Skip to main content

The Rise of Google, Part 2: BackRub: The Algorithm That Ranked the Web

BackRub: The Algorithm That Ranked the Web

🎬 Prefer to watch? Here's the 60-second version:

Watch on YouTube ▶

📚 The Rise of Google — a 20-part series. See all parts »  |  « Part 1: Two Students Who Couldn't Stand Each Other

By 1996, the web had a discovery problem. Millions of pages were coming online, and the search engines of the day—AltaVista, Lycos, Excite—mostly worked like frantic librarians who could read but not judge. Type a query and they would hand back every page that happened to contain your words, ranked by little more than how many times those words appeared. It was easy to game, easy to flood, and increasingly useless. The more the web grew, the worse search got.

Larry Page, then a PhD student at Stanford, was circling a different question entirely. Casting about for a dissertation topic in 1995, he became fascinated not by the content of web pages but by the links between them—the mathematical structure of the web itself. What if the connections mattered more than the words?

Borrowing an idea from the library

The insight Page reached for was older than the web, and it came from academia. For decades, scholars had measured the influence of a research paper by counting its citations—a technique pioneered by Eugene Garfield in the 1950s. A paper cited by hundreds of others was probably important. A paper cited by a single landmark study might be more important still. Citations were a currency of trust, accumulated one reference at a time.

Page realized the web already worked the same way. Every hyperlink was, in effect, a citation—one page vouching for another. Nobody had thought to read the web backwards: instead of asking "what does this page link to?", ask "who links to this page?" Those inbound links, or backlinks, were a vast, untapped map of what the web itself considered worth pointing at.

He gave the project a fittingly cheeky name: BackRub. In March 1996, Page pointed a homemade web crawler at a single seed URL—his own Stanford homepage—and let it wander outward, following links and recording who pointed where. Sergey Brin, the mathematician with an NSF fellowship and a taste for hard problems, joined soon after. The engineering muscle came from Scott Hassan, who wrote much of the early crawler code. Together they began pulling the web's link graph into a Stanford basement.

The math that ranked the web

Counting backlinks alone wasn't enough—spammers could always manufacture a thousand junk links. Page and Brin's real breakthrough was recognizing that not all votes should count equally. A link from a page that was itself important should carry more weight than a link from an obscure one. And the importance of that linking page depended, in turn, on who linked to it.

That circular definition sounds like it should collapse into nonsense, but it's exactly the kind of self-referential system mathematics knows how to solve. The result was an algorithm they named PageRank—a pun that worked on two levels, referring both to web pages and to Larry Page himself. Importance flowed through the web like water finding its level, each page inheriting a share of the trust held by the pages that pointed to it.

The elegance was that the web scored itself. No editor, no directory of hand-picked sites, no human deciding what mattered—just the collective judgment of everyone who had ever bothered to make a link. A single endorsement from a genuinely trusted source could outweigh a hundred desperate ones, which made the whole scheme remarkably hard to fool.

A tool that quietly outclassed everyone

The difference showed instantly. Where rival engines returned bloated lists padded with keyword-stuffed junk, BackRub surfaced the pages people actually wanted—often at the very top. It felt smarter, because in a real sense it was borrowing the accumulated intelligence of the entire web.

By August 1996 the system was running on Stanford's network, indexing on the order of tens of millions of pages and gobbling up close to half the university's bandwidth—an early hint that this "research project" had ambitions no dorm-room server could contain. It ran on a scavenged mix of Sun and Intel machines, held together with academic budgets and stubbornness.

The idea proved valuable enough to patent. Stanford filed for it, was granted the PageRank patent in 2001, and licensed it exclusively to the company its two students would go on to found. In exchange the university took shares—a stake it later sold for roughly $336 million. Not a bad return on a dissertation topic.

In 1998, Page and Brin would lay the whole thing out in a now-famous paper, The Anatomy of a Large-Scale Hypertextual Web Search Engine. But long before the academic write-up, they had a problem of a very different kind: BackRub was outgrowing Stanford, outgrowing its name, and outgrowing its founders' plans to ever finish their degrees.

Next time: the awkward name that came from a spelling mistake, a garage in Menlo Park, and a $100,000 check written to a company that didn't legally exist yet.


🔗 Explore more from Syncster

Comments

Popular posts from this blog

Cursor AI Review: Is the AI Code Editor Worth It?

I've been using Cursor as my main code editor for a while now, and enough people have asked whether it's worth switching to that a proper review felt overdue. Short version: for me, yes — but with caveats. What is Cursor? Cursor is an AI-first code editor built as a fork of VS Code. That means every extension, theme, and keybinding you already use in VS Code works here, but with AI woven directly into the editing experience instead of bolted on as a plugin. It's made by Anysphere and can run models from OpenAI and Anthropic under the hood. What I like Tab completion is uncanny. Cursor predicts your next edit — not just the rest of the line, but the next change across the file. Once you get used to hitting Tab, going back to a plain editor feels slow. The Composer / Agent mode. You describe a change in plain language and it edits multiple files at once, showing you a diff to accept or reject. For refactors and boilerplate, this saves real time. It unde...

MacBook Pro M5 vs M5 Pro: Which One Should You Actually Buy?

Apple's latest 14-inch MacBook Pro comes in two very different flavors: the base M5 and the step-up M5 Pro . On paper they look similar — same gorgeous Liquid Retina XDR display, same design — but under the hood the gap is bigger than the names suggest. Here's a clear, no-hype breakdown, with concrete use cases so you can match the chip to your work. Quick spec comparison Spec M5 M5 Pro CPU 10-core (4 performance + 6 efficiency) Up to 18-core (6 performance + 12 efficiency) GPU 10-core Up to 20-core Neural Engine 16-core 16-core Memory bandwidth 153 GB/s 307 GB/s (roughly double) Unified memory 16 / 24 / 32 GB 24 / 48 / 64 GB Max storage Up to 4 TB SSD Up to 8 TB SSD Battery (video playback) Up to 24 hours Up to 22 hours Media engines Single encode/ProRes engine More encode/ProRes engines (higher configs) What actually changes between them More cores — the M5 Pro nearly doubles CPU cores and adds GPU cores, so sustained, multi-threaded work finishe...

How I used Google Sheets and Apps Script

Google Sheet is one of the most powerful spreadsheet application that exists online, rivaling with Microsoft's Excel. One of the main strengths is its strong support for collaboration with other users, much easier and popular than collaboration tools with Microsoft Office. Aside from plain spreadsheet, it also supports extensions such as macro. If you are familiar with macros on other office tools, they work almost the same. However, the most extension I use and tinker with is the Apps Scipt . Apps Script Extension One of the challenges I faced recently is how do I track or monitor reports in our department if they are submitted on time or worst, forgotten due to lack of better monitoring tools. So I thought if there can be simple applications that can be deployed or use by a more general user to allow reminding periodically what reports are approaching due dates or those that are past dues. Then I looked for a way, instead of creating a full blown app from scratc...