Solution to the LL vs ADS Debate


  • 2018 2017 '16 '11 Moderator

    Okay, as many of you know, we have recently had debates on whether LL is a better benchmark to test strategy or if ADS is better.  Everyone lined up into their respective camps (pro or against LL or ADS) and then bickered for about a month.

    How about a combination of the two?

    Take all the units that hit on a 1 and add them up dividing by 6. (E(Xi)/6)
    *Where E is SIGMA (SUM); Xi is each unit type that attacks/defends at that value (Artillery + Infantry for example).

    Then take all the 2’s, 3’s, 4’s and 5’s (Jets/Super Battleships, whatever.)

    Then roll a die for the remainder as you would in LL for EACH target number.

    So, if you had 20 Infantry, 5 Artillery, 2 Armor, 4 Fighters, and 2 Bombers attacking 60 Infantry, 15 Artillery, 12 Armor, a Fighter and a Bomber you would do the following:

    Attacker:

    15 Infantry @1 = 15/6 = 2 Hits + 1@3
    5 Infantry + 5 Artillery = 20/6 = 3 Hits + 1@2
    2 Armor + 4 Fighters = 18/6 = 3 Hits
    2 Bombers = 8/6 = 1 Hit + 1@2

    Attacker would then throw:

    2@2 and 1@3.  Hits would be added to the 9 Guaranteed hits.

    Defender:

    Bomber = 1/6 = 1@1
    60 Infantry, 15 Artillery = 150/6 = 25 Hits
    15 Armor = 45/6 = 7 Hits + 1@3
    Fighter = 4/6 = 1@4

    So the defender would throw:

    1@1 1@3 and 1@4 adding to the 32 guaranteed hits.


    This would still throw out extreme results, but would still leave enough swing in the results that the game isn’t reduced to an unrealistic formula that will guarantee wins.  It’s, to me, the best of both worlds.


  • Official Q&A

    Interesting.  It does allow for more variance in results than “low luck” does, but the results are still fairly predictable.

    Here’s another approach to reducing the effects of luck that I’ve been using for years:

    Roll each set of attack and defense dice twice and average the results, rounding toward the statistical average number of hits.

    Example: 2 tanks and 6 infantry attack, rolling “3”, “1”, “4”, “1”, “2”, “1” (infantry), “2” and “3” (tanks) the first time (5 hits), and “5”, “5”, “4”, “3”, “2”, “1” (infantry), “3” and “6” (tanks) the second time (2 hits). This averages out to 3.5 hits, rounded down (since the average result is 2) to 3.

    Combat rounds that involve less than 4 total combat points or only one unit on either side are rolled normally.

    The result of this is skewing battle results more to the center of the bell curve.  It cuts down on the influence of luck without eliminating it entirely, and it doesn’t add as much time to the game as one might think.


  • 2018 2017 '16 '11 Moderator

    War Dog (Krieghund),

    I think that might be a little too math heavy for a game.  The idea was to create a method of giving the Low Luckers at least some ability to mathematically assure some battles will go their way but still give the ADSers the thrill of having 1 defending infantry beat 3 attacking infantry and an attacking fighter.

    In LL, that’s an auto hit.

    In ADS, that’s still very high odds of success, but gives that defender a lot more ability to inflict damage (decent shot at killing two infantry, maybe three.)

    In LLADS, just coining the term now, it gives the LL’s the security of high probabilities of still getting the territory in one round with two surviving infantry, but it gives the ADSer’s the thrill of the possibility of inflicting 1-3 infantry kills on the enemy.

    Since (=25%)0.50.5 > 0.1650.1650.1650.5 (=0.224%)

    That means, instead of 100% for the attacker, in LL, you get 25% chance for two hits.  But instead of a fifth of a percent chance for the attacker to get 4 hits, you have a 25% chance to get two hits.

    This also means that 3 inf, fig will be able to do MORE then 1 hit in any given round, which is a benefit to me.  Honestly, I like the chances of killing two infantry defending with 3 infantry, fighter attacking in one round in ADS, but I also like the idea of at least knowing that 3infantry and a fighter won’t go home empty handed.


  • Official Q&A

    There’s not really all that much math involved, Jennifer.  All you have to do is roll the dice twice and take the average.  You don’t even need to calculate the normal average hits unless there’s a fractional result, so half the time (on average) that’s as far as you’ll need to go.  If you do have to calculate the normal average, it’s basically the same process as Low Luck, except you take the leftover pips and add 1 if it’s 4 or more.  You’d have to do that in Low Luck or your system anyway.

    At any rate, I do like your system better than straight Low Luck.  Nice idea!  🙂



  • I think someone already came up with a compromise they posted a while back, something like every turn you choose 2 battles that are low luck, while the rest is ADS (so if you did Russian Triple, two of the attacks are LL, the other one is ADS).

    And the twist is that the opponent gets 3 declarations in which you’re forced to roll it in ADS as opposed to LL. So if it comes down to trying to abuse a strafe, the opponent can force you roll it and you get burned.

    Personally I could probably live with that kind of compromise. You get 2 sure battles per turn so not EVERYTHING can go badly, but also if you try to abuse the low luck nature in strafing the opponent can call you out and force you to roll it ADS.


  • 2018 2017 '16 '11 Moderator

    That rellies too much on keeping track of what battles are LL what are ADS and who’s had how many challenges.

    I think my system here is a better answer.

    When you put your pieces on the battle board, you total up the number of hits per category and roll a die to determine if the remainder for that category hit or not.  It’s simple, it’s sweet, it cuts down on the huge swings that LL’ers fear in ADS games while still having enough statistical “spread” of results to still make the game fun for those who prefer ADS battles.  Sure, you’re two infantry will NEVER kill 5 defending fighters, but you still have a 33% chance to hit one!  Likewise, your 3 infantry, fighter will not always kill a defender on the first attempt either.



  • By the way, I’m not convinced at all that the low luck system gives you a formulaic way to win. There are plenty of lucky/unlucky situations that can add up. A transport can still win against a battleship. The Baltic fleet can die without hitting a single UK plane, or wipe them all out with 2 boats left. The Russian Triple fails almost 50% of the time in LL. I’m not convinced that it’s as formulaic as you make it out to be, because the remainder dice give a lot of leeway. Why would your situation be any different than what LL is already? LL already has swing, you just want to add more swing arbitrarily?


  • 2018 2017 '16 '11 Moderator

    LL has very little dispersion of results.
    ADS has too much dispersion of results (or so claim the LL advocates.)

    My system would still give you dispersion, but not as huge as ADS nor as limited as LL.

    The biggest difference is that you cannot combine the punch of units.  Sure, 1 infantry + 2 fighters always kills a defending infantry with no loss to the attacking fighters.  But 3 Infantry, 1 Fighter does not ALWAYS kill a defending infantry, let alone protect the fighter!

    So those who enjoy low luck because it means they will never go home without at least hitting something in a major battle, would still get that benefit.  However, those who enjoy pure luck because it means their defenders might score in the top 30% or the bottom 30% can still achieve those results.

    All this does is, in effect, chop off the top 15% and the bottom 15% and no, I did not mathematically prove that, I’m pulling those numbers out of my arse to demonstrate my point.

    Think of it this way, for you statisticians out there.

    When you pull a sample population that you wish to run a mean or median test on, do you keep the outliers or just the mainstream data?  Some of you keep the outliers, but I think that distorts the results.  However, if you chop off too MUCH data, instead of just the outliers, you don’t get a good picture either.

    This is just a method of taking away the most extreme results and forcing the strategy to work within the realistic confines of the game for testing strategies.  Any strategy in LLADS will work even better in LL and any strategy in LLADS will almost always work in ADS, except for those outlier results, of course.


  • 2017 '16 '15 Organizer '14 Customizer '13 '12 '11 '10

    How bout everybody either play LL or ADS… and give each player a number of combat events per game where he can choose the other system. It would be like the NFL team challenge to a bad call rule.  That way on the battle that count huge for a player he allocates the system he think works best.

    In fact why cant both side pick their own system and stick to that? everybody would be happy because they are doing what they want.



  • If you’re attempting to “fix” LL more towards ADS, I think you would try to fix the most blatant problem - precision strafing. Personally I think the dispersion is already quite there in LL in small battles, but I can see how in LL you can do ridiculous strafes like 6 inf vs 2 inf that you would never consider in ADS due to a high probability of taking it in that case.

    That’s why the original guy who posted it and Imperious Leader has the idea that I like - keep the base as low luck maybe for all combats instead of just 2 per round, but you get a number of combat events in which you force the opponent to roll it out instead of doing the LL method, which keeps the players honest in terms of not abusing precision strafing. That fixes the most blatant “problem” of LL which is ridiculously accurate strafing, by putting the fear that the opponent will force you to ADS it.


  • 2018 2017 '16 '11 Moderator

    I don’t want to get into flip flopping on dice methodology though.  I’m looking for a better way to determine strategy viability where you cannot do 60 infantry vs 11 infantry and know you will hit 10 in the first round and can retreat leaving him with 1 guy left, but where you can also avoid having 13 attacks all misses while defenders rolling at 1 score 100% accuracy.

    Basically, I want everything within two or three standard deviations of the mean.  I just want to trim off the ridiculously good and bad results, and leave all the stuff in the middle.  If your strategy is too inflexible to handle less then average dice or greater then average dice, then it’s a bad strategy.  But no strategy is going to work against insane dice, no matter how good a player you are.



  • I’m looking for a better way to determine strategy viability where you cannot do 60 infantry vs 11 infantry and know you will hit 10 in the first round and can retreat leaving him with 1 guy left, but where you can also avoid having 13 attacks all misses while defenders rolling at 1 score 100% accuracy.

    Your method doesn’t fix this. 60 inf/ 6 = 10 hits. How does your LLADS fix the ridiculous precision strafing? Adding up like numbers doesn’t help this since you can still count enough like numbers to do it precisely.


  • 2018 2017 '16 '11 Moderator

    60 infantry attacking is 60 punch.

    18 infantry + 10 artillery + 4 armor is also 60 punch, however, under LLADS they are:

    1 Hit + 1@2 for Unsupported Infantry
    3 Hits + 1@2 for Supported Infantry
    3 Hits + 1@2 for Artillery
    2 Hits for Armor

    Now you have a range of 9-12 hits instead of exactly 10 hits.



  • It still doesn’t fix the example that you complained about. You complained about you not wanting 60 infantry vs 11 infantry to perfectly strafe out 10 inf, but your solution doesn’t correct this.


  • 2018 2017 '16 '11 Moderator

    It fixes the much more common combination of forces getting you auto hits but it also allows you to net at least a certain number of hits.  I think it’s a perfect solution.  No one’s brought up a valid argument that the LLADS system doesn’t fix the majority of the errors in both ADS and LL when testing over all strategy.



  • I clearly brought up that your solution does not solve the biggest problem - precise strafing. You can still send 60 inf at 11 inf. Or 6 inf at 2 inf. Or 6 inf 4 arm against 4 inf. You can still do the thing where you send 1 inf + 4 figs at 2 inf, something you complained a lot about LL.

    I see where it makes combination units less easy to use, but it’s still not very difficult, as in the examples I gave.

    It’s an interesting solution, but not a perfect solution.



  • There is one thing that is common to all A&A games…
    DICE!

    Remove the dice, remove the CORE CONCEPT of the game.



  • Also reading your example I am more confused than ever. I thought you group all units that hit on a 2 according to your original post, and then in this last example you gave you grouped supported infantry differently than artillery, yet both of them hit on 2.

    So ACTUALLY for 18 infantry 10 artillery 4 armor, that’s

    8 infantry = 8points/6= 1 hit +2
    10 supported inf 10 artillery =40points/6 =  6 hits +4
    4 armor = 2 hits

    So the variance is 9-11, not 9-12.

    And in case you’re going to say I’m missing the forest for the tree, what I would do in that case is send 1 artillery less. That brings the variance to 9-10 hits. Still a very precise strafe, it’s less sure than low luck but it still brings it within 1-2 inf, and there’s still zero chance of overtaking the territory, because there are 11 units, and the max I could hit is 10. Zero chance.

    How do you iron out that in ADS you have to compensate for both good and bad luck in strafing? LLADS is far from perfect in trying to bring that element of ADS into the game, all it does is make LL slightly less precise, but there’s still no risk. You can still do ridiculously precise strafes with zero risk of overtaking the territory.


  • 2017 '16 '15 Organizer '14 Customizer '13 '12 '11 '10

    How bout this:

    you make a crt using the LL results and you roll one die and index a corresponding scale of results that may be something you already got under ADS.

    Example:

    every 6 combat factors = one hit, so if you got (lets make this easy) 36 attack factors you have this CRT (combat results table) where the results will indicate the number of hits with the median in this example at 6…thus…

    1= 4 hits
    2= 5 hits
    3= 6 hits
    4= 6 hits
    5= 7 hits
    6= 8 hits

    thus you have some luck but its limited. The CRT will look like an old Avalon Hill CRT from the 60’s each value will have a spread of 6 possible results either high or low, but their is protection to the high and low dice results.

    At least its out of the box thinking.



  • Would that be a “Modest Luck” option?    :lol:



  • We need a HL option - Half-ass Luck  :roll:


  • 2018 2017 '16 '11 Moderator

    I meant that you total up your units and then figure out your auto hits and your remainder.  Not that armor + fighters = total punch.  Armor have their own punch, fighters their own punch.

    This doesn’t eliminate precision strafing, but it sure knocks it for a loop in most scenarios that one may encounter.  You start to run the risk of actually winning and thus being exposed or quitting before hitting optimal levels.

    Also, it’s stronger for the defense, I think, since your units almost all defend better and thus, would have greater individual group punches then if you totaled them all and dwindled, them, maybe it’s just perceived, but I always thought the defender had it worse in LL.

    Switch, I’m not looking for a replacement for ADS.  ADS is still king.  I’m looking for a replacement for LL.  Something that exterminates the extreme value results like LL does, but doesn’t confine you to only slightly marginal battle shifts.  Something, basically, to simulate an ADS game but without the risk of attacking with 40 fighters against 1 infantry and having that infantry actually surviving more then one round.



  • Something that exterminates the extreme value results like LL does, but doesn’t confine you to only slightly marginal battle shifts.

    I already showed you how LL does not have marginal battle shifts. The Baltic attack ranges from all airforce surviving to all airforce dead with 2 German boats alive. Attacking Belo with 3 inf 2 fig ranges from taking with 2 inf to merely clearing it. Attacking Pearl ranges from taking no casualties to losing 2 boats. A transport can still beat a bomber or a battleship easily. 2 inf 1 fig can easily lose to 1 inf. There is plenty of variability, and no such thing as planning 100% in LL. You can kiss your pre-plan goodbye if the Kwang transport beats up your destroyer/carrier, or if the Baltic whoops your butt, etc.



  • @ncscswitch:

    Remove the dice, remove the CORE CONCEPT of the game.

    couldnt put it better myself too 😉


  • 2018 2017 '16 '11 Moderator

    But you can pre-plan in LL.

    You have 40 infantry, 10 artillery, 19 armor and you want to take out a territory with 10 infantry defending it while keeping the largest possible force in BOTH locations, that means you want to clear it in one round to minimize the damage you take.

    In LL you could send:
    18 infantry + 10 artillery + 4 armor will ALWAYS give you exactly 10 and no more or less then 10 hits in LL when you use them to attack.

    In ADS you may send closer to 20 infantry, 5 artillery, 12 armor to compensate for bad dice.

    In LLADS you could still send 18 I, 10 R, 4 A because you’ll get 9 hits, but it may take another round, so you might send another couple of infantry or just take the risk.

    The Point is, LLADS is inbetween ADS and LL.  LL is obviously a poor method to extrapolate a 15 round game for theory testing because it has very little variation in results and allows you to calculate your defense and offense to the nth degree.  ADS is also a poor method to extrapolate a 15 round game for theory testing because there are the possibilities of battles going insanely good or bad and thus, throwing off the results.  As I used to say “Good Dice Rolls Beat Good Tactics Any Day.”  LLADS is in the middle.  It’s not perfect, but it eliminates the worst failures of both ADS and LL.  There are no insanely good or bad dice rolls, but their’s also more ambiguity to the outcomes of battles then LL.

    Sure, 6 infantry will still kill 1 infantry.  But who’s going to send 6 infantry to attack one infantry, seriously?  3 is about the cap I ever see.  Infantry, Artillery is more common then 3 infantry (and would be rolled ADS style in LLADS.)  2 Infantry is the most common (and would be rolled LL style.)  Assuming your fighters are tied up or dead, that is.


    @Amon:

    @ncscswitch:

    Remove the dice, remove the CORE CONCEPT of the game.

    couldnt put it better myself too 😉

    This isn’t for actual game play, it’s for testing strategies only and getting a feel for how fault tolerant they are.

    Basically, it’s LL with more dice or ADS with less dice.  Depending on your view.

    I still agree the only way to play a real game is with all the dice.  But a diagnostic game?  Do you really want to test a theory when you lose 6 fighters and a bomber to a battleship and all you do is damage the battleship?


Log in to reply
 

20th Anniversary Give Away

In January 2000 this site came to life and now we're celebrating our 20th Anniversary with a prize giveaway of 30+ prizes. See this link for the list of prizes and winners.
Axis & Allies Boardgaming Custom Painted Miniatures
Dean's Army Guys
T-shirts, Hats, and More

Suggested Topics

  • 4
  • 23
  • 17
  • 1
  • 30
  • 70
  • 20
  • 8
I Will Never Grow Up Games
Axis & Allies Boardgaming Custom Painted Miniatures
Dean's Army Guys

75
Online

14.9k
Users

35.7k
Topics

1.5m
Posts