RECON wrote:

Where do you go to see "Luck Stats"?

Game View page (eg http://www.wargear.net/games/view/36119 ) bottom row of the table (just above the Launch Player button)

Then on the Player page (eg http://www.wargear.net/games/player/36119 ) just below the WARGEAR link in the upper left (aka the breadcrumb area) is a Clover and Chart icon which pops out that same information in a window.

FAQ

Thanks Yertle, this "Luck Stats" is cool stuff.

It seems to me though that late in the game when you are attacking with 50 vs 1, over and over that it distorts the Luck stats. 

I like to see the 3 v 2 stats, 3 v 1 stats, 2 v 1 etc as it takes into acct the expectations of winning with the amount you are  attacking with and against.

Does the current system accurately reflect that/those odds?

RECON wrote:

It seems to me though that late in the game when you are attacking with 50 vs 1, over and over that it distorts the Luck stats. 

I like to see the 3 v 2 stats, 3 v 1 stats, 2 v 1 etc as it takes into acct the expectations of winning with the amount you are  attacking with and against.

Does the current system accurately reflect that/those odds?

My understanding is that yes it does take into account number of dice rolled for the attacker and defender, along with non-normal dice (ie 6vs8) and border modifiers.

Yertle wrote:

RECON wrote:

It seems to me though that late in the game when you are attacking with 50 vs 1, over and over that it distorts the Luck stats. 

I like to see the 3 v 2 stats, 3 v 1 stats, 2 v 1 etc as it takes into acct the expectations of winning with the amount you are  attacking with and against.

Does the current system accurately reflect that/those odds?

My understanding is that yes it does take into account number of dice rolled for the attacker and defender, along with non-normal dice (ie 6vs8) and border modifiers.

Your understanding is correct.  All of the dice modifiers and combinations 3v2, 3v1, etc. are taken into account as the luck is being calculated.

I'm interested in a SimulGear luck stat, seems like it would help with transparency of if it's working or not :P

Anything in the works for that tom?

I am also curious about the probabilities associated with various luck scores.  I am guessing that since the stat is "normalized" according to the sizes of the dice there is a 50% chance each roll of it going higher or lower - is this true?

In particular - Say I make an elimination run with 50 units.  At the end I have 17 units left from the 50 (the 17 is the total on all territories I took over.)  Now say my luck increased by 12 over that time - my conclusion is that I wasn't just lucky in my run because with neutral luck I would have had 17 - 12 = 5 units left behind.  Is this true?  Anything else that can be inferred from this?

Amidon37 wrote: 

In particular - Say I make an elimination run with 50 units.  At the end I have 17 units left from the 50 (the 17 is the total on all territories I took over.)  Now say my luck increased by 12 over that time - my conclusion is that I wasn't just lucky in my run because with neutral luck I would have had 17 - 12 = 5 units left behind.  Is this true?  Anything else that can be inferred from this?

This is a correct interpretation, but I have discovered that one thing is lacking from this.

Assume that you had even luck and lost those twelve armies during the run.  You would have needed to attack (at least) six more times to account for this additional loss.  These potential attacks are not considered in the current luck stat and so really a luck of 12 is not the same as twice as lucky as a luck 0f 6.

Here is an example:
14 on 14 (6v6 dice result of total attack)
Attacker lose 0, defender lose 14: 6.44
Attacker lose 2, defender lose 14: ~5.36
Attacker lose 4, defender lose 14: ~4.28
Attacker lose 6, defender lose 14: ~3.08
Attacker lose 8, defender lose 14: ~2.00
Attacker lose 10, defender lose 14: ~.92
Attacker lose 12, defender lose 12: ~-.16
(~ represents aproximately)

The luck of 6.44 says that you should have lost 6.44 more units during your 7 attacks, which would have resulted in a 7 on 7 or 8 on 6.  (These potential attacks to resolve the battle are not included).

Amidon37 wrote:

I am also curious about the probabilities associated with various luck scores.  I am guessing that since the stat is "normalized" according to the sizes of the dice there is a 50% chance each roll of it going higher or lower - is this true?

As for this question: kind of correct.
Based on a trully random distribution of dice rolls this is almost correct.
For 6v6 dice 3 on 2, it works as follows:
attacker kills 2: probability ~.37, luck .92
attacker kills 1: probability ~.34, luck -.08
attacker kills 0: probability ~.30, luck -1.08
On a single attack, there is a .64 chance of going down and .37 chance of going up (ignore rounding error there).
In the long run, there is a 50/50 change of luck going up or going down.

It's nice to see the sword/skull/clover icons make their way over, I had a fondness for them. =]

Was there a recent change to the luck stats?  It seems that now when I look, only rolls for/against me are counted, not rolls for everyone for the entire game.  I think it used to include all rolls (for a non-fog game).

Cona Chris wrote:

Was there a recent change to the luck stats?  It seems that now when I look, only rolls for/against me are counted, not rolls for everyone for the entire game.  I think it used to include all rolls (for a non-fog game).

I'm seeing the same thing, but think that it used to include all rolls too.

Are you looking in a game with fog?

Nope

I see, something definitely changed.

No, nothing should have changed - which game is this?

tom wrote:

No, nothing should have changed - which game is this?

Example: http://www.wargear.net/games/luck/48357

Can't see any problem there...

Here is a better example:
http://www.wargear.net/games/player/51296
There is no fog, but it is only showing my attacks/defends and Viper is not even on the graph.

Again it looks fine to me - Viper is on the graph, there are attacks / defends between players where they have actually attacked each other. Not much has happened in this game yet.

Not necessarily the graph (although that's a bit tough to tell I guess), but the Dice stats below the graph look incomplete.