This grouping of topics is for the various experiments I am running at any given time.  Each Experiment will receive its own serial number, starting with 2001, which follows next.

01 May 2020:  I've been away from here for a while, but, looking back, I think it's time to index the experiments in this first post.  I hope to get this done sometime this weekend.
This message was last edited by the GM at 05:22, Fri 01 May 2020.

UPDATE:  30 Dec 2017.  dice roll links modified to add the memo field.

As an extension of my explorations into the Help File for the Die Roller, I am now going to attempt to create some specialized dierolls for the game system The Petal Hack (which I use in one of my other games).  Specifically, I am using version 1.1 of this rule set as my reference (since that is what I'm using in my other game.)

DISCLAIMER:  Please note that The Petal Hack is published and made available free of charge.  That said, I will only quote enough of the rules here to suitably explain what I'm trying to accomplish.  It is my belief that such quotes fall under Fair Use protections afforded by the laws of the United States and other countries.

The Core Mechanic (given on page 1 of The Petal Hack) states:

"Everything a character might possibly attempt that could result in failure is resolved by testing attribute stats.  In order to successfully test a stat, a player must roll below it on a d20."

Further, almost directly across from this paragraph on the same page, are expressed the rules dealing with Advantage & Disadvantage:

"If a particular course of action or task has a higher or lower chance of success, the Referee may ask a player to roll an additional d20 when making a test - with Advantage the lower result is used and with Disadvantage, the higher."

These rules set the parameters for my experiment:  I am going to attempt to write dice roller links that, when clicked, fire up the dice roller and do a simple stat check, a stat check with Advantage, and a stat check with Disadvantage, all links to display appropriate texts.

Now the stat check is a simple single roll of d20, but if we want the link to it to say "Simple Stat Check", the code needs to look like:

[dice=1d20 memo="Simple Stat Check" text="Roll a Simple Stat Check"]  Roll a Simple Stat Check

Testing of this worked well enough.  On to more interesting dierolls.

In the case of rolling with Advantage, the lower of two d20 rolls is the result, the other being discarded.  Let's see what that might look like.

[dice=2d20 dok=kl drop=1 memo="Stat Check with Advantage" text="Roll a Stat Check with Advantage"]
Roll a Stat Check with Advantage

Well, that works fine, but it only displays the single result. If I wanted to test the roll, or see some debugging data, I might wish to have the die roller display both dice then also display the actual result.  That might look like:

[dice=2d20 record=yes dok=kl drop=1 memo="Stat Check with Advantage - debug."  text="Roll a Stat Check with Advantage, displaying both dice"]
Roll a Stat Check with Advantage, displaying both dice

Creating a roll with Disadvantage is now as simple as changing a couple of parameters in the roll with Advantage, to wit:

[dice=2d20 dok=kh drop=1 memo="Stat Check with Disadvantage" text="Roll a Stat Check with Disadvantage"]
Roll a Stat Check with Disdvantage

and the debug version:

[dice=2d20 record=yes dok=kh drop=1 memo="Stat Check with Disadvantage - debug." text="Roll a Stat Check with Disdvantage, displaying both dice"]
Roll a Stat Check with Disdvantage, displaying both dice

Well, this was fun.  Now to add some extra sauce to this, I'll try something not mentioned in the Help File:  I'll attempt to choose a dice color from the link code.  (Note:  this feature, if it does exist, is undocumented.)

[dice=2d20 color=red record=yes dok=kl drop=1 text="Roll a Stat Check with Advantage, displaying both dice"]
Roll a Stat Check with Advantage, displaying both dice

That didn't work.  Hmm... did the Help file mention flags for secret or private rolls?  No.

Observations:  Certain features that cannot be coded for in a link can still be selected just prior to rolling.  Also, the person rolling must still choose a Character before rolling to avoid throwing an error unless the link pre-configures the correct character ID number for them.
This message was last edited by the GM at 11:16, Sun 11 Feb 2018.

This experiment was prompted by a thread on RPoL Development at Dice Roller:  Deck of Fate

The subject was using dice rolls to simulate a draw from a deck of cards, specifically the Evil Hat Productions Deck of FATE.

I will not be covering how to do a Deck of Fate because I don't own one.  I will instead use a shorter example that is in Public Domain for illustrative purposes.

Some Background:

There are two scenarios for which a deck of cards may be used:  a single card may be drawn from the deck or two or more cards may be drawn on the same shuffle.

In the first case, only one result is needed, and the deck is re-shuffled before the next draw is made.  This is referred to in Statistics as sampling with replacement.

In the second case, all cards needed from the deck are drawn before the deck is re-shuffled.  That is, the deck is sampled multiple times without replacment.

Now dice rolls can be classed in two general types according to the distributions into which their results fall.

The Uniform Distribution:

The roll of a single die, or the roll of multiple dice interpreted as the "digits" of a result, produces a Uniform Distribution.  This is so called because each result has a uniform probability of occurence if the dice are fair (i.e., not misshapen, loaded, or irregularly spotted/numbered).

The Normal Distribution:

Any roll of two or more dice whose sum is taken as their result produces a Normal Distribution.  The roll of two six-sided dice used in many common board games is perhaps the most familiar example to non-gamers, and the roll of three six-sided dice is perhaps more familiar to those who play role-playing games.

In this type of distribution, there are one or two results which tend to occur more frequently because of the way the dice combine to produce those results.  These more frequent results are referred to as norms.

In a roll of two six-sided dice, for example, the norm is 7, and it can occur in the following ways (note that order of the two dice is significant):

1+6, 2+5, 3+4, 4+3, 5+2, 6+1

So a roll of 7 occurs six times out of the thirty-six possible permutations of 2d6.  A mathematician or statistician might reduce this 6:36 down to 1:6 and be numerically correct in stating these as the odds (or the probability) of a result of 7 on any one given roll of 2d6.

Simulating a Deck of Cards:

In simulating a deck of cards, the structure of the deck is usually represented by a table (or, if you're programming such a simulation, an array or similar data structure).  We'll use a deck of Zener Cards to illustrate.  The original Zener Deck was used to test for Extrasensory Perception back in the 1960s and has a rather generic symbol set:  a Yellow Circle, a Red Cross, a set of Blue Wavy Lines, a Black Square, and a Green Five-Pointed Star.  Five of each symbol is included in the deck to make a total of 25 cards.  I chose this example as the symbols are suitably generic and the deck itself is small, but large enough for illustrative purposes.

Before we get into our example, the structure of the deck matters not one whit so long as all cards are represented.  Just to prove the point, I'm going to modify this Zener Deck such that each symbol is represented in each of the five colors.  This also makes it easier to see that each card is unique.

The First Step:  Building a Reference Table

So, for our Modified Zener Deck, the table might look something like


We could scramble these various cards around in any manner and not affect the validity of the outcomes one bit, but this is the arrangement I will use for my example.  In effect, we're arranging first by color (analogous to suit in a more traditional deck of cards), then by symbol (value).

Step Two:  Creating the Dice Roll for the RPoL Dice Roller.

Now that we have our table, we know what our die roll must look like:  a Uniform Distribution with results ranging from 1 to 25.  How do we know this?  Each of the cards is unique (no duplicates), and there are 25 cards.

So, whether we draw a single card or multiple cards from the deck, the same basic die roll template can suffice.  In the template below, N is the number of cards being drawn.

[dice=Nd25 record=yes unique=yes memo="Draw N cards from the Modified Zener Deck" text="Draw N cards from the Modified Zener Deck."]

For a single draw:

[dice=1d25 record=yes unique=yes memo="Draw 1 card from the Modified Zener Deck" text="Draw 1 card from the Modified Zener Deck."]

Try it out here:

Draw 1 card from the Modified Zener Deck.

To make four draws from the same deck before shuffling:

[dice=4d25 record=yes unique=yes memo="Draw 4 cards from the Modified Zener Deck" text="Draw 4 cards from the Modified Zener Deck."]

Try it out here:

Draw 4 cards from the Modified Zener Deck.

Using the Modified Zener Deck:

To use the deck, roll the die or dice needed (one for each player) and consult the table for the result shown on each die.

For example, if I draw four cards using the dice roller, I might get something like:

20:37, Today: Chief Tinker rolled 80 using 4d25, unique dice ((20,15,24,21)).

The "rolled 80" part is irrelevant.  Look at the results in the double parentheses:

20 = "A Black Five-Pointed Star"
15 = "A Blue Five-Pointed Star" (those who remember Galaxina feel free...)
24 = "A Green Square"
21 = "A Green Circle"

The numbers tell us immediately that all four results are unique, and our table results confirm it.  Once this roll is cast, any new roll will result in a draw or draws from a fresh Modified Zener Deck.

NOTES:

I have not yet figured out how to make a deck persist over multiple dice rolls.  That is, each time a roll is called, the "deck" will, in effect, have all its cards available.  It is, therefore, important for the GM using this method to draw all cards for a given event or place in the adventure at one go, i.e. if six cards are needed, draw all six and keep them aside until needed.

The dice roller will come up with all the settings shown configured already, but it is possible using the rc= option to specify a character name in the link.  Please don't forget to select a character to roll before attempting the roll, or it will fail with an error message.

Example:  Draw 1 Zener Card from the Modified Deck for Chief Tinker  [dice=1d25 rc=299435 text="Draw 1 Zener Card from the Modified Deck for Chief Tinker"]

{For more information about the rc= option, see section 1000 - Basic Stuff under message no. 8, titled 1002.24 - Character-Related Options.}

A roll of this type might have the following output in the dice roller log:

22:08, Today: Chief Tinker rolled 5 using 1d25.

which would be the equivalent of the card Yellow Star.

A color for the dice roll text can be selected prior to the roll.  I have not seen ways to add this feature to dice roll links as yet.
This message was last edited by the GM at 14:05, Sat 04 Aug 2018.

This experiment was prompted by:  Private lines that aren't really private  The original post in that thread is quoted below.



This post gave rise to some answers worthy of experimental verification, so here goes.

Firstly, I will attempt to post a private line to everyone:


Next, a private line to no one:

[Private to: I expect this to throw an error.]

The only error it threw was being visible to everyone.  Let's try a non-existent user...


Huh.  Private to Nobody threw no errors, and sent a private line only the GM can see, presumably.  Let's see what Private to Everybody looks like...


As expected, so long as the character name is of a non-existent character, the message goes to the GM.

It would be helpful if I had some players here to help test things.

This addendum is prompted by an artifact I observed during the previous post:

Posting certain private lines spawned the generation of private messages and re-directed the posting of the previous message there.

I have since copied the content of that message, edited it a bit, and cannot seem to replicate this behavior.

For that reason, I'm repeating certain private lines from the previous post to check and see if I can get this artifact to occur again.

Private to Nobody:


Private to Everybody:


So far just previewing this post does not generate private messages...

Private to:

[Private to:  no recipient included.]

Now to post this thingy and see what pops up.

Curious:  so far no private messages were generated.


Hmm... that leaves private to Chief Tinker and Homunculus...


Huh.  Maybe I made a mistake creating the original post (2003), above?
This message was last edited by the GM at 13:22, Wed 30 May 2018.

This sub-experiment is at the request of new player BaxTheSlayer, stemming from a conversation we had in a private message in one of his games.

The purpose of this effort is to gain a better understanding of how private messages in game posts actually work.

To start off, I'll post a private message to my alternate character, Homunculus, here:


As a bonus to the experiment, I'll attempt to display using code tags what this private line looks like in the message editor window:

[Private to Homunculus: The rain in Spain falls mainly on the plain. Don't forget to roll the r.  I'm making this line needlessly long because I want to post more than one line of text.  Right now it's too much effort for me to come up with something meaningful off the cuff.  Sorry.]

Well... that almost worked.  The color for private messages is a unique one not available in the regular color palette that I can find, but orange is close.  Also note the opening and closing square brackets of the message are in the default text color (depending on which theme you have selected in User Preferences.)

BaxTheSlayer should not be able to see the actual private line above, or even tell if it's there without the context cues I've given.
This message was last edited by the GM at 14:16, Sat 04 Aug 2018.

Observations:

The only indication I can see of the private lines is a single, extra, empty line after the 'statement' line.

From my side, all I see is this:



So, savvy players could construe that something is going on, but it could just as easily be an accidental extra line.

Knowing that private lines between GM characters works at all, I suggest posting the following, as a test to see the effects of organizing private lines in various ways:

Test 1 (Private Line Directly Below)
[Private to Homunculus: The rain in Spain falls mainly on the plain. <i>Don't forget to roll the r.</i>  I'm making this line needlessly long because I want to post more than one line of text.  Right now it's too much effort for me to come up with something meaningful off the cuff.  Sorry.]
Test End Line

Test 2 (Private Line Attached)[Private to Homunculus: The rain in Spain falls mainly on the plain. <i>Don't forget to roll the r.</i>  I'm making this line needlessly long because I want to post more than one line of text.  Right now it's too much effort for me to come up with something meaningful off the cuff.  Sorry.]
Test End Line

Test 3 (Private Line Running Into Text)
[Private to Homunculus: The rain in Spain falls mainly on the plain. <i>Don't forget to roll the r.</i>  I'm making this line needlessly long because I want to post more than one line of text.  Right now it's too much effort for me to come up with something meaningful off the cuff.  Sorry.]Test End Line

Test 4 (Multiple Detached Private Lines)
[Private to Homunculus: The rain in Spain falls mainly on the plain. <i>Don't forget to roll the r.</i>  I'm making this line needlessly long because I want to post more than one line of text.  Right now it's too much effort for me to come up with something meaningful off the cuff.  Sorry.]
[Private to Homunculus: The rain in Spain falls mainly on the plain. <i>Don't forget to roll the r.</i>  I'm making this line needlessly long because I want to post more than one line of text.  Right now it's too much effort for me to come up with something meaningful off the cuff.  Sorry.]
Test End Line

Test 5 (Multiple Attached Private Lines)
[Private to Homunculus: The rain in Spain falls mainly on the plain. <i>Don't forget to roll the r.</i>  I'm making this line needlessly long because I want to post more than one line of text.  Right now it's too much effort for me to come up with something meaningful off the cuff.  Sorry.][Private to Homunculus: The rain in Spain falls mainly on the plain. <i>Don't forget to roll the r.</i>  I'm making this line needlessly long because I want to post more than one line of text.  Right now it's too much effort for me to come up with something meaningful off the cuff.  Sorry.]
Test End Line
This message was last edited by the player at 20:35, Fri 08 June 2018.

Your suggestions seem well thought out.  Here goes:


Test 1 (Private Line Directly Below)
Test End Line

Test 2 (Private Line Attached)
Test End Line

Test 3 (Private Line Running Into Text)
Test End Line

Test 4 (Multiple Detached Private Lines)
Test End Line

Test 5 (Multiple Attached Private Lines)

Test End Line

And here's what I see:



Though, interestingly, when I quoted it, this is the actual capture from the quote link:

My Conclusions:

It seems that everything between the brackets of a private line essentially doesn't exist to non-participating viewers.

As long as there were no extraneous returns between the surrounding text lines, there is no indication that there are private lines at all.  The only caveat, apparently, is that two private lines that run together generate an extraneous return for some reason.

I would guess multiple cascading private lines would continue to generate additional returns.  I would also surmise that one could put a private line in the middle of a sentence without breaking up the sentence (essentially, a private parenthetical note), but these would need to be confirmed.

Finally, it seems that the underlying system does see a single return somewhere in the mix for each private line - it added one in test 1, but not in test two or three which each attached to the test line and the end line, respectively, two in test 4 for each private line, and one in test five for both consecutive private lines.

Of course, these only show up when you quote or possibly if there was some other method to access the actual code of the post.