StarDestroyer.Net BBS

What I mean is, is there any periodicity to it? Because otherwise, I'm going to miss launch windows on account of I don't know when they'll happen. Can you include planetary launch windows in your turn-beginning reports?

Yeah, I can include launch windows. I will look into periodicity, stand by.

EDIT: Hmm...I don't see any periodicity. The windows appear to be rolled: you can launch towards Venus and Mercury at all times, but Mars, Saturn and Jupiter appear randomly.

EDIT2: Damn...something's come up, I might not be able to do an update until Monday.

PeZook wrote:Yeah, I can include launch windows. I will look into periodicity, stand by.

EDIT: Hmm...I don't see any periodicity. The windows appear to be rolled: you can launch towards Venus and Mercury at all times, but Mars, Saturn and Jupiter appear randomly.

Crap. Go figure. Well, as long as we know which windows are available in a given season so we can plan, it should be OK.

EDIT2: Damn...something's come up, I might not be able to do an update until Monday.

Darn.

Key phrase being "might"

Depends how fast I can get through a certain mind-numbing task...

BTW, I could hack in Johny Olds.

He could enter the program by assuming the identity of one of the NPC astronauts?

PeZook wrote:BTW, I could hack in Johny Olds.

He could enter the program by assuming the identity of one of the NPC astronauts?

Make it so. *straightens shirt*

Not another paperwork fuck-up!

As it would turn out, Johnny WAS accepted into the program but somehow got filed in the wrong place! ...actually that'd make him really sad because he could've been backing Bob up or something if only he'd been there. But, you know, either way, I'll make up some story so he'll fit.

Though I really wish we could have Svetmaya around too.

You never know. Just how godlike is PeZook? Is changing names the most he can do, or can random events also be altered? PeZook?

doom3607 wrote:You never know. Just how godlike is PeZook? Is changing names the most he can do, or can random events also be altered? PeZook?

I believe the process involves taking a hexeditor to the save file. This would probably not change future random events, though there has been some discussion suggesting that random events are generated in advance and reside in the save file.

Yeah, PeZook has stated earlier that once the game starts, all the random events are set in stone and will happen. So, there's no save-scumming of random events or anything like that.

If it is possible to add Comrade Svetmaya, I would like it to be done.

If not... Nichevo.

!

doom3607 wrote:You never know. Just how godlike is PeZook? Is changing names the most he can do, or can random events also be altered? PeZook?

I have looked into the possibility, however my arcane powers do not include the ability to change the future at the moment.

I will certainly let you know when/if I gain this ability

OOC - Been looking a bit at the numbers and how a "get out of failure free" card changes things.

Ordinarily, for unblemished success, no failures are permissible (I haven't pulled apart the code yet to figure out exactly which balls-up gets assigned), which is your simple product-of-stage-reliabilities.

A GOFF card, strangely enough, allows one failure to go by without impeding mission success. For equal component reliabilities, it's simply the next term in the binomial expansion.

Frex: a manned suborbital shot has 3 stages - launch, suborbital travel, and recovery (I think)

Assuming 90% reliability across the board, the mission would ordinarily have 72.9% chance of unblemished success - 1 in 3.7 chance of less than full success. Add a GOFF card, and that chance shoots up to 97.2% - 1 in 35.7 chance of less than full success. 98% reliability, those numbers become 94.1% and 99.9% respectively - 1 in 17 chance vs 1 in 844.6.

Up the other end, for a manned historical lunar landing with 22 separate stages (A thru V):
90% reliability - 9.8% chance of success normally, vs 33.9% chance with a GOFF - 1 in 1.1 chance of cockup vs 1 in 1.5. At 98% stage reliability, 64.1% chance of full success normally vs 92.9% chance with a GOFF - 1 in 2.8 chance of cockup vs 1 in 14.1.

A direct ascent mission, OTOH, has 19 separate stages. 90% stage reliability - 13.5% chance of full success vs 42.0% chance with GOFF - 1 in 1.15 chance of cockup vs 1 in 1.7. 98% stage reliability - 68.1% chance of success normally, 94.5% with a GOFF - 1 in 3.1 chance of cockup vs 1 in 18.3 with a GOFF.

The most complicated mission in the game, Joint Lunar Landing: Lunar Orbital Rendezvous, has 26 stages (mainly due to each booster flying separately to lunar orbit). With 90% stage reliability, a normal shot would be facing very long odds with a 6.4% chance of unblemished success (1 in 1.06 chance of failure) (I believe Sergei Pavelovich would call that "suicidal"). A GOFF up your sleeve would improve that to 25.1% (1 in 1.3 chance). 98% stage reliability makes those numbers 59.1% (1 in 2.4 chance of failure) and 90.5% (1 in 10.5) respectively

You guys would do well to remember, though, that the numbers generated for rolls are not actually random, but pseudo-random. So the rolls fall on the normal distribution curve, making the actual failure probabilities far higher than what would normally be the case.

Fuck me drunk, PeZook. So each roll's a normal distribution, mean 50, sigma ~ 16 or so?

BTW, RIS 1.0 rc1 went up on source forge a few hours ago, but nothing executable yet.

fnord wrote:Fuck me drunk, PeZook. So each roll's a normal distribution, mean 50, sigma ~ 16 or so?

That's what happens with pseudo random numbers, yeah

You can see it in action in BARIS due to the fact that, well...more than 10% of lunar missions succeed, no?

Either that, or crew skill affects what fraction of failures are catastrophic - still digging through code.

Interesting tidbit - the AI will ALWAYS attempt unmanned missions, and will attempt manned missions if safety is no more than 15% below max R&D.

PeZook wrote:You guys would do well to remember, though, that the numbers generated for rolls are not actually random, but pseudo-random. So the rolls fall on the normal distribution curve, making the actual failure probabilities far higher than what would normally be the case.

...Shit. Shit shit shit shit shit.

There go all my calculations. My calculated risks, they have been calculated WRONG all this time!

Can I get the standard deviation on that pseudo-random number generator? With that, I can start doing things properly again...

The Cape. MASA headquarters hallway

In the hallways of MASA, there was shouting and the sounds of a fight.

"You fucking coward!" Rex could be heard shouting.

"You stupid child." Fox shouted right back.

The two of them were wrestling, giving each other as many punches to each other's gut as possible. Both of them flew at each other, screaming. Due to both of the men's flight training, they started to almost be graceful as both of them bloodied each other.



"I HATE YOU." Rex screamed, he was letting off steam from all the stress he had been through with losing a friend and all the politics that were involved. Fox stopped holding back, and went full force into fighting back. Fox was the older brother, and so had more experience in keeping Rex in line when the man had a temper tantrum. They hadn't fought like this in years, but two or more Modems in the same vicinity caused heated emotions, and eventually violence.

He kicked Rex in the knees, making the man fall to all fours, unable to stand.

Fox ran up to him and kicked him in the jaw.



Once his brother was knocked out, Fox left the hallway, tired of the bullshit of this place. Fox noticed the looks people gave him as he put his hand to his ribs, hoping they weren't broken. He didn't care, he had a nice job now, one that didn't require for everyone to die on the whims of some unseen dice roll. Rex put up with that shit. Fox wouldn't anymore.

More code diggage:

Stage safety is capped at 99%, regardless of crew skill. Still no idea what happens re: failures

The random generator used appears to be a linear function, not gaussian.

I wonder if that means there could be a 1% chance of failure?

Simon_Jester wrote: Can I get the standard deviation on that pseudo-random number generator? With that, I can start doing things properly again...

I don't know the exact specifications, dude It's just that games do tend to use pseudo-random numbers all the time.

Fnord will probably dig up the exact pseudo-random generator used, but I'd doubt it is something extremely complicated like you'd find in,say, a casino simulator.

Heh. Sorry.

It's just that I compulsively calculate odds of failure for missions; if my random numbers aren't really random it's awkward.

Though honestly, I'd think they are really random- when I've played, the numbers have been all over the map from 1 to 100, and I don't think there are a disproportionate number of very low or very high rolls.

I'm back bitches!

And my oh my how things have gone a bit squirrelly in a week. Ah well. At least you only did a couple of seasons so I haven't missed much.

ChaserGrey, I'll "take over" again after the next update, so I can get back into the groove properly:

Colonel Ravensburg paused briefly on his customary long run to consider the tumultuous events of the past year. The success of the EVA, followed by the utter tragedy that befell Johnson.

Edward looked to the sky, remembering the memorial service.

"Bob, you may have been a Navy pilot, but damn if you weren't a good guy. Godspeed my friend."

He continued running before anyone coudl notice his pause. On the up side though, he was one of the best astroanuts rated from the Lunar Module, so there was a very real chance he would be walking on the Moon before the decade was out. How exciting!