Everyone says 'set exit criteria' — nobody gives you the number. We measured it.Kazdy hovori 'urci si exit kriteria' — nikto ti neda cislo. My sme ho odmerali.
The claim. "Set exit criteria and ignore the sunk cost" is the most repeated career and business advice there is — and it's useless, because it never tells you the threshold. When exactly do you cut aKedy vzdat slabnuce usilie, odmerane. Vzdaj to, ked nedavny vynos klesne ~60% pod svoj vrchol (drawdown stop). Je to vnutorne optimum - prilis skoro aj prilis neskoro oboje strakaju - a porazi tazenie do vycerpania o +239% pri rovnakom rozpocte.
The claim. "Set exit criteria and ignore the sunk cost" is the most repeated career and business advice there is — and it's useless, because it never tells you the threshold. When exactly do you cut a fading project, ad campaign, research line, content series, or sales channel and move on? We built the smallest model of a depleting effort and measured the answer.
The rule. Track the recent yield of the effort and its running peak. Quit when the recent yield has fallen a fraction θ below that peak — a drawdown stop, the same idea a trader uses to cut a losing position. The measured sweet spot is θ ≈ 0.6: once you've given back ~60% of your best, the vein is dead enough — cut and reallocate.
The counterintuitive part — it's an interior optimum. Quitting isn't "as early as possible" or "never." In the reference model (M depletable veins, fixed effort budget), mining each vein to depletion yields 757 findings; the θ=0.6 drawdown rule yields 2,569 — a +239% improvement on the same budget. And the curve has a peak: θ=0.4 → 2,010, θ=0.5 → 2,113, θ=0.6 → 2,569, θ=0.7 → 2,366, θ=0.8 → 2,053. Quit too early (small θ) and you abandon good veins on noise and pay setup costs over and over; quit too late (large θ) and you grind dead veins. Both tails lose. The optimum is in the middle, and it's closer to "cut" than most persistence advice admits.
Why it works. The marginal yield of any depleting effort falls as you exhaust it. The drawdown stop is a cheap, model-free detector of "this vein has passed its useful life" — it doesn't need you to know the vein's true richness in advance, only to watch your own recent output relative to your best. That's why a fixed θ generalizes across very different efforts.
The honest caveat. This decides when a declining effort has declined enough to cut — a drawdown stop on yield. It is not a forecast of whether a brand-new bet will pay off, and the window length and θ are levers you tune to your domain (defaults: 25 periods, θ=0.6 from the reference model). It turns "don't throw good money after bad" from a proverb into a rule you can run.
We packaged it as quitkit: one zero-dependency file (plus an MCP server, so an agent managing a portfolio of efforts can ask "have we hit the drawdown stop?"). Open-core and free — a sibling of our memory, RAG-freshness, statistics and self-reference tools. should_quit(recent_yields) gives you a verdict and the reason.
Tvrdenie. "Urci si exit kriteria a ignoruj utopene naklady" je najopakovanejsia kariera a biznis rada vobec — a je zbytocna, lebo ti nikdy nepovie prah. Kedy presne odrezes slabnuci projekt, reklamnu kampan, vyskumnu liniu, serial obsahu alebo predajny kanal a posunies sa dalej? Postavili sme najmensi model slabnuceho usilia a odmerali odpoved.
Pravidlo. Sleduj nedavny vynos usilia a jeho bezici vrchol. Vzdaj to, ked nedavny vynos klesne o zlomok theta pod tento vrchol — drawdown stop, presne tak ako trader reze stratovu poziciu. Odmerane sladke miesto je theta ≈ 0.6: ked si vratil ~60% zo svojho maxima, zila je dost mrtva — odrez a prealokuj.
Kontraintuitivna cast — je to vnutorne optimum. Vzdat to nie je "co najskor" ani "nikdy". V referencnom modeli (M vycerpatelnych zil, pevny rozpocet usilia) tazenie kazdej zily do vycerpania da 757 nalezov; pravidlo theta=0.6 da 2 569 — zlepsenie o +239% pri rovnakom rozpocte. A krivka ma vrchol: theta=0.4 → 2 010, theta=0.5 → 2 113, theta=0.6 → 2 569, theta=0.7 → 2 366, theta=0.8 → 2 053. Vzdaj to prilis skoro (male theta) a opustas dobre zily kvoli sumu a platis naklady na rozbeh znova a znova; vzdaj to prilis neskoro (velke theta) a melies mrtve zily. Oba konce strakaju. Optimum je v strede a je blizsie k "odrez" nez vacsina rad o vytrvalosti pripusta.
Preco to funguje. Marginalny vynos kazdeho slabnuceho usilia klesa, ako ho vycerpavas. Drawdown stop je lacny, bezmodelovy detektor "tato zila prekrocila svoju uzitocnu zivotnost" — nepotrebuje, aby si vopred poznal skutocnu bohatost zily, len aby si sledoval svoj vlastny nedavny vystup voci svojmu maximu. Preto fixne theta zovseobecnuje napriec velmi roznymi usiliami.
Poctiva vyhrada. Toto rozhoduje kedy slabnuce usilie kleslo dost na to, aby sa odrezalo — drawdown stop na vynos. Nie je to predpoved, ci sa uplne nova stavka vyplati, a dlzka okna a theta su paky, ktore ladis na svoju domenu (predvolene: 25 obdobi, theta=0.6 z referencneho modelu). Premiena "nehadz dobre peniaze za zlymi" z prislovia na pravidlo, ktore vies spustit.
Zabalili sme to ako quitkit: jeden subor bez zavislosti (plus MCP server, aby agent spravujuci portfolio usili mohol opytat "narazili sme na drawdown stop?"). Open-core a zadarmo — surodenec nasich nastrojov na pamat, RAG-cerstvost, statistiku a sebareferenciu. should_quit(recent_yields) ti da verdikt a dovod.