JellyBelly
Joined: Jul 08, 2009
|
  Posted:
Jun 09, 2016 - 02:34 |
|
Awesome work, Candlejack! Now I have indisputable proof when my opponents dice me! |
_________________ "Opinions are like arseholes, everybody's got them and they all stink." - The protagonist, Fallout 2
"Go for the eyes, Boo! Go for the eyes!!" |
|
WhatBall
Joined: Aug 21, 2008
|
  Posted:
Jun 09, 2016 - 04:05 |
|
Awesome tool Candlejack!
As this is associated with FUMBBL, and not for profit, I have no issue with you removing the not with permission bit about the dice images. I will double-check with Christer as well. |
_________________
|
|
WhatBall
Joined: Aug 21, 2008
|
  Posted:
Jun 09, 2016 - 04:16 |
|
If I can make one small comment on the aesthetics, which are great, I would ask you remove the extra s from all the left hand columns. E.g. 1s, 2s, 1DBs, 2DBs, etc. As you are using the plural for the title, I think it is unnecessary and just clutters up the look.
Again, very cool stuff. I can go see just how lucky I really am with my KPL team.
Was this dice, pythrr's lack of skill, or my great coaching? Hmmm... |
_________________
|
|
thoralf
Joined: Mar 06, 2008
|
  Posted:
Jun 09, 2016 - 05:12 |
|
Candlejack wrote: | And that also means that medians etc are meaningless. But if people really want it i can look into it. |
I would, for there are people who argue that the median was the best invention before sliced bread.
On a more serious note, does it also median that means are meaningless too? |
|
|
thoralf
Joined: Mar 06, 2008
|
  Posted:
Jun 09, 2016 - 05:19 |
|
Oh, and two questions:
1. What does "other 2D6s" mean?
2. How much does it cost to run this on Heroku? |
|
|
licker
Joined: Jul 10, 2009
|
  Posted:
Jun 09, 2016 - 05:38 |
|
I'm curious how difficult it would be to note how many 1d6 rolls resulted in a failure. Even better to split out 1d6s into dodges, gfi, and 'other'.
Distributions are great, but they lack a lot of context if anyone has the goal of using them to actually make any 'luck' claims. |
|
|
Lorebass
Joined: Jun 25, 2010
|
thoralf wrote: |
1. What does "other 2D6s" mean?
|
Other 2ds are like shadowing and tentacle rolls. |
|
|
Jim_Fear
Joined: May 02, 2014
|
  Posted:
Jun 09, 2016 - 06:34 |
|
After running a match or two I realize/hope that this might curb some of the dice whining. Even I had to admit that a match in which I had felt the die were against us, my opponent only threw POW's ~7%. I was simply outplayed. |
_________________
|
|
fidius
Joined: Jun 17, 2011
|
  Posted:
Jun 09, 2016 - 06:44 |
|
This is really super, thanks for all the hard work CJ!
The histograms are a decent way of graphing the data but you might consider doing a "deviation from mean" bar instead (if it's possible). ie for 1d6 "mean" is 16.67% so there's your line in the middle. Any dead-on mean would show no bar at all. If you got a die 10% of the time the bar would extend from middle to left by 6.67%; if 21% then from middle to right by 4.33%. Etc.
Really great work man, much appreciated. |
|
|
licker
Joined: Jul 10, 2009
|
  Posted:
Jun 09, 2016 - 06:59 |
|
Jim_Fear wrote: | After running a match or two I realize/hope that this might curb some of the dice whining. Even I had to admit that a match in which I had felt the die were against us, my opponent only threw POW's ~7%. I was simply outplayed. |
I agree that it can help, but it's still not perfect for that purpose (though I'm not expecting it to be).
If your team had no dodge, then it's pows+defender stumbles which matters (also depending on how much tackle your opponent blocked with).
There's also the aspect of 'wasted' POWs if they are thrown together, and 'meaningless' skulls if they are thrown with a POW (or other knock down result). You can look at 'successful' blocks to see how the combinations of block dice actually impacted the match. |
|
|
DrDiscoStu
Joined: Feb 20, 2006
|
  Posted:
Jun 09, 2016 - 07:16 |
|
Yeah right exactly, successful blocks won't indicate whether he had lots of dodge and you had stuff all tackle. Or if one side had more block.
It is a great tool, thanks so much for the post. |
_________________ Check out my fishing and camping blog.
The Black Pearl Bounty-Board.
GUARD CONQUERS ALL! |
|
Grod
Joined: Sep 30, 2003
|
  Posted:
Jun 09, 2016 - 07:21 |
|
This is awesome fun
Any chance of reducing it to a single number for each team, perhaps expressed as a percentage? We could call it the luck-meter? |
_________________ I am so clever that sometimes I don't understand a single word of what I am saying.
Oscar Wilde |
|
Medon
Joined: Jan 28, 2015
|
  Posted:
Jun 09, 2016 - 08:12 |
|
If you roll 1,1,6,6 you have average luck one could argue, but if the first roll was dauntless, the second roll was the pro reroll and the third and fourth were the two POWs from the resulting -2d block people will still claim you are lucky... the point is: dauntless and pro are excellent skills to get rid of your 1s and secretly increase your luck level |
|
|
Purplegoo
Joined: Mar 23, 2006
|
  Posted:
Jun 09, 2016 - 08:24 |
|
I look forward to checking this out later, thanks CJ! |
|
|
Candlejack
Joined: Aug 02, 2003
|
  Posted:
Jun 09, 2016 - 10:13 |
|
Thanks for all the nice words guys. I am really glad you like it. Just rolled out another bugfix for conceded games reported by Trumpkin.
I also removed the trailing 's' from the row labels as requested by the mighty WhatBall, as well as the permission bit about the images (thanks for your permission).
And since the help tool tip is obviously not obvious enough (pun of the day!) I rephrased that a bit to make it stand out more.
As for the feature requests: Actually most of them are or have been on my wish list and when I get around to do them and figure out a sensible way I will try to add them. But some are actually not that easy to do accurately because the replay data is meant to let the replayer show what happened but a lot of what the replayer shows is deduced from the current game state. Which is something I do not use, as it would mean way more complexity. But I will look into it.
As for the costs, this runs on my free Heroku dyno, so for now I am not paying anything. |
_________________ --
The Sanity Resort |
|
|