
Gandalf's thread of numbers and probability
Moderator: Area Leader
- DaigaroOgami
- Very Prolific Poster

- Posts: 830
- Joined: Mon Jun 23, 2014 8:55 am
Re: Gandalf's thread of numbers and probability
Thanks for having a look dude 

- lawastooshort
- Very Prolific Poster

- Posts: 915
- Joined: Thu Apr 17, 2014 3:34 pm
- Location: nonsense
Re: Gandalf's thread of numbers and probability
Do you have 2d6 and block dice in there?
Re: Gandalf's thread of numbers and probability
2d6 yes, injuries caused in there as well. No block dice but I do have the percentage of successful and # of turnovers.
Re: Gandalf's thread of numbers and probability
To test a statement I made, and out of interest to get some expected block success rates, I ran the numbers for the mighty undercheesers' last match (season 7 opening match).
What I actually meant was, 60-65% would be about average for the undercheesers to achieve in that match.
Now this is a relatively easy calculation to do, as the dark elves had no block-related skills, so all we need to know is what block-related skills did the undercheesers have, and how many dice they used.
Blocks with wrestle/block... 3 with 1 dice, 25 with 2 dice
Blocks without wrestle/block... 1 with 1 dice, 3 with 2 dice
all versus no block/wrestle/dodge
The multiply that matrix against the chance of success... 0.5, 0.75, 1/3 and 5/9 if you're interested, and divide by the total number of blocks.
This equates to an expected block success rate of.... 22.25/32 = 69.53%
They actually got 22/32 = 68.75%.
So contrary to the initial claim of the success rate being extraordinary, it was in fact slightly less than expected, and my intuition was a bit off. Maybe because I wanted to find a half-way house between one person saying the numbers were crazy and another saying they weren't.
A fistful of boomstick on the other hand do cause to curse Nuffle, as the zons got a 70% success rate even though they rolled more 1D blocks, the undead had some block/dodge skills, and more of the zon's blocks were with players who didn't have block or wrestle.
I'm so excited that my website will hopefully do all these sums automatically....
Gandalf wrote:Instinctively I'd say 60%-65% would be about average for 2D skilled players vs unskilled ones.
What I actually meant was, 60-65% would be about average for the undercheesers to achieve in that match.
Now this is a relatively easy calculation to do, as the dark elves had no block-related skills, so all we need to know is what block-related skills did the undercheesers have, and how many dice they used.
Blocks with wrestle/block... 3 with 1 dice, 25 with 2 dice
Blocks without wrestle/block... 1 with 1 dice, 3 with 2 dice
all versus no block/wrestle/dodge
The multiply that matrix against the chance of success... 0.5, 0.75, 1/3 and 5/9 if you're interested, and divide by the total number of blocks.
This equates to an expected block success rate of.... 22.25/32 = 69.53%
They actually got 22/32 = 68.75%.
So contrary to the initial claim of the success rate being extraordinary, it was in fact slightly less than expected, and my intuition was a bit off. Maybe because I wanted to find a half-way house between one person saying the numbers were crazy and another saying they weren't.
A fistful of boomstick on the other hand do cause to curse Nuffle, as the zons got a 70% success rate even though they rolled more 1D blocks, the undead had some block/dodge skills, and more of the zon's blocks were with players who didn't have block or wrestle.
I'm so excited that my website will hopefully do all these sums automatically....
Re: Gandalf's thread of numbers and probability
Gandalf wrote:So contrary to the initial claim of the success rate being extraordinary, it was in fact slightly less than expected, and my intuition was a bit off. Maybe because I wanted to find a half-way house between one person saying the numbers were crazy and another saying they weren't.
But that's wrong, as it's based on the assumption that every single block la made was with a player with block versus a player without block.
5 of the Dark elves team had block (4 blitzers and 1 Morg) and at least some of the Underworld blocks were against players with Block (morg got knocked over twice. I specifically remember a goblin knocking over a blitzer on a single dice block), and there's at least 3 rolls where la picked push or re-rolled where Both down was an option.
There's no way to tell what other blocks were against players with block, but I'd assume a few more.
For reference when I was talking of 2 Dice blocks which definitely failed I was talking of 2 dice blocks where the result is push or skull only. Which should happen a 1/4 of the time whatever skills you have.
8 of la's 2 dice blocks failed, 3 of those were because he didn't win when both down was rolled (i.e either my player had block or his didn't).
As I said maybe not so 'extraordinary', but la had the above average block dice and excellent D6. Whereas I had below average block dice and good D6. And there were entire turns where he didn;t fail to knock over a player when he blocked.
<Remember to put something witty in here>




Re: Gandalf's thread of numbers and probability
But that's wrong, as it's based on the assumption that every single block la made was with a player with block
I didn't make that assumption.
versus a player without block.
5 of the Dark elves team had block (4 blitzers and 1 Morg) and at least some of the Underworld blocks were against players with Block (morg got knocked over twice. I specifically remember a goblin knocking over a blitzer on a single dice block), and there's at least 3 rolls where la picked push or re-rolled where Both down was an option.
I did make that assumption, I thought no dark elves had block, sorry for that. And I remembered Morg about 5 minutes after turning my computer off last night, doh.
I thought 69% was a bit too high but I couldn't see a problem with my numbers. I should trust my intuition a bit more, or maybe not do maths late at night.
Re: Gandalf's thread of numbers and probability
Not actually blood bowl related but seemed the best place to post this...
http://www.bbc.co.uk/news/magazine-35893628
http://www.bbc.co.uk/news/magazine-35893628
Who is online
Users browsing this forum: No registered users and 1 guest
