aramis
Posts: 606
Joined: Fri 14 Jun 2019, 20:34
Location: Oregon, USA
Contact:

Re: TN16: A in depth computer simulation. How Miserable and Weary affect probability

Note that by doing a randomized value test, you've actually corrupted the probabilities due to the errors of random number generation on electronics.

the precise odds, found by the much simpler and much more mathematically correct methods of either iteration or formulae are what should be used; they're as much a test of your RNG as the math.

all odds below calculated using loop iteration via python
ndS [Fail, simple success, +1Ꞇ, +2Ꞇ, +3Ꞇ,+4Ꞇ, +5Ꞇ, +6Ꞇ] total count
6dS TN16 [10720, 178040, 223740, 112500, 30000, 4500, 360, 12] 559872
5dS TN16 [6958, 31792, 36260, 14990, 3000, 300, 12, 0] 93312
4dS TN16 [3359, 4945, 5220, 1776, 240, 12, 0, 0] 15552
3dS TN16 [1188, 630, 600, 162, 12, 0, 0, 0] 2592
2dS TN16 [314, 60, 50, 8, 0, 0, 0, 0] 432
1dS TN16 [65, 5, 2, 0, 0, 0, 0, 0] 72
0dS TN16 [11, 1, 0, 0, 0, 0, 0, 0] 12

Now, to make it readble in percentagee chances of success with N+ successes; I dumped the above into Libre Calc...
``````Dice/TN   __F%__  __S%__  _S+1Ꞇ_  _S+2Ꞇ_  _S+3Ꞇ_  _S+4Ꞇ_  _S+5Ꞇ_  _S+6Ꞇ_
6dS	TN16   1.915  98.085  66.285  26.322   6.22     0.87    0.066	0.002
5dS	TN16   7.457  92.544  58.473  19.614   3.55     0.335	0.013   0
4dS	TN16  21.599  78.402  46.605  13.04    1.62     0.077	0       0
3dS	TN16  45.833  54.167  29.861   6.713   0.463    0       0       0
2dS	TN16  72.685  27.315  13.426   1.852   0        0       0       0
1dS	TN16  90.278   9.722   2.778   0       0        0       0       0
0dS	TN16  91.667   8.333   0       0       0       0        0       0
``````
(Why Python? because I can define the dice as a variable array easily, and the loop options are great.)
—————————————————————————
Smith & Wesson: the original point and click interface...

gyrovague
Posts: 416
Joined: Tue 28 Apr 2020, 16:52

Re: TN16: A in depth computer simulation. How Miserable and Weary affect probability

Note that by doing a randomized value test, you've actually corrupted the probabilities due to the errors of random number generation on electronics.
Oh, sorry, I forgot to mention that my server is connected to a device I built using black-market plutonium, and my RNG is coming from alpha decay. In case that melts down, my failover server uses a Lavarand.

I figured it was worth the extra effort to get accurate numbers. I don't want Shadow points from the corruption.

Mythicos
Posts: 124
Joined: Tue 16 Jun 2020, 03:46

Re: TN16: A in depth computer simulation. How Miserable and Weary affect probability

Note that by doing a randomized value test, you've actually corrupted the probabilities due to the errors of random number generation on electronics.

the precise odds, found by the much simpler and much more mathematically correct methods of either iteration or formulae are what should be used; they're as much a test of your RNG as the math.

all odds below calculated using loop iteration via python
ndS [Fail, simple success, +1Ꞇ, +2Ꞇ, +3Ꞇ,+4Ꞇ, +5Ꞇ, +6Ꞇ] total count
6dS TN16 [10720, 178040, 223740, 112500, 30000, 4500, 360, 12] 559872
5dS TN16 [6958, 31792, 36260, 14990, 3000, 300, 12, 0] 93312
4dS TN16 [3359, 4945, 5220, 1776, 240, 12, 0, 0] 15552
3dS TN16 [1188, 630, 600, 162, 12, 0, 0, 0] 2592
2dS TN16 [314, 60, 50, 8, 0, 0, 0, 0] 432
1dS TN16 [65, 5, 2, 0, 0, 0, 0, 0] 72
0dS TN16 [11, 1, 0, 0, 0, 0, 0, 0] 12

Now, to make it readble in percentagee chances of success with N+ successes; I dumped the above into Libre Calc...
``````Dice/TN   __F%__  __S%__  _S+1Ꞇ_  _S+2Ꞇ_  _S+3Ꞇ_  _S+4Ꞇ_  _S+5Ꞇ_  _S+6Ꞇ_
6dS	TN16   1.915  98.085  66.285  26.322   6.22     0.87    0.066	0.002
5dS	TN16   7.457  92.544  58.473  19.614   3.55     0.335	0.013   0
4dS	TN16  21.599  78.402  46.605  13.04    1.62     0.077	0       0
3dS	TN16  45.833  54.167  29.861   6.713   0.463    0       0       0
2dS	TN16  72.685  27.315  13.426   1.852   0        0       0       0
1dS	TN16  90.278   9.722   2.778   0       0        0       0       0
0dS	TN16  91.667   8.333   0       0       0       0        0       0
``````
(Why Python? because I can define the dice as a variable array easily, and the loop options are great.)

Are you aware that your numbers and Dorjcal have at worst a 0.15% diifference in Success rates?

Was it really worth it blaming his choice of methodology, given that for all intents and purposes they're identical?

He's not sending a satellite to Mars, after all...

gyrovague
Posts: 416
Joined: Tue 28 Apr 2020, 16:52

Re: TN16: A in depth computer simulation. How Miserable and Weary affect probability

He's not sending a satellite to Mars, after all...
MIDDLE EARTH IS MORE IMPORTANT THAN MARS!

aramis
Posts: 606
Joined: Fri 14 Jun 2019, 20:34
Location: Oregon, USA
Contact:

Re: TN16: A in depth computer simulation. How Miserable and Weary affect probability

Note that by doing a randomized value test, you've actually corrupted the probabilities due to the errors of random number generation on electronics.

the precise odds, found by the much simpler and much more mathematically correct methods of either iteration or formulae are what should be used; they're as much a test of your RNG as the math.

all odds below calculated using loop iteration via python
ndS [Fail, simple success, +1Ꞇ, +2Ꞇ, +3Ꞇ,+4Ꞇ, +5Ꞇ, +6Ꞇ] total count
6dS TN16 [10720, 178040, 223740, 112500, 30000, 4500, 360, 12] 559872
5dS TN16 [6958, 31792, 36260, 14990, 3000, 300, 12, 0] 93312
4dS TN16 [3359, 4945, 5220, 1776, 240, 12, 0, 0] 15552
3dS TN16 [1188, 630, 600, 162, 12, 0, 0, 0] 2592
2dS TN16 [314, 60, 50, 8, 0, 0, 0, 0] 432
1dS TN16 [65, 5, 2, 0, 0, 0, 0, 0] 72
0dS TN16 [11, 1, 0, 0, 0, 0, 0, 0] 12

Now, to make it readble in percentagee chances of success with N+ successes; I dumped the above into Libre Calc...
``````Dice/TN   __F%__  __S%__  _S+1Ꞇ_  _S+2Ꞇ_  _S+3Ꞇ_  _S+4Ꞇ_  _S+5Ꞇ_  _S+6Ꞇ_
6dS	TN16   1.915  98.085  66.285  26.322   6.22     0.87    0.066	0.002
5dS	TN16   7.457  92.544  58.473  19.614   3.55     0.335	0.013   0
4dS	TN16  21.599  78.402  46.605  13.04    1.62     0.077	0       0
3dS	TN16  45.833  54.167  29.861   6.713   0.463    0       0       0
2dS	TN16  72.685  27.315  13.426   1.852   0        0       0       0
1dS	TN16  90.278   9.722   2.778   0       0        0       0       0
0dS	TN16  91.667   8.333   0       0       0       0        0       0
``````
(Why Python? because I can define the dice as a variable array easily, and the loop options are great.)

Are you aware that your numbers and Dorjcal have at worst a 0.15% diifference in Success rates?

Was it really worth it blaming his choice of methodology, given that for all intents and purposes they're identical?
Yes, it is. Bad stats methods are bad stats methods, and depending upon language and hardware, can introduce more significant errors.
—————————————————————————
Smith & Wesson: the original point and click interface...

Posts: 62
Joined: Sat 10 Jul 2021, 16:01

Re: TN16: A in depth computer simulation. How Miserable and Weary affect probability

For what it's worth, the bad stats method generated a much more readable table for those of us using screen readers. Info presentation may possibly be more important than correct use of statistical calculation methods, particularly if the difference is at most 0.15%. I'm just sayin'

Mythicos
Posts: 124
Joined: Tue 16 Jun 2020, 03:46

Re: TN16: A in depth computer simulation. How Miserable and Weary affect probability

For what it's worth, the bad stats method generated a much more readable table for those of us using screen readers. Info presentation may possibly be more important than correct use of statistical calculation methods, particularly if the difference is at most 0.15%. I'm just sayin'

QFT

gyrovague
Posts: 416
Joined: Tue 28 Apr 2020, 16:52

Re: TN16: A in depth computer simulation. How Miserable and Weary affect probability

I asked my 7-year-old to arbitrate this debate. He likes to say things like, "Dad, it's not quarter past four, it's 4:17." And, "Dad, the sun isn't going down, the earth is turning."

I explained to him as best I could about monte carlo simulations, statistics, and what "pseudo-random" means. I'm not sure he followed the whole explanation, but nevertheless he came to a conclusion: he sided with aramis.

I guess I have no choice but to concede. Well played.

Fenhorn
Moderator
Posts: 2996
Joined: Thu 24 Apr 2014, 15:03
Location: Vänerns Pärla

Re: TN16: A in depth computer simulation. How Miserable and Weary affect probability

Moderator Message: On this forum we show respect to and we discuss topics in a decent manner.
“Thanks for noticin' me.” - Eeyore

Dorjcal
Topic Author
Posts: 81
Joined: Sun 11 Jul 2021, 10:22

Re: TN16: A in depth computer simulation. How Miserable and Weary affect probability

Note that by doing a randomized value test, you've actually corrupted the probabilities due to the errors of random number generation on electronics.

the precise odds, found by the much simpler and much more mathematically correct methods of either iteration or formulae are what should be used; they're as much a test of your RNG as the math.

all odds below calculated using loop iteration via python
ndS [Fail, simple success, +1Ꞇ, +2Ꞇ, +3Ꞇ,+4Ꞇ, +5Ꞇ, +6Ꞇ] total count
6dS TN16 [10720, 178040, 223740, 112500, 30000, 4500, 360, 12] 559872
5dS TN16 [6958, 31792, 36260, 14990, 3000, 300, 12, 0] 93312
4dS TN16 [3359, 4945, 5220, 1776, 240, 12, 0, 0] 15552
3dS TN16 [1188, 630, 600, 162, 12, 0, 0, 0] 2592
2dS TN16 [314, 60, 50, 8, 0, 0, 0, 0] 432
1dS TN16 [65, 5, 2, 0, 0, 0, 0, 0] 72
0dS TN16 [11, 1, 0, 0, 0, 0, 0, 0] 12

Now, to make it readble in percentagee chances of success with N+ successes; I dumped the above into Libre Calc...
``````Dice/TN   __F%__  __S%__  _S+1Ꞇ_  _S+2Ꞇ_  _S+3Ꞇ_  _S+4Ꞇ_  _S+5Ꞇ_  _S+6Ꞇ_
6dS	TN16   1.915  98.085  66.285  26.322   6.22     0.87    0.066	0.002
5dS	TN16   7.457  92.544  58.473  19.614   3.55     0.335	0.013   0
4dS	TN16  21.599  78.402  46.605  13.04    1.62     0.077	0       0
3dS	TN16  45.833  54.167  29.861   6.713   0.463    0       0       0
2dS	TN16  72.685  27.315  13.426   1.852   0        0       0       0
1dS	TN16  90.278   9.722   2.778   0       0        0       0       0
0dS	TN16  91.667   8.333   0       0       0       0        0       0
``````
(Why Python? because I can define the dice as a variable array easily, and the loop options are great.)
I mean.. sure? But the point was to have also Weary and Miserable, and I had no idea how to do it.
As long as you get in the ball park of what it is the probability it is not really going to change much.
No real die is perfectly balanced anyway.

Who is online

Users browsing this forum: No registered users and 5 guests