Low-budget many tests vs high-budget few tests — how to spend your test pool

Low-budget many tests vs high-budget few tests — how to spend your test pool
Low per-adset ($10-20 each, run 8 angles) → cheap reads on which ANGLE has legs, but thin data, noisy, false negatives kill good creatives.
High per-adset ($50+ on 2-3 angles) → clean significant data fast, but you test fewer ideas and pay more to learn a loser.
Low budget to find the angle direction → high budget to confirm the survivors. Don't judge a $10 adset on 200 impressions and call it dead.
Giving every test the same $20? You're underfunding winners and overfunding noise.