zakatan
Joined: May 17, 2008
|
  Posted:
May 24, 2016 - 19:17 |
|
sann0638 wrote: | Just had a look at a replay file, all dice rolls have [], so you just need a way to copy and paste the log, then an Excel sheet, job done. Copy and paste is the tricky bit though, any ideas? |
you have to tamper with java security settings to allow the use of the clipboard. I don't remember exactly how it's done, but it used to be feasable. |
_________________
|
|
sann0638
Joined: Aug 09, 2010
|
  Posted:
May 24, 2016 - 19:39 |
|
I've done it into the chat window, but not from the replay. |
_________________ NAF Ex-President
Founder of SAWBBL, Wiltshire's BB League on Facebook and Discord |
|
jdm
Joined: Nov 30, 2011
|
  Posted:
May 24, 2016 - 20:37 |
|
sann0638 wrote: | Just had a look at a replay file, all dice rolls have [], so you just need a way to copy and paste the log, then an Excel sheet, job done. Copy and paste is the tricky bit though, any ideas? |
When I was recording the bonehead tally for one of my ogre teams (don't ask) I used "countif bonehead(1)"... that sort of thing. Dunno if that helps. Just pasted the whole game log at the end of the game into a column in excel and had that column as the input file for the functions. Doesn't work on replays though, for some reason you can't select the text on them |
_________________
|
|
mdd31
Joined: Oct 23, 2014
|
  Posted:
May 24, 2016 - 23:03 |
|
SidneyKidney wrote: | I was wondering if that are any tools akin to BBManager or Ones And Skulls that will work on Fumbbl games? I'm always interested to see how lucky I got in my games. |
You are new so I will help you out. If you win it is because you are really smart and the greatest coach of all time. If you lose it is because the dice were bad and your opponent got lucky.
That's Fumbbl 101. Good luck going forward. |
|
|
PainState
Joined: Apr 04, 2007
|
  Posted:
May 24, 2016 - 23:17 |
|
If you look at my previous post and the stats that I did covered basically all actions that take place. Now in a one shot game, they are meaningless stats on their own. Sure you could say that look at the pick up stats. I had 5 shots to pick up the ball and failed on 4 of them. That cost me the match. Great, now what?
The stats I did though really are the raw data stats. It was not until I got around 50+ matches where long term patterns showed up.
* If you like to pass the ball how are the long term % working? If the % is low, maybe you should stop passing.
If you take my stats you need a second level of stats.
Lets say your team is playing against elves. They are heavy bldoge, well what is your Blocking success rate with only one tackler? What is it with 3 tacklers?
You could go on and on and on with secondary stats based on very specific match ups and skill combos. Granted you need a very large pool of matches for each grouping though.
When the client was made if Kalimar added a stat calculator, lets just say like I did, that would show up at the end of the match and it was automated. Then those stats are uploaded into one huge file of all games....well, stat nirvana.
BUT
Kalimar was more concerned with getting a client up and running and not my desire for a stat log calculator, which I understand.
So I do it by hand when the whim strikes me.
Sidenote: I did these stat sheets only for the Blood Sea Buccaneers. I felt if I did it with other teams it would pollute the stat sampling with different races and styles of play. |
_________________ Comish of the: |
|
licker
Joined: Jul 10, 2009
|
  Posted:
May 24, 2016 - 23:51 |
|
PainState wrote: |
* If you like to pass the ball how are the long term % working? If the % is low, maybe you should stop passing.
|
Huh?
That you were unlucky over time with passing not succeeding as much as the probabilities suggest it should is not a reason to stop passing.
The reason to stop passing is because passing is a really dumb way to plan to move the ball. |
|
|
PainState
Joined: Apr 04, 2007
|
  Posted:
May 25, 2016 - 00:07 |
|
PainState wrote: |
I understand the immediate reaction on jump on using passing as an example but there are hundreds of action combos you could parce. |
You could take total passes, then parce it by range, LB,L,S,Q. Then parce it by pass +Catch results. Then parce it by pass+ catch + TZ/FA on the pass and catch.
Now my jaded side says, great, so what? In the end after 1 million games all this data will show us is that the RNG is not broke and performs to the expected % of "actions" over that big of a sampling. Which means on a one shot match this data is meaningless.
|
_________________ Comish of the: |
|
licker
Joined: Jul 10, 2009
|
  Posted:
May 25, 2016 - 00:33 |
|
That's not what you said though. I got your ultimate point and agree, but your blurb on passing confused me as it seemed worded differently in that you would actually try to use the results to alter your strategies.
Now, if you looked at your W/L record in games where you passed (have to take out farming passes) vs. games where you didn't pass to show that you won more in one set or the other it could actually be useful.
But still, the main point is that statistics on the dice rolls is ultimately not very useful (or interesting I would argue, but that's pretty subjective). Collecting statistics on how your players perform is fun though. I am partial to the CAS/Block stat myself, but that's just for bragging on guys who through sheer chance have a very high number there. |
|
|
Traul
Joined: Jun 09, 2013
|
  Posted:
May 25, 2016 - 14:07 |
|
The same file parser used in the replay tool should allow you to extract all the dice rolls and their target. |
|
|
|
| |