Argument

Universal Probability Bound

Intro

Atheists often answer the design argument by saying, "Given enough time and enough chances, anything can happen. The universe has been here for a long time and has a lot of matter. Improbable things become inevitable when the universe is this big."

The Universal Probability Bound is a careful, math-backed way of testing that reply. The question it asks is: how many chances has the universe actually had, ever, for anything at all to happen? If we knew that number, we could compare it to any specific improbability and see whether the universe really had enough time and matter to expect a given event to happen by accident.

The number turns out to be calculable. Multiply three things together. First, the total number of particles in the observable universe (about 10 to the 80th power, the so-called Eddington number). Second, the maximum number of physical changes any one particle can undergo per second (about 10 to the 43rd, set by the smallest physically meaningful time interval, the Planck time). Third, the total number of seconds since the Big Bang (about 10 to the 17th, that is, 13.8 billion years). Multiplied together, those give 10 to the 140th, which the mathematician William Dembski rounded up to 10 to the 150th to be generous.

That 10 to the 150th number is the entire probabilistic budget of the universe. Every chance any particle has had to do anything, ever. If a specific outcome is less probable than 1 in 10 to the 150th, the universe simply has not had enough time and matter to make it likely by accident even once.

Now apply this to the origin of life. Douglas Axe of Cambridge calculated the odds of getting one functional protein fold by random amino acid arrangement: about 1 in 10 to the 77th. That single protein, by itself, is still within the universal probability bound. But a minimum viable cell needs hundreds of these proteins, and they have to work together. Stack the odds, and the combined probability blows past 10 to the 150th by many orders of magnitude. The same kind of calculation, run on the origin of DNA's coded information, the integrated origin of replication-metabolism-membrane, and the Cambrian explosion of body plans, produces numbers that crush the bound.

The page lays out the math, the standard objections (the multiverse, hidden chemistry, "given enough alternate universes"), and the rebuttals. The take-home is that the "infinite time and chance" reply is not actually available in the real universe. The universe is large and old, but it is not large and old enough to make blind chance a reasonable explanation for the specific kinds of integrated information we find in living things.

◎ See this page's graph and discover linked topics →

In full

The Universal Probability Bound (UPB), formalized by William Dembski (The Design Inference, Cambridge 1998; No Free Lunch, Rowman & Littlefield 2002), is the upper limit of probabilistic resources available in the observable universe: approximately 10^150 events. It is derived from three physical maxima: the Eddington number (~10^80 particles), the Planck-time reciprocal (~10^43 maximum transitions per second per particle), and the age of the universe in seconds (~10^17). The product (~10^140) is rounded up to 10^150 to provide a generous upper bound. Any specific outcome whose probability falls below 1 in 10^150 exceeds the entire probabilistic budget of the universe and cannot plausibly have occurred even once by chance. Emile Borel anticipated this reasoning (Probabilities and Life, 1962) with what is sometimes called "Borel's law" at a lower threshold (~10^-50). Applied to the origin of life, the bound is decisive: Douglas Axe's calculation for the probability of a functional protein fold (~1 in 10^77, Journal of Molecular Biology 341, 2004) is within the bound for one protein but vastly exceeds it for the multiple integrated proteins of a minimal cell. Eugene Koonin's calculation for a self-sustaining replication-translation system (~1 in 10^1,018, Biology Direct 2007) exceeds the bound by ~868 orders of magnitude. Harold Morowitz's estimate for a minimal living cell (~1 in 10^340,000,000) exceeds the bound by hundreds of millions of orders of magnitude. The argument concludes that naturalistic chance explanations for the origin of life are excluded by the bound, and design is the better explanation. This page is structured as debate prep, each premise carries a second-order positive case, anticipated objections, rebuttals, a live-cite kit, and tactical notes.

Argument structure

Form

Deductive in structure with empirical/mathematical premises. P1 is a quantitative claim derived from accepted physics (the Eddington number, the Planck time, the age of the universe). P2 is the logical consequence of P1 combined with the standard probability-theoretic principle that an event with probability p in n trials is expected to occur np times; if np << 1, the event is not expected to occur. P3 is an empirical claim about specific biological probabilities, citing peer-reviewed work (Axe in the Journal of Molecular Biology; Koonin in Biology Direct; Morowitz's classic estimate). If P1, P2, and P3 hold, the conclusion follows with deductive necessity. Soundness depends on the contested empirical premise P3 (critics dispute the specific probability calculations) and on the response to multiverse-rescue moves that would inflate the probabilistic resources beyond the observable universe.

P1, The probabilistic resources of the observable universe total approximately 10^150 events

Affirmative case (second-order arguments)

1. The three factors are physically well-established. The Eddington number (~10^80 particles in the observable universe) follows from cosmological measurements of mass density and the size of the observable universe; it is mainstream physics. The Planck time (~5.39 × 10^-44 seconds) is the shortest physically meaningful time interval, set by the quantum-gravity scale; its reciprocal (~10^43 per second) is the upper bound on physical transitions per particle. The age of the universe (~13.8 billion years, or ~4.35 × 10^17 seconds) is mainstream cosmology, established by multiple independent methods (CMB, redshift, stellar nucleosynthesis). Multiplying gives ~10^140, generously rounded to 10^150. (Dembski, The Design Inference, 1998, ch. 6.)
2. The bound is generous in the design-inference direction. Dembski's 10^150 rounds up from 10^140 to give the chance hypothesis maximum room. Alternative thresholds in the literature are tighter: Borel's law (Probabilities and Life, 1962) used ~10^-50 as a practical cosmological-improbability threshold. Seth Lloyd's "computational capacity of the universe" calculation gives ~10^120 quantum operations. Each calculation lands in the same order-of-magnitude vicinity; Dembski's is the most generous to the chance hypothesis.
3. The bound is independent of the specific cosmological model. Even if the age of the universe were tripled or the particle count were larger by a few orders of magnitude, the bound would shift by a small number of orders of magnitude. The argument does not depend on precise calibration; it depends on the fact that the universe's probabilistic budget is finite and calculable, and that biological probabilities exceed it by hundreds or thousands of orders of magnitude.

Anticipated objections

1. "The observable universe is not the whole universe. There may be vast unobservable regions or other universes (the multiverse) that inflate the probabilistic budget."
2. "The Planck-time reciprocal overstates the maximum transitions per particle. Most particles do not undergo physically meaningful changes every Planck time."

Rebuttals

1. The multiverse is speculative and does not solve the information problem. The multiverse is a postulate (cosmological inflation, string-theory landscape, many-worlds quantum mechanics), not an empirical observation. Even granting an infinite multiverse for argument's sake, two problems remain: (a) the multiverse itself requires fine-tuning of its generating parameters (the inflaton field, the landscape selection mechanism, the wave-function-collapse rules), which pushes the design question up a level rather than solving it; (b) the multiverse inflates probabilistic resources but does not change what unguided chemistry can produce within each universe. Codes still require minds to write them; the multiverse does not change that. See the chance-explanations literature in Fine-Tuning Argument for the parallel treatment at the cosmological scale.
2. The Planck-time reciprocal is the upper bound, not the average rate. Dembski's calculation uses the maximum to be generous. The actual rate of physically meaningful transitions per particle is much lower, which makes the bound more generous to the chance hypothesis, not less. The objection cuts the wrong direction: tightening the bound makes the chance explanation harder, not easier.

Live-cite kit

• Scholarly: William Dembski (The Design Inference, Cambridge 1998, ch. 6; No Free Lunch, Rowman & Littlefield 2002); Emile Borel (Probabilities and Life, 1962); Seth Lloyd ("Computational capacity of the universe", Physical Review Letters 88, 2002); Arthur Eddington (the Eddington number)
• Aphorism: "10^150 is the entire probabilistic budget of the universe. There is no more room in the trial-counter."

Tactical notes

• Present the three factors concretely. Particles × Planck-times per second × seconds. Concrete numbers anchor the argument; abstract probability talk loses the audience.
• Be ready for the multiverse rescue. The philosophically sophisticated atheist response. Redirect to Fine-Tuning Argument for the broader multiverse-response toolkit.

P2, Outcomes below 1 in 10^150 cannot plausibly have occurred by chance

Affirmative case (second-order arguments)

1. The principle is standard probability theory. The expected number of occurrences of an event with probability p in n trials is np. If np << 1, the event is not expected to occur. The universal probability bound is the maximum n (the number of trials available in the universe). Outcomes with p < 1/10^150 have expected occurrence np < 1, even when n is the universal maximum. They are not expected to occur even once.
2. The principle is the standard rejection criterion in statistical hypothesis testing. Statistical significance tests use threshold probabilities (typically 0.05 or 0.01) to reject the null hypothesis of chance. The universal probability bound is the analogous threshold for cosmic-scale chance hypotheses. The principle of rejecting chance below a sufficiently low threshold is uncontroversial in statistics; the dispute is only about the threshold value for cosmic-scale events.
3. Borel's law anticipated the principle. Emile Borel, the mathematician who developed measure theory and probability theory, articulated essentially the same principle in Probabilities and Life (1962): probabilities below a cosmologically calibrated threshold are practically impossible. Borel's threshold (~10^-50) was tighter than Dembski's; Dembski's relaxation makes the bound more conservative. The underlying principle is mathematically standard, not Dembski's idiosyncrasy.

Anticipated objections

1. "You're assuming uniform distribution. In a structured universe with selection pressures, low-probability outcomes can become high-probability outcomes."
2. "The 'specific outcome' framing is post hoc. Any outcome has a low probability if specified in enough detail; that doesn't mean every outcome warrants the design inference."

Rebuttals

1. The argument considers structured chance, not just uniform chance. Dembski's framework explicitly handles chance-plus-necessity (selection biases). The No Free Lunch (2002) argument is that selection requires prior information (a fitness function) to perform better than blind search. If the fitness function itself is the explanandum (which is the case for the origin of life and the origin of the genetic code), selection does not solve the probability problem. The bound applies to the deeper question of where the structured selection landscape came from, not just to pure uniform chance. See Specified Complexity Argument for the formal treatment.
2. The argument requires independent specification, not post hoc framing. Dembski's framework requires the specification to be describable independently of the specific event. A functional protein is specified by the chemistry it catalyzes (independent of any particular sequence). A snowflake is improbable but not specified (no independent target). The objection conflates post hoc specification with independent specification; the bound applies only to the latter. See Specified Complexity Argument for the formal treatment.

Live-cite kit

• Scholarly: Emile Borel (Probabilities and Life, 1962); William Dembski (The Design Inference, 1998, ch. 6); Ronald Fisher (the standard treatment of statistical significance testing)
• Aphorism: "If the trial-count maxes out at 10^150 and the probability is 1 in 10^77 per try, you are out of trials before you stack a single cell."

Tactical notes

• Lead with the statistical-significance parallel. Most opponents are familiar with rejecting null hypotheses at the 0.05 or 0.01 threshold. The universal probability bound is the same logic at the cosmic scale.
• Insist on independent specification. When opponents raise the snowflake counter-example, redirect: "specification is what distinguishes a snowflake from a royal flush."

P3, Biological probabilities far exceed the universal bound

Affirmative case (second-order arguments)

1. Douglas Axe's protein-fold calculation (Journal of Molecular Biology, 2004). Axe's experimental work on enzyme-folding sequence space gives the ratio of functional folds to total possible sequences for 150-amino-acid proteins as approximately 1 in 10^77. A single functional protein already approaches the universal bound. The peer-reviewed venue (Journal of Molecular Biology 341) makes the result mainstream-credentialed.
2. Eugene Koonin's self-replicating system calculation (Biology Direct, 2007). Koonin, a non-theist NCBI evolutionary genomicist, calculated the probability of a self-sustaining replication-translation system emerging by chance at approximately 1 in 10^1,018. This exceeds the universal bound by ~868 orders of magnitude. Koonin's response was to invoke the multiverse, which is an admission against interest: he conceded the within-universe chance hypothesis fails.
3. Harold Morowitz's minimal-cell calculation. Morowitz, a Yale biophysicist, estimated the probability of a minimal living cell assembling by chance at approximately 1 in 10^340,000,000. This exceeds the universal bound by hundreds of millions of orders of magnitude. The calculation considers the joint probability of hundreds of specified proteins, the lipid membrane, and the integrated genetic code.
4. Fred Hoyle's "tornado in a junkyard" intuition. Hoyle (Evolution from Space, 1981) gave the famous accessible analogy: the probability of a functional cell assembling by chance is comparable to the probability of a Boeing 747 assembling from a junkyard tornado. Hoyle's calculation was attacked as a strawman (defenders argue selection is involved), but the underlying intuition is captured rigorously by the Axe-Koonin-Morowitz numbers.
5. The composite estimates blow past the bound by many orders of magnitude. Any of the three estimates (Axe per protein, Koonin per replicator-translator, Morowitz per cell) is sufficient to refute the chance hypothesis at the universal bound. The composite case is overwhelming: the universe's probabilistic budget is exhausted by a single functional protein under the strictest specification; the integrated cell is many orders of magnitude beyond reach.

Anticipated objections

1. "Axe's number is wrong. Newer protein-engineering work shows functional folds are more common than 10^-77."
2. "Morowitz's number assumes simultaneous chance assembly, which is not what naturalistic OOL theories propose. RNA-world scenarios propose stepwise assembly with selection at each stage."
3. "Koonin's invocation of the multiverse is a valid scientific response, not an admission against interest."

Rebuttals

1. Axe's order-of-magnitude conclusion stands. Critics like Hugh Hunt and others have proposed higher functional densities than Axe's 10^-77, but the order-of-magnitude conclusion does not change: functional folds are a vanishingly small fraction of sequence space. Even at the most generous critic estimates (1 in 10^11 or 10^12 for some restricted protein families), the joint probability for an integrated cell remains vastly beyond the universal bound. The challenge is not at the single-protein level; it is at the integrated-system level. See Protein Sequence Space Argument. Failure mode: point-attacking a single estimate without addressing the cumulative scale.
2. Stepwise assembly with selection requires functional intermediates and prior information. The RNA-world scenario hopes to dissolve the simultaneous-assembly problem by proposing stepwise selection-driven assembly. The problem: (a) functional RNA intermediates are themselves vanishingly rare in RNA sequence space; (b) selection requires a target, which requires the very specified information the argument is asking about. See RNA World Failure Argument for the detailed reply. Failure mode: moving the bottleneck without eliminating it.
3. The multiverse rescue does not solve the within-universe information problem. Even granting an infinite multiverse, the structure of biological information (specified by independent function) is not made more probable by multiplying universes; an unguided process in any universe faces the same probabilistic constraint within that universe. The multiverse multiplies probabilistic resources but does not change what unguided chemistry can produce within each universe. See Specified Complexity Argument and Fine-Tuning Argument for the broader multiverse-response toolkit.

Live-cite kit

• Scholarly: Douglas Axe (Journal of Molecular Biology 341, 2004; Undeniable, HarperOne 2016); Eugene Koonin (Biology Direct 2007; The Logic of Chance, 2011); Harold Morowitz (Energy Flow in Biology, 1968, and subsequent works); Fred Hoyle (Evolution from Space, 1981); Stephen Meyer (Signature in the Cell, 2009)
• Aphorism: "One protein at 10^-77 approaches the bound. A cell at 10^-340,000,000 incinerates it."

Tactical notes

• Lead with Axe's number. Peer-reviewed, mainstream venue (Journal of Molecular Biology), single-protein scope. Concrete and defensible.
• Use Koonin's admission against interest. A non-theist evolutionary genomicist computed 10^-1,018 and reached for the multiverse rescue. The implicit concession is decisive.
• Don't get lost in the Morowitz number. It is rhetorically devastating but invites the "simultaneous-assembly" objection. Use it for force; back it up with Axe's number for technical defense.

Conclusion

Naturalistic chance explanations for the origin of life are excluded by the universal probability bound; design is the better explanation. The bound (10^150) is the entire probabilistic budget of the observable universe, derived from physical maxima (particles, Planck time, age). Outcomes more improbable than this exceed the budget and cannot have occurred by chance even once. Biological probabilities (Axe's 10^-77 per protein; Koonin's 10^-1,018 per replicator-translator; Morowitz's 10^-340,000,000 per minimal cell) exceed the bound by hundreds to hundreds of millions of orders of magnitude. The chance hypothesis is dead at the universal-bound threshold; the design inference is what remains.

Probability calculations summary

Master objections to the argument as a whole

1. "The multiverse rescue solves the bound by multiplying probabilistic resources." Reply: the multiverse is speculative; multiplying universes does not change what unguided chemistry can produce within each universe. Codes still require minds to write them. See Fine-Tuning Argument.
2. "Natural selection biases the probability distribution, so chance estimates are too pessimistic." Reply: selection requires a target (a fitness function) which requires specified information; selection does not solve the design problem, it presupposes it. See Specified Complexity Argument.
3. "The 'specific outcome' framing is post hoc." Reply: independent specification through biological function (catalysis, replication, membrane integrity) is grounded in physics and chemistry, not assigned post hoc. See Specified Complexity Argument.
4. "Future research will discover natural mechanisms that bypass the bound." Reply: this is promissory naturalism. Seventy years of origin-of-life research has increased the probability gap, not closed it. See Argument from Origin of Life.
5. "Even granting design, you have not shown the designer is the Christian God." Reply: granted; this is part of a cumulative case. See Christian God is the Only True God.

Tactical opening / closing

Opening line: "Atheists like to say 'given enough time and chance, anything can happen.' Let us calculate exactly how much time and chance the universe has actually had. Particles times Planck-time interactions per second times seconds since the Big Bang. The answer is 10^150 events. That is the entire probabilistic budget of the universe. Now let us compare that budget to the probability of a single functional protein."

Closing landing strip: "The Universal Probability Bound is not an opinion. It is arithmetic. The probabilistic resources of the universe are finite and calculable. The probability of the simplest functional protein, calculated by mainstream peer-reviewed methods, exhausts that budget. The probability of a minimal cell crushes the budget by hundreds of millions of orders of magnitude. The 'given enough time and chance' answer is mathematically dead. What replaces it is the question of what cause-type produces specified information at this scale. The only known answer is mind."

Connection to Scripture

• Psalm 139:13-16, "fearfully and wonderfully made"; the precision of biological design
• Romans 1:20, "invisible attributes... clearly seen, being understood through what has been made"
• Job 38, the divine speech on the wisdom and ordering of creation; the math of creation as God's signature
• Psalm 19:1, "the heavens declare the glory of God"
• Colossians 1:16-17, "in Him all things hold together"; the integration that the bound shows cannot come from chance
• John 1:1, the Logos as the rational ground of reality; the bound is mathematically calculable because the universe is rationally ordered

Patristic / scholarly note

Classical / patristic:

• Augustine (City of God XII), divine wisdom in the ordering of creation; the rational structure that makes the bound calculable
• Thomas Aquinas (Summa Theologiae I.2.3, the Fifth Way), the teleological argument from order

Pre-Dembski mathematical anticipations:

• Emile Borel (Probabilities and Life, 1962), the precursor "Borel's law" with a tighter threshold (~10^-50)
• Fred Hoyle (Evolution from Space, 1981), the "tornado in a junkyard" intuition

Contemporary intelligent-design movement:

• William Dembski (The Design Inference, Cambridge 1998, ch. 6; No Free Lunch, Rowman & Littlefield 2002), the formal derivation of the bound
• Robert Marks and William Dembski (Introduction to Evolutionary Informatics, World Scientific 2017), the active-information extension
• Stephen Meyer (Signature in the Cell, HarperOne 2009; Return of the God Hypothesis, 2021), the broader OOL information case
• Douglas Axe (Undeniable, HarperOne 2016; Journal of Molecular Biology 341, 2004), the protein-folding combinatorics

Mainstream-science engagements:

• Seth Lloyd ("Computational capacity of the universe", Physical Review Letters 88, 2002), the parallel ~10^120 calculation
• Eugene Koonin (Biology Direct 2007; The Logic of Chance, 2011), the self-replicator probability calculation
• Harold Morowitz (Yale biophysicist), the minimal-cell probability calculation

See also

• Specified Complexity Argument, the formal framework that uses the bound
• Argument from Origin of Life, the master case the bound underwrites
• Protein Sequence Space Argument, Axe's per-protein calculation
• RNA World Failure Argument, the OOL scenario the Koonin number engages
• Irreducible Complexity Argument, the molecular-machine sister
• Signature in the Cell Argument, the DNA-as-information sister
• Molecular Machines Argument, the engineering-analogue sister
• Argument from the Genetic Code, the genetic-code sister
• Fine-Tuning Argument, the cosmological-scale parallel (different probabilities, same inference)
• Intelligent Design, parent movement
• Origins, category master hub
• Christian God is the Only True God, cumulative-case home
• Methodological Naturalism Critique, the gatekeeping move the argument confronts
• Anthropic Principle, the observation-selection counter to the bound
• Stephen Meyer, adjacent scholarly work
• Arguments, top-level master index