫᱷᱮᱨ ᱦᱚᱲ᱾ ᱡᱚᱛᱚ ᱦᱟᱹᱱᱤ ᱛᱟᱭᱚᱢ ᱟᱨᱟᱜ ᱨᱮ ᱵᱟᱡᱤ ᱴᱟᱠᱟ ᱵᱟᱨ ᱜᱩᱱ ᱦᱩᱭᱩᱜ-ᱟ᱾
ᱫᱷᱚᱱ ᱫᱟᱣᱞᱚᱛ ᱦᱟᱨᱟᱣ᱾ ᱡᱚᱛᱚ ᱱᱟᱣᱟ ᱦᱮᱸᱫᱮ ᱛᱮ ᱟᱨᱟᱜ ᱵᱮᱥᱤ ᱥᱟᱹᱨᱤ ᱞᱮᱠᱟ ᱵᱩᱡᱷᱟᱹᱣᱚᱜ-ᱟ᱾
ᱞᱟᱜᱟᱛᱟᱨ ᱢᱩᱪᱟᱹᱫᱚᱜ-ᱟ᱾ ᱠᱮᱥᱤᱱᱳ ᱫᱚ ᱢᱤᱞᱤᱭᱚᱱ ᱢᱤᱞᱤᱭᱚᱱ ᱯᱷᱨᱟᱸ ᱮ ᱡᱤᱛᱟᱹᱣ ᱠᱮᱫᱟ ᱚᱱᱟ ᱦᱚᱲ ᱠᱷᱚᱱ ᱡᱟᱦᱟᱸ ᱠᱚ ᱫᱚ ᱢᱤᱫ ᱞᱟᱜᱟᱛᱟᱨ ᱵᱤᱨᱩᱫᱷ ᱵᱟᱡᱤ ᱠᱚ ᱞᱟᱜᱟᱣ ᱞᱮᱫᱟ, ᱡᱟᱦᱟᱸ ᱫᱚ ᱟᱡ ᱞᱟᱜᱟᱛᱟᱨ ᱠᱟᱱᱟ ᱚᱱᱟ “ᱵᱟᱰᱟᱭ” ᱵᱟᱝ ᱫᱟᱲᱮᱭᱟᱜ ᱛᱟᱦᱮᱸᱱᱟ᱾
ᱜᱚᱨᱚᱢ ᱛᱤ
ᱢᱤᱫ ᱨᱚᱝ ᱞᱟᱜᱟᱛᱟᱨ ᱨᱮ ᱢᱮᱱᱟᱜ ᱢᱮᱱᱛᱮ ᱚᱱᱟ ᱨᱮ ᱵᱟᱡᱤ ᱞᱟᱜᱟᱣ ᱫᱚ ᱚᱱᱟ ᱜᱮ ᱵᱷᱩᱞ ᱩᱞᱴᱟᱹ ᱦᱚᱨ ᱛᱮ ᱠᱟᱱᱟ: ᱞᱟᱜᱟᱛᱟᱨ ᱢᱩᱪᱟᱹᱫ ᱵᱚᱫᱚᱞ ᱨᱮ ᱞᱟᱦᱟᱜ ᱥᱮᱱ ᱟᱥᱟ ᱦᱩᱭᱩᱜ-ᱟ᱾ ᱢᱚᱱ ᱵᱤᱜᱭᱟᱱᱤ ᱠᱚ ᱱᱚᱣᱟ ᱫᱚ ᱜᱚᱨᱚᱢ ᱛᱤᱵᱷᱨᱚᱢᱜᱚᱨᱚᱢ ᱛᱤᱡᱤᱛ ᱨᱮᱭᱟᱜ ᱟᱨᱵᱟᱝ ᱢᱤᱫ ᱨᱮᱡᱟᱞᱴ ᱨᱮᱭᱟᱜ ᱞᱟᱜᱟᱛᱟᱨ ᱫᱚ ᱮᱴᱟᱜ ᱵᱟᱹᱲᱛᱤ ᱥᱚᱢᱵᱷᱟᱵᱱᱟ ᱵᱮᱱᱟᱣᱟ, ᱱᱚᱶᱟ ᱵᱤᱥᱣᱟᱥ᱾ ᱪᱟᱠᱟ ᱨᱮ ᱱᱚᱶᱟ ᱫᱚ ᱩᱞᱴᱟᱹ ᱜᱚᱴᱚᱱ ᱡᱩᱣᱟᱲᱤᱭᱟᱜ ᱵᱷᱩᱞ ᱠᱟᱱᱟ᱾ᱯᱩᱨᱟᱹ ᱯᱟᱲᱦᱟᱣ → ᱢᱮᱱᱛᱮ ᱠᱚ ᱢᱮᱱᱟ, ᱡᱤᱞᱚᱣᱤᱪ, ᱣᱟᱞᱚᱱᱮ ᱟᱨ ᱴᱣᱮᱨᱥᱠᱤ ᱨᱮᱭᱟᱜ 1985 ᱨᱮᱭᱟᱜ ᱯᱟᱲᱦᱟᱣ ᱛᱟᱭᱚᱢ, ᱡᱟᱦᱟᱸ ᱫᱚ ᱵᱟᱥᱠᱮᱴᱵᱚᱞ ᱥᱩᱴᱤᱝ ᱨᱮ ᱞᱟᱜᱟᱛᱟᱨ ᱠᱚ ᱥᱮᱸᱫᱽᱨᱟ ᱞᱮᱫᱟ ᱟᱨ ᱥᱟᱹᱜᱩᱱ ᱵᱟᱦᱨᱮ ᱡᱟᱦᱟᱸᱱ ᱵᱟᱝ ᱧᱟᱢ ᱞᱮᱫᱟ᱾ 2018 ᱨᱮ ᱢᱤᱞᱟᱨ ᱟᱨ ᱥᱟᱱᱦᱩᱨᱡᱳ ᱫᱚ ᱠᱤᱱ ᱩᱫᱩᱜ ᱞᱮᱫᱟ ᱡᱮ ᱯᱚᱨᱚᱠᱷ ᱜᱮ ᱛᱮᱨᱪᱷᱟ ᱛᱟᱦᱮᱸᱱ ᱠᱟᱱᱟ: ᱢᱤᱫ ᱦᱩᱰᱤᱧ ᱨᱮᱠᱚᱨᱰ ᱨᱮ, ᱡᱤᱛ ᱛᱟᱭᱚᱢ ᱛᱟᱭᱚᱢ ᱡᱤᱛ ᱨᱮᱭᱟᱜ ᱦᱟᱹᱴᱤᱧ ᱫᱚ ᱟᱣᱥᱚᱛ ᱛᱮ ᱥᱩᱴᱟᱨ ᱨᱮᱭᱟᱜ ᱡᱚᱛᱚ ᱦᱟᱨ ᱠᱷᱚᱱ ᱠᱚᱢ ᱦᱩᱭᱩᱜ-ᱟ, ᱚᱱᱟᱛᱮ ᱡᱟᱦᱟᱸ ᱠᱷᱮᱞᱚᱸᱰ ᱨᱮ ᱡᱟᱦᱟᱸᱱ ᱜᱚᱨᱚᱢ ᱛᱤ ᱵᱟᱹᱱᱩᱜ-ᱟ, ᱩᱱᱤ ᱦᱩᱰᱤᱧ ᱴᱷᱟᱸᱰᱟ ᱞᱮᱠᱟ ᱧᱮᱞᱚᱜ-ᱟᱭ᱾ ᱥᱩᱫᱷᱨᱟᱹᱣ ᱛᱟᱭᱚᱢ, ᱚᱱᱟ ᱜᱮ ᱰᱟᱴᱟ ᱫᱚ ᱦᱩᱰᱤᱧ ᱜᱚᱨᱚᱢ ᱛᱤ ᱩᱫᱩᱜ ᱮᱫᱟ᱾ ᱢᱤᱫ ᱦᱚᱲ ᱫᱚ ᱜᱚᱨᱚᱢ ᱫᱟᱲᱮᱭᱟᱜ-ᱟᱭ; ᱪᱟᱠᱟ ᱨᱮ ᱫᱚ ᱡᱟᱦᱟᱸᱱ ᱫᱚᱥᱟ ᱵᱟᱹᱱᱩᱜ-ᱟ᱾ ᱛᱤᱱᱟᱹᱜ ᱠᱷᱮᱞᱚᱸᱰ ᱫᱚ ᱵᱟᱨᱭᱟ ᱵᱤᱥᱣᱟᱥ ᱥᱟᱶᱛᱮ ᱠᱚ ᱫᱚᱦᱚᱭ ᱟ, ᱡᱮ ᱢᱤᱫ ᱞᱟᱜᱟᱛᱟᱨ “ᱟᱹᱰᱤ ᱡᱮᱞᱮᱧ” ᱦᱟᱹᱵᱤᱡ ᱪᱟᱞᱟᱜ-ᱟ ᱟᱨ ᱮᱱᱠᱷᱟᱱ ᱢᱩᱪᱟᱹᱫᱚᱜ-ᱟ᱾ ᱢᱤᱫ ᱴᱷᱤᱠ ᱪᱟᱠᱟ ᱨᱮ ᱡᱚᱛᱚ ᱨᱩᱯ ᱨᱮ ᱥᱤᱠᱟᱹ ᱪᱟᱞᱟᱜ ᱠᱟᱛᱮ ᱨᱚᱝ ᱵᱟᱪᱷᱟᱣ ᱞᱮᱠᱟ ᱢᱤᱫ ᱞᱮᱠᱟᱱ ᱟᱸᱫᱟᱡᱽ ᱜᱚᱱᱚᱝᱜᱚᱱᱤᱛᱟᱸᱫᱟᱡᱽ ᱜᱚᱱᱚᱝᱡᱮᱞᱮᱧ ᱚᱠᱛᱚ ᱨᱮ ᱢᱤᱫ ᱵᱟᱡᱤ ᱨᱮᱭᱟᱜ ᱟᱣᱥᱚᱛ ᱨᱮᱡᱟᱞᱴ᱾ ᱡᱚᱛᱚ ᱥᱟᱫᱷᱟᱨᱚᱱ ᱭᱩᱨᱚᱯᱤᱭᱟᱱ ᱵᱟᱡᱤ ᱨᱮ ᱵᱟᱡᱤ ᱴᱟᱠᱟ ᱨᱮᱭᱟᱜ −1/37᱾ᱯᱩᱨᱟᱹ ᱯᱟᱲᱦᱟᱣ → ᱢᱮᱱᱟᱜ-ᱟ: ᱵᱟᱡᱤ ᱴᱟᱠᱟ ᱨᱮᱭᱟᱜ −2.70%᱾
ᱢᱤᱫ ᱥᱟᱹᱨᱤ ᱴᱮᱵᱩᱞ ᱨᱮ
Croson and Sundali (2005) watched 18 hours of security video from one double-zero table at a large casino in Reno: 139 players placing 24,131 bets. On the even-money betsᱵᱟᱡᱤᱥᱚᱢᱟᱱ ᱴᱟᱠᱟ ᱵᱟᱡᱤᱟᱨᱟᱜ/ᱦᱮᱸᱫᱮ, ᱵᱤᱡᱳᱲ/ᱡᱳᱲ, ᱢᱟᱨᱟᱝ/ᱦᱩᱰᱤᱧ᱾ 18 ᱮᱞ, ᱵᱷᱩᱜᱛᱟᱱ 1:1᱾ᱯᱩᱨᱟᱹ ᱯᱟᱲᱦᱟᱣ →, players were about evenly divided between following short runs and opposing them. After five of a kind in a row they leaned against the run, and after six or more, 85% of their bets were on the other side. The same players showed the hot hand about themselves: after a win on the numbers they covered 13.6 numbers on the next spin on average, against 9.2 after a loss. The study found no statistically significant evidence of wheel bias. These are observations from one table, and the authors discuss alternative explanations for changes in betting after wins.
ᱠᱮᱥᱤᱱᱳ ᱵᱟᱦᱨᱮ
ᱚᱱᱟ ᱜᱮ ᱟᱹᱜᱩ ᱫᱚ ᱥᱟᱹᱨᱤ ᱯᱷᱚᱭᱥᱟᱞᱟ ᱦᱩᱭᱩᱜ ᱴᱷᱟᱶ ᱨᱮ ᱦᱚᱸ ᱧᱮᱞᱚᱜ-ᱟ᱾ ᱠᱞᱚᱴᱯᱷᱮᱞᱴᱟᱨ ᱟᱨ ᱠᱩᱠ (1993) ᱫᱚ ᱢᱮᱨᱤᱞᱮᱱᱰ ᱨᱮᱭᱟᱜ ᱢᱟᱦᱟᱭ ᱢᱟᱦᱟᱭ ᱮᱞ ᱠᱷᱮᱞ ᱠᱤᱱ ᱧᱮᱞ ᱞᱮᱫᱟ: ᱢᱤᱫ ᱮᱞ ᱩᱫᱩᱜ ᱞᱮᱱ ᱥᱟᱶᱛᱮ ᱚᱱᱟ ᱨᱮ ᱵᱟᱡᱤ ᱴᱟᱠᱟ ᱟᱹᱰᱤ ᱠᱚᱢ ᱮᱱᱟ ᱟᱨ ᱫᱚᱦᱲᱟ ᱵᱟᱲᱦᱟᱣ ᱞᱟᱹᱜᱤᱫ ᱟᱭᱢᱟ ᱪᱟᱸᱫᱚ ᱞᱟᱹᱠᱛᱤ ᱮᱱᱟ, ᱡᱩᱫᱤ ᱦᱚᱸ ᱰᱨᱚ ᱫᱚ ᱡᱟᱦᱟᱸᱱ ᱵᱟᱝ ᱫᱚᱦᱚ ᱢᱚᱱᱮᱭᱟ᱾ ᱪᱮᱱ, ᱢᱚᱥᱠᱳᱣᱤᱴᱥ ᱟᱨ ᱥᱩ (2016) ᱫᱚ ᱯᱮᱥᱟᱫᱟᱨ ᱦᱚᱲ ᱛᱟᱞᱟ ᱨᱮ ᱱᱚᱣᱟ ᱠᱚ ᱧᱟᱢ ᱞᱮᱫᱟ᱾ ᱭᱩᱱᱟᱭᱴᱮᱰ ᱥᱴᱮᱴᱥ ᱨᱮᱭᱟᱜ ᱥᱚᱨᱚᱬ ᱡᱚᱡ ᱠᱚ ᱫᱚ ᱞᱟᱦᱟ ᱢᱚᱠᱚᱫᱽᱫᱚᱢᱟ ᱦᱩᱠᱩᱢ ᱮᱢ ᱞᱮᱠᱷᱟᱱ ᱫᱟᱨᱟᱭ ᱛᱮ ᱟᱹᱨᱤᱡ ᱨᱮᱭᱟᱜ ᱥᱚᱢᱵᱷᱟᱵᱱᱟ 3.3 ᱯᱟᱨᱥᱮᱱᱴᱮᱡ ᱯᱚᱭᱮᱱᱴ ᱦᱟᱹᱵᱤᱡ ᱵᱮᱥᱤ ᱛᱟᱦᱮᱸᱱ ᱠᱟᱱᱟ᱾ ᱵᱮᱥᱵᱚᱞ ᱟᱢᱯᱟᱭᱟᱨ ᱠᱚ ᱫᱚ ᱢᱤᱫ ᱥᱴᱨᱟᱭᱠ ᱢᱮᱱ ᱛᱟᱭᱚᱢ ᱫᱟᱨᱟᱭ ᱛᱮ ᱥᱴᱨᱟᱭᱠ ᱢᱮᱱ ᱨᱮᱭᱟᱜ ᱥᱚᱢᱵᱷᱟᱵᱱᱟ 1.5 ᱯᱚᱭᱮᱱᱴ ᱠᱚᱢ ᱛᱟᱦᱮᱸᱱ ᱠᱟᱱᱟ, ᱟᱨ ᱥᱮᱴᱮᱨ ᱯᱤᱪ ᱨᱮ ᱚᱱᱟ ᱠᱷᱚᱱ ᱵᱟᱨ ᱜᱩᱱ ᱠᱷᱚᱱ ᱵᱮᱥᱤ᱾ ᱢᱤᱫ ᱥᱮᱫ ᱯᱚᱨᱚᱠᱷ ᱨᱮ ᱢᱤᱫ ᱞᱮᱠᱟᱱ ᱯᱷᱤᱥ ᱧᱟᱢ ᱤᱫᱤ ᱞᱚᱱ ᱚᱯᱷᱤᱥᱟᱨ ᱠᱚ ᱛᱟᱞᱟ ᱨᱮ, 9% ᱦᱟᱹᱵᱤᱡ ᱯᱷᱚᱭᱥᱟᱞᱟ ᱱᱚᱣᱟ ᱦᱚᱨ ᱛᱮ ᱩᱞᱴᱟᱹ ᱮᱱᱟ᱾ ᱢᱚᱠᱚᱫᱽᱫᱚᱢᱟ ᱠᱚ ᱫᱚ ᱵᱮᱜᱚᱨ ᱫᱷᱟᱨᱟ ᱛᱟᱹᱞᱤ ᱛᱮ ᱦᱮᱡ ᱞᱮᱱᱟ, ᱚᱱᱟᱛᱮ ᱢᱟᱨᱮ ᱯᱷᱚᱭᱥᱟᱞᱟ ᱫᱚ ᱫᱟᱨᱟᱭ ᱛᱮ ᱵᱟᱵᱚᱛ ᱡᱟᱦᱟᱸᱱ ᱵᱟᱝ ᱞᱟᱹᱭ ᱞᱮᱫᱟ, ᱴᱷᱤᱠ ᱡᱮᱞᱮᱠᱟ ᱢᱟᱨᱮ ᱜᱷᱩᱨᱟᱹᱣ ᱫᱚ ᱫᱟᱨᱟᱭ ᱛᱮ ᱜᱷᱩᱨᱟᱹᱣ ᱵᱟᱵᱚᱛ ᱡᱟᱦᱟᱸᱱ ᱵᱟᱝ ᱞᱟᱹᱭᱟ᱾

ᱩᱞᱴᱟᱹ ᱡᱩᱣᱟᱲᱤᱭᱟᱜ ᱵᱷᱩᱞ
ᱫᱚᱨᱥᱚᱱ ᱵᱤᱜᱭᱟᱱᱤ ᱟᱭᱚᱱ ᱦᱮᱠᱤᱝ (Ian Hacking) (1987) ᱫᱚ ᱯᱮ ᱥᱟᱭ ᱨᱩᱯ ᱨᱮᱭᱟᱜ ᱧᱩᱛᱩᱢ ᱮ ᱫᱚᱦᱚ ᱞᱮᱫᱟ᱾ ᱢᱤᱫ ᱦᱚᱲ ᱠᱩᱴᱤ ᱵᱷᱤᱛᱤᱨ ᱮ ᱵᱚᱞᱚᱜ-ᱟ, ᱵᱟᱨ ᱯᱟᱥᱟ ᱰᱚᱵᱚᱞ ᱛᱩᱨᱩᱭ ᱧᱩᱨᱩᱜ ᱮ ᱧᱮᱞᱟ ᱟᱨ ᱮ ᱵᱤᱪᱟᱹᱨᱟ ᱡᱮ ᱱᱚᱣᱟ ᱠᱚ ᱞᱟᱦᱟ ᱠᱷᱚᱱ ᱟᱭᱢᱟ ᱫᱷᱟᱣ ᱜᱩᱲᱤᱭᱟᱹᱣ ᱟᱠᱟᱱ ᱛᱟᱦᱮᱸᱱᱟ᱾ ᱡᱚᱛᱚ ᱜᱩᱲᱤᱭᱟᱹᱣ ᱨᱮ ᱰᱚᱵᱚᱞ ᱛᱩᱨᱩᱭ ᱨᱮᱭᱟᱜ ᱥᱚᱢᱵᱷᱟᱵᱱᱟ ᱥᱚᱢᱟᱱ ᱜᱮᱭᱟ, ᱧᱤᱫᱟᱹ ᱨᱮᱭᱟᱜ ᱯᱟᱦᱤᱞ ᱥᱮ ᱦᱟᱡᱟᱨ ᱥᱟᱭ, ᱚᱱᱟᱛᱮ ᱚᱱᱟ ᱧᱮᱞ ᱫᱚ ᱠᱷᱮᱞ ᱛᱤᱱ ᱡᱮᱞᱮᱧ ᱪᱟᱞᱟᱜ ᱠᱟᱱᱟ ᱚᱱᱟ ᱵᱟᱵᱚᱛ ᱡᱟᱦᱟᱸᱱ ᱵᱟᱝ ᱞᱟᱹᱭᱟ᱾ ᱨᱩᱞᱮᱴ ᱨᱮ ᱱᱚᱣᱟ ᱫᱚ ᱚᱱᱟ ᱠᱷᱮᱞᱚᱸᱰ ᱠᱟᱱᱟᱭ ᱡᱟᱦᱟᱸ ᱤᱨᱟᱹᱞ ᱟᱨᱟᱜ ᱩᱫᱩᱜ ᱟᱠᱟᱱ ᱵᱳᱨᱰ ᱨᱮ ᱦᱮᱡ ᱠᱟᱛᱮ ᱮ ᱵᱤᱪᱟᱹᱨᱟ ᱡᱮ ᱚᱱᱟ ᱠᱚ ᱵᱮᱱᱟᱣ ᱞᱟᱹᱜᱤᱫ ᱴᱮᱵᱩᱞ ᱫᱚ ᱜᱷᱟᱸᱴᱟ ᱜᱷᱟᱸᱴᱟ ᱜᱷᱩᱨᱟᱹᱣ ᱟᱠᱟᱱ ᱛᱟᱦᱮᱸᱱᱟ᱾ ᱥᱟᱫᱷᱟᱨᱚᱬ ᱵᱷᱩᱞ ᱫᱚ ᱛᱟᱭᱚᱢ ᱠᱟᱛᱷᱟ ᱫᱟᱨᱟᱭ ᱛᱮ ᱜᱷᱩᱨᱟᱹᱣ ᱨᱮ ᱯᱟᱲᱦᱟᱣ ᱮᱫᱟ; ᱱᱚᱣᱟ ᱫᱚ ᱢᱤᱫ ᱜᱟᱱ ᱜᱟᱱ ᱧᱮᱞ ᱠᱷᱚᱱ ᱡᱟᱦᱟᱸᱭ ᱦᱚᱸ ᱵᱟᱭ ᱧᱮᱞ ᱞᱮᱫ ᱛᱟᱭᱚᱢ ᱠᱟᱛᱷᱟ ᱯᱟᱲᱦᱟᱣ ᱮᱫᱟ᱾
ᱢᱤᱫ ᱢᱚᱱ ᱵᱤᱜᱭᱟᱱᱤ ᱡᱟᱦᱟᱸ ᱫᱚ ᱱᱚᱣᱟ ᱮ ᱵᱤᱥᱣᱟᱥ ᱞᱮᱫᱟ
ᱱᱚᱣᱟ ᱵᱷᱩᱞ ᱨᱮᱭᱟᱜ ᱯᱟᱲᱦᱟᱣ ᱦᱚᱲ ᱠᱚ ᱦᱚᱸ ᱥᱟᱯᱚᱲᱛᱟ ᱠᱚ ᱮᱢ ᱟᱠᱟᱫᱟ᱾ ᱡᱟᱨᱢᱟᱱ ᱢᱚᱱ ᱵᱤᱜᱭᱟᱱᱤ ᱠᱟᱨᱞ ᱢᱟᱨᱵᱮ (Karl Marbe) ᱫᱚ ᱯᱩᱱ ᱨᱮᱡᱤᱥᱴᱟᱨ ᱮ ᱯᱟᱲᱦᱟᱣ ᱞᱮᱫᱟ, ᱡᱚᱛᱚ ᱨᱮ 49,152 ᱦᱚᱯᱚᱱ ᱡᱟᱱᱟᱢ, ᱟᱨ ᱮ ᱧᱟᱢ ᱞᱮᱫᱟ ᱡᱮ ᱠᱚᱲᱟ ᱟᱨᱵᱟᱝ ᱠᱩᱲᱤ ᱨᱮᱭᱟᱜ ᱡᱚᱛᱚ ᱠᱷᱚᱱ ᱡᱮᱞᱮᱧ ᱞᱟᱜᱟᱛᱟᱨ ᱫᱚ 17 ᱛᱟᱦᱮᱸᱱ ᱠᱟᱱᱟ, ᱟᱨ ᱮ ᱢᱮᱱ ᱞᱮᱫᱟ ᱡᱮ ᱡᱮᱞᱮᱧ ᱞᱟᱜᱟᱛᱟᱨ ᱫᱚ ᱥᱟᱹᱜᱩᱱ ᱠᱷᱚᱱ ᱠᱚᱢ ᱦᱩᱭᱩᱜ-ᱟ, ᱚᱱᱟᱛᱮ ᱢᱤᱫ ᱞᱟᱜᱟᱛᱟᱨ ᱫᱚ ᱮᱴᱟᱜ ᱨᱮᱡᱟᱞᱴ ᱵᱮᱥᱤ ᱥᱚᱢᱵᱷᱟᱵᱱᱟ ᱵᱮᱱᱟᱣ ᱮᱫᱟ᱾ ᱥᱟᱹᱜᱩᱱ ᱮᱴᱟᱜ ᱠᱟᱛᱷᱟ ᱞᱟᱹᱭ ᱮᱫᱟ: 196,608 ᱡᱟᱱᱟᱢ ᱨᱮ 17 ᱥᱮ ᱵᱮᱥᱤ ᱞᱟᱜᱟᱛᱟᱨ ᱫᱚ ᱢᱚᱬᱮ ᱫᱷᱟᱣ ᱠᱷᱚᱱ ᱠᱚᱢ ᱵᱮᱥᱤ ᱯᱩᱱ ᱫᱷᱟᱣ ᱩᱫᱩᱜᱚᱜ-ᱟ᱾ ᱢᱟᱨᱵᱮ ᱫᱚ ᱴᱷᱤᱠ ᱚᱱᱟ ᱜᱮ ᱮ ᱧᱟᱢ ᱞᱮᱫᱟ ᱡᱟᱦᱟᱸ ᱫᱚ ᱵᱮᱜᱚᱨ ᱫᱷᱟᱨᱟ ᱞᱟᱹᱲᱤ ᱨᱮᱭᱟᱜ ᱧᱮᱞ ᱠᱟᱱᱟ᱾ ᱮᱯᱥᱴᱟᱭᱤᱱ (2013) ᱫᱚ ᱱᱚᱣᱟ ᱠᱟᱛᱷᱟ ᱫᱚᱦᱲᱟ ᱮ ᱞᱟᱹᱭ ᱮᱫᱟ; ᱞᱟᱜᱟᱛᱟᱨ ᱨᱮᱭᱟᱜ ᱥᱚᱢᱵᱷᱟᱵᱱᱟ ᱥᱟᱦᱴᱟ ᱫᱚ ᱨᱩᱞᱮᱴ ᱨᱩᱯ ᱮᱢ ᱮᱫᱟ᱾

ᱚᱠᱟ ᱴᱷᱮᱱ ᱵᱷᱩᱞ ᱵᱟᱝ ᱞᱟᱹᱜᱩᱜ-ᱟ
If the wheel were physically biased, past spins would be informative, because they would be evidence about the pocketsᱥᱟᱡ ᱥᱟᱢᱟᱱᱠᱷᱟᱱᱟᱨᱚᱴᱚᱨ ᱨᱮᱭᱟᱜ 37 ᱮᱞ ᱚᱞ ᱠᱷᱟᱱᱟ ᱠᱷᱚᱱ ᱢᱤᱫ᱾ ᱮᱴᱟᱜ ᱧᱩᱛᱩᱢ: ᱥᱞᱚᱴ, ᱠᱟᱯ᱾ᱯᱩᱨᱟᱹ ᱯᱟᱲᱦᱟᱣ → rather than about luck. That is the Lab’s subject. The sample required depends on the size of the bias, the test and the acceptable error rates. A short streak alone does not distinguish bias from chance; comparing explanations also requires assumptions about how plausible each was before the data.
ᱠᱩᱠᱞᱤ ᱠᱚ
ᱵᱟᱝ᱾ ᱱᱚᱣᱟ ᱫᱚ ᱡᱟᱦᱟᱸᱱ ᱟᱡᱟᱫ ᱫᱚᱦᱲᱟ ᱦᱩᱭᱩᱜ ᱠᱟᱛᱷᱟ ᱨᱮ ᱞᱟᱹᱜᱩᱜ-ᱟ: ᱯᱟᱥᱟ, ᱥᱤᱠᱟᱹ, ᱞᱚᱴᱟᱨᱤ ᱮᱞ᱾
With more independent spins, the proportions tend toward their probabilities; they do not have to be close after any fixed number. Standard single-zero bets have an expected lossᱜᱚᱱᱤᱛᱟᱸᱫᱟᱡᱽ ᱦᱟᱹᱱᱤᱟᱸᱫᱟᱡᱽ ᱜᱚᱱᱚᱝ ᱜᱩᱬ ᱡᱚᱛᱚ ᱵᱟᱡᱤ ᱴᱟᱠᱟ; ᱢᱤᱫ ᱥᱮᱥᱚᱱ ᱟᱨᱵᱟᱝ ᱚᱠᱛᱚ ᱨᱮ ᱟᱣᱥᱚᱛ ᱛᱮ ᱦᱟᱨᱟᱣᱚᱜ ᱴᱟᱠᱟ᱾ᱯᱩᱨᱟᱹ ᱯᱟᱲᱦᱟᱣ → of 2.70% of turnoverᱴᱟᱠᱟᱡᱚᱛᱚ ᱵᱟᱡᱤ ᱦᱤᱥᱟᱹᱵᱢᱤᱫ ᱥᱮᱥᱚᱱ ᱟᱨᱵᱟᱝ ᱚᱠᱛᱚ ᱨᱮ ᱡᱚᱛᱚ ᱵᱟᱡᱤ ᱴᱟᱠᱟ, ᱫᱚᱦᱲᱟ ᱞᱟᱜᱟᱣ ᱡᱤᱛ ᱴᱟᱠᱟ ᱥᱮᱞᱮᱫ᱾ ᱮᱴᱟᱜ ᱧᱩᱛᱩᱢ: ᱦᱮᱱᱰᱮᱞ, ᱮᱠᱥᱚᱱ᱾ᱯᱩᱨᱟᱹ ᱯᱟᱲᱦᱟᱣ →, but your actual result can be far from that average.
ᱟᱣᱥᱚᱛ ᱨᱮᱭᱟᱜ ᱱᱤᱭᱚᱢ ᱫᱚ ᱵᱟᱹᱱᱩᱜ-ᱟ᱾ ᱢᱟᱨᱟᱝ ᱮᱞ ᱨᱮᱭᱟᱜ ᱱᱤᱭᱚᱢᱜᱚᱱᱤᱛᱢᱟᱨᱟᱝ ᱮᱞ ᱨᱮᱭᱟᱜ ᱱᱤᱭᱚᱢᱯᱚᱨᱚᱠᱷ ᱨᱮᱭᱟᱜ ᱮᱞ ᱵᱟᱲᱦᱟᱣ ᱞᱮᱠᱷᱟᱱ ᱟᱡᱟᱫ ᱨᱮᱡᱟᱞᱴ ᱨᱮᱭᱟᱜ ᱟᱣᱥᱚᱛ ᱟᱸᱫᱟᱡᱽ ᱜᱚᱱᱚᱝ ᱴᱷᱮᱱ ᱥᱮᱴᱮᱨᱚᱜ-ᱟ᱾ᱯᱩᱨᱟᱹ ᱯᱟᱲᱦᱟᱣ → ᱢᱮᱱᱟᱜ-ᱟ, ᱟᱨ ᱚᱱᱟ ᱫᱚ ᱞᱟᱹᱜᱟᱹᱣ ᱛᱮ ᱠᱟᱹᱢᱤᱭᱟ᱾
ᱜᱚᱱᱤᱛ ᱥᱟᱶ ᱨᱟᱹᱥᱠᱟᱹ
ᱱᱚᱶᱟ ᱠᱚ ᱯᱩᱭᱞᱩ ᱟᱯᱮ ᱛᱮ ᱠᱩᱨᱩᱢᱩᱴᱩ ᱯᱮ᱾ ᱢᱤᱫ ᱢᱤᱫ ᱥᱚᱢᱟᱫᱷᱟᱱ ᱫᱚ ᱩᱠᱩ ᱢᱮᱱᱟᱜ-ᱟ; ᱛᱮᱭᱟᱨ ᱦᱩᱭ ᱠᱟᱛᱮ ᱡᱷᱤᱡᱽ ᱯᱮ᱾
10 ᱫᱷᱟᱣ ᱪᱟᱞᱟᱠ ᱛᱟᱭᱚᱢ ᱢᱤᱫ ᱥᱤᱠᱠᱟ ᱨᱮ 7 ᱦᱮᱰ ᱟᱨ 3 ᱴᱮᱞ ᱧᱮᱞᱚᱜ-ᱟ᱾ 1,000 ᱫᱷᱟᱣ ᱪᱟᱞᱟᱠ ᱛᱟᱭᱚᱢ 517 ᱦᱮᱰ ᱟᱨ 483 ᱴᱮᱞ᱾ ᱴᱮᱞ ᱫᱚ ᱥᱟᱵ ᱞᱮᱱᱟ ᱥᱮ?
ᱥᱚᱢᱟᱫᱷᱟᱱ ᱩᱫᱩᱜ
ᱛᱮᱞᱟ ᱵᱟᱝ᱾ ᱯᱷᱟᱨᱟᱠ ᱫᱚ 4 ᱠᱷᱚᱱ 34 ᱦᱟᱹᱵᱤᱡ ᱦᱟᱲᱟ ᱮᱱᱟ᱾ ᱮᱠᱮᱱ ᱟᱹᱱᱥᱟᱹ ᱜᱮ ᱥᱚᱢᱟᱱ ᱥᱮᱫ ᱥᱟᱵᱟᱥᱟᱵᱤ ᱦᱮᱡ ᱮᱱᱟ᱾
The ratio of heads to tails fell from 7/3, about 2.33, to 517/483, about 1.07. This example illustrates how a proportion can move closer to one-half while the count difference grows.
The gap between them grew, from 4 to 34. With a fair coin a gap of around 30 after 1,000 tosses is ordinary: the typical gap grows roughly with the square root of the number of tosses, while the ratio settles toward 1.
Nothing pulled the tails back. The early surplus of heads was diluted by the next 990 tosses, not corrected.
ᱢᱤᱫ ᱥᱤᱠᱠᱟ ᱯᱩᱱ ᱫᱷᱟᱣ ᱪᱟᱞᱟᱠ ᱯᱮ ᱟᱨ ᱮᱠᱮᱱ ᱦᱮᱰ ᱛᱟᱭᱚᱢ ᱜᱮ ᱦᱤᱡᱩᱜ ᱠᱟᱱ ᱪᱟᱞᱟᱠ ᱠᱚ ᱧᱮᱞ ᱯᱮ᱾ ᱚᱱᱠᱚ ᱠᱷᱚᱱ ᱪᱮᱫ ᱦᱟᱹᱴᱤᱧ ᱫᱚ ᱦᱮᱰ ᱠᱟᱱᱟ? ᱦᱩᱭ ᱫᱟᱲᱮᱭᱟᱜ 16 ᱥᱤᱨᱤᱡᱽ ᱨᱮᱭᱟᱜ ᱡᱚᱛᱚ ᱨᱮ ᱦᱤᱥᱟᱹᱵ ᱯᱮ ᱟᱨ ᱦᱟᱹᱴᱤᱧ ᱠᱚ ᱨᱮᱭᱟᱜ ᱟᱣᱥᱚᱛ ᱦᱟᱛᱟᱣ ᱯᱮ᱾ ᱚᱱᱟ ᱫᱚ 50% ᱥᱮ?
ᱥᱚᱢᱟᱫᱷᱟᱱ ᱩᱫᱩᱜ
ᱛᱮᱞᱟ ᱵᱟᱝ᱾ ᱚᱱᱟ ᱫᱚ 17/42, ᱡᱷᱚᱛᱚ 40.5%᱾ ᱱᱚᱣᱟ ᱞᱮᱠᱟᱛᱮ ᱧᱮᱞ ᱞᱮᱠᱷᱟᱱ, ᱢᱤᱫ ᱴᱷᱤᱠ ᱥᱤᱠᱠᱟ ᱦᱮᱰ ᱛᱟᱭᱚᱢ ᱴᱷᱟᱸᱰᱟ ᱦᱩᱭᱩᱜ ᱞᱮᱠᱟ ᱧᱮᱞᱚᱜ-ᱟ᱾
16 ᱥᱤᱨᱤᱡᱽ ᱞᱤᱥᱴ ᱯᱮ (H = ᱦᱮᱰ, T = ᱴᱮᱞ)᱾ ᱚᱱᱠᱚ ᱠᱷᱚᱱ ᱵᱟᱨ, TTTT ᱟᱨ TTTH, ᱨᱮ ᱯᱟᱹᱦᱤᱞ ᱯᱮ ᱪᱟᱞᱟᱠ ᱨᱮ ᱡᱟᱦᱟᱱ ᱦᱮᱰ ᱵᱟᱹᱱᱩᱜ-ᱟ, ᱚᱱᱟᱛᱮ ᱦᱮᱰ ᱛᱟᱭᱚᱢ ᱪᱮᱫ ᱦᱚᱸ ᱵᱟᱝ ᱦᱤᱡᱩᱜ-ᱟ; 14 ᱵᱟᱹᱜᱤᱭᱟ᱾
ᱡᱚᱛᱚ ᱨᱮ, ᱦᱮᱰ ᱛᱟᱭᱚᱢ ᱦᱤᱡᱩᱜ ᱪᱟᱞᱟᱠ ᱠᱚ ᱦᱟᱛᱟᱣ ᱯᱮ ᱟᱨ ᱪᱮᱫ ᱦᱟᱹᱴᱤᱧ ᱦᱮᱰ ᱠᱟᱱᱟ ᱚᱱᱟ ᱧᱟᱢ ᱯᱮ᱾ HHHH ᱫᱚ 3 ᱠᱷᱚᱱ 3 ᱮᱢᱟ᱾ HHHT ᱫᱚ 3 ᱠᱷᱚᱱ 2 ᱮᱢᱟ᱾ HTHT ᱫᱚ 2 ᱠᱷᱚᱱ 0 ᱮᱢᱟ᱾ ᱡᱚᱛᱚ 14 ᱧᱮᱞ ᱠᱟᱛᱮ: ᱛᱩᱨᱩᱭ ᱥᱤᱨᱤᱡᱽ 0 ᱮᱢᱟ, ᱯᱩᱱ 1/2 ᱮᱢᱟ, ᱢᱤᱫ 2/3 ᱮᱢᱟ ᱟᱨ ᱯᱮ 1 ᱮᱢᱟ (HHHH, THHH ᱟᱨ TTHH)᱾
ᱦᱟᱹᱴᱤᱧ ᱠᱚ ᱡᱩᱲᱤ ᱯᱮ: 4 × 1/2 + 2/3 + 3 = 17/3᱾ 14 ᱛᱮ ᱦᱟᱹᱴᱤᱧ ᱯᱮ: 17/42 ≈ 40.5%᱾
ᱡᱟᱦᱟᱱ ᱪᱟᱞᱟᱠ ᱦᱚᱸ ᱴᱮᱞ ᱥᱮᱫ ᱵᱟᱝ ᱦᱮᱡ ᱟᱠᱟᱱᱟ; ᱡᱚᱛᱚ ᱜᱮ 50–50᱾ ᱦᱮᱞᱮᱫ ᱫᱚ ᱥᱤᱨᱤᱡᱽ ᱛᱟᱭᱚᱢ ᱥᱤᱨᱤᱡᱽ ᱟᱣᱥᱚᱛ ᱦᱟᱛᱟᱣ ᱠᱷᱚᱱ ᱦᱤᱡᱩᱜ-ᱟ: ᱦᱮᱰ ᱛᱮ ᱯᱮᱨᱮᱡ ᱥᱤᱨᱤᱡᱽ ᱨᱮ ᱧᱮᱞ ᱞᱟᱹᱜᱤᱫ ᱟᱹᱰᱤ ᱜᱟᱱ ᱪᱟᱞᱟᱠ ᱢᱮᱱᱟᱜ-ᱟ ᱢᱮᱱᱠᱷᱟᱱ ᱢᱤᱫ ᱫᱷᱟᱣ ᱜᱮ ᱞᱮᱠᱷᱟᱚᱜ-ᱟ, ᱢᱤᱫ ᱜᱮ ᱦᱮᱰ ᱢᱮᱱᱟᱜ ᱥᱤᱨᱤᱡᱽ ᱞᱮᱠᱟ᱾ ᱢᱤᱞᱟᱨ ᱟᱨ ᱥᱟᱱᱡᱩᱨᱡᱚ (Miller and Sanjurjo) (2018) ᱫᱚ ᱩᱫᱩᱜ ᱠᱮᱫᱟ ᱡᱮ 1985 ᱨᱮᱭᱟᱜ ᱜᱚᱨᱚᱢ ᱛᱤ ᱯᱟᱲᱦᱟᱣ ᱫᱚ ᱟᱡᱟᱜ ᱵᱟᱥᱠᱮᱴᱵᱚᱞ ᱠᱷᱮᱞᱚᱸᱰ ᱠᱚ ᱱᱚᱣᱟ ᱞᱮᱠᱟᱛᱮ ᱜᱮ ᱢᱟᱯ ᱞᱮᱫᱟ, ᱚᱱᱟᱛᱮ ᱡᱟᱦᱟᱸ ᱠᱷᱮᱞᱚᱸᱰ ᱠᱚ ᱢᱚᱱᱮ ᱞᱮᱠᱟ ᱠᱷᱚᱱ ᱵᱟᱹᱲᱛᱤ ᱞᱟᱜᱟᱛᱟᱨ ᱵᱟᱝ ᱧᱮᱞᱚᱜ ᱠᱟᱱ ᱛᱟᱦᱮᱸ ᱠᱟᱱᱟ, ᱥᱟᱹᱨᱤ ᱛᱮ ᱚᱱᱠᱚ ᱪᱮᱫ ᱜᱟᱱ ᱵᱟᱹᱲᱛᱤ ᱞᱟᱜᱟᱛᱟᱨ ᱛᱟᱦᱮᱸ ᱠᱟᱱᱟ᱾
ᱞᱟᱸᱫᱟ ᱞᱟᱹᱜᱤᱫ ᱠᱚᱢᱤᱠ

ᱞᱟᱹᱜᱟᱹᱣ, ᱥᱩᱫᱷᱨᱟᱹᱣ ᱵᱟᱝ᱾ ᱪᱟᱠᱟ ᱫᱚ ᱡᱟᱦᱟᱸᱱ ᱨᱤᱱ ᱵᱟᱝ ᱦᱟᱛᱟᱣ ᱟᱠᱟᱫᱟ᱾
ᱯᱷᱮᱰᱟᱛ
- Feller, W. An Introduction to Probability Theory and Its Applications, Vol. 1.
- Schwartz, D. G. Roll the Bones: The History of Gambling.
- Ethier, S. N. The Doctrine of Chances: Probabilistic Aspects of Gambling.
- Epstein, R. A. The Theory of Gambling and Statistical Logic, 2nd ed.
- Croson, R. & Sundali, J. “The gambler’s fallacy and the hot hand: empirical data from casinos.”
- Chen, D. L., Moskowitz, T. J. & Shue, K. “Decision making under the gambler’s fallacy: evidence from asylum judges, loan officers, and baseball umpires.”
- Clotfelter, C. T. & Cook, P. J. “The ‘gambler’s fallacy’ in lottery play.”
- Gilovich, T., Vallone, R. & Tversky, A. “The hot hand in basketball: on the misperception of random sequences.”
- Miller, J. B. & Sanjurjo, A. “Surprised by the hot hand fallacy? A truth in the law of small numbers.”
- Hacking, I. “The inverse gambler’s fallacy: the argument from design. The anthropic principle applied to Wheeler universes.”