Project 23: Birth Chart Similarity and Career Outcomes
Book: The Big Astrology Book of Research by Renay Oshop
Source: bigastrologybook.com
🌟 Overview — What We Asked
Do individuals in the same professional field have more similar planetary sign placements than individuals from different fields? Can chart similarity predict career category better than chance?
💡 Why This Matters
Most astrological career research asks: "Does sign X predict trait Y?" This project takes a different approach: rather than testing specific astrological rules, it asks whether people in the same profession tend to share chart features — an agnostic, data-driven question that bypasses prescriptive zodiac symbolism entirely.
The methodology — cosine similarity between full one-hot-encoded planetary vectors — preserves more information than traditional sun-sign analysis. Pairwise within-group vs. between-group comparison is the correct statistical framework for clustering questions.
📊 The Data
| Field | Detail |
|---|---|
| Sample | 95 unique individuals across 8 career categories |
| Categories | Tech, Music, Science, Arts, Politics, Sports, Literature, Philosophy |
| Source | Merged dataset: 24 manually selected celebrities + extended celebrity database |
| Planets | 11 bodies: Sun, Moon, Mercury, Venus, Mars, Jupiter, Saturn, Uranus, Neptune, Pluto, North Node |
| Zodiacs | Both Tropical and Vedic (Lahiri ayanamsha) analyzed separately |
Each person's astrological fingerprint: 11 planets × 12 one-hot-encoded sign positions = 132-element binary vector.
Similarity Metric: Cosine similarity between all pairs of vectors. Higher values = more similar planetary sign distributions.
📈 Results
| Zodiac System | Within-Group Mean | Between-Group Mean | p-value | Conclusion |
|---|---|---|---|---|
| Tropical | 0.1015 | 0.0920 | 0.061 | Not significant (near-miss) |
| Vedic (Lahiri) | 0.1123 | 0.0920 | 0.000086 | SIGNIFICANT |
Tropical: Near-Miss
Same-career individuals show slightly higher Tropical chart similarity (0.1015) than cross-career pairs (0.0920), but the difference doesn't reach p < 0.05 (p = 0.061). This is a near-miss that could reflect genuine weak clustering or statistical noise.
Vedic: Statistically Significant Signal
The Vedic (Sidereal) analysis shows a statistically significant difference (p < 0.0001): people in the same career field have more similar planetary sign distributions than people in different fields.
The effect is modest but real:
- Within-group mean cosine: 0.112
- Between-group mean cosine: 0.092
- Difference: +0.020 (approximately 22% higher within-group similarity)
With N=95 and thousands of pairwise comparisons, the signal is robust to sampling variation.
🔍 Interpreting the Vedic Signal — With Caveats
Several important caveats must be addressed before treating this as evidence for astrological career influence:
1. Generational clustering: Some career categories naturally concentrate births in certain decades (Tech pioneers in 1940–1960; modern musicians more recently). Slow-moving outer planets (Jupiter through Pluto) occupy the same signs for years. If Tech founders and Scientists cluster in similar birth decades, their outer planet signs will be similar — not because of career influence, but because of shared birth era. This is the same confound identified in Projects 10 and 26.
2. Selection bias: The celebrity dataset is dominated by historically prominent individuals — not a random sample of professionals. Chart features that cluster among highly successful musicians may differ from those among average musicians.
3. Multiple testing: Two zodiac systems × multiple career-category comparisons = several tests. The Vedic result (p < 0.001) survives a Bonferroni correction for two systems, but not if subcategory analyses are included.
4. What "similarity" measures: Cosine similarity of one-hot-encoded signs is agnostic but imprecise. It doesn't distinguish tight conjunctions from coincidentally shared signs due to generational position.
🔬 What a Better Study Would Do
- Use AstroDatabank — thousands of profession-tagged charts with verified birth data
- Control for birth decade — focus on fast-moving personal planets (Sun, Moon, Mercury, Venus, Mars) that can't be explained by generational cohort effects
- Run permutation tests — shuffle profession labels 10,000 times to build a null distribution that respects the astronomical structure of the data
- Replicate across zodiac systems — confirm whether the Vedic advantage persists with a larger sample
🌟 Conclusion
Pairwise cosine similarity analysis of 95 celebrity charts finds a statistically significant clustering of planetary sign patterns within career categories in the Vedic zodiac (p < 0.0001), with a modest effect (~22% higher within-group similarity than between-group). The Tropical zodiac shows a near-significant trend (p = 0.061).
These findings do not establish astrological causation. The most likely explanation is (a) generational birth-date clustering producing shared outer-planet positions, and/or (b) a genuine but small astrological signal requiring larger, better-controlled replication to characterize.
What is established: the similarity-based methodology is sound, detects structure in the data, and points toward the generational confound as the critical variable to control in future work.