Malimath

[Kael Hate]

I assume there are peopleout there that like me like to know statistical data so they can use the things at their fingertips to their best or at least analyse why they lose.

Let use this thread to discuss Malimath

Firstly, Has anyone made any tools yet? Before I go writing a thousand lines of code to give me the split between decks and odds, I'd like to see what others have done or maybe even use theirs if offered.

Secondly, what system do most people believe in? I am partial to a D13 system because it removes the presence of Jokers and is quicker to calculate during play. The D13 system will not count for spreads or consecutive action because you can't draw a 13 five consecutive times during play without a reshuffle. It does offer however a neutral gesture to play when a deck is currently running how or cold. A hard thing to do when Jokers alone can make a :-fate:-fate better than a :-fate flip to succeed.

[letsallchant]

wouldnt this thing you keep ranting about basically cancel itself out if both jokers are in the deck? equal chance to draw either. assuming both jokers are in the deck.

[Mr. Bigglesworth]

Would cancel, but the error rate would be high enough. Also a d13 system error rate would only increase with each flip. Too many variables because you might also be hunting for a suit. There is a "d4" variable.

Any system you make would require you to input each card seen.

You can use the black joker math and create a variable for suit dependency. But even than most people probably will struggle because of conversations are more prevalent on the gaming table. Really your best tool is a 2-4 line abacus with 13+ stones. If anyone pulled one out while playing I would walk away, doesn't sound like a fun game.

Lastly each master is dependent on certain variables, and some have more. This why a system wouldn't be universally effective without mass record keeping.

But don't let me deter you, I like math so it seems like a fun exercise.

Edited December 20, 2011 by Mr. Bigglesworth

[letsallchant]

sorry that comment sounded a bit blunt, but i have been churning this over in my head for a few days...

so if both jokers are in the deck, the probability of drawing red and black is the same, and if you draw both you lose. so your better off going for the two card flip

but... what if you have the black joker in your hand?!? its probably equally likely you would end up with the red joker in your hand, saving it for when needed. and holding onto a bad card costs you in game terms. so i dont think that makes any difference.

then theres the issue of the size of your draw pile. assuming the best case, you have the black joker in your hand or discard, and then you can fish for the red one... the chances of drawing it get worse depending on how large your draw pile is right? with almost a full draw pile, you would get very little statistical benefit from the triple flip... your probably better off getting the better average number from two cards?

and in the extreme opposite case, if you have three cards in your draw pile, and you know one of them is the red joker, the triple flip is always the best option by far.

so somewhere inbetween is where it becomes feasible to try for the triple flip?

and the target number is also a problem, if you absolutely need a high card then why not go for the triple, since you were likely to fail anyway. conversely if you only need a low or average card, two cards seems better.

anyway, stats is not my thing but it seems like there are very limited scenarios where you want to be drawing three cards instead of two on a negative flip. its like saying that your better off with a double positive flip on damage instead of a triple, in case you draw the black joker...

[Kadeton]

One of the things I like most about this game is that the statistics are really complex to work out, and every card you flip changes the probabilities of every subsequent result. Factor in the control hand, and the calculations are way beyond anything a player can work out in their head at the table.

I like this because it means that players tend to avoid trying to work out which course of action is "statistically optimal" and instead go with what "feels right". Personally, I think that's a much better way to play a game.

However, don't let me dissuade you from your Malimath. There's probably some interesting data to be gleaned from running the numbers.

[EvilGinger]

This is very good because you have to play the game & see what actually happens rather than dither & work out probabilities.

[Mr. Bigglesworth]

Joker is not that big of a deal its odds of popping are low enough where it wouldn't destroy the results. The main issue with this is suit variables. That is the hardest part of adding to the equation. If you solely rely on one suit not so bad but if you rely on two and need certain numbers this will inevitable make most equations worthless with out constant data feed.

Really only reason to bother is if you are going to play on vassals. Which to me would basically be cheating or letting a computer dictate your game.

Edited December 20, 2011 by Mr. Bigglesworth

[Justin]

If you really want to do this (which I have, a few times, for playtest purposes) it's relatively easy to calculate using changing probability. Most any good site on calculating statistics in poker will help you out.

That said, you also have to factor in the notion of "pot odds." Basically, the notion that, even if your course of action is statistically unlikely to succeed, it still may be the best one if the pay off is big enough.

For example, if you only have a 40% chance of winning a poker hand, you should still call if the bet was only 1 dollar and the pot is 100 dollars.

In Malifaux there is no hard dollar amount to analyze, so this concept is very difficult to apply. Making it, essentially, the heart of the game.

"Is leaving cover to shoot worth it?" Etc.

[nilus]

I've worked out the numbers a few ways for playtesting purposes. Its good for analysis and theoryfaux but not really practical in game. Plus when you start talking systems and hot and cold decks you start getting dangerously close to card counting theory which is frowned upon.

[LastDinosaur]

Crunched some of the numbers in a google spreadsheet, here. Took a bit fiddling, but it seems to work now, I know the calculations mightn't seem very intuitive, can elaborate on them if anyone so desires.

It's assumed that the flip is made from a full deck with all 54 cards in it.

Negative twists are included.

As you choose the card to use with positive flips those are a bit different, but if you only care for the numerical value you should just reverse the values for ace and king, two and queen and so forth (at least for the numbers in the first part).

All in all I don't think it's that useful a tool, it really only confirms what a few games worth of experience could give you anyway, had some fun doing it though, so hey.. Maybe working on it further and having opposed flips would be more useful, but still unable to take cheating into account and positive twists going for lower values to get the suit - so don't know if it'd be worth the trouble.

Anyway.. There it is should anyone find it remotely useful then it's good

[Mike3838]

I blogged on this a few months back.

http://malifauxnoob.blogspot.com/2011/03/009-maths-of-malifaux.html

You've got the same numbers as me, which is encouraging

I included positive flips, since more often than not you'll be taking the higher value, but of course choosing lower and suits/triggers makes a mess of any attempt to develop a simple system.

Mike

[Kael Hate]

I blogged on this a few months back.

http://malifauxnoob.blogspot.com/2011/03/009-maths-of-malifaux.html

You've got the same numbers as me, which is encouraging

I included positive flips, since more often than not you'll be taking the higher value, but of course choosing lower and suits/triggers makes a mess of any attempt to develop a simple system.

Mike

I like that you agree with me on Joker Math.

[Kael Hate]

Ok, a test for anyone wanting to math it up.

Nino vs Hans

Show which of the 2 is best against a spread of the Masters. Assume the opponent never has cards and is always in the open. The shooter can choose Focus or 2 shots, whenever approriate (ie vs Zoraida). Who is the better Killer?