Movement Analysis

Origins

Movement Analysis draws on RoboWiki's Movement Weaknesses page, developed and documented by the RoboWiki community as a catalog of the habits that let a gun read a bot's movement instead of just reacting to it.

A bot's movement can look busy and still be predictable in a way that a gun can exploit. RoboWiki's Movement Weaknesses page catalogs the habits that give a bot away. Read it in reverse and it becomes a checklist for judging any bot's movement, including a bot's own, before that bot ever meets an opponent.

Six habits that give a bot away

A bot that switches direction on a fixed timer rather than watching the enemy's energy drop dodges one bullet power well and every other power badly. One that ignores how far a bullet has to travel gets caught by precise or circular prediction. One that repeats the same shape, the extreme case covered on the Pattern & Enemy Dodging Movement page, hands a pattern matcher exactly what it needs.

Three more habits are about how much the bot actually moves. Fewer than 30 units of travel between shots leaves a bot close enough to a stationary target that Head-On Targeting at a 45-degree angle can land hits, which is exactly why Distancing matters. Moving mostly in one direction gives an opponent's gun roughly 50% more chances to hit, per RoboWiki, since it only has to solve for one side. Orbiting an enemy without ever closing in or backing off invites a corner or a wall, and also invites Ramming & Mirror Movement from an opponent that notices the orbit never breaks.

Weak movement repeats one short hop. Strong movement varies both the distance traveled and the direction it travels in.

Reading the list from the other side

The same six items work as a live cue, not only a design flaw. RoboWiki points to rambots like SledgeHammer and Tide, which lean on exactly this: an opponent running low energy and short hops draws ramming attacks that land above 90% of the time while collecting the ramming bonus on top. A bot that notices a short-hop, low-energy opponent has a real reason to close in rather than keep firing from range.

Name the cost

The checklist gives a snapshot, not a running score. A bot can clear one item on the list, say movement distance, while still failing another, say fixed timing. A real segmentation or statistical gun still earns its keep by adapting shot to shot, something a static checklist cannot do. The checklist behaves the same on both platforms, since it describes behavior rather than any platform-specific API.

Further Reading

• Movement Weaknesses - RoboWiki (classic Robocode)
• Ramming Movement - RoboWiki (classic Robocode)