Corner Pocket

How the table actually plays

A physics loop, not a pot percentage

The table is 960 by 520 units with a playing surface inset from the frame and six pockets, the two middle ones sitting fractionally outside the long rails as they do on a real table. Balls have a radius of 11 units against a pocket radius of 24, so a pocket is a little over twice a ball wide — generous on a straight approach, unforgiving on a thin cut.

Every shot is simulated in fixed steps of one one-hundred-and-twentieth of a second, so the same shot produces the same result regardless of your frame rate. Balls lose speed to cloth friction continuously and stop once they drop below a small threshold rather than sliding forever.

Collisions are not all equal. Ball-on-ball contact returns almost all of the energy, while the cushions return substantially less, which is why a two-cushion escape leaves the cue ball much shorter than players expect on their first few attempts.

Shape matters more than potting

Because the cushions absorb far more energy than the balls do, the reliable way to move the cue ball across the table is to send it through traffic or along a rail at pace, not to bounce it repeatedly. Two cushions is a plan. Four is a hope.

The pocket-to-ball ratio rewards straightness disproportionately. A ball approaching the jaws square has more than a full ball-width of margin; the same ball cut thin has almost none, because the effective target narrows with the angle. When you are choosing between two potting options, take the straighter one even if it leaves slightly worse position — the miss you avoid is worth more than the angle you gain.

Against the computer, fouls are the currency. Ball in hand lets a competent opponent place the cue ball wherever it likes, and the two stronger opponents will convert that into a multi-ball run. A safety that leaves them a legal but awkward shot is almost always better than a low-percentage pot that risks handing over free placement.

The three opponents, and what separates them

The computer players are not the same routine with a difficulty number attached. They differ in how many shots they consider, how much error is applied to the shot they choose, whether they simulate the outcome before committing, and whether they will play a deliberate safety at all.

The weakest opponent never plays safeties — it only ever attempts to pot. That is the single biggest behavioural difference and the reason the step up to the middle opponent feels much larger than the accuracy numbers alone suggest.

Common mistakes

Four habits account for most lost frames, and three of them come from misjudging how energy moves around the table.

• Over-hitting the break. Maximum power scatters the rack but leaves the cue ball uncontrolled; a firm break with a plan for where the cue ball finishes wins more frames.
• Planning long multi-cushion escapes. Each cushion contact costs a large share of the remaining speed, so the fourth rail is usually reached at a crawl, if at all.
• Treating a thin cut as a normal pot. The usable pocket width collapses as the cut angle grows, which is why thin cuts along a rail miss far more often than they look like they should.
• Giving up ball in hand to attempt a difficult pot. Both stronger opponents plan with the cue ball placed exactly where they want it, and that is where their long clearances come from.

What is different about this version

There is no hidden pot roll. Whether a ball drops is decided entirely by where it travels, so a shot you aimed well goes in and a shot you did not does not — the outcome is reproducible from the same starting positions and the same strike.

This is a solo table with computer opponents. There are no live players, no rooms, no chat, and nothing to wager. Match progress and any equipped cosmetics use the existing Table Game profile, and the physics run entirely in your browser rather than on a server.

Written by the Silverweb Games team from this game’s own implementation. Last reviewed July 31, 2026.