Aconite
Joined: Jun 03, 2004
|
  Posted:
Dec 08, 2004 - 20:12 |
|
Could also be a bit misleading saying that 6 out of 10 GFI's failed, what about everything else that went on in the game?
You can't really just pull out individual things you try and talk about or the numbers aren't fairly spread. Need to look at the games as a whole to see how bad (or good) they really were... Could be that loads of the 3+ rolls you had to make for dodges (for example) were all 6's, but you don't take notice of that, only that it was passed... |
_________________ Dejected, he hangs up the phone. "Air Holes. Remember the air holes", the online pet store owner mutters while slapping his forehead.
(Brad Osberg) |
|
Christer
Joined: Aug 02, 2003
|
Mully wrote: |
The vets gets yelled at for mocking the "random" number generator. You will find it is anything BUT random. The same numbers are rolled about 20% of the time, which statistically speaking is very un-random (average would be 16%). I coiuld live with 15-17%.
The other side of the arguement is that its "random" for both coaches. So I've resigned myself to the fact its average number generator but its the best we got. |
You've got to be kidding me...
The java random number generator is likely to be more random than your average BB player rolling blocking dice on the tabletop.
Ofcourse, you have obviously analyzed the random number generator and came up with the number 20%. I'm very much interested in viewing that analysis and how you came up with the number. I'm guessing you've written out a short string of rolls and then assumed your set is completely representative of the randomness.
Your 20% figure is most definately not correct. If it was, I'm quite convinced that the people who are likely smarter than both of us (Donald Knuth for one) would have picked it up and the Java engineers would have chosen another algorithm for the random number generator.
Thsi issue has been discussed over and over and over again and I am quite convinced that people will keep bringing it up after unlucky games. But the fact is that the random number generator that's used passes statistical analysis without any sign of differing from "true randomness" in such a magnitude that you're implying. |
|
|
nazerdemus
Joined: Nov 02, 2004
|
  Posted:
Dec 08, 2004 - 20:27 |
|
[quote="Mully"]
The vets gets yelled at for mocking the "random" number generator. You will find it is anything BUT random. The same numbers are rolled about 20% of the time, which statistically speaking is very un-random (average would be 16%). I coiuld live with 15-17%.
[quote]
so atypical rollling of 100 random generated numbers would get
20 its 1s
20 its 2s
20 its 3s
20 its 4s
20 its 5s
20 its 6s
i supose out of 100 rolls thats not that bad
if you rolled 100 dice and did get
16 of each number there would be something exceptionally wrong with your number generator .
Chaos Theory anybody |
|
|
Ironik
Joined: Jun 28, 2004
|
  Posted:
Dec 08, 2004 - 20:36 |
|
If my memory does serve me well there should be a tread where the most exemplary cases of bad luck had been moaned about. Now if anybody would trace it back... Inquisitorpustus do you think you can track that one?
bad luck sure does come in droves! I had a 100 game bad luck streak heh!
Browwnrob_ni we should see who got most jinx...
Ps an italian term for bad roll is sfiga...
Propose to be adopted to mean bad roll streak |
|
|
Delta
Joined: Aug 02, 2003
|
  Posted:
Dec 08, 2004 - 20:52 |
|
naz.....erm isn't that 120 rolled dice?
Besides, 20 of each would never be random. 1 x 6's and 119 x 1's. That's random!!
Surely the fact that some people have awful games, and some have truly brilliant ones would prove that things are random?
Anyway, let's try and put a stop to this, as this topic has been mentioned almost as many times as fouling.
BTW, Fouling...truly random or Broken? |
_________________ Cain is for Charlie and Delta is for Cain |
|
BadMrMojo
Joined: Aug 02, 2003
|
  Posted:
Dec 08, 2004 - 21:55 |
|
So I had a few minutes and I started working on the script mentioned before.
It's not ready for prime time, yet... but here's a quick sample.
Sample output wrote: | The results, in order. * represents a re-roll (either a skill or a team reroll) was used between the two rolls.
1 5 1 6 5 1 6 5 2 1 6 1 1 2 4 4 1 2 5 4 4 4 3 1 5 3 3 3 6 2 1 1 2 6 3 5 3 1 2 3 3 1 1 * 5 4 1 3 * 2 5 1 6 5 4 6 3 1 3 1 6 * 1 3 3 5 2 * 6 1 5 4 4 3 3 3 1 2 2 5 2 5 5 2 1 4 * 2 6 * 2 6 1 5 5 2 5 3 * 2 3 3 1 4 * 3 4 5 1 4 2 5 5 4 3 * 3 * 5 3 5 6 6 4 4 1 4 1 6 1 3 6 2 5 4 5 5 1 3 2 6 * 1 2 1 1 6 2 2 6 5 2 1 6 * 5 * 1 4 2 1 2 * 3 3 5 4 2 6 * 2 4 1 1 3 5 6 2 2 * 4 1 4 4 1 1 * 2 4 5 4 6 1 5 4 6 3 5 4 6 4 1 6 1 4 * 3 4 5 5 6 4 6 1 1 4 4 4 3 3 5 4 1 3 6 6 5 1 5 3 3 6 5 5 6 1 3 2 6 6 2 5 6 2 5 1 4 2 4 4 6 2 1 3 6 4 2 6 2 5 2 6 5 3 5 2 6 1 1 4 3 5 3 3 2 2 3 4 6 4 6 6 1 1 2 4 3 1 5 2 5 3 3 3 2 1 3 2 6 2 6 1 6 1 |
Looks like a whole lot of doubles, right? Well, read on...
Quote: | Your opponent rolled 57 ones, 46 twos, 47 threes, 44 fours, 47 fives and 45 sixes.
Your opponent's rerolls came up the same 1 times and different 17 times. That means they rolled the same number consecutively 5.55 % of the time.
All your opponent's rolls came up the same 46 times and different 240 times. That means they rolled the same number consevutively 16.08 % of the time.
|
|
_________________ Ta-Ouch! of BloodBowl
Condensed Guide for Newbies |
|
Glomp
Joined: Jan 04, 2004
|
  Posted:
Dec 08, 2004 - 22:02 |
|
I would Ironik, but there are so many of these threads that even my intrest in the topic has waned.
This is possibly the most pointless repeated thread here, even surpassing fouling, skill picks and whining about the administration.
I like to yabber as much as the next chimp, but this is getting incredibly annoying. |
_________________ Forum terrorist. |
|
Mully
Joined: Aug 02, 2003
|
  Posted:
Dec 08, 2004 - 22:17 |
|
Mojo - Yes - I did the same thing but manually (it was a painfull process using the log files)
And yes - it looked at every roll, agility plus armor plus infury, etc. The only thing i coiuldnt see were block dice cause i believe from memory it just showed pow, skull, etc, and i wouldnt be able to tell if a pushback was a 3 or 4.
But basically I did the same thing as you, and over a 4 game sample the same number was rolled 19.6 or 19.7% (cant remember here at work).
Granted a 4 game sample is not a representative amount. But in my already biased view it seemed to support the arguement that back to backj numbers were rolled more often than statistically normal. But I like your script very much. |
_________________ Owner of the REAL Larson
Come join the CCC League |
|
Mully
Joined: Aug 02, 2003
|
  Posted:
Dec 08, 2004 - 22:20 |
|
PS - i dont care what Bunny says about you, you are a bright fellow |
_________________ Owner of the REAL Larson
Come join the CCC League |
|
BadMrMojo
Joined: Aug 02, 2003
|
  Posted:
Dec 08, 2004 - 22:23 |
|
Cool. The source code will be on there as soon as I tidy things up, comment it and add a few extras (like noting when turns end, etc...). Maybe I'll have it throw results into a database to collect large statistical samples.
Having played you on both unlucky days and a lucky one or two, I wholehearedly concur with one thing:
Four games is not nearly enough to establish a representative sample.
ps. No one cares what Bunny says. |
_________________ Ta-Ouch! of BloodBowl
Condensed Guide for Newbies |
|
Seppuku
Joined: Oct 12, 2003
|
  Posted:
Dec 08, 2004 - 22:46 |
|
Fayeed wrote: | In one game i made a totalt of 10 GFI..and 6 times i got a 1..i rerolled 3 times (all the rerolls i had) and one of those i failed too...so a total of 7 1 on 13 rolls. |
Really, moan about bad luck before you make one about moderate luck In one run I rolled 11 consecutive GFI failures, quite a few with re-roll. The 12th I managed to pass (with re-roll if i recall) but then of the next 11 GFI's I attempted, I failed 5 more.
16/22 failed GFI's....how you like them apples?
I rarely GFI now, and when I do, I expect a failure, try to keep a re-roll in hand, and pray. Half the time I'm elated, half moaning like a $5 tart. It really does pay to expect the worst, you won't get upset as often.... |
_________________ Life ain't scary...making a 2+ Dodge with my luck...THAT's scary...
Last edited by Seppuku on %b %08, %2004 - %22:%Dec; edited 1 time in total |
|
monboesen
Joined: Aug 02, 2003
|
  Posted:
Dec 08, 2004 - 22:46 |
|
BadMrMojo
That would be an excellent addition to the stats vault. It should firmly shut up the whining crowd of "The generator is broken" whiners. And for stat nerds (present company included) it would be very interesting to have a large sample of fumbbl games broken down in that way.
Cudos to you for being able to write out such a program |
|
|
borderline
Joined: Nov 05, 2003
|
  Posted:
Dec 08, 2004 - 23:38 |
|
*bangs head against the wall* |
|
|
MorbidDeath
Joined: Mar 13, 2004
|
  Posted:
Dec 09, 2004 - 00:12 |
|
I LOVE this topic because I know what each side (and those inbetween) feel like and think.
I think the best way to sum up fumbbl - is to accept that the random generator DOES work correctly (thus being broken - ´cause anything that works to a specific goal cant be random - haha only kidding!!!!).
I guess what makes many people feel that it does not work correctly is ONE thing - that when things get "bad" they are consistently bad - and thus feels like something aint right. Of course at the other end of the battle field - we see things going consistently "right" or "good" - which aint right. Therefore it must be working.......
I think you know where i´m coming from......Although many times I curse a bad game - ask Paul hicks - I´m in the DMU room screaming at my bad luck - but also I know that a bad game is a blessing - as it means that sooner or later I HAVE to have a good one....well thats what keeps me playing (apart from I´m obsessed - and yes my girlfriend hates christer and fumbbl )
Keep your chin up buddy - soon you will get a roll where you have great dice - always comes (just seems so short when they end ) |
_________________ Believe in yourself - not in the others. |
|
Bruno
Joined: Sep 21, 2003
|
  Posted:
Dec 09, 2004 - 00:18 |
|
"the probability that one of you will remember a good roll are inversely proportional to you bitching about a poor roll"
Im just longing for the day when we get to have "Wow, I was just so lucky, the random generator got to be broken"-threads.
So, how long do you think ill have to wait for that?
And hey, dont you just go and create one right now just to prove me wrong!
Hmm guess I should make a meaningless poll about it? Or not... |
|
|
|