The takeaway: Minecraft was never designed to behave like a calculator. Its world is built entirely from cubes, with no smooth curves and no native concept of continuous geometry. That makes it an awkward place to tackle a number like pi, which depends on the properties of a perfect circle. But two mathematicians have shown that, even in a blocky universe, you can still get surprisingly close.
Molly Lynch of Hollins University and Michael Weselcouch of Roanoke College approached the problem less like programmers and more like experimentalists. Instead of recreating a traditional algorithm inside Minecraft – a process that typically involves building elaborate in-game logic systems – they leaned on probability and the game's existing mechanics to do the work for them.
The game has already been proven to be Turing-complete, meaning it can theoretically run any computation. But actually pulling that off usually requires massive, intricate builds that simulate computer hardware at a granular level. Lynch and Weselcouch deliberately avoided that route. Their goal was to find a method that worked with the game, not against it.
They settled on the Monte Carlo method, a statistical approach that uses randomness to approximate values. In this case, the idea is to estimate pi by comparing the area of a circle to the square that contains it. If you randomly scatter points across the square, the fraction that lands inside the circle should approach pi divided by 4 over time.
Translating that idea into Minecraft required some improvisation. The researchers built a square boundary using blue blocks and, inside it, a rough approximation of a circle using red blocks with a radius of 11 units. Because everything in Minecraft sits on a grid, the circle is visibly jagged, but it serves as a workable stand-in.
The more interesting challenge was generating randomness. Rather than coding a random number generator, Lynch and Weselcouch used in-game creatures. Slimes, which wander unpredictably, became their source of randomness. As they explain, "slimes continue moving when no players are nearby and they change direction at random," making them a natural fit for the role.
To turn movement into measurable data, the researchers introduced zoglins – hostile mobs that attack and kill slimes. Each time a slime is killed, it effectively marks a random point within the square. Where that happens determines whether the "point" falls inside or outside the circle.
They then used hoppers, which automatically collect dropped items, to track the results. By covering the circular area with hoppers and counting total deaths across the entire square, they could calculate the ratio without manually tracking each event. Every item collected represented a single data point in the experiment.
In one test, 619 slimes were killed, with 508 of those deaths occurring inside the circular area. Plugging those numbers into the Monte Carlo formula produced an estimate of pi equal to 3.283. It isn't especially accurate, but that is beside the point. The system works, and it does so entirely within the rules of the game.
There are clear ways to improve the result. A larger circle would better approximate true curvature, and a higher number of slime encounters would reduce statistical noise. Both changes would push the estimate closer to the real value of pi. But even with those adjustments, this method is far less efficient than conventional computation.
Efficiency was never the goal. Lynch and Weselcouch were more interested in demonstrating that mathematical ideas can take shape in unexpected places. By using Minecraft's built-in systems – entity behavior, item drops, and simple block structures – they created a model that is easy to understand and replicate without specialized knowledge.
The experiment stands out for how little it relies on traditional computing concepts. There is no code in the usual sense, no in-game processors or memory registers. Instead, randomness comes from creature movement, and data collection happens through the game's physics and item systems.