a multiplayer game of parenting and civilization building
You are not logged in.
Shrinking it down to a min of 25 would introduce way too much noise and fluctuation in Gene score. You can never tell what each life will give you, which is why the aggregate over time gives a better picture than a mere handful of lives.
If it were to keep a 'reecent' record only for Gene fitness, I'd prefer it taking the aggregate of at least 100, or just the way it is. It may look weird for people to see their Gene score drop as their 101st and older siblings drop off their Gene list.
Avatar by Worth
Offline
Expected and fine.
Offline
Two ideas to twerk on it.
Raise the numerical value to deplete rng factor.
Instead of recent lifes count XX "best" lifes.
I am Sheep, the lord of kraut, maker of the roads, professional constructor, master smith, bonsai enthusiast, arctic fisher, dog whisperer, naked nomad and an ORGANIZER. Nerf sharp stone it's op.
"BAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA" -Jaleiah Gilberts
"All your bases are belong to us"-xXPu55yS14y3rXx-
Offline
As far as evaluating only recent lives... hmm.... that's definitely possible.
That would prevent scores from ever getting too high. It's an interesting idea.
The fitness server currently keeps your last 25 lives and offspring in the database, and flushes the rest.
What if, to compute your current score, it starts from 30 and then adds in the effect of JUST these most recent lives and offspring? In the client, these are the ones that are shown to you..... so you could look there, and see the ENTIRE history (25 offspring and lives) that made up your current score.
I don't like this at all. This makes it so that only your last ~2 lives matter, and that just doesn't feel right.
Offline
Count those 25 recent lives but exclude the worst and the best ones. Average matters more than getting good/bad luck.
Offline
Count those 25 recent lives but exclude the worst and the best ones. Average matters more than getting good/bad luck.
+1
Instead of recent 25 your score will be defined by your 25 most generic lifes XD
Ofcourse this would bring some fast paced fluctuation to the leaderboards, but after some time the score would settle on some set value of how well your offspring generically manage to navigate world.
This would also create:
1. A ingame learning curve of a character as the new players usually start the game by dying a lot thus having low slot. After playing bit by bit they would learn more and more tool slots as their average lifespan woud raise up.
2. Prevent noobs from picking up too many random tools and destroying less everything they touch with em.
3. More value on actually being really good eve on average.
4. pay to win system, where by buying a new account after mastering a game and then trying to go for 25 really good lives. Imagine the cash money for account buyes XD
Last edited by arkajalka (2019-11-24 03:47:32)
I am Sheep, the lord of kraut, maker of the roads, professional constructor, master smith, bonsai enthusiast, arctic fisher, dog whisperer, naked nomad and an ORGANIZER. Nerf sharp stone it's op.
"BAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA" -Jaleiah Gilberts
"All your bases are belong to us"-xXPu55yS14y3rXx-
Offline
As far as evaluating only recent lives... hmm.... that's definitely possible.
That would prevent scores from ever getting too high. It's an interesting idea.
For your consideration: another idea.
Just use a regression to 30 effect. After every played game, take the difference from the current score and 30 (so let's say you have a score of 70; the difference is 40) and regress by 10%. So in the example given, before anything else happens, the score regresses to 66 (70 - (40 * 10%) ) Then add the scores as normal, whatever that ends up being.
The effect is that older lives matter, but they matter less and less as you play. The amount of the effect can be adjusted by simply adjusting that 10%. Use 5% to keep older lives relevant for more time or use 20% to make older lives contribute less.
Perhaps "/die" could use a slightly adjusted percent. So you lost 10% normally, but only 5% if you use /die. Obviously this needs some fine-tuning to find the right balance between giving people born into hopeless situations a break, but not encouraging its use too much.
Another advantage is that this is practically the definition of simple, both to understand and to implement.
Finally, this introduces a theoretical cap on how high the scores can go, but in a more natural way. At some point, the regression is going to take away as many points as any life can theoretically give you. This alone might not be enough to prevent too high of scores (I think some weight should be given to how long you were alive during any particular person's lifetime. So if you were alive 2 years for your grandson's life you should get significantly less of an effect than for your daughter, where you lived 40 years together).
Offline
25 is way too few to have much meaning. At least 250; 500 would be even better. So far I'm liking the way it is now though.
Offline