Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

Muster up remaining strength and complete the current chapter. Bring the game to a logical close, then reboot it.

32 (37.2%)

Same as 1, but a whole new game will be started instead, still Multiworld, but no longer Madness.

9 (10.5%)

Get just this last turn done and reboot the game. No point wasting energy if updating takes so long.

31 (36%)

Get the last turn done and start a new game project.

14 (16.3%)

Total Members Voted: 86

[Zako]

Not to mention you can piece yourself back together, unlike most of us. 

[Akhier the Dragon hearted]

Oh My GOD! This thread is epic. I have finally churned through this massive heap of win and declare it fit for my reading. Please put me on the waiting list though from the way things are going and how long the list is it would probably be faster for me to just start my own rtd and have fun that way.

[Sean Mirrsen]

Thanks!

I'll try to compile the first two worlds into a readable form sometime in the near future, so that people can enjoy the thing without having to click through two hundred pages. Meanwhile, all players are in and I've begun the next turn.

Hmm, actually...

Turns 0-100
Turn 101-150
All praise Dropbox!

Both are Excel spreadsheets, which aren't what I would call "readable form" since they contain all of the forum syntax as well, but you also use these as a framework if you want to run your own RTD. For best results though, you'd need to use the latest version of my layout, which you can obtain from the latest turn.

[Sir Speshkitty]

=[6-2=4] From the depths of your mind, you summon forth a great power. As if fueled by a benevolent god, you rise from your underwearly prison, your eyes still closed, and gently float towards one of the tables. In your dream, you see streams of en

The cells are getting cut off after 255 characters in both OpenOffice and MS Office 2007.
Is that just me?

[Sean Mirrsen]

Maybe. Maybe not. I'll check.

edit: yep, problem on my end. I was using Office XP (or whatever was installed here) which doesn't like cells over 255 characters for some reason. It should only be a problem for the turns 0-5 though (which were done before I started using the sheet), turn 6 has all the lengthy stuff intact. I'll go edit the files and tell you when you can redownload them.

[Sean Mirrsen]

OK, should be alright now. You can redownload the 0-100 file from the same link. (dropbox is a wonderful thing)

[Dermonster]

Right, you're the one with the longest running game, and friend of mine wanted over on mspa wants to start one, with some... odd rules.

he wants to use the Doomsday dice cascader for rolls. ... somehow.

Problem with that being, he doesn't know any rollers that go up into the septillions
Random.org apparently only goes from -1bil, to one billion.
Though i highly doubt it, do you know of any?

[qwertyuiopas]

Simply halve the number of the first roll(the number of dice used) or have it weighted(1=1, 2,3=2, 4,5=3, 5=4)
So that the number can't go past some still crazily high number that is more tolerable.

[Sean Mirrsen]

Well, making one in Visual Studio would be a fairly straightforward task...  I think... Though you'd have to make an octaword variable to hold the friggin' result, since even an unsigned int64 won't be enough for "septillions". I might make one, but since the built-in random function likely won't handle it, it'll involve lots and lots of coinflips. Which will royally wreck the probability curve. I might think of something. I don't know of any dice rollers capable of int128 return variables.

(qwerty, splitting one massive roll into several smaller rolls will funk up the probability curve. even two rolls instead of one will make half the values less probable than the others)

[qwertyuiopas]

I mean how the first roll in the source was a 6 that corresponded to using 6 dice overall. Limit that to 4, and you only have to tolerate

(((6²)²)²)²  (2821109907456)
Possibilities, rather than

(((((6²)²)²)²)²)² (6.3340286662973277706162286946812e+49)


Could an int128 go up to ~6e49 digits?

[Dermonster]

i suggested that before i asked you guys. sorta.
"Keep rerolling till you get something that's not five or six"
in a PM.
I haven't gotten a reply yet though >.>

[Sean Mirrsen]

Nah, that defeats the purpose of having that thing. You're wrong on the maximum number though, it turns out to be 7.95 octillions, as in a figure with 25 digits (more precisely, 7958661109946400884391936). Thankfully, that's within the capabilities of an int128. Unfortunately, I can't even make an int128 variable in C#.Net2008...
(which is odd, considering the built-in windows XP calculator showed me this result with no problem)

[Dermonster]

wow. that's a whole other illion than he thought it was.
i have no idea what int128 is. 

[qwertyuiopas]

Well..
Of 6 dice at maximum:
First roll: 6.
6*6=36.
Second roll: 36
36*36
third roll: 36²
36^4
or
6^8
...

6^64
or
6^(2^n), n=6
=

6.3340286662973277706162286946812e+49


3.3219280948873623478703194294894 bits per digit (log(10)/log(2), equivalent to a log₂(10))

so
49*3.3219280948873623478703194294894


162.77447664948075504564565204498
bits to store that number.

Going by the powers of two sizing, you would need an int256

To fill it, get 4 int64s from random.org, as it's generator is not predictable, and string the bits together.

That is the worst case scenario, fortunately.


And I hope that my math was wrong there, because that worst case scenario is quite bad.

[Sean Mirrsen]

Hmph. I don't know what you did there, but here's my progression:

6 <- Master die

6<-First roll maximum
36<-Second roll maximum
1296<-Third roll maximum
1679616<-Fourth roll maximum
2821109907456<-Fifth roll maximum
7958661109946400884391936<-Sixth roll, and final - its value is the maximum obtainable

170141183460469231731687303715884105728<-limit of a signed int128.

18446744073709551615<-C#'s ulong, or uint64. If I were to simply add them together, I'd need 431440 operations.

There's no good way to do this - even just making this many coinflips will take too bloody long, and you'd need a devious plan to counteract the stupendous bell-curving of the result.
I suppose you could get away with the Xd10 method, if you do something funky so that if the result exceeds the required target value, it rerolls until it's within limit. Technically, that's still very very wonky, but at least it provides a believable probability distribution. Yeah, that might work. Roll up the required amount of digits and reroll until what you get is within range. It's like rolling a d10000000000000000000000000 until you get a passable result. Noice.