Stratego, new victim of artificial intelligence

DeepNash, an artificial intelligence has learned to play, and to win a board game, the Stratego.

Stratego VS artificial intelligence

A few days ago, we revealed to you that an artificial intelligence had managed to win at Diplomacy, a game of bluffing and negotiation, against human beings. DeepNash, an artificial intelligence (AI), has just beaten another board game, Stratego.

DeepNash, the latest artificial intelligence from researchers at DeepMind Technologies Ltd., has just developed a few days ago an AI capable of learning, and winning against the Stratego, that old classic board game. Described as more complex than chess and go, and trickier than poker, with a hint of bluffDeepNash now masters Stratego.

DeepMind announced that its artificial intelligence DeepNash has learned, from zerohow to play Stratego. Considered difficult to master, the AI ​​learned game strategies by playing against itself 5.5 billion times. DeepNash went on to beat almost every human female player in the game.

The history of Stratego

The Stratego, you have certainly already heard of it. You’ve probably played it as a kid. Stratego is a two-player strategy board game.

Each person has 40 pieces representing individual soldiers in their army. The goal is to capture the opponent’s flag or capture enough enemy pawns to block their ability to make another move. Opponents cannot see the ranks of opponent’s pawns, making this a limited information game. I know, I see, I discover. But I don’t know everything.

Stratego was created by Mogendorff during World War II. It was registered as a trademark in 1942 by the Dutch company Van Perlestein & Roeper Bosch NV.

In 1958 the license was granted to Hausemann an Hotte. The first version of Stratego was distributed by Smeets and Schippers in 1946. In 1961 the game was sublicensed to Milton Bradley, the game publisher which was later acquired by Hasbro in 1984. Stratego was and first published in 1961 in the United States.

Chinese ancestors

The origins of Stratego date back to the traditional Chinese board game Junglealso known as “Game of Fighting Animals” (Dou Shou Qi) or “the game of animal fighting”.

The game Jungle also has counters, but of animals rather than soldiers, with different ranks and higher rank counters capture lower rank counters. The plateau, with two lakes in the middle, also resembles that of Stratego.

The main differences between the two games are that in Jungle, the pawns are not hidden from the adversary and the initial configuration is fixed. Dou Shou Qi is very similar to chess.

1670309353 281 Stratego new victim of artificial intelligence

A modern and more elaborate Chinese game known as “land battle game” (Luzhanqi) is a descendant of Jungleand a cousin of Stratego : the initial configuration is not fixed, the two players keep their pieces hidden from their opponent. The objective is to capture the enemy’s flag.

The basic mechanics of the Luzhanqi are similar, although differences include “missile” counters and a board layout resembling the XiangqiChinese chess with the addition of railroads and defensive “camps”.

A third person also acts as a neutral arbiter to decide battles between the pawns, without revealing their identity.

European ancestors

In its current form, Stratego appeared in Europe before World War I as a game called L’Attaque. The Attack very probably developed at the beginning of the XXe century or perhaps as early as 1880 in France. A patent issued by the French Patent Office (Patent No. 396.795) has been discovered. [10]) dropping by Hermance Edan in 1909 based on a game she developed in the 1880s.

Hermance Edan had not given a name to his game but as early as 1910, a French manufacturer was selling the game under the name L’Attaque.

A patent on another game, called war game, was registered in 1907 by Julie Moller for a similar game. The ancestor of Stratego was therefore created by women.

After World War I, the game was also released asattack by the London firm HP Gibson & Sons, Ltd.

The 1910 version of the game divided the armies into red and blue colors. The rules for L’Attaque were basically the same as those of the game we know today as Stratego.

It featured rectangular pieces of standing cardboard, printed in color with soldiers who wore contemporary uniforms (until 1900), not Napoleonic uniforms.

The attack game is played with 36 pieces each on a 9×10 square board. Each piece has a numerical value, which is hidden from the opponent and only revealed when an attacking piece moves onto a square occupied by an opposing piece. In most cases, the piece with the highest value wins and the losing piece is removed from the board.

Play continues until one player finds the opponent’s flag piece and takes it.

Classic Stratego

The modern game of Strategowith its Napoleonic imagery, was originally manufactured in the Netherlands by Jumbo, and was licensed by the Milton Bradley Company for American distribution, and introduced in the United States in 1961.

The pawns were originally made of printed cardboard. After World War II, painted wooden counters became standard, but from the late 1960s all versions had plastic parts.

The switch from wood to plastic was done for economic reasons, as was the case with many products at that time. But with Stratego, the change also had a structural function: unlike the wooden pawns, the plastic pawns were designed with a small base. The wooden pawns did not have one, which often caused the pieces to tip over and fall.

This posed a real concern for the person who revealed his rank in this way. And who risked bringing down other pieces, in domino mode.

1670309353 400 Stratego new victim of artificial intelligence

DeepNash explodes Stratego

Back to our sheep, artificial.

Artificial intelligence already masters chess, Go, poker, Diplomacy. But the game Stratego is much more complex. According to calculations, Go presents ten to the 360 ​​power of possibilities of movements, much more than chess or poker. But Stratego has it ten to the 535th power. So… a lot!

With DeepNash, this new advancement in artificial intelligence surpasses and exceeds the game. This technological development will now allow AIs to maneuver real-world scenarios with limited information. DeepNash could then be used in uncertain situations to develop balanced results for complex problems. We don’t know everything. But the AI ​​is able to scramble to find a solution.

Stratego is a bluff game but also involves strategy to play and win. Each player must place the 40 pawns that are hidden from her opponent at the start of the game. This is where the strategy comes into play (it is the case to say it). Each person must visualize all possible outcomes when setting up the board and play in turn with limited information. We see the opposing pawns. We do not know their rank, their “secret”.

As in poker, the ability to bluff is vital in Stratego. What artificial intelligence had to learn. But unlike poker, a game of Stratego lasts much longer than a hand in poker, sometimes with hundreds of moves before a person is declared the winner.

A surprising factor in this discovery is that the artificial intelligence has developed an “unpredictable strategy”. This allowed the AI ​​to guess its human opponent, which is essential in this game. Vincent de Boer, the co-author of a published article in the journal Science this 1er December 2022, said he had never heard of an AI from Stratego able to win against a human opponent.

DeepMind reported that its artificial intelligence, DeepNash, has become so advanced in gaming Stratego that she ranked in the TOP 3 among human experts on gravel, the platform to play Stratego online. The AI ​​played 50 games, earning a record 84% win.

Diplomacy. Now Stratego. What’s the next AI-exploded board game?

👉 To know how DeepNash works in detail, everything is explained here.

Article written by Gus. Editor-in-chief of Gus&Co. Teaches atGraduate School of Comics and Illustrationhas worked in the gaming world since 1989 as an author and journalist.

Your reaction to the article?

We want to say thanks to the author of this short article for this remarkable content

Stratego, new victim of artificial intelligence

We have our social media pages here and other pages related to them here.