Best Board Games by Age: How to Pick for Your Table

The Völuspá describes the first age of the world in one small scene. The gods are on the field of Iðavöllr, they are building, and in the evening they play at tables with golden pieces and there is no shortage of anything.

That is the golden age. Not a war, not a feast, not a hunt. A board game played by people who have not yet made any mistakes.

Amazon’s board game chart has found its way back to that morning, and it looks a lot like a preschool.

The top six

Amazon’s board game chart, read this morning:

1. Candy Land, 1949.
2. Sorry!, 1934.
3. Trouble, 1965.
4. Connect 4, 1974.
5. Connect 4 again, a 2025 edition of the same game.
6. Monopoly, 1935.

Six of six are games you can teach in under a minute, four of six require no reading, and the newest design idea in the group is fifty-one years old. Chutes and Ladders is at nineteen, Operation at sixteen, Guess Who at ten, Twister at fourteen.

This is not a chart of what people play. It is a chart of what people buy for a child’s birthday, and there are a great many child birthdays.

Rank here is an age, not a verdict

Someone here ran a game shop for four years, and the thing that shop taught fastest was that the person holding the money is almost never the person who will play the game. A best-seller chart in this category is a gift chart. Gifts get bought for children, gifts get bought by adults who are not confident about games, and both of those buyers reach for the box they recognise.

So the ranking is doing something specific and worth naming: it sorts by how young the intended player is and how little the buyer wants to think. Candy Land at one is not a claim that Candy Land is the best board game. It is a claim that three-year-olds have a lot of birthdays and their aunts have very little information.

Once you read the chart that way, the useful part becomes obvious. Do not shop the top. Shop the middle.

Where family game night actually starts

Number seven is where the chart changes character, and everything interesting on it sits between seven and eighteen.

• Catan at seven — the game that started the modern hobby.
• Hues and Cues at eight — up to ten players, no reading.
• Guess Who Plus at ten — two-player deduction.
• Tapple at twelve — fifteen-minute word game.
• Codenames at thirteen, Twister at fourteen, Wavelength at seventeen, Herd Mentality at eighteen — all party or social-deduction games from the last fifteen years.

That band is where a household actually shops, because it is the only part of the chart that distinguishes between a table of four adults, a table of two, and a table with a six-year-old at it. The top six cannot make that distinction. They are all the same answer.

Pick the table, then the game

The single most useful question is not “what is good.” It is “who is sitting down, how many of them, and for how long.”

• Two players — Guess Who at ten. Almost nothing else on this chart is designed for exactly two.
• A big mixed group, including small children — Hues and Cues at eight. Ten players and no reading is a rare combination.
• Fifteen minutes before dinner — Tapple at twelve.
• A proper evening with players over ten — Catan at seven.

Four questions, four answers, none of them in the top six.

Where these picks came from, honestly

All four read off Amazon’s own board game chart this morning, rank, rating and review count together, best-sellers first. All four are research picks as listings — we have not bought these particular boxes. The reading of the chart is not a research pick: it comes from four years behind a game shop counter, where the same pattern showed up every December.

The one honest warning

Look at the review counts in the top thirteen and you will find three numbers that do not belong: the Connect 4 2025 edition at number five has around three hundred ratings. A new game at number nine has under three hundred. The second edition of Codenames at thirteen has around two hundred and fifty.

Their neighbours have twenty thousand, thirty thousand, forty thousand.

A new edition of an old game starts its review count at zero and keeps none of the old game’s history. The chart rank is real — those boxes really are selling — but the star rating beside them is being calculated from a couple of hundred early buyers, and early buyers of a new edition are disproportionately fans. It is the thinnest possible evidence dressed in the same 4.8 as a game with forty thousand ratings.

This applies directly to a pick above. The Catan listing at number seven is the sixth edition, freshly listed, and its rating sits on a small number of reviews. We are recommending Catan on thirty years of the game existing, not on that listing’s star count — and the honest version of that sentence is worth saying out loud, because the star count is what you will actually see on the page.

When you see a low review count high on this chart, check whether you are looking at a new game or an old game in a new box. If it is a new game, that rating is a guess. If it is an old game, judge the game and ignore the number.