New Dice Please

[Nehetsrev]

LynnJuniper - 10/13/2006  7:36 AM

Which is why when I make a bronze sheild at 15% on Shri, I shake my head sadly and think

"some poor sod somewhere just lost a soul strand"

That was probably Melanna...  5 SS's in her last 8 deaths and she's only at level 9.  Quit making shields!  You're killing her!  lol!

But in all seriousness, sometimes it seems like the only rolls we're getting are the low ones, but as someone else pointed out, we easilly forget all the successes we have when the RNG is rolling good for us.  Maybe it is because we only see the numbers for the failing rolls that our attention is more focussed on them, or maybe it's just Human nature to see the negative more than the positive.

[Hellblazer]

My only beef with the dice rolling evening out through the whole server is, that its not realy logical. If you are let say at 70% you should logicaly get close to 7 out of 10 but I have had time where i had 1 out of 10 even at 70% Now people will say that it evened out because someoen else got the dice thing right at the same moment ( thats what I understand out of it) But they did not dice my thing I did so why do they get my sucess  
  I'm ranting lol No truly even witht he explanation L or Dorg gave the other time I'm still unsure of the syste as by pure logic 70% is 70%

[Dorganath]

I'll remind everyone of this counter page: http://orth.ca/cgi-bin/cnr_rolls.pl
  Which shows the die roll stats for CNR attempts.  Sadly at the moment, it is only showing rolls from today, which is not enough of a sample.  I sent a message to orth to have that get the data from the archived rolls instead.  When it gets changed, you'll see the distribution of rolls is pretty even across the board.  
  @Hellblazer: The thing about  statistics is that a 70% success rate is something that falls out over a very large sampling.  Expecting 7 successes in 10 is not practical.  Though possible, it is far from guaranteed. In 10,000 attempts, however, you should expect to see something very close to 7,000 successes, give or take.
  The thing is that rolling dice is a random event.  As such: a) in order to compensate for the randomness, only a very large sampling of rolls will reflect the ideal statistical success rate b) (this is important) in any given set of rolls, there is nothing preventing a certain roll from appearing more than once before another result appears at all.  For a d20, in 20 rolls, the number 5 (for example) might appear 2-3 times and the number 17 may not appear at all.
  And of course, they're not your dice as such, but everyone's dice.  Everyone's using effectively the same random number generator, so while you may roll two 5's in a row, there's probably some other rolls in between that (again) average out statisitcally over a large sample.
  EDIT: This the above link has been fixed to show the overall die rolls

[Faldred]

Hellblazer - 10/13/2006  9:50 AM    My only beef with the dice rolling evening out through the whole server is, that its not realy logical.  If you are let say at 70% you should logicaly get close to 7 out of 10 but I have had time where i had 1 out of 10 even at 70%

Sometimes, you just have a bad day.  Other times, you'll have a great day -- for example, Zug hit 3 straight on a 50% chance for someone who wanted some silver smelting done.    What I find humorously peculiar is the way the luck seems to run... I seem to do worse than my averages at weapon crafting, but hit on low- to mid-percentages at alchemy more often than I should.

[Stephen_Zuckerman]

That's only one in four, Faldred.

Nailing seven poison sacks at 30% is... .02187% odds.

[Faldred]

Stephen_Zuckerman - 10/13/2006  11:17 AM

That's only one in four, Faldred.

Nailing seven poison sacks at 30% is... .02187% odds.

One in eight, actually.  It was the most recent example of a "lucky streak" that I could recall.

[orth]

Dorganath - 10/13/2006  10:19 AM    I'll remind everyone of this counter page: http://orth.ca/cgi-bin/cnr_rolls.pl
  Which shows the die roll stats for CNR attempts.  Sadly at the moment, it is only showing rolls from today, which is not enough of a sample.  I sent a message to orth to have that get the data from the archived rolls instead.  When it gets changed, you'll see the distribution of rolls is pretty even across the board.

 Fixed this just for reference.

[Hellblazer]

Dorganath - 10/13/2006  10:19 AM  @Hellblazer: The thing about  statistics is that a 70% success rate is something that falls out over a very large sampling.  Expecting 7 successes in 10 is not practical.  Though possible, it is far from guaranteed. In 10,000 attempts, however, you should expect to see something very close to 7,000 successes, give or take.
 

 aye I know for having work with statistic before, I was just having fun   the bigger tha sample the more accurate result you will get but always keep in mind a 2 to 5% of error.
  Still having some difficulty to understand that page but thats just me, althought would be nice to see the dice/result (thats already tehre) / #attemps.
  That might give a more caccurate feel of the succes rates.