Moderate luck option?



  • Sorry if this has been asked and discarded, but it seems like there should be a choice that falls between LL and dice. Call it "Moderate Luck" or "Reduced Luck."

    There are any number of ways to reduce dice volatility without making it as predictable as LL.

    In the boardgame days, we tried some of the following (not necessarily together in the same game):

    --All 1s in combat were automatically combined into one, like a separate LL. Everything else was rolled as normal.

    --Combine all factors as in LL, except divide into units of 5 and roll dice to make 5s. Or, with 4s. Not as sure-thing as LL, but it dramatically reduced the volatility of the results.

    --Each side (Axis or Allies) had a bowl of chits numbered from 1 to 6, used only for combat rolls (not tech) where 1 is the best result and 6 the worst. You draw from your own side's bowl and, after reading them, threw your "luck" (or bad luck) into the other side's bowl.

    I'm sure there are other ideas, any of which could be tried and tested with monte carlo sims and see which one created a "luck level" midway between LL and dice.


  • Admin

    @David_VanDyke Yeah, this is come up a bunch of times in the past. I think one of the challenges is coming to some consensus on what additional "Moderate Luck" option most folks would prefer/use. Out of the ones you suggest I'd vote for something along the lines of:
    --Combine all factors as in LL, except divide into units of 5 and roll dice to make 5s. Or, with 4s. Not as sure-thing as LL, but it dramatically reduced the volatility of the results.



  • Well, rather than worrying too much about consensus, give something a try and let people try it out, and ask for modification. It will become popular or not depending on how good it is. Personally, I'd suggest the "roll for 4s" option myself.


  • Admin

    @redrum Wouldn't it be practical to look at what is even possible... I mean it could really narrow down your options first looking at what makes the most sense AND is reasonably feasible.

    Personally I think it would make a lot of sense to look at grouping units by their attack/defense value...

    So you would...
    add up all the units that roll at 1.... roll for whatever remainder
    add up all the units that roll at 2... roll for whatever remainder
    So on and so on for whatever number of dice sides you have in the game.

    But perhaps making a list of the different options might be a good start.



  • The problem with that option, Hepps, is that it's very little different from LL. There needs to be enough swing to make people make strategic, warlike decisions, i.e., make sure you send in enough to win if that's what you need to do, rather than constantly calculating out the perfect efficiency of attack like in LL--but at the same time, reduce the odds to almost zero to have rounds where you get triple the expected hits, for example. In other words, pick an option as close to directly between the two current options as possible.


  • Admin

    @Hepps Well anything is possible :)

    But yeah level of effort to implement would be a factor but wanted to collect up some ideas first then consider weeding those out that may take too much effort.

    @David_VanDyke Oh? I actually thought your suggestion was exactly what Hepps described and was what I was thinking. Group units by attack/defense value then do essentially LL for each group. Can you clarify what you mean then by:
    --Combine all factors as in LL, except divide into units of 5 and roll dice to make 5s. Or, with 4s. Not as sure-thing as LL, but it dramatically reduced the volatility of the results.


  • Admin

    @David_VanDyke Remember that whatever is to be implemented needs to be applicable to more than just D6. There are a number of games here that use dice sides other than 6.


  • Admin

    @David_VanDyke said in Moderate luck option?:

    The problem with that option, Hepps, is that it's very little different from LL.

    This would actually be quite a bit different from traditional LL.

    Lets look at an example say based on some fairly normal stats... ignoring support (for simplicity sake).

    Say I am attacking with...

    5 Inf att. @ 1
    2 Art att. @ 2
    5 Tanks. att. @ 3

    So under traditional LL ...

    you'd roll 1 die at 5 or less with 4 auto hits.

    Under what I suggested...

    You'd roll one die at 5 or less for the Inf.
    You'd roll one die at 4 or less for the art.
    You'd roll one die at 3 or less for the tanks with 2 auto hits.

    That seems like a roughly medium swing between dice and LL.

    I don't know... perhaps there is a flaw in my logic that I simply cannot see.


  • Admin

    @Hepps yes he has big maps as does @Frostion


  • Admin

    @prastle He's got big maps... he's got big maps... but I've got the biggest maps of them all!!!
    =)


  • Admin

    @Hepps off topic but i want a new server!!!!
    poke them please!


  • Moderators

    The biggest issue with the current Low Luck is you can often predict exactly how many hits you will get, never more. This leads to a massive change in how hit and run attacks work; attacks I would never consider become very appealing. It can be a bit annoying because I like LL a lot of certain maps, like age of tribes. The reason is there are so many rolls on 1/12 that sometimes the decisions players make feel arbitrary compared to the dice.

    I'm not sure how much the grouped LL would address this. You can easily pair units with 2/6 or 3/6 attack in the right sized group to get an exact amount of damage. I suppose it helps, but IDK how much. But for 10 or 12 sided it would work well


  • Admin

    @CrazyG Yes you could still pair groups to achieve a guaranteed hit. But that's only if you have the ability to mass units in those denominations all over a map.

    I agree that on a larger Dice sides the system would work optimally (as does the game in general)... but it would still operate well on a D6 system.

    What I like about this is it would also reward (presumably) more costly units by giving them more auto hits as you have more of them in a given attack or defense. While cheaper and less effective units would need to be massed in greater numbers.

    I don't know... I kind of just threw the idea out there... I hadn't done any in-depth statistical analysis.


  • Moderators

    @redrum said in Moderate luck option?:

    @David_VanDyke Yeah, this is come up a bunch of times in the past. I think one of the challenges is coming to some consensus on what additional "Moderate Luck" option most folks would prefer/use. Out of the ones you suggest I'd vote for something along the lines of:
    --Combine all factors as in LL, except divide into units of 5 and roll dice to make 5s. Or, with 4s. Not as sure-thing as LL, but it dramatically reduced the volatility of the results.

    Premitting that, on maps that I repute to be well made, I don't personally feel the need of having less luck than what dice provide, and it is mostly a matter of good mapmaking to avoid extreme luck driven dynamics, unless wanted, of course, which is true for LL too (like the round 1 attack on Lilybaeum in 270BC or the round 1 attack in 10 Sea Zone in World War II v4 or, again, the round 1 attack on Cyrene in 270BC, that is averagely clearly convenient in LL, but, still in LL, you have a less than 1% probability of failing it, which would be terrible) I tend to agree as well that this would be the cleanest way to do something like what wanted, but I want to point out that we are not talking of "Low Luck" anymore.

    If you have a map in which all units roll at 5 and you put on an option to make them roll at 4, then you are increasing the variability of the dice in that map. Of course, this besides the fact that not all maps use 6 sides dice.

    So, this proposal would be just better configured as a numeral property like:

    Special Luck = true / false
    Special Luck Dice = X

    (the first would enable the option; the second would set it and be effective only if the first one is true)

    Where you would sum up your total power, divide it by X and roll an equal amount of dice, plus rolling another dice, in case, for the remaining part.

    Thus, for example, if you have 14 total power and X=4 then you would roll 3 dice at 4 plus 1 dice at 2.

    Such a property would cover the "Low Luck" option itself as, in a 6 dicesides map setting it at 6 would be the same thing as setting "Low Luck" true.

    On the other hand, if you set it at 1, then you would have all rolls made at 1; thus if you have power 14 you would roll 14 dice at 1, no matter what else. In this case, coming from a situation in which not all units roll at 1, this option would increase luck, and rather be the equivalent of a "High Luck" option, that if true makes you roll only dice at 1 equal to your total power.

    Thus this was just to say that this proposal can be taken as a property in which you can set the number at which to roll all but the remainder (and if you set it equal to the dice sides you have just Low Luck) but the name should be such as this property would not be hinting at being about lowering luck, unless the property restrict itself at not accepting any values lower than the highest strength option set for that game, which would imply that, in most games, you would be able to set it only at 4, 5 or 6, with 6 meaning just Low Luck (and be on the outlook out for things like the Classic Jet Power tech, increasing the defence of fighters to 5).

    In this case, for example, in a map like WAW, you could remove the "Low Luck" option and use this option, instead, setting it true and at default equal to 6, so that people can either set it false, to play dice, or set it at a lower number, coprising setting it at 1, to play high luck.

    But, again, it is mostly a matter of the map. If the map has any critical round 1 battles with extreme swinging outcomes, that can be won or lost closely and have a massive impact on the whole game, then Low Luck won't save you either (for example, if you play 270BC and concentrate in Lilybaeum max defence and put a factory in it, then the opposite side will have around 80% to take that factory on round 1, using Low Luck, which is modestly good if it succeeds, but very bad if it fails, and it can be argued that here LL would make it even more extreme, as you are not coming back against a good player, if you do that and get the 20% fail (which is just a map "problem", and the main reason why I made my variant of 270BC)).

    So, we are talking of dice or low luck, and low luck surely reduces the impact of luck a lot, but I also want to point out that some maps with low luck might be more luck driven than other maps with dice (I'm particularly not confortable with maps like World War II v4, in which on round 1 you can send 1 submarine to kill or not to kill 1 cruiser + 2 transport, and that is exactly the same both with dice and low luck).


  • Admin

    @Cernel jesus ya still make my head hurt :)



  • I've always felt that the underlying issue (the frustration people have with dice) has more to do with the design of a map's economy and the replacement cost of units relative to starting TUV, than with the combat mechanics per se.

    I think LL actually magnifies the issue sometimes, esp in smaller engagements, since it's easier to predict the likely outcomes of a given combat in LL, and when the "luck" (which is still involved) disrupts the expected result here, it can be almost just as damning to your plans for unit attrition as a string of weird ass rolls would be in a regular dice game. It's just that the upsets typically happen incrementally in LL instead of all at once as in dice.

    Unfortunately this insight isn't all the useful, since you usually can't redesign a game's economy once its established. But typically the ratio between starting TUV and starting income is like 10 to 1. That's pretty fucking high. It means that in an average A&A game, it takes 10 rounds of purchasing just to replace what you're given at the outset, and any swing that knocks off a sizeable chunk of TUV (particularly in the opening round, say via scripted attacks) has a huge cascading effect over the entire game. LL doesn't really do anything to address that aspect of the gameplay, it just makes the losses in a given round or in a key battle more predictable. And of course it makes the strafe and the airblitz extremely potent compared to dice.

    I don't know what a happy medium would look like. Hepps suggestion seems interesting, even if it allows for a similar bean counting dynamic as LL, at least the mental math involved would be somewhat more challenging than just counting up 6 pips for the entire attack/defense force. Still, trying to think of something that works for a game like AA50 seems kind of daunting. I guess the 5s or 4s idea suggested above might be workable. I wasn't entirely sure if I read it correctly, but I was thinking the suggestion was to somehow create a system without auto-hits? Auto-hits and rolling for the remainder seems to be the defining characteristic of LL, but also the thing that makes LL basically anathema to normal dice players. So it's hard to see how you can excite one set of players without losing the other. Unless you change something about how the pips actually translate into hits.

    I agree though, it would be cool if we could find some sort of alternative between the crushing randomness of dice and the exacting slog of LL. Of the two playstyles I favor the former, so I'd probably be biased regardless. But I think a lot of people have articulated the issues they have with both, so a third way might gain traction if only we could figure out what that looks like and then popularize it. Tall order though, definitely beyond my pay grade haha


  • Moderators

    @Black_Elk I mostly agree, but I believe the average is rather about 8 or 9, for the classic games.
    I believe, amongst the classic TripleA games, the one with lowest TUV / Production ratio is 270BC, at 6.2.
    I've also recently made a variant of that, called 270BC 40% (you can find in 270BC Variants), that goes as low as 5.45 of TUV / Production ratio.
    With this said, I've not a problem with TUV / production ratios of even 12 or more (like in World War II v6, that I like), and in a way I actually really like your forces being very valuable. Definitely you need to like dice, in this case, tho.
    Really, rather than LL, the best solution, on many levels, would be having maintainment costs (already possible), in my opinion. Tho this reduces the importance of initial TUV swings but increases the importance of initial production swings a lot.


  • Admin

    @Black_Elk said in Moderate luck option?:

    I'm not going to speak to some of the other stuff in your reply because while I agree with what yourself and @Cernel are saying about game design affecting the perceived and real impacts of either dice or LL outcomes on a game... I think that is really a separate matter from the topic at hand. The question isn't whether the design of a game should take into account whether you want it played with dice or Low Luck... we already know this to be true.

    What we really want to sink or teeth into is whether there is a viable middle ground that could potentially offer players and designers a better/different option.

    I don't know what a happy medium would look like. Hepps suggestion seems interesting, even if it allows for a similar bean counting dynamic as LL, at least the mental math involved would be somewhat more challenging than just counting up 6 pips for the entire attack/defense force. Still, trying to think of something that works for a game like AA50 seems kind of daunting.

    As I have already acknowledged... I realize that what I am suggesting would retain a certain level of calculability for straffing attacks... that isn't really necessarily a bad thing... since we are trying to find a middle ground... and anything beyond a one combat round strafe would become increasingly more variable than LL currently is.


  • Moderators

    Here is an idea. Half Low Luck. Its calculated the same as Low Luck, but for every guaranteed hit, it just rolls 2 dice at 50% (3/6, 5/10, 6/12, etc.)

    So if I attack with 8 infantry, instead of having 1 guaranteed hit + a roll on 2/6. I would get 2 rolls on 3/6 and 1 roll on 2/6.

    If I attack with 16 total power, I would roll 4 dice at 3/6, and 1 dice at 4/6

    You could potentially even have a few guaranteed hits on really large battles to drop variance further. Reduces the variance of dice without creating overly predictable battles. My experience is that most of the time a player wants to avoid dice, they want to avoid large numbers of rolls on 1/6 or lower, because this is where the variance starts to get really high.


  • Admin

    Yeah, after thinking about this for a while I think the 2 most straightforward options would be:

    1. @Hepps suggestion of grouping units by attack/defense value then doing LL on each group
    2. @CrazyG suggestion of doing essentially half LL by totaling all values up but dividing by half the dice max (3 for 6d, 6 for 12d) then rolling for all. @Cernel Also pointed out it could be done more generally by having a parameter to define what do divide it up by rather than hard coding to half of the dice max.

    I think both could be useful and have pros/cons. Both do a pretty good job of minimizing the extreme randomness of having to roll many dice @ 1 or 2.

    Option 1: still allows for auto hits but introduces more randomness than LL in most situations. The main edge case that you could argue doesn't work great is if you then attack with units that have all the same value with then ends up functioning just like LL. I don't think this would be that common since usually you want max attack power for strafing and for doing weird retreat stuff you usually want to use most units (think switzerland in NWO). I'd say this option leans a little more towards LL than dice.

    Option 2: works well in most situations but has some cases that end up with more randomness than dice. The main edge case I thought of is attacking with a single high value unit like a battleship. Instead of it attacking with 1d @ 4, it would instead roll for 2d @ 3 & 1. I think this mostly affects sea battles as they tend to have higher value units as ships and planes on many maps have greater than half of the max dice.


Log in to reply
 

Looks like your connection to TripleA Forum was lost, please wait while we try to reconnect.