Strategy guides › Color Flood
How to beat par in Color Flood, nearly every time
▶ Play Color Flood🎯 Try board 7 (par 27)
Move nineteen. Your blob covers most of the board, five colours are left, and somewhere in the bottom-right corner a single purple square is hiding behind a wall of black. You will need two more moves for it whatever you do. If you had gone that way ten moves ago, it would have come along for free.
Color Flood looks like a game about grabbing as much as you can. It is really a game about distance.
What par actually is
The par on our boards is not a target someone set by hand. The game computes it when the board loads, by playing a simple greedy strategy: every move, pick the colour that makes your region biggest right now, with ties going to the colour further left on the swatch bar. Par is how many moves that strategy needs.
Greedy is a decent player but a short-sighted one, and par is easy to beat once you know why. To measure by how much, we took the first 200 daily boards (12Ă—12, six colours) and ran several strategies on each, plus a wide search that keeps the 100 most promising positions at every move. Here is what they needed on average.
| Strategy | Average moves | Beats par | Worse than par |
|---|---|---|---|
| Par (greedy) | 22.5 | – | – |
| Greedy, but always finish a colour when you can | 21.5 | 130 | 0 |
| Look two moves ahead | 20.4 | 160 | 11 |
| Head for the centre, finish colours | 20.6 | 143 | 28 |
| Shrink the farthest corner, finish colours | 19.9 | 172 | 14 |
| Wide search | 17.9 | 198 | 0 |
On 198 of 200 dailies a line shorter than par exists, and the typical saving is four moves; on three boards the search found a line ten moves shorter. The other sizes tell the same story. On 10Ă—10 boards par averages 18.8 and the search 15.3, with a shorter line on 193 of 200 boards; on 14Ă—14 it is 26.3 against 20.7, and a shorter line turned up on all 200.
Finish a colour when you can
The single cheapest improvement: if one of the colours can be wiped off the board entirely with one move, because every square of it left touches your region, play it. A colour that is gone can never cost you a move again, and greedy often leaves a few stragglers of a colour that it then has to come back for.
This rule alone never did worse than par on our 200 dailies, and it beat par on 130 of them. You can apply it without counting anything: look at the swatch bar, and for each colour ask whether any of it is still out of reach.
Count distance, not squares
The deeper idea is that the board has a shape. For any square, count how many colour changes you would cross to walk to it from your region; that is the fewest moves before it can possibly be yours. The square with the biggest number sets a hard floor on your score.
Greedy ignores that floor. It happily spends moves grabbing a fat patch next to the corner it already owns, while the far corner stays twelve changes away. The better habit is to drive a channel toward the farthest part of the board early, picking up whatever is on the way, and to sweep the near side later, when your region already touches most of it and each move swallows a lot.
The strategy “shrink the farthest corner” does exactly that: each move, pick the colour that brings the most distant square closest, and use size only to break ties. With the finish-a-colour rule on top, it beat par on 172 of 200 dailies. On board 7 it needs 18 moves, nine fewer than par, and our wide search found 17.
Don’t count the squares you touch
A tempting shortcut is to pick the colour that has the most squares along the edge of your region. We tested it, and it is worse than plain greedy: on 132 of 200 dailies it took more moves than par, and it averaged 24.3. The problem is that touching squares are only the front of each colour patch. A colour with three squares touching you might be a long snake that runs to the far corner; another with six touching squares might be a dead end. Greedy at least counts the whole patch you would swallow. Distance counts where it gets you.
When you look ahead, look at what a move opens up. A move that is small now but puts you next to a big patch of a second colour is often better than the biggest move on offer, which is why two-move look-ahead beat par on 160 of 200 boards.
Practice
Board 7 (12×12) is a good place to feel the difference: par is 27, a corner-first plan gets 18, and 17 is possible. Before your first move, find the bottom-right squares and plan a route there. Then try today’s daily and see whether you can get under par by three.
Go far first. The near side will still be there.
