Constructive Malifaux Feedback

[Gruesome]

I think a full breakdown would be more interesting as we get into the SS costs where the vast majority of models are.

I think that rezzers have:

10SS: Dead Rider

9SS: Bete and Molly

8SS: Shikome and Hanged

I don't think I would use those as representative of the average defense of rezzers... But, I guess we will see.

EDIT: And no, I will not count Von Schill as a 10 point rezzer model.

[Silas Cordell]

I find comments like this intriguing, and they always make me want to look further into it. Thankfully, RottingPanda has a whole spreadsheet and has shared some of this information on the boards previously.

I found these very helpful overall when I was looking at comparisons, and I look forward to the completion of the 6ss through 2ss models. It may change once those are out, but on a quick look at the data he currently has:

(snip)

10ss models - Res have the 2nd highest average defense.

9ss models - Res have the highest average defense.

8ss models - Res have the 2nd highest, beating guild but tied with all other factions but NB

7ss models - Res have 2nd lowest, beating out Guild.

Is the terrible Df score from the lower cost models or is it perception?

Without doing any calculations (so purely perception), I would say that Resser models don't have terrible defense, but are not likely to have high defense, either. They seem to sit more in the 4-5 range, rather than some other factions where an 8 and 2 may balance out in that range, but it can make the average misleading.

I'm probably too lazy to confirm this myself, I'll admit.

[edonil]

Ressers don't have a lot of models in those ranges. We have a lot of 2-4 SS models that make up the majority of lists. An 'elite' resser list is pretty rare, because we don't have a lot of that stuff. Unless I'm remembering the faction wrong, which is totally possible.

[Boscotopia]

You have, let's say, 20% good cards and 80% bad cards. You flip one card - you still have the same ratio; you flip twenty cards, you still have the same ratio. Statistically speaking. There are minute differences, but we aren't talking about singular cases here, but trends. Also note that there is an infinite number of good cards in the deck, since if you run out, you shuffle it anew.

If what you just said was true, Vegas would have nothing to fear from card counting. Flipping a card from the deck changes the probability of all other possibilities (suit, high/low, specific value). Statistics is merely taking data and analyzing the frequency of past events, so please provide the equations and data to support your claim.

The basic rule of calculating probability states that you take the number of favorable outcomes and divide it by the total number of possible outcomes. So, let's use your example: say there are 20 cards left in the deck, 4 are high 16 are low - we want high. So, we see we have 4/20 , or 1 in 5 chance of pulling our high card. Now, the next card flipped changes the probability 1 of 2 ways. Pull the high card you wanted, that lowers your chance to 1 in 6.33. Pull a low card and your odds increase to 1 in 4.75.........it does not stay the same.

[Omenbringer]

I just cant resist...

[Malovane]

Personally I like the red joker mechanic, but i very much dislike that it can be responsible for killing your model when it should, in theory, be safe.

This is where I have to disagree. I don't think the model should ever be absolutely safe. I don't like the idea of guarenteed safety. It makes the game feel too calculated to me. I love the lack of safety and the idea that bad things will happen. Opinions will certainly vary.

I think a full breakdown would be more interesting as we get into the SS costs where the vast majority of models are.

I think that rezzers have:

10SS: Dead Rider

9SS: Bete and Molly

8SS: Shikome and Hanged

I don't think I would use those as representative of the average defense of rezzers... But, I guess we will see.

I would certainly like to see the full breakdown as well. So far, I've yet to feel like ressers have a substantially lower def. Lots of 4's and 5's. Feels fairly comparable to guild, neverborn probably ranks higher. The h2w seems to compensate for speed and general focus on melee. H2W gives the potential to survive while you close, but it doesn't guarentee it.

Regardless, I've always felt that the ressers play a meat shield & recycle kind of game. Use the cheap fodder to bring in the heavy hitters like bete noire. Then raise your losses back up again. Of course, that is often easier said then done. Trying to play the attrition game seems hit or miss. Generally feels like the aren't enough ap free or enough high crows available to do it effectively.

[edonil]

I just cant resist...

[ATTACH=CONFIG]6297[/ATTACH]

Wish I had thought of throwing that up first...

[Boscotopia]

Wish I had thought of throwing that up first...

+1

That is why probability will always be superior to statistics.

[Math Mathonwy]

If what you just said was true, Vegas would have nothing to fear from card counting. Flipping a card from the deck changes the probability of all other possibilities (suit, high/low, specific value). Statistics is merely taking data and analyzing the frequency of past events, so please provide the equations and data to support your claim.

The basic rule of calculating probability states that you take the number of favorable outcomes and divide it by the total number of possible outcomes. So, let's use your example: say there are 20 cards left in the deck, 4 are high 16 are low - we want high. So, we see we have 4/20 , or 1 in 5 chance of pulling our high card. Now, the next card flipped changes the probability 1 of 2 ways. Pull the high card you wanted, that lowers your chance to 1 in 6.33. Pull a low card and your odds increase to 1 in 4.75.........it does not stay the same.

Yes, after one card, in real life, the distribution changes, but across a million instances of flipping this one card from the deck of twenty it changes according to distribution of the deck. In other words, in 20% of the cases it changes negatively, in 80% of the cases it changes positively. In a single instance it therefore affects the chances, but we are talking about the effects of the rule across the whole game, not across a single instance.

Essentially the same thing as a dice producing an average result of 3.5 though it can never produce that result, the distribution remains the same on average though in actuality it can never remain the same in a singular real life instance.

Maybe it's easier to understand with five cards? When you pull five cards, you are, statistically, again in the same situation. Obviously in a singular situation in real life you probably aren't, since you have pulled five bad cards or two good cards and three bad cards or whatever, but across a million pulls of five cards from a deck of twenty, you are at the same situation on average.

Card counting has absolutely nothing to do with this.

To add to my resume in hopes of keeping me from posting proof (especially since that might get me banned ), I play CCGs and have done quite a bit of statistical analysis of different deck ratios, so I'm rather well-versed in deck statistics.

Edit: I do agree however, that the schemes and strategies are in more dire need of fixing, so maybe we should talk about how one should go about fixing them?

Edited July 6, 2012 by Math Mathonwy

[DeleteAccount]

Got to agree with the strat and scheme fixing. A couple of quick ones I'd love is to add to Grudge and Kill Protege that if the model that is being targetted is killed or sacrificed by his own band, it also counts to avoid too much chickanery. I'd also add to deliver the message that if no master is alive, you have to deliver it to a henchmen, if none alive, the highest costing model in play and so on.

[magicpockets]

Got to agree with the strat and scheme fixing. A couple of quick ones I'd love is to add to Grudge and Kill Protege that if the model that is being targetted is killed or sacrificed by his own band, it also counts to avoid too much chickanery. I'd also add to deliver the message that if no master is alive, you have to deliver it to a henchmen, if none alive, the highest costing model in play and so on.

Strongly disagree with this, that's half of the point of Malifaux. If I want to sacrifice my 10ss model to avoid you getting 2vp and play with a handicap that should be my decision. Also, me doing that to you is the risk you take for announcing the scheme for the extra VP - don't want the risk, you only get 1vp my friend.

[nilus]

Strongly disagree with this, that's half of the point of Malifaux. If I want to sacrifice my 10ss model to avoid you getting 2vp and play with a handicap that should be my decision. Also, me doing that to you is the risk you take for announcing the scheme for the extra VP - don't want the risk, you only get 1vp my friend.

While I don't agree with the killing you own stuff strategy on principle I don't think it needs to be changed as bad as people think. In theory these things can be done but in practice I rarely see it happen. Killing your own model depletes you of resources and can easily backfire on you.

[Gruesome]

Strongly disagree with this, that's half of the point of Malifaux.

For you.

If I want to sacrifice my 10ss model to avoid you getting 2vp and play with a handicap that should be my decision. Also, me doing that to you is the risk you take for announcing the scheme for the extra VP - don't want the risk, you only get 1vp my friend.

It seems like the debate of deliver a message(and a couple other schemes) and red jokers are just a series of people repeating themselves.

At least both sides seem to understand what the other is saying, we all just seem to be stuck in this loop pretending like if we word things a different way that the other side will change their minds.

I think that as someone that likes fluff and trying to keep things that sort of "make sense" to me that I would like the current schemes fine, even with the sacrificing, if we just changed their names and descriptions.

Instead of "Steal Relic". What if it was "Steal Soul"?

I would totally kill myself rather than get my soul stolen. (Please, no philosophy debates, Malifaux is a world with Soul Stones )

What if instead of "Kill Protege", it was "Interrogate" ?

If I were the master, especially a rezzer master, I would absolutely kill a minion before I let them divulge my plans.

(Well, as a rezzer master, I never let them KNOW my plans, but thats nit-picking)

etc...

My point is, I know its a game, but I still like when things make some sort of sense and I just cannot see my seamus or nicodem saying to themselves, "This relic sure is cool, I would rather DIE than let someone take it".

I get that Magic Pockets and others do not care about that sort of point or think that in a world with Giant Murdering Teddy bears and such that its a silly point, but I would hope that they could recognize that not everyone feels that way.

[magicpockets]

I would hope that they could recognize that not everyone feels that way.

I do, but Wyrd wrote the rules as well as the fluff. If they don't match in your opinion you should be taking that up with Wyrd, not complaining about ME when I do something in a tournament which is perfectly legit.

Sorry if that seems ranty but I don't think it's fair to criticize me for playing fairly *wink*

[nix]

I just cant resist...

[ATTACH=CONFIG]6297[/ATTACH]

There's a better than even chance you made that up.

*grin*

[Math Mathonwy]

I do, but Wyrd wrote the rules as well as the fluff. If they don't match in your opinion you should be taking that up with Wyrd, not complaining about ME when I do something in a tournament which is perfectly legit.

Sorry if that seems ranty but I don't think it's fair to criticize me for playing fairly *wink*

Strangely enough I think that Gruesome kinda is taking it up with Wyrd by posting here.

[Gruesome]

I do, but Wyrd wrote the rules as well as the fluff. If they don't match in your opinion you should be taking that up with Wyrd, not complaining about ME when I do something in a tournament which is perfectly legit.

Sorry if that seems ranty but I don't think it's fair to criticize me for playing fairly *wink*

Its unfortunate if you think I am criticizing you. I believe that my comments have been reflective of wanting the schemes mechanics to be reflective of their "fluff" descriptions and I do not believe that most masters are killing themselves to save relics.

I also offered suggestions on how you can keep the mechanics that you like in a way that makes the descriptions/fluff of the schemes match much better.

If you think I intend any comments as critical of what you like in the game, then I need to think of better ways of making my points because that is certainly not my intent.

[DeleteAccount]

Sorry if that seems ranty but I don't think it's fair to criticize me for playing fairly *wink*

May be so, but the shared deliver message thread shows that it being completely legal does not equate to people not ending with a bitter taste in their mouth. Talking about your Levi-collodi game you yourself admitted that it was not a fun game. And well, that should be the point. Also, frankly, if you get a 2 VP lead and the 8-10 soulstone model you brought with you isn't really needed to complete your conditions, then it's a no brainer.

And it's not like you are being a master tactician doing this, you just blow a turn into soulstone action which literally costs you only one action doing something your opponent has no way whatsoever of hoping to counter. With the changes I propose, you at least have to actually be tactful about what you do with your targeted model, be it using it sparingly or outright retreating while at the same time not leaving a complete sensation of "nothing I can do" towards your opponent just because he dared announce a scheme, having you perfectly know what I'm gunning for so you know where you can and can't head to is enough of an additional penalty to get the extra point.

I'm not saying you are an "evil powergamer" or anything of the sort and I agree it is completely legal, but like hell that's going to stop me from campaigning for more tweaked strategies and schemes that can't become denied in a single action.

[magicpockets]

I'm not saying you are an "evil powergamer" or anything of the sort

Am certainly not evil

And like I said, it was a GT so all bets were off. I'd never pull that in a friendly game (unless it was against UKRocky cos he deserves it)

[Boscotopia]

First, cards and dice are not the same. Apples and oranges.

Second, this is how your post reads, " Yes, after one card, in real life, the distribution changes,........but I don't want to talk about real life. "

If you do not want to show your work and your data sets, your statements have no substance and we cannot see how you arrived at your opinion.

If whatever you are doing works for you, great! Have fun and don't get stuck in the numbers, the fluff is way more enjoyable.

---------- Post added at 10:45 AM ---------- Previous post was at 10:43 AM ----------

Strongly disagree with this, that's half of the point of Malifaux. If I want to sacrifice my 10ss model to avoid you getting 2vp and play with a handicap that should be my decision. Also, me doing that to you is the risk you take for announcing the scheme for the extra VP - don't want the risk, you only get 1vp my friend.

+1

We actually agree on something, MP!

[Math Mathonwy]

I'd never pull that in a friendly game

Doesn't that kinda indicate that it is not optimally fun at the moment and that a change might be a good thing?

---------- Post added at 05:59 PM ---------- Previous post was at 05:48 PM ----------

First, cards and dice are not the same. Apples and oranges.

Second, this is how your post reads, " Yes, after one card, in real life, the distribution changes,........but I don't want to talk about real life. "

An extremely simple task for you: if you have a deck of 10 cards with two (20%) good ones and eight (80%) bad ones and you flip one card and repeat this one million times (flipping one card always from a fresh deck of ten). What is the average distribution of the nine card deck? Please, consider this with an open mind and do the math (it's very, very simple).

If you do not want to show your work and your data sets, your statements have no substance and we cannot see how you arrived at your opinion.

Work and data sets? The "dataset" is the deck. And "work" is very simple statistical analysis.

If whatever you are doing works for you, great! Have fun and don't get stuck in the numbers, the fluff is way more enjoyable.

Huh?

[Malovane]

...and repeat this one million times...

This is where things break down for me. I'm generally not concerned where the data will trend over a large/infinite sample size. I'm concerned on the effect within a very small sample size - the course of a single game - or even more often, the results within a single turn. I'm not concerned with the average distribution. I'm concerned with that immediate result and how it affects the next flip i'm going to make. I'm not going to be able to play an infinite number of games. Odds are I'm not going to flip a million cards anytime soon.

Concerning ourselves with a single case during a single turn. 10 cards remain. 2 good. 2 in 10 to get a good card - 20%. I flip a card and its not good. My next flip is now a 2/9 chance of being good - 22.2%. Clearly, a single flip will effect the probability of the following flip(s). Now look at a single possible effect of H2W-2. That first card flipped is still bad, but the following two cards are good. I'm now left with no "good" cards in my deck - all 7 remaining are "bad". H2W has had a vast impact on my ability to succeed the next flip. Every card I flip, every card I burn is having an immediate and calculable effect on the probability of what I flip next. Going up against H2W is going to either help or hinder me. I might flip 3 weaks, I might flip 3 severe, I might flip both the black & red joker. Its effect can be even more drastic on crews that require specific suits - say ressers looking for that high crow. Loosing a pair of high crows to a negative flip, or really any flip other than when needed, will hinder the odds of pulling off a summon.

There are minute differences, but we aren't talking about singular cases here, but trends.

This is exactly where I disagree, and where I believe we hit an impasse. I'm far more concerned with that singular case and not remotely concerned with the trend.

Anyway, getting away from the mathematics, since I can't see further discussion bearing any fruits, I rather enjoy the option to kill your own model. I adds a rather deep layer of strategy and risk analysis to the game. Will I be able to win this game if I remove my master? I certainly don't see it as a full-proof plan, though that rather depends on the specific match up in question. I certainly don't think it makes for the most enjoyable game. Yet, I like that the option is present. I'm all for having more strategic options and risky moves, even if it is perceived as a rather underhanded strategy. Simply playing as Hamlin could be considered an underhanded strategy by some :1_Happy_Puppet2:

[Math Mathonwy]

This is where things break down for me. I'm generally not concerned where the data will trend over a large/infinite sample size. I'm concerned on the effect within a very small sample size - the course of a single game - or even more often, the results within a single turn. I'm not concerned with the average distribution. I'm concerned with that immediate result and how it affects the next flip i'm going to make. I'm not going to be able to play an infinite number of games. Odds are I'm not going to flip a million cards anytime soon.

Concerning ourselves with a single case during a single turn. 10 cards remain. 2 good. 2 in 10 to get a good card - 20%. I flip a card and its not good. My next flip is now a 2/9 chance of being good - 22.2%. Clearly, a single flip will effect the probability of the following flip(s). Now look at a single possible effect of H2W-2. That first card flipped is still bad, but the following two cards are good. I'm now left with no "good" cards in my deck - all 7 remaining are "bad". H2W has had a vast impact on my ability to succeed the next flip. Every card I flip, every card I burn is having an immediate and calculable effect on the probability of what I flip next. Going up against H2W is going to either help or hinder me. I might flip 3 weaks, I might flip 3 severe, I might flip both the black & red joker. Its effect can be even more drastic on crews that require specific suits - say ressers looking for that high crow. Loosing a pair of high crows to a negative flip, or really any flip other than when needed, will hinder the odds of pulling off a summon.

You don't have a priori knowledge of the order of the cards. If you decide that the first card is good, then you can of course do whatever you want with the probabilities. We can also decide that you always flip Red Joker on Seamus, but that doesn't really mean anything. I'm very close to pulling out modal logic, muddy children, possible worlds and the diamond operator, but I'm afraid that it would take a bit too long to explain that way, really

10 cards, 2 good, 8 bad. You flip a card and have a 20% chance that it is good and 80% chance that it's bad. Therefore you are down 0.2 good cards and 0.8 bad cards leaving the distribution untouched. In other words, it doesn't matter how many cards you flip, the distribution is the same. It works with one case, it works with a million cases.

If you choose that you have a 100% chance of a good card, then the distribution changes. But unless you know the order of the cards, you can't know in advance whether the card will be a bad one or a good one.

The original claim was that flipping lots of cards is bad because it worsens the distribution since there are fewer good cards than bad cards. But that is not so - the distribution is unchanged no matter how many cards you flip.

I'm flabbergasted that people find deck probabilities so utterly alien.

[Guy in Suit]

Dude - just because you write long posts doesn't make you more correct.

Everyone else is talking about how to win actual games with actual cards. Boscotopia and Malovane have it.

When your count is high, you want to cycle cards. When it is low - you do not. The jokers of course add some additional nuance.

The cards left in your deck determine how you try and situate your duel totals. It's not the probability of a single flip that matters, it is how many cards you choose to flip based on the sequential probabilities that result. That's all that's really relevant.

[dgraz]

Where's the BJ in all of this? Not in the deck we're playing with?