Justin Blumenstiel

Natalie Portman meme. Panel 1: Check out this amazing correlation. Panel 2: You controlled for phylogeny, right? Panel 3: Blank stare. Panel 4: Right

Am I meme'ing correctly?....I am, right?

They need to start saying he is the actual president. When people starting saying it about Elon it really got under Trump's skin and ended it.

This woman has a truly remarkable talent in distilling things.

Fisher's Fundamental Theorem of Natural Selection—A Philosophical Analysis | The British Journal for the Philosophy of Science: Vol 59, No 3 Abstract This paper provides a philosophical analysis of the ongoing controversy surrounding R.A. Fisher's famous ‘fundamental theorem’ of natural selection. The difference between the ‘traditional’ a...

I found this paper really helpful on Fisher’s concept of the environment

www.journals.uchicago.edu/doi/abs/10.1...

Submitted: December 3, 2025 Published: November 13, 2025

OMG. And not only that. You were able to PUBLISH IT INTO THE PAST!

Is the Fundamental Theorem of Natural Selection of any use? Abstract. There have been many recent discussions of the Fundamental Theorem of Natural Selection, with an emphasis on its mathematical accuracy. It is arg

"There is still life in the old dog."

Brian Charlesworth on Fisher's Fundamental Theorem in @journal-evo.bsky.social

doi.org/10.1093/evol...

EXACTLY. Without dom., the Fisher and p' approaches were giving the same. But not w/dominance. I never really conceptualized the change in the dominance effect when allele frequencies change as part of the "changing environment" before. Just shows I didn't really understand the Price Equation.

YES!

I know, I'm lucky as hell. But I don't want to wear out my welcome!

Shout out to John Kelly??? I'll tell him you say hi, but I won't tell him it was because I was struggling with quant gen and broadcasting it all over the internet..LOL.

In his 1972 paper, "Fisher's ‘fundamental theorem’ made clear," George Price provided the rigorous mathematical proof that Fisher’s theorem was not an approximation, but an exact statement—provided one understands that Fisher was only talking about the "partial change" in fitness.The Exact QuotePrice clarifies the distinction between the total change in fitness and the specific component Fisher was measuring:"The theorem characterizes the partial change in the average fitness of a population that is directly ascribed to the action of natural selection... it does not give the total change in average fitness. This total change is the sum of two parts: a part due to natural selection (which is the part Fisher’s theorem is about) and a part due to 'environmental change'."The Price Equation ContextPrice demonstrated that the total change in mean fitness ($\Delta \bar{w}$) can be partitioned into two distinct terms. This is often represented conceptually as:$$\Delta \bar{w} = \text{Cov}(w_i, z_i) + E(w_i, \Delta z_i)$$The first term ($\text{Cov}$): This represents the Fisherian part. It is the change in the trait (fitness) due to the covariance between the trait and fitness (selection). For fitness itself, this covariance equals the additive genetic variance ($V_A$).The second term ($E$): This represents the "Environmental" part. In this context, "environment" includes everything that changes the relationship between genes and fitness from one generation to the next—including dominance, epistasis, and external environmental shifts.Why this explains your discrepancyIn your calculation, when you used $\Delta p$ to find the new mean fitness, you were calculating the Total Change.When you used $V_A$, you were calculating only the Partial Change.

OMG. I have actually read some stuff by Steve Frank on the Price equation (whether I understood it is another matter). But this from Gemini is great. More reading to do I guess. But at least I am not losing my mind anymore (well, on this specific matter at least)

So, you find it useful?

Gemini is WAY better in the past couple months. But Trust But Verify still is critical...

I *think* Fisher was talking about an instaneous rate - not a prediction for the next, non-overlapping generation. For this dumbass, it is great to have a tutor like this. Peace out.

OK, maybe you think I am a dummy. But the fact that it provides this guidance has REALLY opened a door for me to think about this more. And I don't have to harass John Kelly, thus saving myself certain embarrassment.

. Comparing the Two ResultsPrediction from $V_A$ (Fisher): $\Delta \bar{w} = 0.000789439$Actual $\Delta \bar{w}$ (from $\Delta p$): $1.00082435 - 1.0 = \mathbf{0.00082435}$Why are they different?You likely noticed that the actual change ($0.000824$) is slightly higher than the Fisher prediction ($0.000789$). This is a classic "aha!" moment in quantitative genetics:Linear Approximation: Fisher's Theorem is a first-order derivative. It predicts the change in fitness based on the current state of the population. It assumes the "slope" of the fitness increase stays constant.The "Moving" Population: In reality, as $p$ increases, the population becomes slightly more efficient at producing the $A_1A_1$ genotype in the very next generation than the linear prediction accounts for.Non-Zero Higher Moments: Fisher’s theorem technically states that the rate of change is equal to the $V_A$, but it doesn't account for the change in $V_A$ itself as the frequency shifts within that same generation.In your specific case, the discrepancy is small because $p$ is small. If $p$ were $0.5$, the "gap" between the Fisher prediction and the actual change would be more pronounced.

Anyway, I used Gemini to check my work using the two approaches and it made the same calculations. So then I was like: Yo, WTF. Why doesn't Fisher's/Falconer/Quant Gen approach agree with p'. And here is what it spat out: 3/n

Using the standard p' = (p^2(w11)+pq(w12))/w_bar, I have noticed that the change in mean population fitness can go up more than that predicted using Average Effects of alleles to calculate additive genetic variation (Va). I was sure I was doing something wrong. 2/n

Um. Holy %^#K. Gemini is REALLY freaking good at explaining stuff. Using it a few weeks ago, I finally understood branching processes/extinction probabilities. Today, I started working through some stuff I HAVE BEEN LOSING MY mind about for years re: change in mean fitness using Fisher vs. p'. 1/n

Keeping the bureaucracy satiated, one email at a time.

OK - is vaccination status intrinsic or extrinsic? 🤪

If Joseph Goebbels and a rectal polyp had a kid.

Is cancer caused by HPV intrinsic or extrinsic?

Bring back the phenotype When thinking about evolutionary change, many practicing biologists will focus on changes in allele frequencies over time. This gene-centric view of evolution has strongly impacted how evolution (and...

"Life is marvelous, complex, and certainly more than machinery and genetic information"

@amartinezarias.bsky.social would agree.

nph.onlinelibrary.wiley.com/doi/10.1111/...

@iammrbeat.bsky.social You will like this, could be good for a stream:

jacoblum22.github.io/philosophy-q...

DEAR FRIENDS - My son Jake made this online quiz to help you discover your life philosophy, which might not be what you think. Note that the questions are tough and the answers are limited/imperfect. That's life, baby, 😉
Takes about 15-20 minutes. jacoblum22.github.io/philosophy-q...

Okay. Time to call your democratic senators and tell them that Bovino’s demotion and humiliation WILL NOT SUFFICE to fund DHS.

greg bovino next to stalin, in an image from stalin era infamous for removing the person where bovino is standing

if bovino is out and conservatives are turning on noem then there is a real opportunity to demand that miller resign too.