5D Library › 古典卜卦尊贵度积分、恒星二十七宿岁差与维姆肖塔里运限链 · English
第八章 「值得注意之处」的基率与两项预注册研究
一张盘里总有可说之处
推导
# Chapter 8. p15: How often is something "notable" in a chart?
# Over random moments 1900-2049 (p02, p03, p05 rules):
# - at least one of Mercury ... Saturn retrograde
# - at least one planet in its domicile or exaltation
# - at least one mutual reception by domicile
# - at least one planet combust (within 8 deg 30' of the Sun)
# - the Moon "void of course": no Ptolemaic aspect to the Sun or to
# Mercury ... Saturn before it leaves its sign (linear motion)
import importlib.util
import random
def load(name):
spec = importlib.util.spec_from_file_location(name, name + ".py")
m = importlib.util.module_from_spec(spec)
spec.loader.exec_module(m)
return m
eph = load("p02_ephemeris")
dig = load("p01_dignity_tables")
rec = load("p03_receptions")
pro = load("p04_prohibition")
OTHERS = ("Sun", "Mercury", "Venus", "Mars", "Jupiter", "Saturn")
if __name__ == "__main__":
rng = random.Random(73015)
N = 15000
hits = {"retrograde": 0, "dom/exalt": 0, "mutual": 0, "combust": 0,
"void Moon": 0, "any of these": 0}
for _ in range(N):
jd = rng.uniform(2415020.5, 2469807.5)
pos = eph.positions(jd)
w = {b: eph.speed(jd, b) for b in OTHERS + ("Moon",)}
f = {"retrograde": any(w[p] < 0 for p in OTHERS[1:]),
"dom/exalt": any(int(pos[p] // 30) in dig.DOM[p] or
int(pos[p] // 30) == dig.EXALT[p]
for p in dig.PL),
"mutual": rec.cycles({p: int(pos[p] // 30)
for p in dig.PL})[0] > 0,
"combust": any(abs((pos[p] - pos["Sun"] + 180) % 360 - 180)
<= 8.5 for p in OTHERS[1:] + ("Moon",))}
left = (30 - pos["Moon"] % 30) / w["Moon"]
f["void Moon"] = all(pro.next_perfection(
pos["Moon"], w["Moon"], pos[x], w[x], left) is None
for x in OTHERS)
f["any of these"] = any(f.values())
for k, v in f.items():
hits[k] += v
for k, v in hits.items():
print(" {:13s} {:.4f}".format(k, v / N))
# retrograde 0.6882
# dom/exalt 0.8179
# mutual 0.3595
# combust 0.4857
# void Moon 0.2237
# any of these 0.9908在 1900 到 2049 年的随机时刻,水星到土星中至少一颗逆行的概率为 0.6882;至少一颗行星入庙或入旺 0.8179;至少一对庙宫互溶 0.3595;至少一颗行星(含月亮)焦伤 0.4857;月亮「空亡」(在离开所在星座前不再与太阳及水星至土星成托勒密相位)0.2237。这五项中至少出现一项的概率为 0.9908。换言之,任何一张盘几乎都带有至少一个「值得注意」的标记。
这与第 71 本第五章的多重比较是同一个现象:事先指定的一项特征有它自己的基率,事后从许多特征中挑出发生了的那一项,基率就接近于一。再加上笼统陈述的覆盖效应与选择性记忆 (Forer 1949; Dickson and Kelly 1985),「这张盘说中了」的经验几乎不需要任何真实的预测力就可以产生。
注
四步核查。 第一,写明计分约定:界的版本、三分主星、游走是否重复扣分、是否计入后天尊贵度(第一、三章)。第二,写明岁差体系与运限年长(第四、七章)。第三,核对出生时间或起盘时间的精度(第五章)。第四,任何「特别的标记」先与本章的基率比较,并把「准不准」换成事先登记的盲法检验。
研究甲:卜卦是非题
收集结果会在三个月内揭晓的是非题。一位从不接触提问者的占星师只根据提问时刻的星盘判断「是」或「否」,在结果揭晓前登记。由于不同问题「是」的基率不同,每题配一个以抛硬币决定的对照判断,机会一致率为 0.5。
推导
# Chapter 8. p16: Two pre-registered studies and their sizes.
# (A) Horary yes/no: questions with an outcome that will be known within
# three months are submitted; an astrologer who never sees the asker
# judges each from the chart of the moment of asking, recording
# "yes" or "no" before the outcome. Base rates of "yes" differ by
# question, so each question is paired with a coin-matched control:
# chance agreement 0.5. Exact one-sided binomial, alpha 0.05, power 0.90.
# (B) Dasha periods: participants receive four unlabelled timelines of
# mahadasha changes (their own and three from other birth data) and
# choose the one that best matches their life. Chance 1/4.
from math import comb
def tail(n, k, p):
return sum(comb(n, i) * p ** i * (1 - p) ** (n - i)
for i in range(k, n + 1))
def plan(p0, p1, alpha=0.05, power=0.90):
for n in range(20, 5000):
k = next(k for k in range(n + 1) if tail(n, k, p0) <= alpha)
if tail(n, k, p1) >= power:
return n, k, tail(n, k, p0)
if __name__ == "__main__":
for p1 in (0.60, 0.65):
n, k, a = plan(0.5, p1)
print(" (A) accuracy {:.2f} vs 0.50: {} questions, at least {}"
" correct".format(p1, n, k))
print(" exact size {:.4f}".format(a))
for p1 in (0.32, 0.35):
n, k, a = plan(0.25, p1)
print(" (B) hit rate {:.2f} vs 0.25: {} participants, at least {}"
" hits".format(p1, n, k))
print(" exact size {:.4f}".format(a))
# (A) accuracy 0.60 vs 0.50: 213 questions, at least 119 correct
# exact size 0.0499
# (A) accuracy 0.65 vs 0.50: 93 questions, at least 55 correct
# exact size 0.0483
# (B) hit rate 0.32 vs 0.25: 359 participants, at least 104 hits
# exact size 0.0485
# (B) hit rate 0.35 vs 0.25: 179 participants, at least 55 hits
# exact size 0.0485若真实准确率为 0.60,单侧 0.05、把握度 0.90 的精确二项检验需要 213 道题,至少答对 119 道,实际第一类错误为 0.0499;若为 0.65,需要 93 道、至少答对 55 道。
研究乙:运限时间线的盲法辨认
每名参与者收到四条不标来源的大运时间线(一条来自自己的出生资料,三条来自他人),选出最符合自己人生的一条。机会命中率为 1/4。若真实命中率为 0.32,需要 359 名参与者、至少 104 人命中;若为 0.35,需要 179 人、至少 55 人命中。时间线只列大运的起止年龄与主星名称,不附任何解释文字,以免巴纳姆效应。
假说 · 尚未验证
预注册的预测(尚未验证): 研究甲的准确率与 0.5 无显著差异;研究乙的命中率与 0.25 无显著差异。什么会证明它错: 研究甲准确率显著高于 0.5,或研究乙命中率显著高于 0.25。
全书小结
先天尊贵度是黄经上的阶梯函数,各行星的期望得分不等;互溶是定位星映射的 2-循环,截夺是首达竞争;「最强行星」随计分约定而变。岁差使 0.8073 的黄经(p07)在两种黄道中属于不同星座;九分盘是宿足序号模 12;维姆肖塔里运限在固定年龄人群中的比例既不是九分之一,也不是周期比例。这些结论都来自规则本身的结构,不涉及任何判断是否应验。
参考文献
Fan, Y. (2026). Horary Dignity Scoring, Sidereal Nakshatra Precession and Vimshottari Dasha Chains: Reception Matrix Algebra, Ayanamsha Drift Kinetics and Non-Uniform Planetary Period Partitions [古典卜卦尊贵度积分、恒星二十七宿岁差与维姆肖塔里运限链] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.24090195 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).