Harshbarger’s team quickly produced a family of four 12‑sided dice that work flawlessly for two, three, or four participants, delivering a statistically fair result and a clear winner each time.

Extending the concept to five players proved far more difficult; early theoretical models required as many as 1,440 faces per die, a size that was clearly impractical. A researcher in Australia later proposed a more realistic configuration of five 120‑sided dice.

The breakthrough arrived when Canadian software engineer Paul Meyer, who approached the problem as a massive computational search, identified a viable design using five 60‑sided dice—far fewer faces than any previous proposal.

The final Go First Dice are five golf‑ball‑sized, 60‑sided dice that guarantee a single winner for a five‑player game with a single roll, and they are now available for purchase.

Harshbarger emphasizes that the underlying mathematics are sound, noting that 30 sides is the theoretical minimum for five players, and his team continues to explore whether fewer sides or larger player groups can be accommodated, even though current tools are insufficient for those extensions.