Bring a token out or move on? What to do with a six
Published · By TG, the site operator
Today's note is a strategy idea for the start and middle of a game: what to do with a six when it could bring a new token out of the yard or move one already on the track. Every figure below comes from code. The opening odds were enumerated through the site's rules code, and each plan was played for 40,000 full games through the same code against the site's Normal computer, taking Red in half of them.
Why the yard is expensive
Only a six brings a token out on this board. With all four tokens in the yard, a turn frees at least one of them 16.7% of the time: one token in 13.9% of turns, two in 2.8%, when a second six follows the first. The first token needs six turns on average, and 40.2% of the time you are still waiting after five turns. A six spent on a token that is already running is a six you may wait a long while to see again.
Three plans against the computer
The plans differ only on a six that offers both an entry and a move on the track, and only when that six cannot capture or bring a token home; those always came first. Every other choice followed the Normal computer's own move scoring. Win rates against the Normal computer, each within about half a point:
- Bring a token out whenever you can: 50.2%, level with the computer playing itself (49.8%).
- Keep two tokens on the track and move those: 38.0%.
- Run with what is out, and enter only when nothing else can move: 21.8%.
Where the games are lost
Not on captures. The plan that keeps tokens in the yard lost fewer of them: 1.21 tokens sent back per game, against 2.20 for the plan that brings them out. It lost on dead rolls instead. A 1 to 5 that no token can use ends your turn, and the run-first plan wasted 20.1 of those per game while its opponent wasted 8.7. The bring-them-out plan wasted 10.1, the same as the computer.
The rules show how those dead rolls pile up. When your only token on the track is hit, all four are back in the yard and every 1 to 5 is useless until the next six. And a lone token in the home column moves only with a number that fits the squares left, so a single runner near the finish stalls on most rolls. A second or third token on the track gives each roll somewhere to go.
Using it
- Captures and finishes first. All three plans took a capture or a finish with a six before thinking about the yard.
- A new token arrives safe. Each colour's entry square is one of the eight marked safe squares, so a token that has just come out cannot be hit there.
- Not a law. The test compared whole plans, not every position. A six that lifts a token out of danger may still be the better move now and then.
The computer already does this
The site's computer scores an entry at 45 points plus 30 for landing on a safe square, while an ordinary move earns 0.65 points per step of progress. In simulated games where a six offered both, Normal and Hard both brought a token out about 97% of the time. To keep up with it, do the same: start a game against Normal on the game page, and read the guide for the six, the yard and the safe squares.
Sources
- Why a six is not just a lucky Ludo roll (rules guide) — ewhite911-game2610-ludo.vercel.app, 2 October 2026
- Ludo game page: how the game works and computer levels — ewhite911-game2610-ludo.vercel.app
- Game code (app.js): rolls, sixes, entries and moves — ewhite911-game2610-ludo.vercel.app
- Computer opponent code (ai-worker.js): legal moves, safe squares and move scoring — ewhite911-game2610-ludo.vercel.app
Written by the operator of this site from the sources listed above. Spotted a mistake? Email warmeve.label@gmail.com and it will be corrected.