We've been working on figure learning for a little while now. The testing over the past few months has pushed us to make even more radical decisions regarding gameplay. Movements that constantly caused friction in the learning process and led to disruptive actions in gameplay have been removed. We've decided to focus 100% on rotations, especially since the tests carried out in free-running mode were not conclusive.

The new features in this version are:

Ne-Mov – a robotic coach assistant that follows the player and provides three types of support: figure reading assistance, figure input guidance, and figure visualization.

Training Mode Chapter 1 – the first chapter of the training mode, designed to teach tricks step by step in a logical sequence.

We've also spent time readjusting a number of key details:

• Dead Zone – tweaked for better feel during small diagonal movements when performing half-turns.
• Flip Snapping – the character's position is now reset along the transverse axis after each jump, preventing cumulative rotation from one flip to the next, which previously led to landing on the stomach or back.
• Default Jump – the height of a normal and a powerful jump is now set to the equivalent of 1 and 2 flips respectively, if the directional stick input is held fully throughout the jump. This helps reduce the risk of falling.

The goal of this version is to provide tools that help players take control of the character and better approach trick learning. We're now working on the next chapters of the training mode, as well as other game modes that will allow players to challenge themselves, compete with others, and win cool stuff.