PDA

View Full Version : Why are some people so bent on using....



TheMeanDM
2007-12-23, 11:03 AM
The rollers on this board?

For me, they have consistently been low, Low, LOW.

I see lots of people creating characters using the board roller, and then say "wow, those are crap"...and they truly, truly are.

Even when your point-buy has been knocked down to 32 (for whatever reason), people seem to take the PB more than the rolls on this board.

Does anyone else find that these board rollers...well...suck...for them too?

Also, sometimes even though the syntax is right, it doesn't work.

So what I can't figure out is why some people are so hellbent on making sure that people roll here?

Invisible Castle is much easier to input, copy paste is sweet, and it's verifiable.

I am not directing this at anyone in particular, just trying to figure out why people want to use this roller. The results always (for me at least) seem below average/terrible.

(4d6b3)[15]

(4d6b3)[10]

(4d6b3)[13]

(4d6b3)[7]

(4d6b3)[11]

(4d6b3)[10]

STR 12
DEX 11
CON 14
INT 6
WIS 13
CHA 13
http://invisiblecastle.com/find.py?a=show&id=1442062

(4d6b3)[14]

(4d6b3)[11]

(4d6b3)[15]

(4d6b3)[9]

(4d6b3)[11]

(4d6b3)[16]

Stat Table
STR 10
DEX 14
CON 12
INT 13
WIS 10
CHA 15

You can link to these results at
http://invisiblecastle.com/find.py?a=show&id=1442068

Everyman
2007-12-23, 11:23 AM
Perhaps you have an unfair amount of bad luck? I doubt that Rich and Co. set up the rollers to roll low, so it would have to be your odds working against you.

Hey! Maybe you're the guy who helps even out those who always seem to roll high. You know...balancing the equation and what not.

SoD
2007-12-23, 11:35 AM
Well, I've found that things that roll for you (internet, calculators, computer programs [I also like to include other people]) do tend to be lower than usual. I've heard that it's also impossible for a truely random thing like that, even though it may seems so.

Where's my proof? I've got lots. Experience, hearsay, and a post-graduate diploma from the What-I-heard-from-some-bloke-in-the-pub University.

Prophaniti
2007-12-23, 11:41 AM
Perhaps you have an unfair amount of bad luck? I doubt that Rich and Co. set up the rollers to roll low, so it would have to be your odds working against you.

Hey! Maybe you're the guy who helps even out those who always seem to roll high. You know...balancing the equation and what not.

Sorry, you're falling victim to the Gambler's Fallacy. No one's rolls on a random die (or with a random number generator) will influence any other rolls. This is a common misinterpretation of the Law of Large Numbers (you can wiki either for the full explanation). Basically, no matter how many bad rolls I get, the odds of rolling higher DO NOT increase.

I rolled up pretty decent stats for my character (in sig). Got high scores enough for the stats I needed, anyway. I haven't really noticed any tendency to roll low, any more than I do with my dice at home. Some days you just have bad dice rolls. 'Course that doesnt mean tomorrow will be any better, since both are independantly random. Isn't that a comforting thought?

Emperor Demonking
2007-12-23, 11:43 AM
How are they at all bad rolls.

If memory serves someone has proof the dice roller is fair.

Drider
2007-12-23, 11:51 AM
How are they at all bad rolls.

If memory serves someone has proof the dice roller is fair.

solo had a sig of his post where he proved it.

Winterking
2007-12-23, 11:57 AM
Sorry, you're falling victim to the Gambler's Fallacy. No one's rolls on a random die (or with a random number generator) will influence any other rolls. This is a common misinterpretation of the Law of Large Numbers (you can wiki either for the full explanation). Basically, no matter how many bad rolls I get, the odds of rolling higher DO NOT increase.

I rolled up pretty decent stats for my character (in sig). Got high scores enough for the stats I needed, anyway. I haven't really noticed any tendency to roll low, any more than I do with my dice at home. Some days you just have bad dice rolls. 'Course that doesnt mean tomorrow will be any better, since both are independantly random. Isn't that a comforting thought?

Heresy! Any die/random number generator has a limited amount of good luck inherent in it; every good roll reduces that, until there is nothing left. Fortunately, there are a variety of propitious actions and rituals which can restore the luck, including, but not limited to,
-storing the dice with the highest number up (so that weight migrates to the bottom, lowest number),
-ritual consecration in the fluid of your choice (whisky, while a sin to waste, is my preferred fluid. And the dice don't seem to mind, either).
-pre-rolling at the start of a session, to "shake the bad numbers out"
-shouts, exclamations, or odes dedicated to improving the morale, spirit, or moral bearing of the dice
-using communal dice, and waiting until someone else has rolled terribly, recharging the high numbers.
-if all else fails, rolling low numbers yourself tends to regenerate the luck, albeit at a terrible cost.

*Yes, I know none of this is actually sensible, as far as academic statistics goes. But I dislike academic statistics for many reasons. I personally think that luck will eventually balance out with any random number generator. No, any single die roll, taken on its own, is no more likely to be low than high, but if you've just rolled 19 1's, it's more than a 1/6 chance that you're going to roll a non-1, because 20 1's in a row is less likely than 19 1's in a row.
At least, that's what common sense tells me, and I'll take that over math any day...

daggaz
2007-12-23, 11:58 AM
Sorry, you're falling victim to the Gambler's Fallacy. No one's rolls on a random die (or with a random number generator) will influence any other rolls. This is a common misinterpretation of the Law of Large Numbers (you can wiki either for the full explanation). Basically, no matter how many bad rolls I get, the odds of rolling higher DO NOT increase.

I rolled up pretty decent stats for my character (in sig). Got high scores enough for the stats I needed, anyway. I haven't really noticed any tendency to roll low, any more than I do with my dice at home. Some days you just have bad dice rolls. 'Course that doesnt mean tomorrow will be any better, since both are independantly random. Isn't that a comforting thought?

I know what you mean... but it needs pretty careful explanation to be accurate. A large number of bad rolls will not influence in any way the odds of the next roll (or the next, or the next..) taken individually, but a large number of rolls period will influence the odds of rolling good at least once in that set. So if you do roll a bunch of low ones many times, you are tending towards a statistical outlier, and you are in fact, "unlucky" (or lucky).

new1965
2007-12-23, 12:36 PM
Looks like nothing more than bad luck to me

Stat Table
STR 16
DEX 13
CON 15
INT 12
WIS 11
CHA 13

You can link to these results at
http://invisiblecastle.com/find.py?a=show&id=1442155

Prophaniti
2007-12-23, 12:44 PM
Heresy! Any die/random number generator has a limited amount of good luck inherent in it; every good roll reduces that, until there is nothing left. Fortunately, there are a variety of propitious actions and rituals which can restore the luck, including, but not limited to,
-storing the dice with the highest number up (so that weight migrates to the bottom, lowest number),
-ritual consecration in the fluid of your choice (whisky, while a sin to waste, is my preferred fluid. And the dice don't seem to mind, either).
-pre-rolling at the start of a session, to "shake the bad numbers out"
-shouts, exclamations, or odes dedicated to improving the morale, spirit, or moral bearing of the dice
-using communal dice, and waiting until someone else has rolled terribly, recharging the high numbers.
-if all else fails, rolling low numbers yourself tends to regenerate the luck, albeit at a terrible cost.

Hmmm... you know, I seem to get it right when I call out the number I want an eerie amount of time... Great list, I laughed.

Icewalker
2007-12-23, 12:46 PM
Don't bring up the idea that no random generator can be truly random as evidence that the forum roller could be flawed. The way random rolls are generated off computers generally is by throwing hugely random equations at the microsecond (or some other unbelievably small measure of time) number on the computer, and coming out with something effectively random in the range. In fact, I believe it would be extremely difficult (or at least take clear intent and a bit of work) to write a roller which seemed random but tended towards lower rolls.

Corlis
2007-12-23, 12:49 PM
-shouts, exclamations, or odes dedicated to improving the morale, spirit, or moral bearing of the diceI recommend "Daddy needs a new pair of Slippers of Spider Climbing!"

Azerian Kelimon
2007-12-23, 12:57 PM
And, of course, the classic: get a patsy to roll a lot of ones for you, and the abuse the 20's.

herrhauptmann
2007-12-23, 01:20 PM
You could 'charge them with your chi'.
The only time I saw someone get triple sixes in Risk was when he shouted, "I charge them with my chi!" He had spent the 20 minutes before, and the 15 minutes after that turning the game around for the win.

VeisuItaTyhjyys
2007-12-23, 01:25 PM
I've always gotten pretty good rolls. Then, I believe in the heart of the cards.
OH NOES DND IS TOO ANIME!

mabriss lethe
2007-12-23, 01:49 PM
I get my fair share of ups and downs. Heck, one of the most fun characters I've ever played was a 1st edition elf that during creation rolled all 10s and 11s for stats. He was a "Mr. Average" sort of guy.

de-trick
2007-12-23, 01:49 PM
i don't now one player in my campaign could not get above a 20 with mod at 6 level he has a +9-12 to attack, and that was on invisible castle. Also once they had to make a strength check took multiply times I believe 8 and the DC was 15. And the party is fighter strong.

ocato
2007-12-23, 01:52 PM
The forum roller makes some fantastic sets of rolls, for other people. I'm usually looking at my 15, 13, 12, 10, 9, 9 while some other guy in the game we're applying to is trying to figure out what to do with his 3rd 18. My suggestion is typically impolite and in regards to an orifice. :smallbiggrin:

Azerian Kelimon
2007-12-23, 01:56 PM
Of course, if you learn your way 'round vbulletin, you can whip out special commands to get full 18's or the like. I remember another forum in which I once used my Aceofspades command. It opened a Media Player of Motorhead's masterpiece, and gave me full 18's. Life was good.

Kaelik
2007-12-23, 01:56 PM
The forum roller makes some fantastic sets of rolls, for other people. I'm usually looking at my 15, 13, 12, 10, 9, 9 while some other guy in the game we're applying to is trying to figure out what to do with his 3rd 18. My suggestion is typically impolite and in regards to an orifice. :smallbiggrin:

Suggest he trade it to you.

Jack_Simth
2007-12-23, 02:06 PM
hmm... how high can the forum roller go?
Let's try a lot of 10d20's...
10d20
10d20
10d20
10d20
10d20
10d20
10d20
10d20
10d20
10d20
Let's see... 10 sets of 10d20 should total something close to 1050 - hmm... but what's the standard deviation? Oh well...
Grr... I can never remember how to get the die-roller to work properly... oh well.

Randel
2007-12-23, 02:46 PM
Hmmm...

A 1st level wizard with 4 int and 4 ranks craft(alchemy) decides to make some money by crafting some flasks of acid. However, some other wizards had just flooded the market with acid a few days ago and nobody will buy (because the acid killed everyone).

So, our wizard decides to craft flasks of Alchemists Fire instead (to sell to adventurers so they can go kill those wizards that flooded the market with acid).

So, he invests 134 gp (he gets some investors) into buying raw materials and gets to work. with a +8 modifier, he needs roll at least a 12 to make it. With fair dice, he has will fail 11 out of 20 times but since he's getting back 3 times what he puts in on a successful roll he figures he can still turn a small profit until he can afford an alchemists lab. (please note he has a penalty on wisdom)

So he tries crafting 9 flasks. each costing 6 2/3 gp and should gain him 20 gp if successful. If the dice are truly fair, will succeed 9 times and lose 11 times. He needs at least 7 successful to break even.

1.1d20+8
2.1d20+8
3.1d20+8
4.1d20+8
5.1d20+8
6.1d20+8
7.1d20+8
8.1d20+8
9.1d20+8
10.1d20+8
11.1d20+8
12.1d20+8
13.1d20+8
14.1d20+8
15.1d20+8
16.1d20+8
17.1d20+8
18.1d20+8
19.1d20+8
20.1d20+8

Glawackus
2007-12-23, 02:47 PM
Clickama clickama (http://www.giantitp.com/forums/showpost.php?p=2501630&postcount=243)

My knowledge of prob/stat is pretty minimal, but this seems to make sense to me. If people really want to push this, I can ask my teacher when I'm back at school in Jan. how to do this sort of test myself, and we can run it against Myth Weavers or IC...although there are only so many ways you can make a pseudorandom number generator, so yeah. :smalltongue:

Randel
2007-12-23, 02:49 PM
wait:

1d20+8
1d20+8
1d20+8
1d20+8
1d20+8
1d20+8
1d20+8
1d20+8
1d20+8
1d20+8
1d20+8
1d20+8
1d20+8
1d20+8
1d20+8
1d20+8
1d20+8
1d20+8
1d20+8
1d20+8

Randel
2007-12-23, 02:50 PM
Maybe rolling 20 times in one post is a bit much.

1d20+8

CASTLEMIKE
2007-12-23, 03:24 PM
Worst and Best rolls out of 50 at invisible castle:

6, 6, 17, 6, 9, 10

12, 15, 18, 17, 11, 13

For most of the rolls point buy would have been a better method to make a PC.

Mark Hall
2007-12-23, 04:13 PM
Maybe your dice at home are biased, which leads to unrealistic expectations on your part?

Solo
2007-12-23, 04:14 PM
The rollers on this board?

For me, they have consistently been low, Low, LOW.

I see lots of people creating characters using the board roller, and then say "wow, those are crap"...and they truly, truly are.




Check the link in my sig, young padawan.

I have re-posted it here for your viewing pleasure.


Here's what I think about whether or not the forum roller is balanced.

Now, the mean for the forum roller should be 10 if I roll a d20, so if i rolled it a hundred times, it should average out to be 10, if the rolls are normally distributed.

In order to ensure that they are normally distributed, I shall conduct a test using 30 rolls of the d30 in order to ensure that the sample population is normally distributed.

I shall then take the average and standard deviation of the sample, and run a T-test to determine if my null hypothesis mean=10 is true or not.

Gimmie a sec to access my calculator.

(30d20)[1][5][10][5][2][20][3][2][1][19][13][5][10][9][5][18][5][1][4][18][20][17][9][5][20][5][13][16][6][13](280)
This is my sample population.

The total of these rolls is 280, which, divided by 30, is 9.3333333.

The standard deviation is calculated through formula to be 6.5828

Now, we do a T-test, with a 0.05 significance level* and 28 degrees of freedom!

T= (9.3333333-10.5)/(6.5828/30)= -0.9707

Looking this T-value up on the T-distribution critical value chart, we find that P= 0.339, which means that the probability of the forum roller giving me a set of rolls with the mean 9.3333333 is 33.97%, which, by the way, is greater than the significance level of 0.05.

Thus, we conclude that the forum roller works just fine and the mean will usually be 10.5

*The significance level of 0.05 means that if the probability of whatever outcome I have happening is less than 5%, is is statistically significant, and therefore unlikely.

I was editing it and got rid of my earlier string of rolls, but I had copied the numbers down to this post, thank the gods....

And now you know about the power of statistics.

May the Force be with you.

Lord_Asmodeus
2007-12-23, 04:15 PM
I usually get either high or low rolls, with few mid-ranging rolls, but maybe I'm just lucky :smallbiggrin:

TempusCCK
2007-12-23, 04:49 PM
The thing is, computers cannot truly do a randomized number, what happens in most cases if that a computer either uses one set of "seed" numbers that rotate through various different calculations to produce a number, and if you change the seed, you change the number produced.

However, many computers use their seed based on various numbers in the clock or any other number that can be taken from the computer. So, if you knew teh exact equations and the exact seed, you could theorhetically guess what number you're going to roll.

It could be that the forum is simply using the "bad" seed numbers when you enter your posts for the rolls. I'm not sure how the forums generators actually works.

Ne0
2007-12-23, 04:53 PM
Nope, I checked it myself recently. If you roll 1000d20, then divide the amount by 1000, you approxomately get 10.5. The average of a d20. Not precisely, but I think if you roll more of them, you'll get closer and closer to that amount.

Also, the die roller doesn't work in the OOC boards.

kamikasei
2007-12-23, 05:11 PM
How's about: if you roll below average, you'll probably fail at whatever you're doing, and therefore notice it.

If you roll above average (not over a long time, just above 10 on a single d20 roll) you've probably succeeded, so you just move on and don't take note.

If you roll super well, like a twenty or close, then you might notice it.

So the set of good rolls that will stick in your mind is smaller than the set of bad rolls; therefore subjectively you feel like you roll badly more often than well.

If you think the forum roller is actually statistically biased towards low rolls, test it. If your test contradicts that, man up and use the forum roller if the game you're playing requires it, and stop trying to game reality; you're not a Fatespinner. :smalltongue:

Panda-s1
2007-12-23, 05:31 PM
You could always use the die roller at Random.org: http://random.org/dice/

Apparently they base it off atmospheric noise, which is one of the most random things known to man. They also have an integer generator if you need to roll something that's not a d6.

15 (+2)
11 (+0)
5 (-3)
11 (+0)
15 (+2)
9 (-1)

Okay, that'd be a reroll (sum of all modifiers is 0).

11 (+0)
10 (+0)
15 (+2)
15 (+2)
14 (+2)
14 (+2)

Wow, not bad. And if this whole atmospheric noise thing is true then this really is a good random generator.

Patashu
2007-12-23, 06:12 PM
Sorry, you're falling victim to the Gambler's Fallacy. No one's rolls on a random die (or with a random number generator) will influence any other rolls. This is a common misinterpretation of the Law of Large Numbers (you can wiki either for the full explanation). Basically, no matter how many bad rolls I get, the odds of rolling higher DO NOT increase.

I rolled up pretty decent stats for my character (in sig). Got high scores enough for the stats I needed, anyway. I haven't really noticed any tendency to roll low, any more than I do with my dice at home. Some days you just have bad dice rolls. 'Course that doesnt mean tomorrow will be any better, since both are independantly random. Isn't that a comforting thought?

Ah, but if it's a roller program and not actual dice, if the random number generator is not sufficiently random past rolls could influence present rolls.

Just throwing that out there.

cupkeyk
2007-12-23, 06:22 PM
Becuase computers can't do an off hand, back hand throw.

Mr.Bookworm
2007-12-23, 06:31 PM
I would like to point out that no current system used for games actually models pure randomness. Probability, yes, but there is no way to not be able to tell how the dice (whether virtual or real) come out.

For real life dice, if one were to model how the dice are held, which face is facing up when thrown, how much force is used, the surface of the table, etc., etc., you could find out how the dice would come out. Theoretically, you could then put the exact amount of force required, along with using an exact table, and such, and be able to roll all sixes.

As has been pointed out, digital rollers aren't random either. A lot easier to mess with, too.

But, unless the program/die has some underlieing problem with it, it creates a close enough approximaty of randomness that for all intents and purposes, it is random.

So, yeah.

[/rant]