Math Mathonwy Posted January 18, 2015 Report Share Posted January 18, 2015 These are approximations, and they ignore the effects of jokers, of stacking your deck with card filtrationAs an aside, my program can take that into account as well 1 Quote Link to comment Share on other sites More sharing options...
Adran Posted January 19, 2015 Report Share Posted January 19, 2015 Sheesh here I am believing in the heart of the cards. If I flip two deuces then I added +2 to my value my tiny little mind does not comprehend well my Sh 5 is sorta like shoot 7 since I have a . To me it's shoot 5 that has a better chance at flipping higher cards. Or mitigates cover. Damn technobabble Personally the I like better than +2 stat. Mitigating cover is amazing. Being sh7 is fine but burning a stone to be put on negative to damage (generally speaking. Yeah not worth it. Specially when damage profile is 2/3/5. Also ryle is tankier and hits harder when engaged. So there is that. Flip Sidirs Sh 54 times, add up all the final scores, and divide by 54, and you will end up with a mean value of 14, giving a mean flip of 7. It has a range of values from 7-21. and 8-20 are all equally likely, but half the time you will get a result of 14 or better. Flip Ryles sh 54 times, add up the final scores and divide by 54, and you end up with a mean value of 14.a little bit. Giving a mean flip of 9.something It has a range of values between 5 and 19, but rather than an even distributuion that we had on the straight flip, you are much more likley to get the higher numbers as your final total than the lower numbers. But just over half the time you should get a result of 14 or better. Math could probably put up a distribution curve of the likely results, but many people like to work on the rule of thumb that they will get their mean value( Which they should half the time). This obviously doesn't take into account the ability to cheat, it is just based on flipping randomly from the deck. 1 Quote Link to comment Share on other sites More sharing options...
The Godlyness Posted January 19, 2015 Report Share Posted January 19, 2015 And you brainiacs take all this into account before choosing what you are going to do? How long are your games lol. I see something I want to attack I flip cards and hope for the best. This seems to win me games. So why yes I will never really want to comprehend mathhammer Or mathifaux I just sling dice (or used to) and flip cards. My strength lies in positioning sacrifice and profit. So to me the value of a is: Mitigating cover, flipping more cards for higher ones or burning two out of my deck. The only numerical value I see is what's flipped. 1 Quote Link to comment Share on other sites More sharing options...
Myyrä Posted January 19, 2015 Report Share Posted January 19, 2015 Math could probably put up a distribution curve of the likely results, but many people like to work on the rule of thumb that they will get their mean value( Which they should half the time).The distribution isn't symmetrical so the mean isn't actually same as the meadian. Quote Link to comment Share on other sites More sharing options...
LunarSol Posted January 19, 2015 Report Share Posted January 19, 2015 And you brainiacs take all this into account before choosing what you are going to do? How long are your games lol. Nope, which is why its helpful to narrow things down into gross simplifications like using a D13 or averaging things down to a +2ish concept. Neither of these things are exactly right but they're within proximity to where they're roughly as useful. After all, when talking odds, it's hard to come across anyone that immediately recognizes a 65% chance of success as failing more than a third of the time. Calculating the odds is more about providing guidelines to what constitutes a good use of limited resources. With a little experience, its easy to see when spending AP to have a low stat model attack a high df value is futile and you'd be better off ignoring them and coming up with another plan. Having a statistical guideline for less straightforward probabilities like makes it easier to properly utilize those models, as well as resources like Soulstones and the Focus action. 1 Quote Link to comment Share on other sites More sharing options...
Math Mathonwy Posted January 19, 2015 Report Share Posted January 19, 2015 And you brainiacs take all this into account before choosing what you are going to do? How long are your games lol.I see something I want to attack I flip cards and hope for the best. This seems to win me games.So why yes I will never really want to comprehend mathhammer Or mathifaux I just sling dice (or used to) and flip cards. My strength lies in positioning sacrifice and profit.So to me the value of a is: Mitigating cover, flipping more cards for higher ones or burning two out of my deck. The only numerical value I see is what's flipped.People who deny that they don't do any "mathhammer" simply don't understand the concept. Of course you do math - otherwise you would really suck at the game. But you've internalized the simple stuff and now you're refusing to internalize the slightly more complicated stuff because of some misinformed belief that math is bad and intuition is good. When they actually are essentially the same thing.If you didn't do any "mathhammer" you would be equally happy shooting into hard cover as not into hardcover. After all, you might do just as much damage! It's just a matter of luck! Or you would attack Lilith's Df as happily as her Wp - after all, you might score a hit either way, so what's the difference?Of course the difference is in the likelihood of you succeeding. And that's all this is. All that "mathhammer" is. Understanding the power of a is just that.Am I better off taking the or the +2 to Ca as Sonnia? When should I choose which? What do I have to consider when making that decision? You will likely know the answer when presented with a situation. All thanks to the awesome power of "mathhammer!" 3 Quote Link to comment Share on other sites More sharing options...
Mastershake Posted January 19, 2015 Report Share Posted January 19, 2015 The value depends on a lot of factors, but mainly the defense of the model you're trying to hit. Other than that, you have to take into account how you want to use your cards. If it is a duel you intend or are willing to cheat, then the higher stat is better. If it's a duel you're not likely to cheat, the + is better. When you take into account things like the defenses of your target (both cover and Df value itself), a lot more gets thrown into the mix. For example, trying to shoot Lilith (Df 7). If you're trying to kill her for, say, assassinate, then you're willing to throw high cards to hit her. In this case, Sidir is better because the Sh 7 is much more likely to hit on the cheat than Sh 5. However, if she's just in your way and you don't need to kill her, you might not intend to cheat. In that case, Sh 5 + is better because you're more likely to have a high starting value (forcing Lilith to cheat) without having to cheat yourself. Honestly, there's just a lot that goes into those considerations in a game like this. I tend to prefer the high stats, but part of that is because I tend to run crews with very few models (and therefore it's easier to allocate resources). The more models I run, the fewer cards I have for each individually... and the more I would want + to my flips. /shrug This kind of gets of to the heart of the matter because unlike a game like Warmachine where it's easier to check the math on something like +1D6 remove the highest there's a lot of variables involved that can dramatically improve or degrade the value of a positive twist. An example would be if you had a mob of Guild Guards with Dashel. They can easily get single and even double positives, but their attack stat is only a 4, so anything Defense 6 is going to be problem to hit even with all the bonuses. OTOH, those Guild Guards will start a fair number of their duels beating their targets even at Defense 6, so they can certainly force the other person to give up cards they normally wouldn't. Negative twists also have a huge impact on the value of a positive twist since you can get back to neutral without having to spend AP focusing. Rule of thumb 6-8ss or less, you want positives. Most of these models don't have OMG damage tracks and you usually aren't relying on them to carry your game so positives give them a good chance of participating even if you don't feel like spending cards for it and in turn give you a better chance of having a good return on investment. 9+ss are usually models you buy for their ability to wreck other models (obviously there are always exceptions), so a higher base stat often has more value (if they also get a positive great), but I've seen Rail Golems fall apart a lot on key models because of their somewhat mediocre Ml stat. Anything that gives out twist though, a simple is worth it's weight in gold, just ask anyone who's ever played against Yin or Jack Daw. Quote Link to comment Share on other sites More sharing options...
LunarSol Posted January 19, 2015 Report Share Posted January 19, 2015 This kind of gets of to the heart of the matter because unlike a game like Warmachine where it's easier to check the math on something like +1D6 remove the highest there's a lot of variables involved that can dramatically improve or degrade the value of a positive twist. An example would be if you had a mob of Guild Guards with Dashel. They can easily get single and even double positives, but their attack stat is only a 4, so anything Defense 6 is going to be problem to hit even with all the bonuses. OTOH, those Guild Guards will start a fair number of their duels beating their targets even at Defense 6, so they can certainly force the other person to give up cards they normally wouldn't. This isn't actually all that true. Both games actually run on what is a 2DX system. Warmachine operates on (2D6 + MyStat) vs (YourStat + 7) where Malifaux operates on something kind of like a (D13 + MyStat) vs (YourStat + D13). Both produce bell curve results: Warmachine's is just from 2-12 and relies on the 7 point offset while Malifaux's is curved from -12 to +12. For what its worth, Malifaux has a much higher variance, but much tighter bounds on the effect of that variance. In the Guild Guard example, the Guard succeeds in all combinations of MyCard - YourCard that total +2 or higher. There are 169ish combinations of cards on a straight flip and 66 of them result in MyCard being 2 higher than your card. I can flip the 3, but you can still flip the Ace after all. When you add a or to the mix you're still looking at a set of combinations; that set has just risen to 2197 or 28561 with a higher percentage of the combinations adding up to +2 due to the number of times a bad card is replaced with a much higher card. Cheating, however, is the true wildcard to chance here. The number of cards pulled adding to +2 drops dramatically if your opponent is willing to spend a resource to set their value to a constant like 10 or 11. 1 Quote Link to comment Share on other sites More sharing options...
The Godlyness Posted January 19, 2015 Report Share Posted January 19, 2015 All that to me equates to guild guard suck at shooting. Why do they suck well they have a sh of 4. Most things have df 5. So already lower chance of hitting. So with them my target priority is to shoot at things with pure luck maybe I will hit foci do great. If I don't pot shots are pot shots for a reason. @math yes math is integral part of this game. I understand that. And yes I will shoot into hard cover if I have a . Since now I am on a straight and can cheat it. Not because it gives me an ephemeral number that doesn't apply. If I don't have that then I can focus if the target is priority. Or flip cards. And hope for the best. Which sometime pays unforseen dividends. @Jack daw I smile all the time with rifleman. What ? I have seen way to often flip low and miss consistently. Also ca9 to anything means I have good chance to hit. But as you said simple math. I don't go well the probability of me flipping xyz to hit them is this... since I am +3-5 on any card. Which is huge. Simple. Making it into mathematical formulas which don't intrinsically effect what your doing seems like an exercise in futility. Lol. So at the original poster. Need a tanky model that has built in to attack and generally damage. Hits like a truck in melee and has some movement shenanigans and a heal. And this complements your crew take Ryle. But if you a more accurate gun that will suffer from cover can take an upgrade to take hits has a decent melee but not super tanky. Take sidir if he complements your crew in a way you like. Don't take one or the other cause of magic numbers !! Yu-Gi-Oh that shit and believe in the heart of the cards. (That show/card game is terribad but it did leave it's Mark.) Also learn target priority positioning and sacrifice. Deploying poorly will cost you the game more often then cards. Quote Link to comment Share on other sites More sharing options...
Mastershake Posted January 19, 2015 Report Share Posted January 19, 2015 This isn't actually all that true. Both games actually run on what is a 2DX system. Warmachine operates on (2D6 + MyStat) vs (YourStat + 7) where Malifaux operates on something kind of like a (D13 + MyStat) vs (YourStat + D13). Both produce bell curve results: Warmachine's is just from 2-12 and relies on the 7 point offset while Malifaux's is curved from -12 to +12. For what its worth, Malifaux has a much higher variance, but much tighter bounds on the effect of that variance. In the Guild Guard example, the Guard succeeds in all combinations of MyCard - YourCard that total +2 or higher. There are 169ish combinations of cards on a straight flip and 66 of them result in MyCard being 2 higher than your card. I can flip the 3, but you can still flip the Ace after all. When you add a or to the mix you're still looking at a set of combinations; that set has just risen to 2197 or 28561 with a higher percentage of the combinations adding up to +2 due to the number of times a bad card is replaced with a much higher card. Cheating, however, is the true wildcard to chance here. The number of cards pulled adding to +2 drops dramatically if your opponent is willing to spend a resource to set their value to a constant like 10 or 11. In Warmachine there's a static value you're trying to hit, in Malifaux it changes with every attack. Even if you don't factor cheating in a model with a higher stat can simply flip unbeatable numbers off the top of the deck. Where in WM I know I have an X% chance of hitting a DEF X model with a RAT X model and modifiers change that by a static amount, in Malifaux a positive twist working out to a roughly a +2 could still mean missing every attack whereas an actual +2 may be able to hit...if that makes any sense. Quote Link to comment Share on other sites More sharing options...
decker_cky Posted January 19, 2015 Report Share Posted January 19, 2015 Malifaux also relies on a series of dependent flips (flip A affects the probability of flip , whereas dice games are series of independent events. It is completely correct in Malifaux to think "my opponent just flipped a pile of face cards, so he's likely to flip lower with the remainder of his deck" (assuming you don't only notice the high cards). 1 Quote Link to comment Share on other sites More sharing options...
LunarSol Posted January 19, 2015 Report Share Posted January 19, 2015 In Warmachine there's a static value you're trying to hit, in Malifaux it changes with every attack. Even if you don't factor cheating in a model with a higher stat can simply flip unbeatable numbers off the top of the deck. Where in WM I know I have an X% chance of hitting a DEF X model with a RAT X model and modifiers change that by a static amount, in Malifaux a positive twist working out to a roughly a +2 could still mean missing every attack whereas an actual +2 may be able to hit...if that makes any sense. Yes.... and no. Opposed rolls create the impression this is true, but opposed rolls are still combinations of 2 individual random results using subtraction instead of addition. There's a sense that your opponent's result controls the probability of your success, but this is no more true than rolling and combining dice. If I need an 8 on 2d6 and I roll them individually, what's my chance of success if the first die is a 1? Quote Link to comment Share on other sites More sharing options...
Mastershake Posted January 19, 2015 Report Share Posted January 19, 2015 Yes.... and no. Opposed rolls create the impression this is true, but opposed rolls are still combinations of 2 individual random results using subtraction instead of addition. There's a sense that your opponent's result controls the probability of your success, but this is no more true than rolling and combining dice. You are incorrect. If I have Ml 4 and my Df 6 opponent flips a 13, the duel is over, his result did affect my chances of success and in fact made it impossible. Having a moving target along with the fact that no deck will ever be complete (some cards will be in your hand, flipped for initiative on previous duels or trashed from last turn etc) means trying to apply a static value to something like a positive is, at best misguided. Quote Link to comment Share on other sites More sharing options...
LunarSol Posted January 19, 2015 Report Share Posted January 19, 2015 You are incorrect. If I have Ml 4 and my Df 6 opponent flips a 13, the duel is over, his result did affect my chances of success and in fact made it impossible. Having a moving target along with the fact that no deck will ever be complete (some cards will be in your hand, flipped for initiative on previous duels or trashed from last turn etc) means trying to apply a static value to something like a positive is, at best misguided. The difference here is that you're focusing on the probability of only your half of the flip. I'm stating the probability of the duel succeeding incorporating both flips. Opposed rolls create the impression of my luck vs your luck but from the moment of the last decision point (declaring the action) both individual rolls are cast to determine the outcome. It's no more a moving target than rolling a pair of dice 1 at a time. It's simply a matter of perception. Yes it's possible the second card is high enough to negate the first but it's also possible the second die is too low for the first to matter. It's all a matter of combinations and the distribution of results among them. Quote Link to comment Share on other sites More sharing options...
doctorlucky Posted January 19, 2015 Report Share Posted January 19, 2015 really interesting conversation- thanks for sharing all... I'm gonna go with 'rule of thumb, positive twists ~+2' Assuming a negative twist is the same? Quote Link to comment Share on other sites More sharing options...
Mastershake Posted January 19, 2015 Report Share Posted January 19, 2015 The difference here is that you're focusing on the probability of only your half of the flip. I'm stating the probability of the duel succeeding incorporating both flips. Opposed rolls create the impression of my luck vs your luck but from the moment of the last decision point (declaring the action) both individual rolls are cast to determine the outcome. It's no more a moving target than rolling a pair of dice 1 at a time. It's simply a matter of perception. Yes it's possible the second card is high enough to negate the first but it's also possible the second die is too low for the first to matter. It's all a matter of combinations and the distribution of results among them. A RAT 5 model shoots at a DEF 13 model. You need an 8 and always will need an 8 unless outside factors increase that. This means you can apply a relative value to things like boosting or adding a die and removing the lowest/highest that suitably incorporates your increased probability. A Ml 5 attacks a Df 6 model. The number your card needs to show to hit will vary with each attack and your likelihood of landing that number will vary with each card removed from the respective decks even before factoring in cheating. None of this is perception related. I appreciate the point you're trying to get at and the need of the internet to quantify things, but at the end of the day I don't think there's a good numerical value to apply to a positive twists. Trying to think of a Sh 5 model with a positive as being effectively Sh 7 is going to lead to a lot of disappointment when you're popping shots at Df 6 or 7 models. Quote Link to comment Share on other sites More sharing options...
Myyrä Posted January 20, 2015 Report Share Posted January 20, 2015 really interesting conversation- thanks for sharing all... I'm gonna go with 'rule of thumb, positive twists ~+2' Assuming a negative twist is the same? Single negative is very close to -2 yes. Mean is 4.9602 Median is 4 Win probability is closest to stat-2I calculated without taking into account that you can't flip the same card twice, but the error caused by this is minimal. Edit: Fixed. I even drew you a picture to show shape of the distribution 1 Quote Link to comment Share on other sites More sharing options...
LunarSol Posted January 20, 2015 Report Share Posted January 20, 2015 A RAT 5 model shoots at a DEF 13 model. You need an 8 and always will need an 8 unless outside factors increase that. This means you can apply a relative value to things like boosting or adding a die and removing the lowest/highest that suitably incorporates your increased probability. A Ml 5 attacks a Df 6 model. The number your card needs to show to hit will vary with each attack and your likelihood of landing that number will vary with each card removed from the respective decks even before factoring in cheating. None of this is perception related. I appreciate the point you're trying to get at and the need of the internet to quantify things, but at the end of the day I don't think there's a good numerical value to apply to a positive twists. Trying to think of a Sh 5 model with a positive as being effectively Sh 7 is going to lead to a lot of disappointment when you're popping shots at Df 6 or 7 models. So what number does a die need to show to to hit a DEF 13 with RAT 5? If you're answer is 8, then that's pretty much impossible on a D6. The total of the two dice needs to be 8, which can manifest itself as 5/3, 2/6, 4/4, 3/5, or 6/2, not including the ability to hit on 9, 10, 11, or 12. A Ml 5 swinging at a Df 6 needs a 1. It's still a combination of two random totals, simply using subtraction instead of addition (or treating the second number as a negative value if you want to stick with addition). It can manifest itself as 13/12, 12/11, 11/10, etc not to mention all the the ways you can get a total of 2 to 12. Results based on the combination of two individual random events don't change just because I "own" one die and my opponent "owns" the other. If you look at it from the perspective of "what is the chance my action succeeds" they are fundamentally the same, whether I own both dice and add them vs a static total or whether we each own a result and subtract one from the other. 2 Quote Link to comment Share on other sites More sharing options...
The Godlyness Posted January 20, 2015 Report Share Posted January 20, 2015 I even drew you a picture to show shape of the distribution I can honestly say I have no idea what I am looking at. I know it's graph. I see it has numbers. But how does it correlate to malifaux? How does shoot 5 become pseudo shoot 3? Take the lowest card add it to 5 you can't cheat, they can lol. Just knowing you can't cheat will be at the forefront of the players mind. Not supreme technobabble. Hmm yes your in cover so my 5 is really like 3 so when trying to shoot your df 5 I am going to need a card 2 higher than you to tie. But I can't cheat. And you can. This is how people think when playing this game? You can keep it. I will make a decision based on target priority cards in hand a general knowledge of what has been flipped and if that model can score me any ap or if it's pot shooting. Why is it 400? Why is there a 400? I am intrigued. But don't expect me to comprehend right away. Quote Link to comment Share on other sites More sharing options...
Math Mathonwy Posted January 20, 2015 Report Share Posted January 20, 2015 @math yes math is integral part of this game. I understand that. And yes I will shoot into hard cover if I have a . Since now I am on a straight and can cheat it. Not because it gives me an ephemeral number that doesn't apply. If I don't have that then I can focus if the target is priority. Or flip cards.And hope for the best. Which sometime pays unforseen dividends.Of course. We're talking statistics. That's how it works. I have seen way to often flip low and miss consistently.I've seen 3+ rolls on D6 fail a lot! None of this is perception related. I appreciate the point you're trying to get at and the need of the internet to quantify things, but at the end of the day I don't think there's a good numerical value to apply to a positive twists. Trying to think of a Sh 5 model with a positive as being effectively Sh 7 is going to lead to a lot of disappointment when you're popping shots at Df 6 or 7 models.But it is perception related!You getting into an impossible situation when an opponent flips high enough that you can't win is a perception. The real part is how many times you win or lose. And that is about the same when going Sh 5 against Df 7 as when going Sh 7 against Df 7. 2 Quote Link to comment Share on other sites More sharing options...
Myyrä Posted January 20, 2015 Report Share Posted January 20, 2015 I can honestly say I have no idea what I am looking at. I know it's graph. I see it has numbers. But how does it correlate to malifaux? How does shoot 5 become pseudo shoot 3? Take the lowest card add it to 5 you can't cheat, they can lol. Just knowing you can't cheat will be at the forefront of the players mind. Not supreme technobabble. Hmm yes your in cover so my 5 is really like 3 so when trying to shoot your df 5 I am going to need a card 2 higher than you to tie. But I can't cheat. And you can. This is how people think when playing this game? You can keep it. I will make a decision based on target priority cards in hand a general knowledge of what has been flipped and if that model can score me any ap or if it's pot shooting. Why is it 400? Why is there a 400? I am intrigued. But don't expect me to comprehend right away. It's a historgram of all the possible combinations that can appear in flip from a full deck. There are 2862 possible card combinations there. X-axis shows the numerical value of the flip and y-axis shows how many combinations produce that value. There are nearly 400 different combinations that produce 1 as result. Quote Link to comment Share on other sites More sharing options...
Csonti Posted January 20, 2015 Report Share Posted January 20, 2015 An interesting point on top of the value of a : What is the value of :+fate? I don't have my math at hand but I'm pretty sure it is close to 10 (or +3 compared to the straight flip). So it is only 1 better than a simple positive. Quote Link to comment Share on other sites More sharing options...
Myyrä Posted January 20, 2015 Report Share Posted January 20, 2015 I don't have my math at hand but I'm pretty sure it is close to 10 (or +3 compared to the straight flip). So it is only 1 better than a simple positive. Mean increases from 9.0300 to 9.6794 and median from 10 to 11. Win probabilities fall short from +3. Quote Link to comment Share on other sites More sharing options...
moxypoo Posted January 20, 2015 Report Share Posted January 20, 2015 I've been following this thread with interest. I'm not a good enough statistics guy to work this out myself, so special thanks to Myyra and Math Mathonwy for doing the simulations (sorry if I missed anyone else!). You can keep it. I will make a decision based on target priority cards in hand a general knowledge of what has been flipped and if that model can score me any ap or if it's pot shooting. I find your attitude a little confusing since you've pretty much described doing what everyone is discussing in this thread. In this and an earlier post you mention factoring all sorts of different variables to determine your chance of success in an action. That's exactly what this thread is about, just using a more complex mathematical system to show it. The main point is that a Sh5 will on average, end up with the same final duel total as a Sh7. In game, cheating fate and -:fate modifiers and all sorts of other things have to be taken into account as you mentioned, but this is a good general rule of thumb. Consider a player who doesn't use an statistics and math to determine his/her actions; that player would essentially act randomly, because there's always a very rare chance that a flip will be a statistical outlier and pay off. I've met players like this, but they're few and far between. I also wanted to note the concept of confirmation bias since it's relevant to this thread. There are a few posts that kind of hint at it but none that really describe it. Confirmation bias is usually used in science to describe an experiment that's set up in such a way that the only possible outcome is the one that the scientist expects (and therefore, a poor experiment). In wargaming, this tends to manifest itself as humans remembering the flips/rolls/whatever on the outside regions of the bell curve. For example, we tend to remember the time we flipped 3 Aces on a :+fate duel or two 13s on the duel over all the times we flipped a 5 and a 9 or other more "average" events. If your deck is sufficiently randomized (dice aren't loaded, etc.) then you will flip the statistical average over a large number of randomized events. Keep track of every flip during your next game, and I bet you'll end up with something in the middle of the bell curve with a few outliers. 2 Quote Link to comment Share on other sites More sharing options...
Adran Posted January 20, 2015 Report Share Posted January 20, 2015 Just to muddy the waters once more, a postivie twist is more likely to get the Ram required for the trigger. Whilst this may not be the higher of the 2 cards, you can pick the lower card if you want, especially if it will still get you winning the duel, so you will have an increased chance of getting getting the trigger on the flip. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.