Meet Chessity’s teaching bots: your one-on-one chess tutor

One-to-one tutoring has long been thought the most effective approach to chess teaching, or any teaching for that matter. In a traditional classroom or group setting, however, this is simply not feasible. That’s where artificial intelligence comes to your aid. We used it to create teaching bots that serve as intelligent tutor systems. Chessity’s teaching bots help you in your chess learning by providing one-to-one tutoring.

Chessity’s bots in action: scaffolding to the max

Time to show you how it works. Let’s take a lesson from the Knight Level as our example: #24 Checkmate by clearance.

This is one of the puzzles:

The aim is to checkmate the black king by moving one piece to check the king and in moving it, clearing the way for a piece behind it to take away flight squares next to the king.

Once you have located the black king, you decide to move your white rook. Suppose you move it to c8 (check). Look what happens:

1. You hear a disapproving sound, indicating that your move is incorrect. Simultaneously, a yellow arrow and a startled purple emoticon appear to show you why it is wrong: the black rook can take.
2. Next, your move is undone and you can have another try at this puzzle. But before you do, the educational engine points out what you need to consider: a green arrow from White's queen all the way to the c8-square, indicating that the king has no flight squares on the c-file.

Two different things are at work here: direct feedback and scaffolding.

Number 1 offers feedback in various ways. It tells you that you made a mistake, and lets you know why your answer is wrong. The feedback is offered immediately, which makes it highly effective.

Number 2 offers scaffolding: you get a little bit of support to explain the essentials of the lessons. This helps you to find the correct answer, without giving it away.

Now, you may give the chess puzzle another try. With the feedback and the pointers, you can probably figure out that the correct move is rook takes the pawn on b5:

Well done, you have checkmated the king and solved the puzzle. It earns you a green check mark. Yippee!

But let's have a closer look at the row of check marks:

The first two have a white circle surrounding the green dot, whereas the third and fourth check mark are circled in red. This may seem insignificant details, but they are not. The white circle indicates that the puzzle was solved in one attempt. The red circles indicate that you have solved the puzzle, but needed a second (or sometimes even a third) attempt. Therefore, the answer is worth only 50 points (instead of 100), which is not enough to earn you three stars in the end.

If you don't succeed in solving the puzzle in three attempts, you are allowed to give up. You now get the choice whether you want to continue with a new puzzle or rather want to see the solution of this puzzle:

This feature adding an extra opportunity to process the feedback.

What you have seen so far is what you can see at the frontend of Chessity: the system provides timely guidance, feedback, and explanations to the learner. Artificial intelligence offers a way to experiment and learn in a relatively judgment-free environment. You can deal with trial-and-error without being put on the spot in front of your peers or a teacher. It also offers a way to get direct feedback and support when no teacher is at hand, which makes it ideal for self-tuition.

The most advanced way to learn chess

At the backend, there's other exciting stuff going on. This is where Chessity's teaching bots analyze your learning behavior, measure your chess level and adapt the level of difficulty correspondingly:

• if you solve a puzzle in one attempt, the next one will be more difficult. 
• if you need more than one attempt, the next puzzle will be easier

Sophisticated algorithms determine the exact level to fit your personal learning needs. This means that you will always learn at the level of difficulty that is most appropriate for you, following a gradient scale of difficulty.

Chessity has used machine learning to make lessons personalized since the start. We are convinced that artificial intelligence is a game changer in education. As argued by Woolf, et al., (2013), artificial intelligence can contribute to essential educational goals, including:

• Virtual mentors for every learner.
• Addressing 21st century skills like self-direction and self-assessment.
• Analysis of interaction data, to better understand students and the settings in which they learn.
• Provide opportunities for global classrooms.
• Lifelong and lifewide technologies, taking learning outside of the classroom and into the learner’s life outside of school.

Chessity's teachings bots combined with the positive effects of gamified learning make Chessity the most advanced way to learn chess. 

Wrap-up:

All chess lessons from Pawn to Queen Level are structured in the same way. You need to solve five puzzles in a row, all in one attempt, in order to earn three stars. Chessity's intelligent tutor system understands and reacts to your solutions: 

• Direct feedback, more scaffolding and the chance to try again.
• No immediate red crosses to underline your failures, but a green check to celebrate success (even though it is red circled). After three failed attempts you are allowed to give up and continue with a new puzzle.
• Level of difficulty is adapted, so that you get the content that is most appropriate for you.

Any thoughts or questions? Please leave a comment below! Your feedback will help us to improve.

Update: May 5, 2017