Reed's Law, proposed by David P. Reed, states that the value of networks that allow group formation (like social networks and messaging platforms) scales even faster than Metcalfe's Law — at 2^N rather than N^2. This is because the number of possible subgroups (communities, group chats) grows exponentially.

Why does Reed's Law hit infinity so quickly?

Reed's Law (2^N - N - 1) grows doubly exponential. At N=10, there are 1,013 subgroups. At N=50, the number exceeds 10^15 (a quadrillion). JavaScript's floating point hits Infinity around N=1024. In practice, most subgroups have zero value — Reed's Law represents theoretical maximum, not realized value.

Network Value Calculator

Compare Metcalfe's Law (N² connections) vs Reed's Law (2^N subgroups) to understand how network value scales.

Number of Users (N)

Metcalfe Value (connections)

N(N-1)/2

Reed Value (subgroups)

1,013

2^N - N - 1

Reed / Metcalfe Ratio

22.5x

Reed's Law generates 22.5x more potential value

Comparison at Various Network Sizes

Users (N)	Metcalfe	Reed	Ratio

How to Use the Network Value Calculator

Enter the number of users — use the input field or slider to set your network size (2 to 1,024 users).
Compare the values — Metcalfe's Law shows point-to-point connections while Reed's Law shows possible subgroups.
Check the ratio — see how dramatically Reed's Law outpaces Metcalfe's Law as N grows.
Review the comparison table — see values at benchmark sizes (5, 10, 15, 20, 30, 50 users).

Understanding Network Value Laws

Network effects are the driving force behind the most valuable companies in the world. When a product becomes more valuable as more people use it, a positive feedback loop creates explosive growth. Two mathematical models capture this dynamic: Metcalfe's Law for communication networks and Reed's Law for community-forming networks.

Metcalfe's Law: N² Scaling

Robert Metcalfe, inventor of Ethernet, observed that the value of a telecommunications network is proportional to the square of its users:

V_Metcalfe = N(N - 1) / 2

With 10 users, there are 45 possible connections. With 100 users, there are 4,950. This quadratic growth explains why a telephone network with 10 million users is vastly more valuable than 10 separate networks of 1 million users each — the combined network has 100x the connections.

Reed's Law: 2^N Scaling

David P. Reed extended the analysis to networks that allow group formation — platforms where users can create subgroups, communities, and teams:

V_Reed = 2^N - N - 1

This exponential growth dwarfs Metcalfe's Law. At 20 users, Metcalfe predicts 190 connections while Reed predicts over 1 million subgroups. This explains why social platforms (Reddit, Discord, Facebook Groups) grow in value so much faster than pure communication tools — every possible combination of users represents a potential community.

Real-World Implications

Network value laws explain several phenomena in technology markets. They explain why platforms pursue growth at all costs (more users = exponentially more value), why market leaders tend toward monopoly (the largest network offers disproportionately more value), and why platform switching costs are so high (leaving means losing access to all your connections and communities).

Practical Limitations

Neither law perfectly describes reality. Metcalfe's Law overestimates because not all connections are equally valuable — you care more about connecting with friends than strangers. Reed's Law overestimates even more because most theoretical subgroups are meaningless — no one needs a group of exactly user #47 and user #2,891. The actual value of a network lies somewhere between Sarnoff's linear model and Reed's exponential model.

Frequently Asked Questions

Metcalfe's Law states that the value of a telecommunications network is proportional to the square of the number of connected users. With N users, the number of possible connections is N(N-1)/2. This explains why networks become exponentially more valuable as they grow.

Reed's Law, proposed by David P. Reed, states that the value of networks that allow group formation scales at 2^N rather than N^2. This is because the number of possible subgroups (communities, group chats) grows exponentially with network size.

Reed's Law grows doubly exponential. At N=10, there are 1,013 subgroups. At N=50, the number exceeds 10^15. JavaScript hits Infinity around N=1024. In practice, most subgroups have zero value — Reed's Law represents theoretical maximum, not realized value.

Modern platforms like Facebook, Reddit, and Discord exhibit both laws. Metcalfe's Law captures person-to-person connections (messaging, following). Reed's Law captures group formation (subreddits, Discord servers, Facebook groups). The actual value lies between the two models.

Because network value scales faster than linearly with users, the largest network tends to dominate. A network with 2x users has 4x the connections (Metcalfe) or exponentially more subgroups (Reed). This creates natural winner-take-all dynamics.