The future is probabilistic

When do you stop looking for something better and commit to the best option you’ve found so far?

People seem to struggle with this — in many facets of life, but perhaps nowhere more so than in their romantic relationships. 

This has grave consequences: Indecisiveness in choosing a life partner is contributing to an epidemic of loneliness, a demographic timebomb of falling birth rates, and a spiraling cost-of-housing crisis (inflated by single-person households).

Fortunately, there’s an easy fix for all this: probabilistic thinking. 

Instead of constantly second-guessing our romantic decisions, endlessly questioning whether someone better is just around the corner, agonizing over whether the person we’re with really is THE ONE… we can just do some math instead. 

The biggest decision you’ll ever make is really very simple: Plan to go on dates with as many as 100 people. After the 37th date, stop as soon as you find someone that’s the best so far. Because that’s your ONE.

This is the 37% rule, and it’s mathematically guaranteed to work.

In theory.

Meeting 37% of your pool of 100 potential partners is the mathematical sweet spot where you've gathered just enough information to set meaningful standards without burning through so many options that you’ve already rejected your ideal match.

Here’s the math to prove it: P(x) = -x ln(x), maximized when x = 1/e.

Where P(x) is the probability of choosing the absolute best person from your dating pool, x = 1/e represents your maximized chance of ending up with your soulmate.

Real life is messier than any equation, of course, so we need to be a bit more flexible than the 37%-rule algorithm would suggest.

That’s why, in addition to thinking more probabilistically, we also have to think more like a Bayesian — which is to say, we have to constantly update our "priors" as new information comes in.

The favorite word of Bayesian thinkers, "prior" is simply your initial assumption about how the world works — and the key to thinking like a Bayesian is being willing to revise those assumptions as you gather new evidence.

For example: When using the 37% rule to search for a life partner, if the first ten dates you go on are all disasters, it will be past time to update your priors. 

The sad truth is, you’re not likely to find 37 plausible life partners to go on a date with, let alone 100.

That might make the first ten dates a dispiriting reality-check for you. 

But you’ve gathered some useful information!

Now, with your priors duly updated, if you go on an 11th date and it happens to be terrific, you’ll know exactly what to do: propose.

That, at least, is what a good Bayesian would do, because new information requires updated thinking.

But as intuitive as that sounds, few of us do it.

Instead, we do the opposite, taking every opportunity to confirm our priors. 

We actively search out information that matches what we already think; we twist ambiguous information to support our preexisting view; we selectively remember information that reinforces our beliefs and conveniently forget information that undermines it.

Even worse, we actively resist being corrected: When presented with evidence that contradicts our beliefs, we perversely double down on them (the "backfire effect").

Fortunately, there’s a fix for all of this, too: prediction markets.

Humanity’s last chance?

Losing a hand of poker with three aces because someone else had four kings is bitterly disappointing. 

But it doesn’t mean you should have folded the three aces. It just means you lost a hand that you’re likely to win 9 times out of 10. 

It happens — approximately 10% of the time. But betting with three aces is still the right bet to make, 100% of the time (nearly).

And yet hardly anyone thinks this way. 

Instead, we retroactively judge our decisions by their outcomes: If something worked, it was smart; if it didn't work, it was stupid.

Poker players call that "resulting" — a habit of mind that even a low-stakes poker game should break you out of. Losing a few hands of poker when you have, say, the second-best possible cards, teaches us to think probabilistically: The future is uncertain, and the only way to plan for it is by thinking in percentage probabilities. 

Prediction markets have the potential to spread this lesson far beyond poker — because, as Vitalik puts it, they show us "probabilities that reflect genuine uncertainty in the world."

Vitalik contrasts the uncertainty of prediction markets favorably with social media, where every prediction is presented as something close to a sure thing. 

It usually requires just a glance at Polymarket to learn that the probability of the latest sure-thing happening is more like, say, 18%.

18% probabilities do also happen (approximately 18% of the time, as it turns out).

But there’s an 82% chance that whatever’s trending on X won’t happen. Knowing this might 1) lower our collective blood pressure and 2) redirect our scarce attention in societally beneficial ways. 

The best thing about prediction markets, Vitalik explains, is that they "favor truth seeking" — and they do so at the expense of social media commentators who are otherwise rewarded for their predictions of doom with "monetizable clout."

If we all start watching prediction markets instead of social media sites, or the news, even, the doom-mongerers will have less clout to monetize.

If we all start participating in prediction markets, the shift will do even more. 

With money on the line, prediction markets will incentivize us to actively seek out reasons we may be wrong (something we’re naturally wired to avoid).

They will make us better Bayesians, by training our brains to see new information not as a threat to our ego or identity, but as a reason to update our priors.

Regularly losing prediction-market bets will encourage us to hold our opinions more lightly.

Watching others lose bets will get us to stop listening to attention-seeking pundits. 

Eventually, governments might be encouraged to fund projects based on their "expected value" (probability x impact) rather than their political optics.

Individuals, thinking in terms of expected outcomes, might evaluate their personal decisions more accurately — from medical procedures to career changes to finding a life partner to have babies with.

And what could be more important to the world than finding someone to have babies with?

Probabilistic thinking could literally save humanity from extinction.

— Byron Gilliam

As crypto markets mature into a capital-efficient financial system, the sUSDe term structure emerges as a forward-looking signal to gauge market sentiment and forecast market regimes.

sUSDe has found its strongest product-market fit on Pendle v2, a yield-stripping exchange that lets the market speculate on the product’s underlying yield into set maturities. With multiple expirations for sUSDe on Pendle, the market can reveal the implied yield for sUSDe, and in turn the basis, across various points in time out into the future.

Overall, the sUSDe term structure on Pendle reveals the market’s expectations channel, offering a forward-looking signal to changes in the cost of carry, price level of BTC, and all metrics downstream of these.

_{Brought to you by:}

Arkham is a crypto exchange and a blockchain analytics platform that lets you look inside the wallets of the best crypto traders — and then act on that information.

Arkham’s Intel Platform has a suite of features including real-time alerts, customisable dashboards, a transaction visualization tool, and advanced transaction filtering — all of which is accessible on all major blockchain networks, and completely free.

Arkham’s main product is the exchange, where users can express their trade ideas against the market.

Crypto’s premier institutional conference is back this March 24–26 in NYC.

Don’t miss SEC Chairman Paul S. Atkins’ keynote on Day 1.