Dr. Zach Hancock, a population geneticist who allegedly teaches the subject as an Assistant Professor at Augusta University, left a comment defending himself after Will Duffy repeatedly called him out for his lackluster and uninformed response to Probability Zero in Will’s recent appearance on the Gutsick Gibbon channel. At least, I think that was supposed to be the purpose, although it appears to be more of a public exhibition of self-destruction than self-defense.
My thoughts: 1. At 33:42, Will quotes Vox Day being upset no one wants to read his book. It’s simple. If we (scientists) stopped to read every crackpot book that claims to overturn our field, we’d never get any science done. There are real scientists writing groundbreaking work that I sadly haven’t gotten around to reading yet. The reason we don’t want to read his book is because he’s irrelevant. Furthermore, Will claims to present Vox’s argument. Thus, any misunderstanding on mine or Erika’s part is a fault on Will, not our willingness to read the book. If a student of mine doesn’t understand natural selection after I present a lecture on it, I don’t say, “Well, it’s because you didn’t read Darwin’s Origin!” I own that I did a bad job teaching the argument.
That is a blatant lie. Will did not quote me about being upset. He did not quote me as being upset because I never said that I was. I am not even remotely upset that my credentialed critics haven’t read Population Zero. To the contrary, as Will literally quoted me, I predicted on September 23, 2026 that they would not read the book.
“I do not expect anyone to actually (a) talk to me or (b) read the book. I don’t know what it is about biologists, population geneticists, and other science enthusiasts, but they always seem to imagine they can effectively critique anything without bothering to actually learn anything about it.”
This does not upset me. To the contrary, it leads me to believe they are tremendously stupid for saying one single word about a book they haven’t read.
Back in the day, I was a game reviewer nationally syndicated by Universal Press Syndicate and I also reviewed games for Computer Gaming World. I would occasionally go without food or sleep in order to play through an entire game in order to meet a CGW deadline, because neither Johnny Wilson nor Chris Lombardi permitted a reviewer to write a review for a game without completing it first. The real challenge was to do so without using the provided cheat codes. I wrote the cover review for id Software’s Heretic after playing it through in 18 hours straight and it took a week before I could close my eyes without strafing. So I have nothing but deep contempt for these fake reviewers who observably and self-admittedly have no idea what they’re talking about.
There is nothing wrong with being too busy to engage with new ideas. The problem is that Hancock is trying to simultaneously engage and not-engage with these new ideas, which is both nonsensical and pointless. And if he’s so busy, then why is he wasting all these hours on what amounts to nothing more than a pseudo-scientific Maoist struggle session? I suspect the actual reason is that he suspects the mathematical structures of his field are shaky and he’s afraid to look too closely at precisely how precarious its foundations are.
Will is not responsible for teaching anyone anything. Dr. Zachary Hancock, PhD, Assistant Professor in the Department of Biological Sciences at Augusta University, is not his student. If Dr. Hancock isn’t going to read the book, that’s absolutely fine, but then he should sit down and shut the fuck up because all he’s going to accomplish is to demonstrate that he’s not only ignorant, but stupid and unprofessional as well.
At 34:38, Will states that Day’s d is “still necessary.” If true, then he shouldn’t have set it equal to 1 when he presented it. Furthermore, the reason that “d” is not a variable in population genetics is because it’s nonsensical. How fast the entire gene pool turns over is not meaningful, because alleles are not turning over and appearing all at once, but are staggered through time. Will flashes the “d formula” on the screen, claims there are variables not accounted for, but doesn’t define any of the variables in the equation to support this claim. Yet again, this is a “read Vox’s book” moment. No. You are making the argument. Explain it. If you can’t, don’t make it.
The necessity of a variable is not determined by what it happens to be set at when it is presented. Hancock also dismisses the selective turnover coefficient by asserting that “how fast the entire gene pool turns over is not meaningful, because alleles are not turning over and appearing all at once, but are staggered through time.” This is a non sequitur. Rates are by definition averages of staggered events. The SNP mutation rate is an average of mutations that don’t all happen at once. The fixation rate is an average of fixations that don’t all happen at once. Dismissing a rate because the underlying events are staggered would invalidate every single rate in population genetics. Hancock either needs to show that d is internally incoherent or that the generational overlap it measures is adequately captured by existing parameters. He does neither.
His complaint about Will not defining d for him is ludicrous because there is both an entire chapter and an appendix that not only defines the selective turnover coefficient, but shows its derivation, provides the empirical evidence for it, as well as explains why it is necessary and how it is different from the various things that Hancock incorrectly believed do the same thing.
Ironically, Hancock himself illustrates the need for d later in his comments. This simply underlines how stupid he is for attempting to criticize a book that he hasn’t read. Sit down and shut the fuck up, Zachary. You are clearly and observably too short for this ride.
At 37:00, Will makes the claim, apparently first pointed out by his popgen teacher, Griffiths, that the fixation equation I showed is derived explicitly for single nucleotide polymorphisms (SNPs), not structural variation. I have no idea where either of them came up with that. At no point does the fixation equations I presented distinguish between mutational types, whether explicitly or implicitly. The mutation rate I initially use is for SNPs, but the adjustment to 76.8 for a haploid genome, following the correction, modifies that rate to include structural variants as well. That is, u = (3e9)*(76.8) = 2.56e-8, not the 1.2e-8 I used for the SNP calculation.
Amusingly, Hancock makes a major error in the process of claiming his critics don’t understand the math. He writes: “The mutation rate I initially use is for SNPs, but the adjustment to 76.8 for a haploid genome, following the correction, modifies that rate to include structural variants as well.”
He’s saying he swapped mutation rates mid-calculation. He started with the SNP rate (~1.2×10⁻⁸ per site per generation), then switched to 76.8 mutations per haploid genome per generation (from Perubski et al.), which includes structural variants. He claims the fixation equation is indifferent to mutation type in an attempt to justify his switch as being valid.
But it’s not. The ~35 million divergent sites between humans and chimps that are the 1.09% divergence figure are measured in single nucleotide differences. If you inflate your mutation rate to include structural variants, then you need to compare your predicted fixation count against the total divergence, not against the SNP-only divergence. You cannot use a total mutation rate as the input and then validate the output against a fraction of the genome that omits 98.91 percent of the total. And being a mathematician, Griffiths was absolutely correct to flag it, as I previously did as well. Zach’s claim that “at no point does the fixation equation distinguish between mutational types” is totally irrelevant with regards to the empirical application of where all those expected neutral mutations happened to fix.
Again, Zachary really needs to sit down and shut the fuck up. Both a game designer and a billiards player clearly and observably understand his job better than he does.
4. At 38:04, Will states I use the wrong generation number (252,000 instead of 201,600). Again, I used the number that Will presented. What Will should be doing here is saying, “I was wrong, I presented the wrong number.” Again, this would be as if I taught a concept incorrectly and then blamed my students for not reading the textbook and catching my mistake. You are the one making the argument, Will. Own it and get it right.
That’s exactly what d does, you absolute fucking moron. When d=1, as in the Wright-Fisher model, there are 252,000 generations separating the CHLCA from modern humanity. When d=0.80, as the weighted time-average that corrects for overlapping human generations over that period works out to be, that leaves 201,600 effective generations available. This is one of the many reasons why Hancock should have read the book before ever opening his mouth in the first place. Notice that contrary to what Hancock previously said, neither Wright-Fisher nor Moran accounts for this correction, which is precisely why the selective turnover coefficient was developed in the first place.
5. Lastly, at 38:11, we get the most bizarre claim yet. He seems to believe that the 1.09% of divergent sites are the *only ones that could have diverged neutrally*. About 90% of the human genome lacks signs of selective constraint. This means that since we diverged from chimps, any of those sites could have been the ones that alternatively fixed between us. But which ones actually fix is entirely random. Rerun the clock, and a different 1.09% would fix. The fact that only 1.09% have fixed is a testament of the recency of our divergence, not that those were the only one’s that could fix. There’s nothing special about those sites. This is like being shocked when a random number generator spits out a specific number. For me, the most frustrating thing here is that no one has yet addressed why the classical population genetic fixation formulas are incorrect. There have been quibbles about mutation rates, whether or not we read Vox’s book, and what his coefficients mean. But the fact is that Vox’s model is refuted by basic population genetics from the 1930s that I derived already. And we have seen clearly that both Will and Griffiths don’t understand them, or they wouldn’t make the “explicitly relies on SNP” argument. Will seems to suggest that Griffiths, Day, and myself are going to have some meeting of the minds to work this all out and get to the bottom of it. This is frankly insulting. It’s like telling a surgeon that your accountant buddy is going to join at the operating table to help figure out the surgery.
At this point, we have no choice but to conclude that Dr. Zach Hancock is quite literally retarded. Was he some sort of diversity hire or something? First, WE are the ones who pointed out that the neutral mutations have to apply to the entire genome. HE is the one who tried to apply all of them to the 1.09 percent SNP divergence. Second, he’s flat-out lying again because I have specifically addressed why the classical population genetic fixation formulas are incorrect. You can read the papers for yourself; and don’t let the titles mislead you, as each one directly addresses a specific population genetic formula, from Wright-Fisher and Haldane to Kimura and Ohta.
k ≠ μ: The Finite-Time Collapse of Neutral Substitution Theory
The Hard Limits of Fixation Through Genetic Drift
Relictation: A Third Evolutionary Mechanism
Independent Confirmation of Haldane’s Limit: Empirical Validation Through Observed Fixation Rates
The Bernoulli Barrier: How Parallel Fixation Violates the Law of Large Numbers
While I am perfectly willing to engage in a public discussion of the issues raised by Population Zero, I very much doubt there is any purpose in including Dr. Zach Hancock in any discussion of evolution that involves mathematics. He is neither intelligent enough nor well-read enough to meaningfully participate in one.
And Dr. Zachary B. Hancock’s PhD in Ecology and Evolutionary Biology from Texas A&M clearly isn’t worthy of serving as toilet paper in a roadside Taco Bell.
