Thanks for the game!

[mordant23]

Thanks for the game and thank you to folks who provided hints on this
mailing list. I just finished the normal level without attempting the
easier levels. I didn't win any records with speed, but I feel like I
accomplished something. Part of the fun was figuring out the game
mechanics. It took me a while to figure out Construction CPUs, It was
interesting how well the game captured the theme of the singularity. I
was looking for a game that was challenging yet had simple mechanics
with an interesting world. Endgame: Singularity hit the spot.

[Phil Bordelon]

Always glad to hear positive feedback. Did you have any particular
sticking points in play?

P

[Charlie Nolan]

> It did help that I could google information when I got stuck. The tips and
> strategies posted in this group were helpful. Now I'm using
> http://brededor.narod.ru/text/Endgame-Singularity-IG/Endgame-Singularity-IG.html
> for suggestions on Impossible level.

Wow, that was a really strange-feeling document to read, seeing as how
I'm the one who wrote the Impossible difficulty level. I did spot a
few small errors, but otherwise it's a pretty good guide.

* The previous-discovery detection penalty can be 1.32 (1.1 * 1.2) if
a location has two bases discovered in a row.
* It's possible to build a new base during the grace period by
destroying the old one and building a new one in the same moment. You
won't gain much from this, since you have to earn 100 first, but every
bit helps.
* A lot of the discovery/suspicion math seems to forget that luck is
an important factor. Average gains and losses are useful predictors,
but you can lay low at the start successfully or get killed off by a
string of discoveries after Advanced Intrusion. I've done both,
probably in the same game. I wouldn't really say the game gets safe
until you have at least two time capsules, and maybe not even then.
* I'm not aware of any bugs with the maintenance cost displayed,
though there may be some. In any event, the exact values can be read
out of the maint_list value in data/bases.dat. The list's contents
are cash | CPU | time, with the last one always 0 as time makes no
sense for maintenance. (It's only kept for consistency with other
cost lists.)

Also, I strongly suggest beating Ultra Hard before you attempt
Impossible. Ultra Hard is basically "Impossible Lite", requiring you
to learn a lot of the same information, but without Impossible's utter
lack of pity for the player. The new difficulties came about when I
realized that our old hardest difficulty was trivial if you understood
the game, so I adjusted them with the idea that Very Easy was for
beginners, Normal for people who understood the basics, Ultra Hard for
people who knew the game in depth and wanted a challenge, and
Impossible for masochists. Ideally, I'd like to see near-perfect play
on Ultra Hard have a 90% win rate, and on Impossible have a 10% win
rate.

Hmm, there's a thought. It might not be a terrible idea to have a
"max difficulty" setting in prefs.dat. Start it with only Very Easy
unlocked, winning that unlocks up to Normal, Normal unlocks up to
Ultra Hard, and Ultra Hard unlocks Impossible. Naturally, there's
nothing preventing someone from altering the prefs to unlock it early,
and a "cheat code" that unlocks it from the menu isn't a bad idea
either. Even something as simple as "shift-click lets you select
locked difficulties" would accomplish letting veterans play whatever
they like while guiding newbies through the progression.

-FM