## Re: [pcgen] Add Common Stat Roll methods to gameModes

Expand Messages
• ... Well, if you re-roll the ones until you can something that is not a one, you have a 20% probability for each 2, 3, 4, 5 and 6 result. That s what I assume
Message 1 of 22 , Oct 1, 2005
At 13:37 2005.10.01, Stefan Radermacher wrote:
>Éric Beaudoin wrote:
>>> ROLLMETHOD:4d6, reroll ones, then drop lowest [tab] METHOD: ???
>>
>> That would be 4 (d5+1) and keep the three highest. I don't know how to put the +1 in the but for the d5, you just have to replace the 6 with a 5.
>
>That is not identical from a statistical point of view. 4d6 and re-roll
>ones has 2 at 33% probability and 3-6 at 16% each; your method has all
>numbers 2-5 at 20%.
>
>Regards,
>Stefan.

Well, if you re-roll the ones until you can something that is not a one, you have a 20% probability for each 2, 3, 4, 5 and 6 result. That's what I assume the user wanted because that's the home rule I use.

If you have only one chance to re-roll the one, you right, the probability is not the same. You have 1/36 to still get a one and 11/30 to get each one of the other results. You are in for a pretty messed up expression four your dice.

I don't see how re-rolling the one a particular dice make it more probable to get a two. No mater if you re-roll the one 1 time or multiple time, the probability to get any other result on the dice will be evenly distributed.

-----------------------------------------------------------
Éric "Space Monkey" Beaudoin
Founding Member of the Hidden-in-the-Trench Club
Release Monkey and Syntax Watchdog
>> In space, no one can hear you sleep.
>> Camels to can climb trees (and sometime eat them).
<mailto:beaudoer@...>
• ... 11/30 is not accurate (I need coffee). It should be 7/36. So if you drop the 1, you roll a d35 and divide the result by 7 and this is the same as rolling a
Message 2 of 22 , Oct 1, 2005
At 13:51 2005.10.01, Éric Beaudoin wrote:
>If you have only one chance to re-roll the one, you right, the probability is not the same. You have 1/36 to still get a one and 11/30 to get each one of the other results. You are in for a pretty messed up expression four your dice.

11/30 is not accurate (I need coffee). It should be 7/36. So if you drop the 1, you roll a d35 and divide the result by 7 and this is the same as rolling a d6 and re-rolling once if you get a one. So in rnd(35) return a random value in [ 0 .. 34 ], the dice expression would be floor(rnd(35)/7)+1.

-----------------------------------------------------------
Éric "Space Monkey" Beaudoin
Founding Member of the Hidden-in-the-Trench Club
Release Monkey and Syntax Watchdog
>> In space, no one can hear you sleep.
>> Camels to can climb trees (and sometime eat them).
<mailto:beaudoer@...>
• ... I really need that cooffee, I forgot to account the 1/36 to get the one and nothing pop up right. I ll try again later after my coffee. ... Éric Beaudoin
Message 3 of 22 , Oct 1, 2005
At 13:58 2005.10.01, Éric Beaudoin wrote:
>11/30 is not accurate (I need coffee). It should be 7/36. So if you drop the 1, you roll a d35 and divide the result by 7 and this is the same as rolling a d6 and re-rolling once if you get a one. So in rnd(35) return a random value in [ 0 .. 34 ], the dice expression would be floor(rnd(35)/7)+1.

I really need that cooffee, I forgot to account the 1/36 to get the one and nothing pop up right. I'll try again later after my coffee.

-----------------------------------------------------------
Éric Beaudoin <mailto:beaudoer@...>

CONCORDE: It goes twice as fast as a bullet and you get smoked salmon.
-- From A Scientific Encyclopedia for the Enquiring Young Nome
by Angalo de Haberdasheri
(Terry Pratchett, Wings)
• ... Ha coffee, let s do this again. We want a formula that has the same probabilistic result as the event rolling a d6 and re-rolling the d6 once if the result
Message 4 of 22 , Oct 1, 2005
At 14:01 2005.10.01, Éric Beaudoin wrote:
>At 13:58 2005.10.01, Éric Beaudoin wrote:
>>11/30 is not accurate (I need coffee). It should be 7/36. So if you drop the 1, you roll a d35 and divide the result by 7 and this is the same as rolling a d6 and re-rolling once if you get a one. So in rnd(35) return a random value in [ 0 .. 34 ], the dice expression would be floor(rnd(35)/7)+1.
>
>I really need that cooffee, I forgot to account the 1/36 to get the one and nothing pop up right. I'll try again later after my coffee.

Ha coffee, let's do this again.

We want a formula that has the same probabilistic result as the event rolling a d6 and re-rolling the d6 once if the result for the first roll is 1.

The probability to get a result of 1 is 1/36. The probability to get any of the other result (2, 3, 4, 5 and 6) is 7/36.

Let rnd(36) returns a random value that is >= 0 and < 36

customroll() = floor((rnd(36)+6)/7)+1

If we use d36 as a function that return a value that is >= 1 and <= 36, we have the following

customroll() = floor((d36+5)/7)+1

I think I got it this time. It's really easy so it makes you wonder what I was thinking earlier. Another proof that coffee is indeed a drug.

Value of d36 Result
------------ ------
1 1
2 2
3 2
4 2
5 2
6 2
7 2
8 2
9 3
10 3
11 3
12 3
13 3
14 3
15 3
16 4
17 4
18 4
19 4
20 4
21 4
22 4
23 5
24 5
25 5
26 5
27 5
28 5
29 5
30 6
31 6
32 6
33 6
34 6
35 6
36 6

-----------------------------------------------------------
Éric "Space Monkey" Beaudoin
Founding Member of the Hidden-in-the-Trench Club
Release Monkey and Syntax Watchdog
>> In space, no one can hear you sleep.
>> Camels to can climb trees (and sometime eat them).
<mailto:beaudoer@...>
• ... If you keep dropping the 1 and reroll until it s not a 1 then that statistically amounts to 1d5+1 right? If you keep a 1 on the reroll then the odds get
Message 5 of 22 , Oct 1, 2005
David M. Bebber scribed:

> I use a varient of the first method: 4d6, reroll ones, then drop lowest.
> how would you code that?

If you keep dropping the 1 and reroll until it's not a 1 then that
statistically amounts to 1d5+1 right? If you keep a 1 on the reroll then the
odds get complicated, I'm not sure how that would be formulated here.

But, I don't know how integer additions fit into the syntax of this tag (or
even if they do), I tried this:

ROLLMETHOD:4d6, reroll ones, then drop lowest[tab]METHOD:roll(4,5+1,[2,3,4])

And it seems to work. But maybe not, I changed it to
METHOD:roll(4,5+10,[2,3,4]) just to see what I'd get, I expected results 30
points above average but instead I got numbers from the teens on into the
40's so I really don't know what this is doing.

One thing I have noticed is that the preference window does not remember
which one you have selected, every time you open the preferences the menu
defaults to the first method encountered in the miscinfo.lst file which is
somewhat inconvenient.
--
~ Eddy Anthony (MoSaT)
~ PCGen Content Silverback
• ... METHOD appears to accept djep expressions. For example, the 3d6 is written simply 3d6. Is there anywhere we can find the full doc for djep or even just for
Message 6 of 22 , Oct 1, 2005
At 14:47 2005.10.01, Eddy Anthony wrote:
>But, I don't know how integer additions fit into the syntax of this tag (or
>even if they do), I tried this:
>
>ROLLMETHOD:4d6, reroll ones, then drop lowest[tab]METHOD:roll(4,5+1,[2,3,4])

METHOD appears to accept djep expressions. For example, the 3d6 is written simply 3d6.

Is there anywhere we can find the full doc for djep or even just for the roll function? My memory tells me that it is a modification of the standard jep parser but I don't remember if it was done in-house or if we got the code from somewhere. There is no license text for djep.jar in the lib folder.

Anybody knows?

-----------------------------------------------------------
Éric "Space Monkey" Beaudoin
Founding Member of the Hidden-in-the-Trench Club
Release Monkey and Syntax Watchdog
>> In space, no one can hear you sleep.
>> Camels to can climb trees (and sometime eat them).
<mailto:beaudoer@...>
• ... it was contributed by one of our coders I think it was Ross Lodge. If it s useing djep, you can do all sorts of crazy stuff, even (1d2)d6 or 3d(1d4)+5
Message 7 of 22 , Oct 1, 2005
Éric Beaudoin wrote:

>At 14:47 2005.10.01, Eddy Anthony wrote:
>
>
>>But, I don't know how integer additions fit into the syntax of this tag (or
>>even if they do), I tried this:
>>
>>ROLLMETHOD:4d6, reroll ones, then drop lowest[tab]METHOD:roll(4,5+1,[2,3,4])
>>
>>
>
>METHOD appears to accept djep expressions. For example, the 3d6 is written simply 3d6.
>
>Is there anywhere we can find the full doc for djep or even just for the roll function? My memory tells me that it is a modification of the standard jep parser but I don't remember if it was done in-house or if we got the code from somewhere. There is no license text for djep.jar in the lib folder.
>
>Anybody knows?
>
>
>
it was contributed by one of our coders I think it was Ross Lodge. If
it's useing djep, you can do all sorts of crazy stuff, even (1d2)d6 or
3d(1d4)+5

it's pretty cool ;)

Devon
• ... I ve look into the doc and I saw some of the coolness you describe but I can t find any reference to the roll() function. Any clue on this? Also, if it was
Message 8 of 22 , Oct 1, 2005
At 16:10 2005.10.01, Devon Jones wrote:
>it was contributed by one of our coders I think it was Ross Lodge. If
>it's useing djep, you can do all sorts of crazy stuff, even (1d2)d6 or
>3d(1d4)+5
>
>it's pretty cool ;)
>
>Devon

I've look into the doc and I saw some of the coolness you describe but I can't find any reference to the roll() function. Any clue on this?

Also, if it was contributed by someone in the PCGen family, we should put out a license text for it and 1) recognize the contribution 2) give it a license that is compatible with the jep license if it is based on the jep source code. Do we have the source for djet BTW?

-----------------------------------------------------------
Éric "Space Monkey" Beaudoin
Founding Member of the Hidden-in-the-Trench Club
Release Monkey and Syntax Watchdog
>> In space, no one can hear you sleep.
>> Camels to can climb trees (and sometime eat them).
<mailto:beaudoer@...>
• ... I asked Greg Bingleman about this recently, he was able to pass me a couple ... This was for the KIT FUNDS tag which can do random starting gold, its about
Message 9 of 22 , Oct 1, 2005
Éric Beaudoin scribed:

> I've look into the doc and I saw some of the coolness you describe but I can't
> find any reference to the roll() function. Any clue on this?

of examples:

> Try QTY:roll("2d4")

> HARP uses:
> FUNDS:GP QTY:roll("10+1d10")

This was for the KIT FUNDS tag which can do random starting gold, its about
the only place in PCGen (besides the stat METHOD) where this would be
useful right now because any other place you would use this in a formula
PCGen would roll a new number every time you changed tabs.

As always, its on my agenda to get this in the docs along with the random
stat rolling stuff and all the other cool toys Greg's thrown in lately.
--
~ Eddy Anthony (MoSaT)
~ PCGen Content Silverback
• ... Happy to ear that. If you point me to the code where the roll function is define, I m willing to try to see what it does. I don t write Java but I can read
Message 10 of 22 , Oct 1, 2005
At 16:34 2005.10.01, Eddy Anthony wrote:
>Éric Beaudoin scribed:
>
>> I've look into the doc and I saw some of the coolness you describe but I can't
>> find any reference to the roll() function. Any clue on this?
>
>of examples:
>
>> Try QTY:roll("2d4")
>
>> HARP uses:
>> FUNDS:GP QTY:roll("10+1d10")
>
>This was for the KIT FUNDS tag which can do random starting gold, its about
>the only place in PCGen (besides the stat METHOD) where this would be
>useful right now because any other place you would use this in a formula
>PCGen would roll a new number every time you changed tabs.
>
>As always, its on my agenda to get this in the docs along with the random
>stat rolling stuff and all the other cool toys Greg's thrown in lately.

Happy to ear that.

If you point me to the code where the roll function is define, I'm willing to try to see what it does. I don't write Java but I can read it sometimes :-).

In this case, I'm very interested to find out how I could make PCGen use (roll) a dice expression a number of time and keep the x greatest (or lowest) rolls out of n. This is what is needed to implement the funky house-rules methods that some users (including me) use.

-----------------------------------------------------------
Éric "Space Monkey" Beaudoin
Founding Member of the Hidden-in-the-Trench Club
Release Monkey and Syntax Watchdog
>> In space, no one can hear you sleep.
>> Camels to can climb trees (and sometime eat them).
<mailto:beaudoer@...>
• ... Ask Greg, I can t even read it. ... Me too, me know what you learn and I ll include it in the doc entry. -- ~ Eddy Anthony (MoSaT) ~ PCGen Content
Message 11 of 22 , Oct 1, 2005
Éric Beaudoin scribed:

> If you point me to the code where the roll function is define, I'm willing to
> try to see what it does. I don't write Java but I can read it sometimes :-).

> In this case, I'm very interested to find out how I could make PCGen use
> (roll) a dice expression a number of time and keep the x greatest (or lowest)
> rolls out of n. This is what is needed to implement the funky house-rules
> methods that some users (including me) use.

Me too, me know what you learn and I'll include it in the doc entry.
--
~ Eddy Anthony (MoSaT)
~ PCGen Content Silverback
• ... Yes, the source is *inside* the jar ;) Devon
Message 12 of 22 , Oct 1, 2005
Éric Beaudoin wrote:

>At 16:10 2005.10.01, Devon Jones wrote:
>
>
>>it was contributed by one of our coders I think it was Ross Lodge. If
>>it's useing djep, you can do all sorts of crazy stuff, even (1d2)d6 or
>>3d(1d4)+5
>>
>>it's pretty cool ;)
>>
>>Devon
>>
>>
>
>I've look into the doc and I saw some of the coolness you describe but I can't find any reference to the roll() function. Any clue on this?
>
>Also, if it was contributed by someone in the PCGen family, we should put out a license text for it and 1) recognize the contribution 2) give it a license that is compatible with the jep license if it is based on the jep source code. Do we have the source for djet BTW?
>
>
>
>
Yes, the source is *inside* the jar ;)

Devon
• ... Let s pretend for a second that the only thing I know about the .jar file generation is that I have to type ant clean build test when I want to build the
Message 13 of 22 , Oct 1, 2005
At 17:28 2005.10.01, Devon Jones wrote:
>>Also, if it was contributed by someone in the PCGen family, we should put out a license text for it and 1) recognize the contribution 2) give it a license that is compatible with the jep license if it is based on the jep source code. Do we have the source for djet BTW?
>>
>>
>>
>>
>Yes, the source is *inside* the jar ;)
>
>Devon

Let's pretend for a second that the only thing I know about the .jar file generation is that I have to type "ant clean build test" when I want to build the PCGen .jar files. How can I get the source out of a .jar that I do not compile myself?

-----------------------------------------------------------
Éric "Space Monkey" Beaudoin
Founding Member of the Hidden-in-the-Trench Club
Release Monkey and Syntax Watchdog
>> In space, no one can hear you sleep.
>> Camels to can climb trees (and sometime eat them).
<mailto:beaudoer@...>
• The file you want is src/java/pcgen/util/PJEP.java Do a search for: 10+d10 That ll get you to the definition of the function, but won t be entirely helpful.
Message 14 of 22 , Oct 1, 2005
The file you want is src/java/pcgen/util/PJEP.java
Do a search for: 10+d10

That'll get you to the definition of the function, but won't be
entirely helpful. What you need to look at is
src/java/pcgen/util/DiceExpression.jj. This file defines the grammar
used to parse the dice expressions. It gets run through javaCC (Java
Compiler Compiler). The .jj file is credited to binkley, so if you'd

Byngl

--- In pcgen@yahoogroups.com, Eddy Anthony <eddyba@m...> wrote:
> Éric Beaudoin scribed:
>
> > If you point me to the code where the roll function is define,
I'm willing to
> > try to see what it does. I don't write Java but I can read it
sometimes :-).
>
>
>
> > In this case, I'm very interested to find out how I could make
PCGen use
> > (roll) a dice expression a number of time and keep the x greatest
(or lowest)
> > rolls out of n. This is what is needed to implement the funky
house-rules
> > methods that some users (including me) use.
>
> Me too, me know what you learn and I'll include it in the doc entry.
> --
> ~ Eddy Anthony (MoSaT)
> ~ PCGen Content Silverback
• ... Here s what I can tell about the roll function: We can invoke the function multiple ways. What the function do is based on the number and type of the
Message 15 of 22 , Oct 1, 2005
At 19:01 2005.10.01, Byngl wrote:
>That'll get you to the definition of the function, but won't be
>entirely helpful. What you need to look at is
>src/java/pcgen/util/DiceExpression.jj. This file defines the grammar
>used to parse the dice expressions. It gets run through javaCC (Java
>Compiler Compiler). The .jj file is credited to binkley, so if you'd
>
>Byngl

Here's what I can tell about the roll function:

We can invoke the function multiple ways. What the function do is based on the number and type of the parameters.

The type can be integer, array of integer or string.

The roll functions appear to be all defined in \src\java\pcgen\util\DiceExpressionFunctions.java and they all call RollingMethods.roll in \src\java\pcgen\core\RollingMethods.java

When using the roll function, numbers are treated as integer, list of coma separate integer enclosed in [] are treated as array of integers, characters enclosed in "" are treated as strings. I do not know for sure how a string without the "" is interpreted but I think the parser somehow tries to convert it to integer before feed the functions (e.g.. 3d6 without the "" would be seen as 3 or 6 or 36, 5+1 would probably be 6).

The possible ways to call the roll functions are:

General note: all the roll function returns the total of all rolls performed

roll(times,sides) times => integer, number of time to roll
sides => integer, type of dice to roll
e.g. roll(3,6) == 3d6

roll(sides) sides => integer, type of dice to roll
e.g. roll(6) == 1d6
note: this is one form that do not call
RollingMethods.roll but call Globals.getRandomInt

roll(times,sides,[keep])
times => integer, number of time to roll
sides => integer, type of dice to roll
[keep] => array of integer, index of the dice
we want to keep. The dices are sorted
from the lower result (index = 1) to
the largest result (index = times)
e.g. roll(4,6,[2,3,4]) == 4d6, keep the three best

roll(times,[sides]) times => integer, number of time to roll
[sides] => array of integer, possible dice to choose
from randomly
e.g. roll(4,[4,6]) == roll four times, each time you
randomly choose a d4 or a d6
note: this is one form that do not call
RollingMethods.roll but call Globals.getRandomInt
Whoever code this called it bizarre dice
rolling with good reason.

roll(times,[sides],[keep])
times => integer, number of time to roll
[sides] => array of integer, possible dice to choose
[keep] => array of integer, index of the dice
we want to keep. The dices are sorted
from the lower result (index = 1) to
the largest result (index = times)
e.g. roll(4,[4,6],[3,4])
== roll four times, each time you randomly
choose a d4 or a d6, and you keep only the
two best results
note: this is one form that do not call
RollingMethods.roll but call Globals.getRandomInt
Whoever code this called it bizarre dice
rolling with good reason.

roll("djep") "djep" => string, djep expression that gets fed to the
djep parser
e.g. roll("3d6") == 3d6

And the last but certainly not the least clear

roll(times,sides,modifier,keep,reroll_value)
times => integer, number of time to roll
sides => integer, type of dice to roll
modifier => integer, value that gets added to the total
after all the rolls are done
keep => integer, number of dices to keep, the lower
values are dropped
reroll_value => integer, if the result is not greater
then reroll_value, roll again (multiple
times is necessary)
e.g. roll(4,6,10,3,2)
== roll 4d6, reroll any 1 or 2, keep the 3
highest values, and add 10 to the total

Well, this is a bit ugly and it doesn't allow us to do everything we need. So I ask for three FREQ.

1) Please, put all the existing way to call roll in the documentation before we
forget. I really don't want to do this again (my nose is bleeding!)

2) Two new form of rolls (really, we don't nearly have enough of them)

roll(times,"djep")
roll(times,"djep",[keep])
times => integer, number of time to roll
"djep" => string, djep expression that gets fed to the
djep parser
[keep] => array of integer, index of the dice
we want to keep. The dices are sorted
from the lower result (index = 1) to
the largest result (index = times)
e.g. roll(3,"1d5+1") == evaluate the djep expression
1d5+1 three times and return the total
e.g. roll(4,"1d5+1",[2,3,4])
== evaluate the djep expression 1d5+1 four
times and return the sum of the three
highest values

maybe roll(times,"djep",keep) could also be done
times => integer, number of time to roll
"djep" => string, djep expression that gets fed to the
djep parser
keep => integer, number of dices to keep, the lower
values are dropped
e.g. roll(4,"1d5+1",3)
== evaluate the djep expression 1d5+1 four
times and keep the three highest values

I'm not sure though since it add complexity and if we have to choose, I prefer we keep the more versatile [keep] version. I think we should be OK with only three form that use "djep".

3) We deprecate all the forms that do not use djep. Let's face it, roll("3d6")
or roll(3,"d6") are way clearer then roll(3,6). The djep expressions allow
us to do way more.

-----------------------------------------------------------
Éric "Space Monkey" Beaudoin
Founding Member of the Hidden-in-the-Trench Club
Release Monkey and Syntax Watchdog
>> In space, no one can hear you sleep.
>> Camels to can climb trees (and sometime eat them).
<mailto:beaudoer@...>
• ... Oh, heh :) It s a zip file. just rename it to .zip, and you can open it with your favorite zip tool (or, if you use a command line program, you don t even
Message 16 of 22 , Oct 2, 2005
Éric Beaudoin wrote:

>At 17:28 2005.10.01, Devon Jones wrote:
>
>
>>>Also, if it was contributed by someone in the PCGen family, we should put out a license text for it and 1) recognize the contribution 2) give it a license that is compatible with the jep license if it is based on the jep source code. Do we have the source for djet BTW?
>>>
>>>
>>>
>>>
>>>
>>>
>>Yes, the source is *inside* the jar ;)
>>
>>Devon
>>
>>
>
>Let's pretend for a second that the only thing I know about the .jar file generation is that I have to type "ant clean build test" when I want to build the PCGen .jar files. How can I get the source out of a .jar that I do not compile myself?
>
>
>
Oh, heh :) It's a zip file. just rename it to .zip, and you can open it
with your favorite zip tool (or, if you use a command line program, you
don't even need to rename it to .zip for many of them).

Devon
Your message has been successfully submitted and would be delivered to recipients shortly.