Alright, so crits work on a paired system. If one of the pair rolls a 50, and the other rolls an even number, it is a crit. If it is an odd number, no crit.
[50, x] [x, 50]
would both be crits if x = 6, and neither would be crits if x=7. Basically, crits are being kept at 1 percent as best I can.
Hm.
There is a one percent chance of getting [50, x] for even values of x.
There is a one percent chance,
separately and independently, of getting [x, 50] for even values of x.
This would increase crit chance to 1.98% instead of 1% (since the [50, 50] case is doubled).
Critfails on the other hand, require pairs of ones, and are pool based. So if you roll a single die, you only get a critfail if you roll 2 1s. If you roll eight dice, or 16d50, if any two of them are 1s that is a critfail. So small dice investments should be very very safe, while larger dice pools are increasingly unsafe.
Let's see, what's the breakpoint... and I COULD be wrong, so anyone who feels confident in probability theory and is willing to show their formula, here's a chance to show me up!
Representing the change in odds of a *single* critfail as (old probability of critical failure) -> (new probability of same), I believe it's
[ (1-(.99^X)) ] -> [ (2X-1)*(1-(.98^(2X-1))*.02 ]
1 die: 1% -> 0.04%
2 dice: 1.99% -> 0.35%
3 dice: 2.97% -> 0.96%
4 dice: 3.94% -> 1.85%
5 dice: 4.90%-> 2.99%
6 dice: 5.85% -> 4.38%
7 dice: 6.79% -> 6.01%
8 dice: 7.82% -> 7.84%
So for projects rolling eight dice, the risk of a critfail is now
infinitesimally higher, as in "two such megaproject rush rolls out of 10000 will critfail when this would otherwise not have happened," when I doubt we've done more than a few dozen eight-die rolls in the entire game so far.
For projects of five to seven dice, the risk is at least half of what it once was (about 60% as much risk at five dice, about 85% at seven dice).
For projects of two to four dice the risk is greatly reduced (one sixth to just under one half of what it was, respectively), but not truly negligible.
The risk of failure on one die is now negligible.
So we do have an incentive to slow-walk and diversify projects where we can, but the risk of critfails when we rush dice to the tune of 7-8 dice hasn't actually gotten
meaningfully worse than it was before. At nine dice or above on a single project, the risks would start to get noticeably worse now than they were before.
In short, literal meme plans are now courting grave danger, but almost everything else we do is unaffected or positively affected.
We were told, I believe, that the more lunar mines we did the easier it would be to do lunar settlements due to the preexisting infrastructure for the mines.
Hmm...
Lunar mining income contributes to our income goal, right?
It might be worth leaving the red zone alone for a bit and do some lunar mines instead to see if that helps lunar settlements.
Focus on refineries and other stuff in tiberium for a turn or two.
I don't think we can afford to ignore Red Zone operations completely, we've already been expanding at a much slower pace for some time now. And we need to keep abatement up if we want the Red Zones to go on shrinking reliably, since we aren't super-over-abundant with that to the same degree.
But trying one phase of moon mines alongside the first round of habitat construction isn't a bad idea, just so we learn what the scope of the effects is now.
Our new tunnel boring tech should allow us to more easily build subterranean habitation on the moon, to say nothing of expanding the lunar mines.
By the way
@Simon_Jester I thought you were going to write up the latest SCEDQuest post. Are you okay? Has anything changed on that front?
I'm okay, but I've been distracted by other projects and relatively petty personal problems. I do still think I can do it, though if anyone else is feeling eager, I'd be glad to let them have a go at it.
I still hope that in the next month or so I can get moving and get us caught up; we may have to desynchronize and do two SCED turns in one GDI turn to catch up, if Ithillid doesn't mind.
