🤑😋Behold and despair!!! 😜🤑(Who needs bless pregnancy anyway?)

[icebrand]

I just rolled a bona-fide 102 points character!!!

15 STR, 13 CON, 12 SIZ, 17 INT, 16 POW, 17 DEX, 13 CHA

Clearly this NPC just became my new PC if i ever get to play 😂😂😂 (lanbril thief btw!)

Anyway, just wanted to share, never got even close to this insanity!!!

Edited September 9, 2021 by icebrand

[Baron Wulfraed]

I'm sure the +20 manipulation/weapon modifier will be put to good use...

[Nick Underwood]

20 minutes ago, icebrand said:

I just rolled a bona-fide 102 points character

Congrats!!!

I bet you didn't even spam the automatic dice roll in Phil Hibs Excel character generator!

Psychologically interesting - and I share this double bind - that we desire cheaty outcomes without cheating ourselves by actually cheating... That's some self-inflicted torment. 

[icebrand]

16 minutes ago, Nick Underwood said:

Congrats!!!

I bet you didn't even spam the automatic dice roll in Phil Hibs Excel character generator!

Psychologically interesting - and I share this double bind - that we desire cheaty outcomes without cheating ourselves by actually cheating... That's some self-inflicted torment. 

I rolled by hand, a grand total of 3 times (first two were 76 and 81 pts, and really ugly). I wouldn't be posting if i spammed a dice roller, that's not legit!!!!

[icebrand]

25 minutes ago, Baron Wulfraed said:

I'm sure the +20 manipulation/weapon modifier will be put to good use..

You are forgetting stealth!!!

[Nick Underwood]

13 minutes ago, icebrand said:

I wouldn't be posting if i spammed a dice roller, that's not legit!!!!

I know. I was hinting at my cheatass tactics, not yours. 😉

[icebrand]

29 minutes ago, Nick Underwood said:

I know. I was hinting at my cheatass tactics, not yours. 😉

Well, now that I think about it, if every npc i roll can be potentially used as a PC... I'm a cheatass too, FML 😞

[PhilHibbs]

I know a guy at poly (mid '80s) who used a home computer to randomly roll characteristics sets and print them out all day on a dot matrix printer and manually looked through them for the best ones, and claimed that it was legitimate.

[HreshtIronBorne]

53 minutes ago, Nick Underwood said:

1 hour ago, icebrand said:

I wouldn't be posting if i spammed a dice roller, that's not legit!!!!

I know. I was hinting at my cheatass tactics, not yours. 😉

This is why we have Players roll 4d6, ignore the lowest and reroll 1s. SIZ and INT are 3d6, lowest roll becomes a 6. We wanna play budding heroes most of the time. Not wee babies spending seasons and thousands of lunars on getting solid stats where you need them. A potential Rune-Lord that rolls 8 CHA has a helluva slog to get +10 CHA. 

[icebrand]

47 minutes ago, HreshtIronBorne said:

This is why we have Players roll 4d6, ignore the lowest and reroll 1s. SIZ and INT are 3d6, lowest roll becomes a 6. We wanna play budding heroes most of the time. Not wee babies spending seasons and thousands of lunars on getting solid stats where you need them. A potential Rune-Lord that rolls 8 CHA has a helluva slog to get +10 CHA. 

Completely agree!!! I had my players roll 2d6+6 for all (and they could swap 1 stat for another with my permission), but my other option was exactly like yours, but INT and SIZ were 3d6+6 drop lowest (to be honest your method seems better, i just used BGB options)

I also gave them INTx10 personal skill points (on top of the previous XP) because borderlands recommended 4 characters with 60-80% weapon skills (borderline impossible to generate with the rules) and i started with 3 players, and also gave them CHA/2 battle magic points instead of the 0-6 points you get in CE depending on previous experience.

To top it off i use 1 skill for both attack and parry, as per BRP default for a couple decades now, but again, from BGB.

Also using RQG-like rune points, but i gave everyone 3 points to start and im regretting it HARD.

[Nick Underwood]

56 minutes ago, PhilHibbs said:

I know a guy at poly (mid '80s) who used a home computer to randomly roll characteristics sets and print them out all day on a dot matrix printer and manually looked through them for the best ones, and claimed that it was legitimate.

As long as he was convinced, maybe it was legitimate... 

[Arcadiagt5]

Congrats on the lucky rolls. 🙂

It’s a matter of taste. I’ve adopted an array approach - 18-16-16-14-14-12-12 - that my players can distribute as they see fit. It’s possibly at the high power end of the spectrum but it’s consistent across all characters and I do find that it makes it easier for me to balance encounters* if I have that consistency in capability.

* Perverse Roll20 dice that make fights way more exciting than they’re supposed to be notwithstanding… 

[Baron Wulfraed]

2 hours ago, HreshtIronBorne said:

This is why we have Players roll 4d6, ignore the lowest and reroll 1s. SIZ and INT are 3d6, lowest roll becomes a 6. We wanna play budding heroes most of the time. Not wee babies spending seasons and thousands of lunars on getting solid stats where you need them. A potential Rune-Lord that rolls 8 CHA has a helluva slog to get +10 CHA. 

Wow... Generous...

Yes, I do use the Best n from n+1... But no rerolls. The dice were very favorable when I generated Varg Fenrirson (yes, another stupid name; I have a whole pack of canid named characters I've generated for practice -- Would never attempt that name in a Norse setting [Wolf, son of the Fenris Wolf?!]). He's the one that managed a +25 manipulation/weapon modifier.

[Bill the barbarian]

I must be the last person on the planet that figures to be heroes you have to be heroic, not stack the dice. I still (as I have for decades) have the players roll 3d6 for five stats (in order) and 2D6+6 for 2 stats, (also in order) and add the three bonus points if the rolls are too low and add another three for Rune affinities. One can always increase five of the stats later. If the player insists I will allow a reroll of the entire character.

[HreshtIronBorne]

Meh. Simple monkey brain like smash smash, big number, many dice!

[Dissolv]

56 minutes ago, Bill the barbarian said:

I must be the last person on the planet that figures to be heroes you have to be heroic, not stack the dice

You are not the last.   I have consistently seen in my campaigns over the......decades.......that the players who stat focus will inevitably falter during actual game play.  Those who focus on role play typically wind up with stronger characters in the long run, even, maybe even especially, when their characters are fairly meh to start with. 

Two explanations:  Those with weaker characters I see explore the rules, and the world more aggressively, while those with stronger starting characters tend to sit pat.   More importantly, the world intrinsically rewards heroic, even sometimes downright sacrificial actions.   At least my Glorantha.

[Rodney Dangerduck]

We let players assign 92 points worth of stats.  Then add 6 (3 from runes) in accordance with the rules.

[French Desperate WindChild]

2 hours ago, Rodney Dangerduck said:

We let players assign 92 points worth of stats.  Then add 6 (3 from runes) in accordance with the rules.

the amount may vary but I agree, it is better (for me) to define the same amount for all players.

 

 

[Baron Wulfraed]

18 hours ago, Baron Wulfraed said:

Wow... Generous...

Yes, I do use the Best n from n+1... But no rerolls. The dice were very favorable when I generated Varg Fenrirson (yes, another stupid name; I have a whole pack of canid named characters I've generated for practice -- Would never attempt that name in a Norse setting [Wolf, son of the Fenris Wolf?!]). He's the one that managed a +25 manipulation/weapon modifier.

Spent the morning generating 10,000 random rolls of 3D6, best 3 of 4D6, 2D6+6, and best 2 of 3D6 + 6. No, I wasn't sitting at a table with dice and a log sheet <G> Python program that randomly generated the N+1 rolls, summed the first N for the nD6 case, THEN sorted the four and selected the last N (ascending sort) for the n of n+1 cases. Wrote them to CSV file, and pulled in R statistics package (I could likely have generated the data using R, but I'm not that familiar with its programming language).

At the simplest view, the means for the best n from n+1 only went up about 1.5 points (3D6: 10.5 => B3of4: 12.2) (standard deviations actually tightened up). However, histograms do show some skew, so that mean may not be the best representation. The skew is most visible in the best 2 of 3D6 + 6 case, not as visible in the best 3 of 4D6 samples)

For the best 2 of 3, there is about a 15% chance of getting an 18, vs 33% for a 15. The straight 2D6+6 shows about 5% for 18, and 35% for a 13.

For the best 3 of 4, about a 6% chance for an 18, 14% for 14; straight 3D6 shows 1.5% for 18, and 13% for 12. (for some reason -- these plots are a pain; either the min 3 is not plotted, or the max 18 isn't).

Boxplot attached (I hope). Left two are the 3D6 variants, right two are the 2D6+6. The thick bar is the median (not mean), and the boxes span the 25-75% quartiles. Shows the spread tightened up for both "best of", and also shows how skewed the 2D6+6 became -- the median is at what was the 75% level of the straight roll mode.

[Shiningbrow]

7 hours ago, Baron Wulfraed said:

Spent the morning generating 10,000 random rolls of 3D6, best 3 of 4D6, 2D6+6, and best 2 of 3D6 + 6. No, I wasn't sitting at a table with dice and a log sheet <G> Python program that randomly generated the N+1 rolls, summed the first N for the nD6 case, THEN sorted the four and selected the last N (ascending sort) for the n of n+1 cases. Wrote them to CSV file, and pulled in R statistics package (I could likely have generated the data using R, but I'm not that familiar with its programming language).

At the simplest view, the means for the best n from n+1 only went up about 1.5 points (3D6: 10.5 => B3of4: 12.2) (standard deviations actually tightened up). However, histograms do show some skew, so that mean may not be the best representation. The skew is most visible in the best 2 of 3D6 + 6 case, not as visible in the best 3 of 4D6 samples)

For the best 2 of 3, there is about a 15% chance of getting an 18, vs 33% for a 15. The straight 2D6+6 shows about 5% for 18, and 35% for a 13.

For the best 3 of 4, about a 6% chance for an 18, 14% for 14; straight 3D6 shows 1.5% for 18, and 13% for 12. (for some reason -- these plots are a pain; either the min 3 is not plotted, or the max 18 isn't).

Boxplot attached (I hope). Left two are the 3D6 variants, right two are the 2D6+6. The thick bar is the median (not mean), and the boxes span the 25-75% quartiles. Shows the spread tightened up for both "best of", and also shows how skewed the 2D6+6 became -- the median is at what was the 75% level of the straight roll mode.

Of these, can you give totals? To compare to a point-buy.

[Nick Underwood]

On 9/10/2021 at 2:38 AM, Bill the barbarian said:

I must be the last person on the planet that figures to be heroes you have to be heroic, not stack the dice.

You got me thinking and trying to unpack that (perfectly packaged!) summation.

From a storytelling perspective the GM can reward heroic intention with heroic outcomes, and the GM can make failure every bit as rewarding as success (or more so).

But from a rules perspective, heroic intentions mean nothing and failure has no reward. Failed rolls only hinder or prevent your progresion along the path to rule-defined herodom - becoming a shaman or rune lord, getting huge stats and a CHA-full of RPs.

Of course, GMs and gaming groups blend to their liking the process of storytelling with the framework for deciding outcomes that the rules provide. But the more a group relies on the rules to decide outcomes, the more dice stacking (in one form or another) is the only path to avoiding failure and ensuring the emergence of heroic characters. (Isn't this exactly what we see in the beefed-up starting characters of RQG?) And isn't the GM storytelling technique of interpreting heroic intentions by the players as having heroic consequences just dice stacking under another guise?

[Hmm... Did I really spend the last 30 minutes to come to the conclusion that if there weren't any rules, no-one would cheat?]

Edited September 11, 2021 by Nick Underwood
clarified an ugly sentence

[Bill the barbarian]

36 minutes ago, Nick Underwood said:

[Hmm... Did I really spend the last 30 minutes to come to the conclusion that if there weren't any rules, no-one would cheat?]

Yep, and very well summarized too!

[Baron Wulfraed]

13 hours ago, Shiningbrow said:

Of these, can you give totals? To compare to a point-buy.

No, as I didn't generate characteristic sets -- only raw 3D6 (and modified 3D6), and same for 2D6+6. The first 36 (I'm sure no one wants to see all 10000) rolls are:

Quote

"r3d6"    "b3d6"    "r2d6p6"    "b2d6p6"
7    11    17    17
10    10    13    17
11    13    15    15
15    15    14    17
14    14    15    16
13    13    16    16
6    9    17    17
7    10    15    18
13    13    14    16
7    10    12    16
12    13    13    13
13    15    14    14
14    15    18    18
11    14    8    12
10    10    15    18
14    16    11    11
13    16    16    16
13    17    15    15
11    12    10    10
6    9    11    14
15    16    15    15
9    9    18    18
10    11    16    17
12    15    10    13
16    16    16    17
11    11    15    15
8    13    18    18
9    14    13    18
14    14    13    17
9    13    17    17
14    14    10    11
10    11    13    14
11    11    15    15
3    7    13    18
8    13    10    11
6    8    17    17

The headers use "r" for straight roll, "b" for "best n out of n+1". Of course, only INT and SIZ use the 2d6p6, everything else is the 3d6 -- and as the skewed plots point out, it is INT and SIZ that gets the most benefit from the "b" version. One could take five rows of the 3D6 and two rows of 2d6p6 from the above, and repeat a few times going down the list. For the shortness of the list, randomly selecting ("with replacement") a row for each characteristic would allow better results to be determined (mean and standard deviation for each pair of sets).

Okay, you got me curious.

Using the data set I'd generated for the first comparison, I just scratched up a Python program to load that set, then select rows randomly to fill five 3D6 and two 2D6+6 slots (for both straight and modified -- using the same row for each type), summed the "characteristics" and generated a few statistics on them. I generated 10000 such sets of characteristics

Straight Rolls - Mean: 78.6146  Standard Deviation: 7.463379733318169
                Max: 103        Min: 49
Modified Rolls - Mean: 90.1627  Standard Deviation: 7.1495412652862145
                Max: 114        Min: 64

I've run it a few times now, so picking different sets of rows, but the results are coming up fairly consistently -- upper 70 for straight, and around 90 for modified. Histograms attached. The results do feel a bit low, almost as if the die was 0..5, not 1..6 -- but...

def rollnd6(n):
    return [random.randint(1, 6) for x in range(n)]

For comparison, two more runs (no histograms).

Straight Rolls - Mean: 78.4442  Standard Deviation: 7.4754581536826725
                Max: 107        Min: 51
Modified Rolls - Mean: 89.9858  Standard Deviation: 7.144501574656908
                Max: 115        Min: 63

Straight Rolls - Mean: 78.4146  Standard Deviation: 7.437152544071984
                Max: 109        Min: 50
Modified Rolls - Mean: 90.003   Standard Deviation: 7.171536383049963
                Max: 118        Min: 61

Note that those results for "straight" rolls are mathematically valid. The average for a D6 is 3.5, 3D6 => 10.5, 5x 3D6 => 52.5; 2D6+6 => 13; 2x 2D6+6 => 26; 26 + 52.5 => 78.5.

If one is rolling fair dice per the rules on page 53 of the book, practically everyone is going to be adding 3pts (which only brings the average to 81.5 -- still much below the "92 points or less" clause). Even using the Modified rolls, 50% of the characters will qualify for the 3pt addition!

.1

Edited September 11, 2021 by Baron Wulfraed
Added purple text

[Bill the barbarian]

Just rolled a character using 3d6 and 2D6+6, but replace any 1s with a 6... might be the first time that I broke a 92 characteristic points ever (I was confused at not adding 3 points like always). It does create a very good character.

[Baron Wulfraed]

1 hour ago, Baron Wulfraed said:

Straight Rolls - Mean: 78.6146  Standard Deviation: 7.463379733318169
                Max: 103        Min: 49
Modified Rolls - Mean: 90.1627  Standard Deviation: 7.1495412652862145
                Max: 114        Min: 64

Just some more musing on that data set.

It is commonly accepted that, for a normal distribution, 95% of the data falls within +/- 2 standard deviations of the mean. That means each tail holds 2.5%. So...

78.6 + 2x7.5 => 93.6 -- pretty close to that 92 point cut-off in the rules for adding 3 pts. Any character with such a sum is already in the upper 2.5% of the population (and anyone with stats <63.6 is in the bottom 2.5%)

The sidebar suggests rerolling if characteristics average 12 or less -- but 7x12 is still only 84

On 9/10/2021 at 1:21 AM, Rodney Dangerduck said:

We let players assign 92 points worth of stats.  Then add 6 (3 from runes) in accordance with the rules.

So 95 points total (BEFORE RUNE modification)... Basically everyone is "Batman" (I'm not going to load up R and check books on how to extract probabilities matching values) -- probably in the top 2% if not higher.

Think I'll stick with my modified rolls (which is what my GM used in the mid 80s). They allow enough variation in characters to make some interesting to play (the occasional wimp vs the beach bully -- if anyone remembers those advertisements), while averaging just below the 92 point range (ie; an average below "Batman").  Sure, there is a potential for a "Hulk" (to cross comic franchises) to pop up. Like my "Varg Fenrirson" light cavalry at 114 (with Rune modifiers) vs "Dhi Ngo" scribe at 90 (with Rune assist) [or "Phydeaux" philosopher/sorcerer also at 90].