Pointless optimisation game
The Pointless Optimisation Game (or Pointless Optimization Game in non-British English) is a common term for a particular type of pattern optimisation. Specifically, it involves improving the bounding box area, usually of a glider gun. It often happens that small improvements in bounding box area can be made at the expense of significant increases in the population and/or complexity of the gun.
| The original form of the period-27 glider gun, which has a much smaller population of 156 cells, but a slightly larger bounding box. It is also easier to find a glider synthesis for this form. (click above to open LifeViewer) RLE: here Plaintext: here |
| The bounding-box-optimised form of the period-27 glider gun, which has a slightly smaller bounding box at the expense of a much larger population of 323 cells. This form has no known glider synthesis. (click above to open LifeViewer) RLE: here Plaintext: here |
History
The reason that the Pointless Optimisation Game exists as a concept is that the bounding box metric was chosen to determine the "smallest" guns in Dieter and Peter's gun collection from the 1990's. That convention continued to be followed in later collections of glider guns maintained by Jason Summers, Chris Cain, and Adam P. Goucher. Bounding-box-optimized guns started getting an increasing amount of attention around the year 2000, when the focus began to shift toward creating a complete collection of optimized guns for all periods between 14 and 1000. That project was completed using Hersrch in September 2003.
The slightly disparaging tone of the phrase "pointless optimisation game" is due to the fact that obsessively optimising a gun to reduce its rectangular bounding box to the lowest possible area does not necessarily produce the most generally useful gun. An edge-shooting gun might well be much more useful in larger patterns. For example, a gun with a small bounding octagon or narrow diagonal diameter may allow many copies to be packed more efficiently into a larger shotgun pattern like the ones found in bounding-box-optimized guns for constructible spaceships. Alternatively, a gun optimised for population could allow for easier reduction of patterns such as SKOPs and sawteeth.
However, the bounding box is very convenient in terms of ease of calculation, for comparing different guns to determine new record-breaking entries for gun collections. So the bounding box metric has never been supplanted by any other measurement that might be technically more appropriate but would be significantly harder to calculate, such as the bounding diamond, bounding octagon or minimum covering polyplet size.
In the current incarnation of the glider-gun collection on Catagolue, any existing gun can be replaced at any time, simply by submitting a new gun pattern that fits into a smaller bounding box area.