# Is evolution mathematically impossible? ![[images/Evolution-Mathematically-Impossible.webp]] **Creator:** Joel is Curious · **Published:** 2026-11-01 · **Length:** 2:59 (YouTube Short) · [Watch on YouTube](https://www.youtube.com/watch?v=zPMU24P62EY) > Caption source: **YouTube auto-generated English captions (en-orig)**, fetched with yt-dlp (android_vr client). No TubeAlfred credits were used. Timestamps are from the captions. The short alternates a vertical talking-head take outdoors (a gray "dinosaurs:" hoodie) with black panels of a dot simulation labeled "simulation – made-up odds" ("generation 128", "generation 1,326"), a "C → T, one letter of… 6,000 years ago" lactase panel, and a "220,000 years ago… billions of changes… at least 100,000 new changes, about 2 kept (each line)" panel marked "illustration, numbers from the papers." Burned-in captions throughout. Screenshots are one about every 20 s. ## Summary "Evolution is mathematically impossible, right? It's a reasonable question." A comment he gets a lot: "show me one example of a cat giving birth to a dog or a monkey changing into a human." The odds of one species turning into another in one generation are astronomical; humans and chimps are cousins, but our DNA differs by about 35 million letters. "So, they're right. A dog never turns into a cat and a monkey never gives birth to a human. But, no scientist claims that evolution works that way." Populations change a little at a time. Each individual is born with mutations their parents didn't have; "you have about 60." Multiply by everyone alive and one generation brings many times more mutations to our collective DNA than all the letters in one copy of yours. As a picture: every dot is a creature, and every generation each one rolls the dice. Most changes do nothing; some hurt; the picture shows only the ones that help. "This one helped and it was lost anyway. That happens to most helpful changes. This one didn't. It had more kids and within a few hundred generations, it's nearly everywhere." Harmful changes sometimes stick around too. Then another helpful change lands in a creature that already carries the first, and it spreads; the next change starts from there. "It's not one dice roll." In a real population it's hundreds of billions of new mutations every generation. A real case: one DNA letter that keeps the milk-digesting gene switched on in adults was rare in Europe 6,000 years ago but is common in northern Europeans and their descendants today. Seven million years ago there were at least 50,000 or so of our ancestors; 60 new changes each, generation after generation for 7 million years, is "hundreds of billions of changes, maybe a trillion." Almost all vanished, and most that stayed do little or nothing. "But, we only need 35 million to account for the changes between us and chimps in the last 7 million years." "Asking the odds of that is the wrong question." What were the odds last night's game would go exactly as it did, every shot and every life that led to it? What were the odds you'd be exactly here watching this? "The odds argument looks at where we landed and asks how likely it was. That assumes that there was a target that evolution was shooting for. There wasn't." Small changes every generation across a huge population, and the ones that help stick around: "That's evolution." ## Outline ### 0:00 The claim - [0:00](https://www.youtube.com/watch?v=zPMU24P62EY&t=0s) "Evolution is mathematically impossible, right? It's a reasonable question." - [0:04](https://www.youtube.com/watch?v=zPMU24P62EY&t=4s) "A comment I get a lot is show me one example of a cat giving birth to a dog or a monkey changing into a human." - [0:09](https://www.youtube.com/watch?v=zPMU24P62EY&t=9s) "The odds of one species turning into another in one generation are astronomical." - [0:13](https://www.youtube.com/watch?v=zPMU24P62EY&t=13s) "For example, humans and chimps are cousins, but our DNA differs by about 35 million letters." - [0:23](https://www.youtube.com/watch?v=zPMU24P62EY&t=23s) "But, no scientist claims that evolution works that way." ![[images/Evolution-Mathematically-Impossible-t0003.webp]] ### 0:25 A population, a little at a time - [0:25](https://www.youtube.com/watch?v=zPMU24P62EY&t=25s) "Populations change a little bit at a time, generation after generation." - [0:31](https://www.youtube.com/watch?v=zPMU24P62EY&t=31s) "Each individual in that population is born with mutations that their parents didn't have. You have about 60." - [0:37](https://www.youtube.com/watch?v=zPMU24P62EY&t=37s) "Multiply that by everyone alive and one generation brings many times more mutations to our collective DNA than all the letters in one single copy of yours." ![[images/Evolution-Mathematically-Impossible-t0023.webp]] ### 0:47 The dot picture - [0:47](https://www.youtube.com/watch?v=zPMU24P62EY&t=47s) "Here's that as a picture. Every dot is a creature and every generation, every one of them rolls the dice." - [0:53](https://www.youtube.com/watch?v=zPMU24P62EY&t=53s) "Most of those changes do nothing. Some of them hurt. This picture only shows the ones that actually help." - [0:59](https://www.youtube.com/watch?v=zPMU24P62EY&t=59s) "This one helped and it was lost anyway. That happens to most helpful changes." - [1:03](https://www.youtube.com/watch?v=zPMU24P62EY&t=63s) "This one didn't. It had more kids and within a few hundred generations, it's nearly everywhere." - [1:11](https://www.youtube.com/watch?v=zPMU24P62EY&t=71s) "Then, another helpful change lands in a creature that already carries the first one and it spreads, too." - [1:18](https://www.youtube.com/watch?v=zPMU24P62EY&t=78s) "It's not one dice roll." - [1:19](https://www.youtube.com/watch?v=zPMU24P62EY&t=79s) "In a real population, it's hundreds of billions of new mutations every generation." ![[images/Evolution-Mathematically-Impossible-t0043.webp]] ![[images/Evolution-Mathematically-Impossible-t0063.webp]] ### 1:29 Milk, and 7 million years - [1:29](https://www.youtube.com/watch?v=zPMU24P62EY&t=89s) "One letter of DNA that keeps the milk-digesting gene switched on in adults was rare in Europe 6,000 years ago, but it's common in northern Europeans and their descendants today." - [1:39](https://www.youtube.com/watch?v=zPMU24P62EY&t=99s) "7 million years ago, there were at least 50,000 or so of our ancestors." - [1:43](https://www.youtube.com/watch?v=zPMU24P62EY&t=103s) "60 new changes each, generation after generation for 7 million years. That's hundreds of billions of changes, maybe a trillion." - [1:51](https://www.youtube.com/watch?v=zPMU24P62EY&t=111s) "Almost all of those mutations have vanished." - [1:58](https://www.youtube.com/watch?v=zPMU24P62EY&t=118s) "But, we only need 35 million to account for the changes between us and chimps in the last 7 million years." ![[images/Evolution-Mathematically-Impossible-t0083.webp]] ![[images/Evolution-Mathematically-Impossible-t0103.webp]] ### 2:09 The wrong question - [2:09](https://www.youtube.com/watch?v=zPMU24P62EY&t=129s) "Asking the odds of that is the wrong question." - [2:12](https://www.youtube.com/watch?v=zPMU24P62EY&t=132s) "What were the odds that last night's game would go exactly as it did?" - [2:24](https://www.youtube.com/watch?v=zPMU24P62EY&t=144s) "Astronomically low odds, and yet it happened, exactly as it did." - [2:29](https://www.youtube.com/watch?v=zPMU24P62EY&t=149s) "And you, everything that's happened in your life and in the lives of your ancestors, what were the odds you'd be exactly here, right now, watching this video?" - [2:37](https://www.youtube.com/watch?v=zPMU24P62EY&t=157s) "Unquantifiable, and yet here you are." ![[images/Evolution-Mathematically-Impossible-t0123.webp]] ![[images/Evolution-Mathematically-Impossible-t0143.webp]] ### 2:40 No target - [2:40](https://www.youtube.com/watch?v=zPMU24P62EY&t=160s) "The odds argument looks at where we landed and asks how likely it was." - [2:44](https://www.youtube.com/watch?v=zPMU24P62EY&t=164s) "That assumes that there was a target that evolution was shooting for." - [2:47](https://www.youtube.com/watch?v=zPMU24P62EY&t=167s) "There wasn't. It just happened." - [2:49](https://www.youtube.com/watch?v=zPMU24P62EY&t=169s) "Small changes every generation across a huge population, and the ones that help stick around." - [2:55](https://www.youtube.com/watch?v=zPMU24P62EY&t=175s) "That's evolution." ![[images/Evolution-Mathematically-Impossible-t0163.webp]] ## Anti-evolution argument *Stated fairly, from Henry M. Morris (Institute for Creation Research) and the comments the video answers.* - **The commenters' version:** "show me one example of a cat giving birth to a dog or a monkey changing into a human"; one species becoming another is too improbable to happen. - **Morris's version:** in "The Mathematical Impossibility of Evolution" (first in ICR's *Acts & Facts* Issue No. 2, 1972; reprinted 2003), Henry Morris imagines "a very simple putative organism composed of only 200 integrated and functioning parts," built one successful mutation at a time. Granting each step a one-in-two chance, "the probability of 200 successive mutations being successful is then (½)^200, or one chance out of 10^60" [R1]. - **Not enough tries:** allowing a billion mutating systems on every square foot of Earth for 10^18 seconds, each restarting after failure, gives 10^39 attempts, so success "just once, anywhere in the world, in all the assumed expanse of geologic time, is less than one chance out of a billion trillion" [R1]. - **No beneficial mutations:** Morris adds that "no one has ever actually observed a genuine mutation occurring in the natural environment which was beneficial (that is, adding useful genetic information to an existing genetic code)" [R1], and concludes evolution is "mathematically and logically indefensible." ## Answer *Point by point. Numbers refer to the References table.* ### 1. "A cat giving birth to a dog" - Agreed, it never happens, and evolution doesn't predict it. Change is spread over many generations and individuals. ### 2. The numbers - About 35 million single-letter differences between human and chimp: "approximately thirty-five million single-nucleotide changes" [R2]. Accurate. As the description notes, these built up on both lines since the split. - About 60 new mutations per person: Kong et al. measured 1.20 × 10⁻⁸ per letter per generation, rising with the father's age [R3]. Across a diploid genome of about 6 billion letters that's roughly 70; "about 60" is the commonly quoted figure (the exact 60 isn't in the abstract). - More new mutations per generation across everyone alive than letters in one genome: 8 billion people × ~60 is about 480 billion, roughly 150 times a 3.1-billion-letter genome. The description cites Shendure & Akey 2015 [R4]; its abstract doesn't give the figure. - At least 50,000 ancestors: Burgess & Yang found ancestral hominoid populations "5-10 times larger than modern humans" [R7]; the description's 100,000 isn't in the abstract. - 7 million years: Langergraber et al. date the split to "at least 7-8 million years" [R8]; see also [R9, R10] on the mutation rate behind such dates. - "Hundreds of billions, maybe a trillion": 50,000 × 60 × about 280,000 generations (25-year generations over 7 million years) is about 840 billion. The arithmetic holds. ### 3. Selection isn't one dice roll - Morris's model has each system "start over for a new try" after any failure [R1]: it throws away every successful step, which is the opposite of selection keeping helpful changes and building on them. The video's dot picture shows the alternative: changes kept and stacked. - The chimp genome paper finds both lineages' evolution "dominated by the fixation of neutral and slightly deleterious alleles" [R2], matching "most of those that stuck around don't do much." Under neutral theory, the long-run substitution rate equals the mutation rate [R11], which is why mutation counts and genome differences can be compared at all. (Simplified in the video: the 35 million are fixed differences, not raw mutations.) ### 4. Beneficial mutations are observed - Lactase persistence: Bersaglieri et al. found "genetic signatures of strong recent positive selection at the lactase gene," a common haplotype (~77%) in northern European-derived populations [R5]. Evershed et al. call it "the most strongly selected monogenic trait to have evolved over the past 10,000 years" [R6]. This answers Morris's "no one has ever actually observed" a beneficial mutation [R1]. The description's 4,700 BCE earliest carrier and rarity until about 2,000 BCE aren't in the abstract and weren't checked. ### 5. Odds after the fact - The video's point is that calculating the probability of one specific outcome after it has happened (a game, a life) says nothing about whether the process works. Morris's calculation targets a specific 200-part organism [R1]; the video argues there was no target. ## Key takeaways - Nobody claims one species turns into another in one generation. - Each person carries about 60 new mutations; across a population that's an enormous supply every generation. - Selection keeps helpful changes, so steps accumulate; probability models that restart from zero don't describe evolution. - Beneficial mutations are observed, e.g. lactase persistence. - The human-chimp difference of about 35 million letters is small next to the hundreds of billions of mutations available over 7 million years. ## Things mentioned ### Things and methods - De novo mutations; mutation rate; effective population size; natural selection and genetic drift (helpful changes lost by chance); fixation; selective sweeps; lactase persistence (a single C → T letter change, per the video); molecular clock ### Places - Europe and northern Europe (lactase persistence) ### Ideas - Probability arguments against evolution; post hoc probability (the "last night's game" analogy); no target in evolution ### Works (from the description) - Chimpanzee Sequencing and Analysis Consortium 2005; Kong et al. 2012; Shendure & Akey 2015; Bersaglieri et al. 2004; Evershed et al. 2022; Langergraber et al. 2012; Moorjani et al. 2016; Scally & Durbin 2012; Burgess & Yang 2008 ### Links from the description - Papers: https://doi.org/10.1038/nature04072 · https://doi.org/10.1038/nature11396 · https://doi.org/10.1126/science.aaa9119 · https://doi.org/10.1086/421051 · https://doi.org/10.1038/s41586-022-05010-7 · https://doi.org/10.1073/pnas.1211740109 · https://doi.org/10.1073/pnas.1600374113 · https://doi.org/10.1038/nrg3295 · https://doi.org/10.1093/molbev/msn148 - https://joeliscurious.com · https://www.instagram.com/joeliscurious/ · https://www.tiktok.com/@joeliscurious · https://x.com/JoelisCurious · https://www.youtube.com/@joeliscurious ## Claims check | Claim in the video | Verdict | Source or note | |---|---|---| | Human and chimp DNA differ by about 35 million letters | **Accurate** | [R2] | | Each person has about 60 new mutations | **Accurate** | [R3] (rate; ~60–70 implied) | | One generation of everyone alive adds more mutations than letters in one genome | **Accurate** | Arithmetic from [R3]; description cites [R4] | | Most helpful changes are lost by chance | **Accurate** | Standard population genetics (drift); not checked against a cited source | | A lactase-persistence letter was rare in Europe 6,000 years ago, common in northern Europeans today | **Accurate** | [R5, R6]; dates not in abstracts | | At least 50,000 ancestors 7 million years ago | **Consistent** | [R7]: 5–10× modern human size | | Human-chimp split about 7 million years ago | **Accurate** | "at least 7-8 million years" [R8] | | Hundreds of billions of mutations, maybe a trillion, over 7 million years | **Accurate arithmetic** | ~840 billion on stated assumptions | | Most mutations that stuck do little or nothing | **Accurate** | [R2] | | "We only need 35 million" | **Simplified** | Fixed differences, on both lineages, not raw mutations [R2, R11] | | The dot picture uses made-up odds | **Disclosed** | On screen and in the description | ## References | # | Source | Exists? | Use in the video / role | Link | |---|---|---|---|---| | R1 | Henry M. Morris, "The Mathematical Impossibility Of Evolution," ICR *Acts & Facts*, 2003-11-01 (reprint of a 1972 article) | **Exists** (read) | Added: the creationist case; quotes accurate | [icr.org](https://www.icr.org/article/493/) | | R2 | Chimpanzee Sequencing and Analysis Consortium, "Initial sequence of the chimpanzee genome and comparison with the human genome," *Nature* 437:69–87 (2005) | **Exists** (abstract read) | Description source: 35 million letters | [doi:10.1038/nature04072](https://doi.org/10.1038/nature04072) | | R3 | Kong et al., "Rate of de novo mutations and the importance of father's age to disease risk," *Nature* 488:471–475 (2012) | **Exists** (abstract read) | Description source: about 60 per person | [doi:10.1038/nature11396](https://doi.org/10.1038/nature11396) | | R4 | Shendure & Akey, "The origins, determinants, and consequences of human mutations," *Science* 349:1478–1483 (2015) | **Exists** (abstract read) | Description source: mutations across everyone alive; figure not in abstract | [doi:10.1126/science.aaa9119](https://doi.org/10.1126/science.aaa9119) | | R5 | Bersaglieri et al., "Genetic signatures of strong recent positive selection at the lactase gene," *Am J Hum Genet* 74:1111–1120 (2004) | **Exists** (abstract read) | Description source: lactase selection | [doi:10.1086/421051](https://doi.org/10.1086/421051) | | R6 | Evershed et al., "Dairying, diseases and the evolution of lactase persistence in Europe," *Nature* 608:336–345 (2022) | **Exists** (abstract read) | Description source: the lactase letter in Europe; dates not checked | [doi:10.1038/s41586-022-05010-7](https://doi.org/10.1038/s41586-022-05010-7) | | R7 | Burgess & Yang, "Estimation of hominoid ancestral population sizes under bayesian coalescent models…," *Mol Biol Evol* 25:1979–1994 (2008) | **Exists** (abstract read) | Description source: ancestral population size; 100,000 not in abstract | [doi:10.1093/molbev/msn148](https://doi.org/10.1093/molbev/msn148) | | R8 | Langergraber et al., "Generation times in wild chimpanzees and gorillas suggest earlier divergence times in great ape and human evolution," *PNAS* 109:15716–15721 (2012) | **Exists** (abstract read) | Description source: 7 million years | [doi:10.1073/pnas.1211740109](https://doi.org/10.1073/pnas.1211740109) | | R9 | Moorjani et al., "Variation in the molecular clock of primates," *PNAS* 113:10607–10612 (2016) | **Exists** (metadata read) | Description source: mutation rate and split dates | [doi:10.1073/pnas.1600374113](https://doi.org/10.1073/pnas.1600374113) | | R10 | Scally & Durbin, "Revising the human mutation rate: implications for understanding human evolution," *Nat Rev Genet* 13:745–753 (2012) | **Exists** (metadata read) | Description source: mutation rate | [doi:10.1038/nrg3295](https://doi.org/10.1038/nrg3295) | | R11 | Motoo Kimura, "Evolutionary Rate at the Molecular Level," *Nature* 217:624–626 (1968) | **Exists** (metadata read) | Added: neutral substitution rate | [doi:10.1038/217624a0](https://doi.org/10.1038/217624a0) | ### Not found or not checked - The exact "about 60" per person and the "more than letters in one genome" figure aren't stated in the Kong and Shendure & Akey abstracts; they're consistent with the measured rate. - Evershed et al.'s 4,700 BCE earliest carrier and "rare until about 2,000 BCE" (description) weren't checked (abstract only). - Burgess & Yang's "about 100,000" (description) isn't in the abstract. - The commenters the video answers aren't identified. ## Added context - **Series:** a standalone response short. Previous upload: [[Mount St. Helens dated at 350,000 years (The Age of the Earth, Part 6)]]. - **Related:** [[Microevolution is macroevolution (Is evolution true Part 7)]] (the dog and non-dog argument); [[Is evolution still happening (Is evolution true Part 5)]]; [[One drought. Bigger beaks. (How evolution works, Part 1)]] (selection in action). - **On the description's plea:** "Please be kind; the argument is the point, not the people." ## Biographies ### Joel is Curious (the creator) YouTube channel **Joel is Curious**, all Shorts, from 2026-10-04. In his own words (video description): "I'm Joel, a software developer and former tech executive who likes to learn stuff. Science, history, Buddhism, Stoicism, and the thinkers behind them." He lists his sources in each description. (Sources: the video and its description.) ### Henry M. Morris (added context) Creationist author at the Institute for Creation Research (ICR). R1's 2003 introduction says *Acts & Facts* began in June 1972 and that "Issue No. 2" carried "The Mathematical Impossibility of Evolution," reproduced in 2003.