I have one constant companion—doubt. With every keystroke I make, there is a devil perched on my shoulder questioning whether I deserve to write. I now earn more from writing than I did from my fancy tech job, which, while remarkable, feels undeserved. 

If I wanted to dispel that feeling, I have a couple of options—Oreos, Prozac, pounding Diet Coke until my eyes bleed—but I think the best solution is Napkin Math™. Accordingly, this post is the Napkin Math on whether I’m good at my job. I’ll go through the outcomes I predicted in my essays last year, score their accuracy, and see what I got wrong.

All writers should do this exercise, as too many skate by on reputation and cultural cache. Public reckonings with accuracy hold you accountable to the audience on whether the product they are paying for is worthwhile. My goal is that you pay me for accuracy. This piece is about whether the investment is worth it. 

To prepare, I read everything I’ve written over the last 12 months and scored myself from 1-20 on the accuracy of the forecasts I made. A 1 is terrible, awful, Quibi-is-the-future-of-entertainment-level of take. A 20 is a spooky, 100% correct forecast.

I’d estimate that a lifetime average score of 12 is the minimum acceptable level of performance for a newsletter writer. My hope was for 15. If I got a result below 9, I was going to quit and go back to the farm.

Last year, I scored an average of 16.45 average and a median of 17.5, which translates to a roughly 80% accuracy rate on my calls. This year saw a marked improvement in my score, with a median of 19.5 and an average of 18.5. In 2023 I published over 50 pieces, with most containing one or two predictions, so it’s a decent sample size. The math would say that I am a much-improved analyst. When coupled with the increased virality of my pieces and our reader survey showing more subscribers rating my work highly, I had a very good year. 

While my averages were high, a few things brought them down:

• I was too generous in my estimation of humanity. I thought that this was the year that we finally got serious about understanding algorithmic feeds in content. Very predictably, no progress past tribalism was made. 
• In 2022, I proposed some business strategy theories that were proven true relatively quickly. This year, I proposed similarly grand theories about the internet, but I wasn't lucky enough that evidence immediately proved me right. Unfortunately, evidence does not yet adhere to a fiscal year schedule, though hopefully some obvious proof points will surface soon. 

I also had some pretty strong wins:

• The creator economy is going to crash. Significant layoffs and shutdowns have occurred in the six months since I made this prediction. 
• Media layoffs will continue as the ad model suffers. The winners will be multi-SKU and community-first. Nearly all media companies had layoffs this year, but those that had multiple revenue streams, like the Daily Wire, did much better. 
• AI looks like a bubble. Pricing is out of control, and I’m hearing backroom whispers that growth rates for most AI startups are stagnating. The initial adoption curve has ebbed, and VCs are reeling from somehow doing 100x ARR rounds again.   
• Pornhub’s acquisition was a scam. This piece engendered my first real flirtation with mainstream media coverage. I spoke with Bloomberg and some other outlets. The piece and accompanying Twitter thread got retweeted by Bill Ackman and racked up millions of views. 

You would think with the increased accuracy, virality, and income, I would be satisfied. That is not the case. I’m frustrated with my performance and think that next year merits an evolution in my writing.