Jump to content



Photo

New Tantive IV Crew Cards...


  • Please log in to reply
175 replies to this topic

#161 AdmiralThrawn

AdmiralThrawn

    Member

  • Members
  • 278 posts

Posted 20 May 2014 - 03:43 PM

 Don't know about you guys, but I'm flying my Corvette with the Tantive title and leia on board whether I use the abilities or not...


  • ixdta39 likes this

"Information always matters. Bad information leads to bad tactics. Incomplete information leads to flawed strategy. Both can lead to defeat."


#162 MajorJuggler

MajorJuggler

    Member

  • Members
  • 1,831 posts

Posted 20 May 2014 - 03:47 PM

With remarks like "Will someone get this walking carpet out of my way?", and replies like "No reward is worse this!", you would think that she would actually give stress to your own troops!


  • evanger and AdmiralThrawn like this

#163 Frazio

Frazio

    Member

  • Members
  • 190 posts

Posted 20 May 2014 - 03:50 PM

Leia is actually the same (ish) as the lcg version

#164 Disgruntled

Disgruntled

    Member

  • Members
  • 207 posts

Posted 21 May 2014 - 07:49 AM

Did anyone figure out the average damage for using crew han? I know the average damage for a single die is 87.5, but additional dice should reduce that a bit. A single die is simply: is it a bank? Reroll, for a 50% chance of making it a hit. But with more dice, its more complex. If you have 2 focus and a blank, you should focus the roll. But with 2 blanks and a focus, you should reroll eveything. I have no idea how to figue out exactly how many average hits a 3 or more attack die roll will have.

#165 MajorJuggler

MajorJuggler

    Member

  • Members
  • 1,831 posts

Posted 21 May 2014 - 09:24 AM

Did anyone figure out the average damage for using crew han? I know the average damage for a single die is 87.5, but additional dice should reduce that a bit. A single die is simply: is it a bank? Reroll, for a 50% chance of making it a hit. But with more dice, its more complex. If you have 2 focus and a blank, you should focus the roll. But with 2 blanks and a focus, you should reroll eveything. I have no idea how to figue out exactly how many average hits a 3 or more attack die roll will have.

 

You really have to brute-force all the different permutations. I could do it with my scripts but I haven't yet. Been way too busy.



#166 dvor

dvor

    Member

  • Members
  • 887 posts

Posted 21 May 2014 - 09:27 AM

Did anyone figure out the average damage for using crew han?

A start is on page 3: link


X-wing is played over a series of game rounds. Turn is a type of maneuver.


#167 Aminar

Aminar

    Member

  • Members
  • 2,898 posts

Posted 21 May 2014 - 09:30 AM

Did anyone figure out the average damage for using crew han? I know the average damage for a single die is 87.5, but additional dice should reduce that a bit. A single die is simply: is it a bank? Reroll, for a 50% chance of making it a hit. But with more dice, its more complex. If you have 2 focus and a blank, you should focus the roll. But with 2 blanks and a focus, you should reroll eveything. I have no idea how to figue out exactly how many average hits a 3 or more attack die roll will have.

 
You really have to brute-force all the different permutations. I could do it with my scripts but I haven't yet. Been way too busy.
Well, .875 to the third is .6699 That's the best it can ve on 3 dice. a little over 2/3rds of the rolls at max damage. But I don't believe it's quite that good.

#168 dvor

dvor

    Member

  • Members
  • 887 posts

Posted 21 May 2014 - 09:36 AM

 

 

Did anyone figure out the average damage for using crew han? I know the average damage for a single die is 87.5, but additional dice should reduce that a bit. A single die is simply: is it a bank? Reroll, for a 50% chance of making it a hit. But with more dice, its more complex. If you have 2 focus and a blank, you should focus the roll. But with 2 blanks and a focus, you should reroll eveything. I have no idea how to figue out exactly how many average hits a 3 or more attack die roll will have.

 
You really have to brute-force all the different permutations. I could do it with my scripts but I haven't yet. Been way too busy.
Well, .875 to the third is .6699 That's the best it can ve on 3 dice. a little over 2/3rds of the rolls at max damage. But I don't believe it's quite that good.

It is not.

The above assumes that the best option to choose for each individual die is identical. Which is not always true.


X-wing is played over a series of game rounds. Turn is a type of maneuver.


#169 NotBatman

NotBatman

    Member

  • Members
  • 97 posts

Posted 21 May 2014 - 09:51 AM

 "Will someone get this walking carpet out of my way?"...

 

Princess Leia:

 

Worst.

 

Ambassador.

 

Ever.

 

:)



#170 MajorJuggler

MajorJuggler

    Member

  • Members
  • 1,831 posts

Posted 21 May 2014 - 10:46 AM

dvor is right. Normally you only care about 2 outputs, [hit or critical or focus] vs [blank] (for the example of attack with focus token). This is a binomial distribution so you can use convolution to quickly find the hit PDF.

In this case, we really do care about three ouputs: [hit or crit] vs [focus] vs [blank]. You need an NxN matrix (N = #dice + 1) to uniquely define the probability of each possible outcome after the first roll. Then you need some logic to determine the best course of action for each case, and you can then use the binomial distribution from there (if you target lock).

Despite sounding very complicated its actually not that hard to do, just takes some time to code it correctly, or alternatively compute everything by hand (ouch).

Edited by MajorJuggler, 21 May 2014 - 10:47 AM.

  • ixdta39 likes this

#171 Aminar

Aminar

    Member

  • Members
  • 2,898 posts

Posted 21 May 2014 - 10:48 AM

dvor is right. Normally you only care about 2 outputs, [hit or critical or focus] vs [blank] (for the example of attack with focus token). This is a binomial distribution so you can use convolution to quickly find the hit PDF.
In this case, we really do care about three ouputs: [hit or crit] vs [focus] vs [blank]. You need an NxN matrix (N = #dice + 1) to uniquely define the probability of each possible outcome after the first roll. Then you need some logic to determine the best course of action for each case, and you can then use the binomial distribution from there (if you target lock).
Despite sounding very complicated its actually not that hard to do, just takes some time to code it correctly, or alternatively compute everything by hand (ouch).

For attacks with 2,3,4, and 5 dice... :P. Have fun.

#172 VanorDM

VanorDM

    Member

  • Members
  • 2,804 posts

Posted 21 May 2014 - 11:01 AM

For attacks with 2,3,4, and 5 dice... :P. Have fun.


I really appreciate the work that guys like MajorJuggler and Khyros do around here, crunching the numbers for us.

But there's also times I'm happy I'm not a math guy, because I didn't understand most of what MajorJuggler said above, so I'm not even remotely tempted to try it myself. :D
  • ixdta39 likes this

#173 Marten

Marten

    Member

  • Members
  • 25 posts

Posted 21 May 2014 - 11:32 AM

If you have Targeting Coordinator on the Transport you can use it to give a Corvette a TL?



#174 Aminar

Aminar

    Member

  • Members
  • 2,898 posts

Posted 21 May 2014 - 11:38 AM

For attacks with 2,3,4, and 5 dice... :P. Have fun.

I really appreciate the work that guys like MajorJuggler and Khyros do around here, crunching the numbers for us.But there's also times I'm happy I'm not a math guy, because I didn't understand most of what MajorJuggler said above, so I'm not even remotely tempted to try it myself. :D

I understand what he's saying. I just don't want to do the work, as I'd have to do it by hand. And it would be a PITA.

#175 MajorJuggler

MajorJuggler

    Member

  • Members
  • 1,831 posts

Posted 21 May 2014 - 12:18 PM

 

dvor is right. Normally you only care about 2 outputs, [hit or critical or focus] vs [blank] (for the example of attack with focus token). This is a binomial distribution so you can use convolution to quickly find the hit PDF.
In this case, we really do care about three ouputs: [hit or crit] vs [focus] vs [blank]. You need an NxN matrix (N = #dice + 1) to uniquely define the probability of each possible outcome after the first roll. Then you need some logic to determine the best course of action for each case, and you can then use the binomial distribution from there (if you target lock).
Despite sounding very complicated its actually not that hard to do, just takes some time to code it correctly, or alternatively compute everything by hand (ouch).

For attacks with 2,3,4, and 5 dice... :P. Have fun.

 

 

Not a problem. [edit: by scripts, obviously not by hand] My scripts actually do all of the above already except for the Han logic. I specifically write them to brute-force the problem with an arbitrary number of dice, because I knew that in some cases I would care about all 4 kinds of outputs (or 3 on defense dice). I suppose I could see if MATLAB can do multidimensional convolution, so I don't have to use a bunch of nested loops.

 

My scripts are actually even much more complicated than the above: I can compute the exact hit probability distribution for multiple attackers attacking the same defender, even if the defender has focus or evade tokens (yes, plural), and/or stealth device. Each attack is allowed a unique range (add attack or defense dice), and each attacker an have unique attack properties (# of dice, focus, or it can even be any kind of ordnance). I can input any number of attack and defense dice, so technically I could see what the PDFs look like, for, say, 10 attack dice vs 20 defense dice.

 

If I were more web savvy and hosted a site I would probably re-write it in a different language and make it a publicly available tool that outputs all the results both numerically and graphically. Maybe I'll do this someday.

 

OK enough techno-babble. TL;DR: I'll get around to the Han Solo odds, hopefully in the next 7 days.


Edited by MajorJuggler, 21 May 2014 - 12:22 PM.

  • Disgruntled and ObiWonka like this

#176 ObiWonka

ObiWonka

    Member

  • Members
  • 200 posts

Posted 21 May 2014 - 04:45 PM

Han still doesn't bypass the requirement to spend a Focus token to fire the Blaster Turret. So, Dark Curse and Carnor Jax still laugh at the Blaster Turret.

I realized that after the fact. At least you can still "focus" against Carnor.

 

Also, count me in being impressed by all the MathWingers. I love messing with numbers, but that kind of statistics work is something I never learned.


Rebel: 2x A-Wing, 2x B-Wing, 1x E-Wing, 2x HWK-290, 4x X-Wing, 2x Y-Wing, 1x YT-1300, 1x Z-95 Headhunter

Imperial: 1x Firespray-31, 1x Lambda Shuttle, 1x TIE Defender, 6x TIE Fighter, 1x TIE Advanced, 2x TIE Bomber, 6x TIE Interceptor, 1x TIE Phantom

BEARFACE





© 2013 Fantasy Flight Publishing, Inc. Fantasy Flight Games and the FFG logo are ® of Fantasy Flight Publishing, Inc.  All rights reserved.
Privacy Policy | Terms of Use | Contact | User Support | Rules Questions | Help | RSS