Skip to content
Tool Corner

Board Game Square Odds Heatmap

How often a token ends its move on each of the 40 squares of a property-trading board game, worked out exactly and then checked by rolling the dice a million times.

Illustration of a wooden board game with a winding path of coloured tiles, a pawn and a die
Illustration
Built and verified by Jogeswara, Founder and Engineer, MSc, PMP — Tool CornerModel checked against the sources listed belowReviewed

{{ simTxt }} {{ simPct }}

Result

{{ sentence }}

  • {{ it.t }}

{{ srTxt }}

Square {{ selNum }}
{{ selName }}
Exact landing chance
{{ selPct }}
Rank of 40
{{ selRank }}
Simulated
{{ selSim }}
Against flat
{{ selFlat }}
Chance per turn
{{ selPerTurn }}

Jail / just visiting is shown as one square. It includes every landing that ends in jail ({{ jailIn }}: being sent there, plus each turn spent waiting) and every landing that only passes through as just visiting ({{ jailVis }}).

Colour groups, stations and utilities

Chance of landing on at least one square of the group during a turn, and the average number of landings per lap of the board. A turn is on average {{ rollsTxt }} rolls, and a lap takes on average {{ lapTxt }} turns.

{{ g.name }}
squares {{ g.squares }}
{{ g.turnTxt }}
{{ g.lapTxt }}
per turnper lap

All 40 squares, most landed-on first

RankSquareExactSimulatedvs flat
{{ r.rank }}{{ r.ex }}{{ r.si }}{{ r.vf }}

What your result means

Each square's percentage is the share of all dice rolls, over a very long game, that end with the token on that square. A plain board would give every square 2.5%, so anything above 2.5% is landed on more than its fair share, and anything below it less. The heatmap colours each square by that share; switch on the flat-board comparison to colour it by how far each square sits above or below 2.5% instead.

Three conventions matter. First, a landing is counted where the token ends up: if a card sends you from a card square to Start, the landing counts for Start, not the card square. Second, Jail / just visiting is one square that includes both passing visits and every turn spent in jail; the detail panel shows the two parts separately when you select square 10. Third, "Go to jail" is never ended on while it is active, so it shows 0%.

The group chart answers a different question: the chance that, in one turn, you end at least one roll on any square of that group. Doubles give extra rolls, so a turn can touch two squares, which is why the per-turn chance for a group is not simply the sum of its squares.

The model, in plain words

  • 40 squares numbered 0 to 39 around a ring. You roll two six-sided dice and move forward by their total.
  • A double gives you another roll in the same turn. Rolling doubles three times in a row sends you straight to jail (square 10) without moving for the third roll. You can switch this rule off.
  • Ending on square 30, "Go to jail", sends you to jail. You can switch this off too.
  • Card square B (squares 2, 17 and 33) has a deck of 16 cards: one sends you to Start, one sends you to jail and the other 14 leave you where you are.
  • Card square A (squares 7, 22 and 36) has 16 cards: one each to Start, square 24, square 11, square 39 and square 5, one to the next utility, two to the next station, one to jail, one that moves you back three squares and six that leave you where you are. Going back three from square 36 lands on square 33, a type B card square, so you draw again there.
  • Each card draw is treated as independent, as if every card were shuffled back in before the next draw. This is the standard simplification for long-run odds; with real decks the order is remembered until the deck is reshuffled.
  • In jail, with "wait for doubles", you roll once per turn. A double lets you out and you move by that roll, with no extra roll. After a third failed turn you pay, leave and move by that roll. With "pay to leave straight away" you leave at the start of your next turn and take a normal turn from square 10.

What a Markov chain is

A Markov chain is a system that hops between states, where the chance of the next hop depends only on where you are now, not on how you got there. A token on this board is almost that already: where it goes next depends on its square and the dice. The two things it also needs to remember are how many doubles it has rolled this turn, because a third double sends it to jail, and, if it is in jail, how many turns it has waited. So the chain here has 123 states: each of the 40 squares with 0, 1 or 2 doubles rolled so far, plus three "in jail" states for the first, second and third turn inside.

From every state the page lists every possible dice roll and card draw and the state each one leads to, with its probability. That table is the transition matrix. The long-run share of time spent in each state is the stationary distribution: the one set of probabilities that the matrix leaves unchanged. The page finds it by solving those 123 equations directly, then adds up the three doubles states for each square. That is why the numbers are exact for these rules, not estimates, and why they add up to 100% in every combination of the switches.

Why some squares are hotter

Take every special rule away (no "Go to jail", no cards, no three-doubles rule) and every square comes out at exactly 2.50%. That is not a coincidence of the dice. Every square can be reached by the same set of rolls from the squares behind it, so probability flows in and out of each square evenly and the long-run pattern has to be flat. Try it: switch all three rules off and the heatmap goes a single colour.

Jail is what breaks the pattern. Square 30, a card in each deck and three doubles all send tokens to square 10, so jail collects about one landing in nine. Every one of those tokens then leaves jail with a roll from square 10, and two dice most often total 6, 7 or 8. That is why the squares 16 to 20 are warmer than average. The cards add their own hot spots, because the type A deck sends tokens directly to squares 24, 0, 11, 39 and 5 and to the nearest station or utility. The cold stretch is squares 1 to 9: a token sent from 30 to jail jumps over them, and one leaving jail starts beyond them.

The card squares themselves are cold for the opposite reason. When you land on a type A square, 10 of the 16 cards move you on, and the landing is counted where you end up.

How the simulation relates to the exact answer

The simulation plays the same rules with real random dice and real card draws, one turn at a time, and counts where each roll ends. It knows nothing about the matrix. After enough turns its counts settle on the same percentages, which is the most direct way to check that the exact solution and the rules agree. The narrow strip along the bottom of each square in the heatmap is coloured by the simulated value, so you can watch it converge on the colour above it.

Its error shrinks with the square root of the number of turns, so each tenfold increase buys about three times the accuracy. At 1,000 turns some squares can be a percentage point out; at 1,000,000 every square lands within about 0.05 percentage points of the exact value. The simulation runs in small slices between screen frames, so the page stays responsive while it works.

Hot squares, rules and groups in numbers

These are the exact long-run figures the page solves for, not simulation results: a 123-state chain, 40 squares times 0, 1 or 2 doubles so far plus three jail states, solved directly. Each percentage is the share of all rolls that end on that square.

The hottest squares with every rule on (wait for doubles)
SquareShare of landingsAgainst a flat 2.50%
Jail / just visiting (10)11.53%4.61 ×
Square 243.00%1.20 ×
Start (0)2.92%1.17 ×
Station 3 (25)2.89%1.16 ×
Free square (20)2.82%1.13 ×
Square 182.82%1.13 ×
Square 192.81%1.12 ×
Station 1 (5)2.80%1.12 ×

Jail leads by a wide margin, at 11.53%, because three things send tokens there: the Go to jail square, a card in each deck and a third double. Square 24 follows partly because one card in the A deck sends the token straight to it. The hottest ordinary squares sit at 2.8% to 3.0%, which is 12% to 20% above a flat board: noticeable, but nothing like jail.

The coldest squares with every rule on
SquareShare of landingsAgainst a flat 2.50%
Go to jail (30)0.00%0.00 ×
Card square A1 (7)0.82%0.33 ×
Card square A3 (36)0.82%0.33 ×
Card square A2 (22)1.04%0.42 ×
Card square B1 (2)1.78%0.71 ×
Square 12.01%0.80 ×
Square 32.04%0.82 ×
Tax square 2 (38)2.06%0.82 ×

The coldest squares are the card squares, because the landing is counted where the token ends up and most A cards move it on, and Go to jail, which is never ended on. The cold stretch in squares 1 to 9 is milder, between 2.01% and 2.19% for the property squares there.

Which rule matters most

How much of the game is spent on the jail square
Go to jail squareCardsThree doublesWait for doublesPay straight away
OnOnOn11.53%6.22%
OnOnOff10.80%5.90%
OnOffOn9.53%5.33%
OnOffOff8.76%5.00%
OffOnOn5.70%3.78%
OffOnOff4.83%3.44%
OffOffOn3.42%2.85%
OffOffOff2.50%2.50%

With every rule off, the board is exactly flat at 2.50% on every square, which the page confirms and which is a useful check. Taking away only the Go to jail square cuts the jail share from 11.53% to 5.70%, the biggest single change, and removing the cards cuts it to 9.53%. The three-doubles rule is the smallest of the three.

Groups and the jail rule

Groups of squares, ranked by the chance of landing in one turn
GroupSquaresSum of square sharesChance in one turnLandings per lap
Stations5, 15, 25, 3510.63%12.2%0.65
Orange16, 18, 198.31%9.6%0.51
Red21, 23, 248.18%9.5%0.50
Yellow26, 27, 297.50%8.7%0.46
Green31, 32, 347.36%8.5%0.45
Pink11, 13, 147.16%8.3%0.44
Light blue6, 8, 96.50%7.5%0.40
Utilities12, 285.27%6.1%0.32
Dark blue37, 394.55%5.3%0.28
Brown1, 34.05%4.7%0.25

The group chart answers a different question from the squares. The orange squares sum to 8.31% of rolls, but the chance of landing on at least one of them in a turn is 9.6%, because doubles give extra rolls: there are 1.166 rolls per turn on average.

The jail rule, with every other rule on
MeasureWait for doublesPay straight away
Jail / just visiting, total11.53%6.22%
Of which sitting in jail9.39%3.95%
Of which just visiting2.14%2.27%
Rolls per turn1.1661.187
Turns per lap of the board5.244.86
Orange squares, chance in a turn9.6%10.4%

Paying straight out of jail roughly halves the time spent there, from 9.39% to 3.95% of landings, and lifts the orange group from 9.6% a turn to 10.4%, because tokens lose fewer turns in jail and so make more trips round the board.

Six mistakes when reading the board odds

  • Reading a percentage as the chance of landing there next roll. It is a long-run share of all landings. The chance on any particular roll depends on where the token is now.
  • Treating the jail square as mostly visitors. Of the 11.53% on that square, 9.39% is time spent inside and only 2.14% is a token passing through.
  • Adding up a group's squares and calling it the chance per turn. Orange squares sum to 8.31%, but the chance of a hit in one turn is 9.6%, because doubles give extra rolls.
  • Expecting the card squares to be busy. They are the coldest on the board, as low as 0.82%, because the page counts a landing where the token ends up, and most cards move it on.
  • Expecting one short game to match the table. The figures are exact for a very long game under these rules. A single game is noisy, and the simulation strip on the page shows the counts settling towards the exact figure only gradually.
  • Using the figures under different house rules. Change the jail rule and jail falls from 11.53% to 6.22%. Change the cards and the hot squares move. Switch to the rules your table plays before trusting the heatmap.

Assumptions & limitations: what the model leaves out

  • Card decks are drawn with replacement. Real decks are drawn in order until reshuffled, and some cards are kept by players, which changes the short-run odds slightly.
  • A token on a card square is only moved by the movement cards. Cards that charge or pay money are treated as "stay where you are".
  • Jail strategy is one fixed rule for the whole game. Real players change strategy as the game develops, and the board may have rules about using a card to leave jail.
  • Nothing about money, ownership or other players affects movement here, which matches the real game: where you land depends only on the dice and the cards.
  • The numbers are long-run averages. One game is a few dozen turns per player, far too few to show these percentages reliably.

How to use this simulator

  • Click or tap a square on the ring, or select it from the table, to see its exact chance, its rank and how it compares with a flat board. With the ring focused, the arrow keys step round the board.
  • Change the jail strategy and the three rule switches; the exact answer updates at once and any running simulation restarts with the new rules.
  • Pick a simulation length and press Simulate. Start small to see how noisy 1,000 turns is, then try a million.
  • Copy link keeps every setting in the address, and Save picture downloads the heatmap with your settings and the top squares printed underneath.

Frequently asked questions

Which square is landed on most often?

Jail, by a long way. With the default rules the combined jail square takes 11.53% of all landings, more than four times a flat board's 2.50%. That figure counts every turn spent in jail as well as passing visits: 9.39% is time in jail and 2.14% is just visiting. The most landed-on ordinary square is square 24, a red property, at 3.00%, followed by Start at 2.92% and square 25, the third station, at 2.89%.

Why are some squares hotter than others if the dice are fair?

On a plain loop with two dice every square would come up exactly 2.5% of the time. Three rules break that symmetry. Being sent to jail from square 30, from a card or after three doubles moves many tokens to square 10 at once. Cards then pile tokens onto a few target squares. And the squares a dice roll beyond jail, especially 16 to 20, collect the next landing after each visit there.

Does paying to leave jail straight away change the answer?

It changes the jail square a lot and the rest of the board only a little. Paying at once cuts the combined jail square from 11.53% to 6.22% of landings, because you no longer sit in jail for up to three turns. The order of the hottest ordinary squares stays almost the same: square 24, Start and the stations still lead.

How many turns does the simulation need to match the exact answer?

The simulation gets close quickly and then improves slowly. Its error shrinks roughly with the square root of the number of turns, so ten times as many turns gives about three times the accuracy. At 1,000 turns individual squares can be a whole percentage point out. At 1,000,000 turns every square here lands within about 0.05 percentage points of the exact value.

Does this tell me which properties to buy?

Only partly. It tells you how often each square is landed on over a long game. A property's value also depends on what it costs, what it earns and how many opponents are moving around the board, none of which is modelled here. Treat the heatmap as one input, not a strategy.

Related calculators

Sources and references

The method used here, a transition matrix and its stationary distribution, comes from the following.

Last updated
Was this result right?
What next?
A probability model, not a strategy guide

These are long-run landing frequencies for the rules stated above. A single game is too short to follow them closely, and they say nothing about prices or earnings. This page is not affiliated with any board game publisher.