Going first in checkers matters much less than most players think. With perfect play it doesn't matter at all: checkers was solved in 2007, and the result is a draw. In real games it's a small edge at most. Between two copies of our computer player at the same level, the first mover scored 49.4% at Easy, 55.7% at Medium and 50.6% to 53.0% at Hard, where 50% is an even game. One level of playing strength was worth about 45 points; going first, about 6 at most.
Every number below comes from seeded games between the computer players behind Poozee's Checkers, 36,814 games in all, plus an exact search of every position with one man a side. A draw counts as half a win for each player. If you're new to the game, start with how to play checkers.
With perfect play, going first doesn't matter
In 2007, Jonathan Schaeffer and his colleagues published Checkers Is Solved in Science. Their computers had worked on the problem almost continuously since 1989, and the paper's abstract states the result plainly: "Perfect play by both sides leads to a draw."
So the first move isn't a winning advantage. A player who moves first can't force a win, and a player who moves second can always hold the draw, as long as neither of them makes a mistake. Every edge you'll see below comes from mistakes: the question is whether going first makes the other player's mistakes more likely, or more costly.
How much going first is worth in real games
We had two copies of the computer player at the same level play each other. Easy looks one move ahead with a lot of noise, Medium looks 4 plies ahead (two moves by each side) and picks at random among its best moves, and Hard searches much deeper. In the game, Hard thinks for 0.8 seconds a move; for these tests we held it to exactly 8 plies so that every game replays exactly.
Easy and Medium make random choices, so their games from the start all differ. Hard at a fixed depth plays almost the same game every time from the same position, so we started its games from every opening four plies long instead: all 1,469 ways the first two moves by each side can go, one game each. We played Medium from the same openings too, four games each, for a fair comparison. In the table, "from openings" means those 1,469 starts, and the score and won-drew-lost are counted for the first mover.
| Both players | Score | Won-drew-lost |
|---|---|---|
| Easy from the start | 49.4% | 5,775‑8,226‑5,999 |
| Medium from the start | 55.7% | 1,656‑1,142‑1,202 |
| Medium from openings | 51.5% | 2,327‑1,400‑2,149 |
| Hard at 8 plies from openings | 50.6% | 445‑597‑427 |
| Hard at 10 plies from openings | 53.0% | 533‑492‑444 |
Half of the games from the start had Black moving first and half had Red (more on that below). What the numbers say:
- At Easy, going first made no difference. Over 20,000 games, the first mover scored 49.4%, give or take half a point (a 95% margin of error).
- At Medium, going first was worth about 6 points from the start: 55.7%, give or take 1.3 points, with 1,656 wins to 1,202 losses.
- From varied openings, the edge shrank to 3 points or less. Medium's first mover scored 51.5%, and Hard's 50.6% searching 8 plies and 53.0% searching 10. Hard's margin of error is about 2 points either way, so at 8 plies its result is too close to call, and at 10 plies it's a small edge, about half of Medium's from the start.
Why does Black win more in checkers?
Because Black moves first. The 500 Medium games in our rules guide found Black winning 221 games to Red's 143 between Medium computers, which raises the question of whether that's about Black or about the first move.
So we swapped. The starting position looks the same from both sides of the board, so we let Red make the first move in 2,000 more Medium games. Red, moving first, scored 55.2%, against 56.1% for Black when Black moved first. At Easy, the first mover scored 49.3% as Black and 49.6% as Red. The color doesn't matter; the first move does, a little, at Medium.
The color that moves first depends on the set. On Poozee, as in most home sets, Black (the darker side) moves first. Under tournament rules, the pieces are red and white and Red moves first, so in a tournament game "Red wins more" would be the same effect.
Where Medium's edge comes from: the second player's first reply
If moving first were an advantage in itself, it would show up whatever the opening. It doesn't. Our checkers openings guide played 1,000 Medium games from each of the 49 two-move openings (Black's first move and Red's reply). With every opening counted equally, Black scored 49.7%. When Medium chose Red's reply itself, Black scored 56.3%. (Black's first move was spread evenly over all seven in both tests, 2,000 games each when Medium picked the reply.)
That points at Medium's reply as Red, not at Black's first move. Medium only sees 4 plies ahead, and a few of its favorite replies turn out badly for Red later. The worst is its answer to b6-c5 (9‑14 in book numbers): it always replies e3-d4, after which Black scored 70.7%. Spread the openings out further, over all 1,469 four-ply openings, and Black's score was 51.5%. Moving first gives your opponent the first chance to set you a problem, but the starting position itself doesn't hand them anything.
Going first or being the better player?
To see how the first move compares with skill, we matched each level against the next one up, 1,000 games each way from the start, with the stronger player moving first in one match and second in the other. Hard again searched exactly 8 plies.
| Match | Stronger player moves first | Stronger player moves second |
|---|---|---|
| Medium against Easy | Medium scores 96.2% | Medium scores 96.5% |
| Hard against Medium | Hard scores 96.5% | Hard scores 92.4% |
- Going first doesn't rescue the weaker player. Moving first against Hard, Medium won 24 games, drew 104 and lost 872. Moving first against Medium, Easy won 10 of 1,000.
- One level of strength was worth about 45 points above an even game: Medium took 96.3% of the points against Easy over both matches, and Hard 94.5% against Medium. Going first was worth at most about 6 points, in the equal Medium games.
- The first move helped the weaker side a little against Hard, raising Medium's score from 3.5% to 7.6%. Against Easy it made no real difference.
So if you lose most of your games against someone, moving first won't change that. Getting better will.
When having the move hurts
Having the move sounds like pure advantage, but in checkers it can lose, because there's no passing: if every move you have walks into a jump, you still have to make one. To see how often that happens, we solved every position with one man on each side exactly, with each side to move: 760 positions in all. The search follows each game to the end, through any kings the men become.
- In 548 of them, who moves makes no difference: both ways, it's a draw.
- In 148, having the move is better.
- In 64, having the move is worse, and in 30 of those, whoever has to move loses.
All 30 of those losing-to-move positions have the two men on the same column, facing each other:
If Black is to move, d6-c5 is jumped by d4xb6 and d6-e5 by d4xf6, and Black has nothing left. If Red is to move, the same thing happens to Red. With more pieces on the board the counting gets harder, but the idea is the same: near the end of a game, the player who runs out of safe waiting moves first has to give way. Who that is depends on every move played before, not on who moved first.
What to do with this
- Don't worry about the color you get. With perfect play it's a draw, and in our games going first was worth 6 points at most, a fraction of what one level of skill is worth.
- Switch colors every game. Tournament rules alternate colors from game to game, as our rules guide explains, and Poozee's computer games do it for you, so any small edge evens out over a match.
- When you move second, take your first reply seriously. Medium's edge as the first mover came from Red's early replies. Our openings guide lists Red's best and worst answer to each first move.
- In the endgame, count who has to move. When two men face each other with one empty square between them and no other piece nearby, the one that has to step forward gets jumped. Keep a spare waiting move, a man or king you can move without giving anything away, for that moment.
Pro tip: Before you trade down into an ending, picture the position after the trade and ask whose move it will be. Having the move is usually good, but in 64 of the 212 one-man-each positions where it mattered, being the one to move was worse.
Keep playing
- Play Checkers against the computer, which switches sides with you every game
- How to win at checkers: back-row defense, forced-jump shots and two kings against one
- Checkers opening moves for beginners: every first move and Red's best replies
- Checkers shots: the two-for-one tricks that decide most games
- How long does a game of checkers last?: moves per game at each level, from many of the same computer games
Frequently asked questions
Is checkers a draw with perfect play?
Yes. Jonathan Schaeffer and his colleagues solved checkers in 2007, with computers that had worked on the problem almost continuously since 1989, and showed that perfect play by both sides leads to a draw. Neither player can force a win.
Who wins more in checkers, Black or Red?
Whoever moves first, by a little, and only between players who make some mistakes. Between our Medium computers, the first mover scored 56.1% as Black and 55.2% as Red. At our Easy level, the first mover scored 49.4%, an even game.
Who goes first in checkers?
The darker side. On Poozee that's Black; in tournament sets, where the pieces are red and white, it's Red. Players switch colors from one game to the next.
Is it better to go first or second in checkers?
It barely matters. In our tests, moving first was worth at most about 6 points out of 100, while one level of playing strength was worth about 45. Playing better moves beats picking a color.