35,000+ generation methuselah

For general discussion about Conway's Game of Life.
Post Reply
User avatar
Lewis
Posts: 320
Joined: March 17th, 2009, 5:26 pm
Location: UK
Contact:

35,000+ generation methuselah

Post by Lewis » March 28th, 2010, 5:51 am

Before Andrzej Okrasinski's website closed, the following methuselah was added, but only listed for 'largest final population' and not its lifespan.
It runs for ~35200 generations and doesn't appear to be listed anywhere on the internet. According to my backup version of the site, it was discovered on September 4th 2008 and only noted for its high final population.

Code: Select all

x = 20, y = 20, rule = B3/S23
4o2b2o3b2ob2o3bo$2obo2b2ob3obobo2b2o$bob2o4bo5b2ob2o$bo4bo4bo4bob2o$o
5bobobo4bobo$b3obo2b5obobob2o$ob2o2b4obob2ob4o$2bo6b2obobobobo$b4o4bo
4b2obobo$b2o6b3o3b3o$bo3bo2bo6b5o$3o3bo2bo4bob3o$b2obobob2o4b4o$obo2b
2obo2bobob4o$2o4bobob2obo2bobo$b2o3bobob7o$o3bobo6b3o2b2o$2b2o2bo2bobo
b2obo2bo$3bobobobobo4b2o$2o2b2obobo2bo!
I don't know if it can be reduced in terms of cells, like most other methuselahs have.

User avatar
calcyman
Posts: 2096
Joined: June 1st, 2009, 4:32 pm

Re: 35,000+ generation methuselah

Post by calcyman » March 28th, 2010, 8:08 am

It can be reduced to 44 cells, at least:

Code: Select all

........**............
.......**.............
......................
......................
................*.....
................*.....
................*.....
......................
......*.*.............
.....*...*..*.......*.
......*...**.*.....*.*
......*............**.
.......*....*.*.......
........*...*.........
..........**..........
............***.......
......................
............*..*......
............*.*.......
.............*........
......................
......................
......................
......................
......................
................*.....
***............**.....
...............*......
A dedicated backtracking program might be able to reduce it further.
What do you do with ill crystallographers? Take them to the mono-clinic!

User avatar
dvgrn
Moderator
Posts: 5888
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: 35,000+ generation methuselah

Post by dvgrn » March 28th, 2010, 11:37 pm

Wow! What a thing to have escaped notice all this time! At the rate the Online Soup Search is going, it won't run into a methuselah that good until it's tested ten or twenty times as many candidate patterns as it has processed so far.
calcyman wrote:It can be reduced to 44 cells, at least: ...
The original methuselah stabilizes at 35,201 ticks, and Calcyman's version has the side effect of cutting that down to 35,154.

34 cells in a 27x27 bounding box can produce the same final result -- the population alternates between 5818 and 5823, which is one glider less than the original 20x20 produced, but it doesn't seem to make sense to add that glider back in! The best time I can manage is 35,180 ticks, and I have to be pretty tricky to get that much:

Code: Select all

#C 35,180-tick methuselah with 34 ON cells
x = 27, y = 27, rule = B3/S23
7b2o$8b2o3$16bo$16bo$11bo4bo$5b2o4b2o$4b2o$5b2o$6bo2$21b2o$12b2o7bo$
12bo2$25b2o$25bo3$10b2o$9bobo$10bo4$3o!
Or here's another 3 cells removed at the cost of 20 ticks:

Code: Select all

#C 35,160-tick methuselah with 31 ON cells
x = 31, y = 27, rule = B3/S23
11b2o$12b2o3$20bo$20bo$2bo17bo$obo$3o$o3$25b2o$16b2o7bo$16bo2$29b2o$
29bo3$14b2o$13bobo$14bo4$4b3o!
It would be nice to have an methuselah scoring system so I can decide which one of these "wins". But it's been thirty years now, and I think no two people have really quite agreed on a definition yet that can unequivocally choose a winner between all these variations on survival time, bounding box, and starting population.

Sokwe
Moderator
Posts: 1480
Joined: July 9th, 2009, 2:44 pm

Re: 35,000+ generation methuselah

Post by Sokwe » March 29th, 2010, 12:55 am

dvgrn wrote:34 cells in a 27x27 bounding box can produce the same final result
That can be reduced to 33 cells in a 26x28 bounding box, lasting for an extra generation:

Code: Select all

#C 35,181-tick methuselah with 33 ON cells
x = 26, y = 28, rule = B3/S23
6b2o$7b2o4$14b3o$5bo4bo$4bo5b2o$4bo$4b3o3$20b2o$11b2o7bo$11bo2$24b2o$
24bo3$9b2o$8bobo$9bo3$o$o$o!
Edit: The 31-cell methuselah can be reduced to 30 cells, adding an extra generation to its lifespan (at the cost of a slightly increased bounding box):

Code: Select all

#C 35,161-tick methuselah with 30 ON cells
x = 32, y = 28, rule = B3/S23
12b2o$13b2o4$20b3o$4bo$2bobo$2o$2bo3$26b2o$17b2o7bo$17bo2$30b2o$30bo3$
15b2o$14bobo$15bo3$6bo$6bo$6bo!
The following 7-cell Herschel grandparent can be used to create a 31-cell 35,162-tick methuselah:

Code: Select all

.......
....**.
.*.*...
.*.....
.*.*...
.......
Is there a 6-cell pattern that takes longer than one generation to become a Herschel?
-Matthias Merzenich

Post Reply