Page 4 of 5 FirstFirst ... 2345 LastLast
Results 46 to 60 of 62

Thread: humans ns or prop?

  1. #46
    Post Fiend marve's Avatar
    Join Date
    Feb 2012
    Posts
    277
    there is too much probability in prop to make a blanket statement like that persain. fail

  2. #47
    Forum Addict
    Join Date
    Sep 2011
    Posts
    1,014
    Someone who is either a halfer or has a halfer in their KD, please sacrifice 1 stealth point in war to see exactly the damage that NS does. When that person posts the % damage done to specs and elites, this convo will end quickly(atleast the raw damage conversation).

  3. #48
    Strategy Moderator
    Join Date
    Jan 2012
    Posts
    4,201
    Quote Originally Posted by marve View Post
    there is too much probability in prop to make a blanket statement like that persain. fail
    Its not to much probability its a very simple calculation
    I only make 3 assumtions

    1.20% chance on each op to get each category.
    2.There are 5 categories not 6 where off spec/def spec are sperate
    3. ns v prop fail rate is the same

    You can then calculate damage done in each category if u land there on a prop

    x1 damage done if leets hit
    x2 damage done if wizzards hit
    x3 damage done if theifs hit
    x4 damage done if specs hit
    x5 damage done if solds hit

    sum over x of xi*0.2= total damage done per propaganda(expected value). over say 50-100 ops your not going to vary more that a few % off of 20% in each category.

    U then say is
    Damage done per ns >=< 0.2* (x1+x2+x3+x4+x5).

    the only thing u then need to look at are the values of xi to see how they compare.

    for 99% of the time we can ignore x5, sold damage as its not really important to simplify things.

    Thus we ask in terms of damage DONE on an ARMY IN,
    #leets ns'd+#specs ns'd Vrs. 20%spec if prop+20% leet if prop+20% theif if prop +20% wizzard if prop.

    Based on the data i've gathered and no one has refuted you can cancel out the prop+leet damage to say the average damage done comparison yeilds a comparison that gives us
    1 prop better than 1 NS if
    20% max(theif+wizzard) damage per op> 1/3 ns damage on leets/dpsecs.


    I argued that the only time that is true is if u wanted to get wizards any way and thus were targeting a A/M. otherwise i personally feel that 1/3 ns LOSS of damage to the army makes NS always better than in terms of dealing damage.
    Last edited by Persain; 14-03-2012 at 18:26.

  4. #49
    Post Fiend marve's Avatar
    Join Date
    Feb 2012
    Posts
    277
    yes, there is still a lot of probability involved. your calcs are all fine and good, and i did learn something new in your post, however, if you somehow got 10 props on specs in a row, and got for max or near max gains, then it would be better than ns. which is why i said it depends on how you hit. yes its unlikely for that to happen, but there is a chance of it. so you can say definitely one is always better than the other.

  5. #50
    Strategy Moderator
    Join Date
    Jan 2012
    Posts
    4,201
    @ marve your correct if u hit 10 specs in a row your right you'd be ahead, however that the chance of that is VERY low as in <<1% chance. As well planing a war/ops with the hope that u will get lucky to win not a good strategy.

    edit.....let me add in the math

    using http://en.wikipedia.org/wiki/Binomial_distribution

    u have a
    20% chance of getting 1 on specs
    4% chance of getting 2 in a row on specs
    0.8% chance of getting 3 in a row on specs
    0.016% of 4 in a row.

    see where this is going.

    lets say u want to cover the numbers i gave where instead of being 1/3 lower u want both leets and specs to be 33% higher chance over say 50 ops
    that is instead of 10 specs and 10 leets u want 13 specs and 13 leets hits of propeganda.

    in this case using the wiki link we have
    n=50 (number of attempts at prop)
    p=.4 (since that was the base rate of leets+specs
    k=26 number of hits we want.

    that means we have [50!/(26!*(50-26)!)]*(.4^26)*(0.6^(50-26))=0.0259. ~2.6% chance that prop will deal more damage to leets and specs than a given ns.
    Last edited by Persain; 14-03-2012 at 19:11.

  6. #51
    Forum Addict
    Join Date
    Sep 2011
    Posts
    1,014
    Quote Originally Posted by Persain View Post
    Its not to much probability its a very simple calculation
    I only make 3 assumtions

    1.20% chance on each op to get each category.
    2.There are 5 categories not 6 where off spec/def spec are sperate
    3. ns v prop fail rate is the same

    You can then calculate damage done in each category if u land there on a prop

    x1 damage done if leets hit
    x2 damage done if wizzards hit
    x3 damage done if theifs hit
    x4 damage done if specs hit
    x5 damage done if solds hit

    sum over x of xi*0.2= total damage done per propaganda(expected value). over say 50-100 ops your not going to vary more that a few % off of 20% in each category.

    U then say is
    Damage done per ns >=< 0.2* (x1+x2+x3+x4+x5).

    the only thing u then need to look at are the values of xi to see how they compare.

    for 99% of the time we can ignore x5, sold damage as its not really important to simplify things.

    Thus we ask in terms of damage DONE on an ARMY IN,
    #leets ns'd+#specs ns'd Vrs. 20%spec if prop+20% leet if prop+20% theif if prop +20% wizzard if prop.

    Based on the data i've gathered and no one has refuted you can cancel out the prop+leet damage to say the average damage done comparison yeilds a comparison that gives us
    1 prop better than 1 NS if
    20% max(theif+wizzard) damage per op> 1/3 ns damage on leets/dpsecs.


    I argued that the only time that is true is if u wanted to get wizards any way and thus were targeting a A/M. otherwise i personally feel that 1/3 ns LOSS of damage to the army makes NS always better than in terms of dealing damage.
    Assumption 1 and 2 might not be true. This entire calculation is ridiculous since you(or anybody for that matter) haven't posted the actual damage done to troops by ns.

  7. #52
    Strategy Moderator
    Join Date
    Jan 2012
    Posts
    4,201
    Quote Originally Posted by Topsy View Post
    Assumption 1 and 2 might not be true. This entire calculation is ridiculous since you(or anybody for that matter) haven't posted the actual damage done to troops by ns.
    According to the wiki and what i remember of the data assumptions 1 and 2 are correct. if you would like to prove me wrong go ahead and show your data points. I'd hope to see at least 100-200 prop ops. (old excel file had about 1100 prop ops)

    Ns damage/op is available in plenty of calculators for this game.

  8. #53
    Forum Addict
    Join Date
    Sep 2011
    Posts
    1,014
    Quote Originally Posted by Persain View Post
    According to the wiki and what i remember of the data assumptions 1 and 2 are correct. if you would like to prove me wrong go ahead and show your data points. I'd hope to see at least 100-200 prop ops. (old excel file had about 1100 prop ops)

    Ns damage/op is available in plenty of calculators for this game.
    "Attempts to convince enemy peasants, military, or wizards to revolt and join your province. Only Rogues can master this difficult operation." and " Effect: not fully determined"

    Maybe you're looking at a different wiki?

    And I know collecting how much damage ns does shouldn't be hard, but I know angel is wrong(and it doesn't give the actual breakdown anywase) and I don't have any others. Just post the damage by ns(it might actually not be the same for home/away, not totally sure but would like to find out) and then we'll have it here.

  9. #54
    Post Demon
    Join Date
    Feb 2009
    Posts
    1,407
    this is exactly what elurin meant when he said not to optimize every op.

    You need to look at things from perspective of a KD strat not of a single province. Comparing prop/NS in terms of damage done NS will win most of the time hands down EVEN if human prop deal more damage than NS, NS still wins. Why ?
    Because its reliable, not afected by wt/pat, troops in/out etc etc. NS runs will typically be run in 12hs cycles as you prep the targets for chain and there is no much space to get screwed over by the fact oponent casted Pat or you cant catch his troops home or whatnot, you want focused and concentrated damage done to his deff in order to chain him down easier because you only got 1 wave to get the job done properly.
    The way thieves are mostly used dictates that NS>prop most of the time.

    That said there are ways to use prop that are beneficial as pointed out. Weather its to maintain tpa on fast growing hybrids, disabling ub's, opening no wt's target to NS runs from your attackers with decent tpa or whatever other creative way you can think of.

    Numbers dont tell the whole story and you shouldnt base your decisions on them. ie if you want to prep for chain defhog hybrid with moderate off and no wt's prop is the way to go, and if you want to chain avian tact suicider runing 25 dpa with 30% gs's and some wt's you actually better off burning gs's in the first place than ns'in or proping ... or maybe your kd is runing crapy enough off to actually warant NS run on said avian ?

    I'd say ns>prop in 90% of cases but I still consider it my prerogative to decide on case to case basis what to use.

  10. #55
    Strategy Moderator
    Join Date
    Jan 2012
    Posts
    4,201
    Quote Originally Posted by Topsy View Post
    And I know collecting how much damage ns does shouldn't be hard, but I know angel is wrong(and it doesn't give the actual breakdown anywase) and I don't have any others. Just post the damage by ns(it might actually not be the same for home/away, not totally sure but would like to find out) and then we'll have it here.
    i'll see if i can find exact ns later tonight, i know it used to exists. http://ultima.thedragonportal.eu/ had/has a calculator to determine it but i cant get ultima to work at all on this pc.

  11. #56
    Forum Addict
    Join Date
    Sep 2011
    Posts
    1,014
    Thanks for the link. So if that's right, in 10 ns you will get about 5% of all specs killed, and about 2.5% of all leets.

  12. #57
    Post Demon
    Join Date
    Jul 2011
    Posts
    1,726
    Quote Originally Posted by KuhaN View Post
    Ok, im going to ask this question and i want a serious reply.

    Whens the last time you guys have played rogue or have done a prop wave yourselves? Instead of reading other ppls copy and pastes?

    Every single time ive done prop waves against people around my networth ive gotten results very similar to the data displayed in this thread. Every single time. The only difference when their army is home is that you get ospecs if they have any. Elite gains are always sh*t regardless.

    -Wizards usually two digits.
    -Soldiers are irrelevant (wasted op whenever u get this).
    -Thieves can be good or bad depending on targets tpa.
    -Elites usually two digits.
    -Specialists is the best bit, but remember u get both ospecs and dspecs (ospecs is irrelevant prior to chaining as well).

    So even if we be realistic and say you ns'd 250-300 specs each op, its much better than prop results (he obviously doesnt have 10k solds, obviously not even 1k, did u see the propd solds). Be realistic.

    if you find someone with a huge soldier stash it certainly isnt worthless. soldier aid bombs are a crucial part of keeping provinces strong in war.

    i would say the most important parts of prop are, in this order

    1. specs - incredible to re-pop ur specs after sending at dragon, or to save up more in order to send at the dragon
    1. soldiers - great great great for putting together aid packages
    1. thieves - thieves you usually get in fairly large chunks, and knocking off a bunch of thieves from an elf that is trying to A/M opens him up BIG-TIME for others

    2. wizards. propping an elf and knocking off some thieves and then grabbing chunks of his wizards is great. AW is better but if ur doing prop ur hoping for one of the first 3, but getting wizards isnt a fail.

    3. elites, you get a low amount so its typically not worth it.


    NS is all around useful. and i Kuhan is basing most of his perception on the fact that you should be opping TO SET UP CHAINS, which is typically correct!!

    Really the biggest advantage of prop is that you GET the troops, but you have to weigh if its really something you need. If ur setting up a chain on a target, its not the best op.... but their are times when u want to get those troops, and specifically when taking large chunks of thieves to allow for your A/t to get involved can be game-changing.

  13. #58
    Veteran
    Join Date
    Mar 2011
    Location
    The Netherlands
    Posts
    503
    nm this post
    Last edited by Humanoize; 15-03-2012 at 04:59.

  14. #59
    Post Demon lastunicorn's Avatar
    Join Date
    Aug 2011
    Posts
    1,458
    Quote Originally Posted by Persain View Post
    According to the wiki and what i remember of the data assumptions 1 and 2 are correct. if you would like to prove me wrong go ahead and show your data points. I'd hope to see at least 100-200 prop ops. (old excel file had about 1100 prop ops)
    I just propped 3 times, 2 were elites, the other I got 28 specs. When I looked at my throne I had gained 10 ospecs. This seems to validate the assumption that ospecs and dspecs are the same category.

  15. #60
    Post Fiend Agronaut's Avatar
    Join Date
    Sep 2011
    Posts
    140

    hmm

    i agree that nightstrikes are better then prop.i say that because of the damages i inflict on the enemy. prop is all over the place in the damages you do. generaly i hit an enemy attacker with heavy night strikes post his new som then move on to the next one once hes beaten down enough.propaganda sure has some up sides but it never inflicts the damages the night strikes do.i may use it when i need some specs to toss at a dragon or when i want to honor hump some poor guy. other then that thats about it.
    i think halflings are weaker in the thief department then before. would be nice if they added something like targeted propaganda. that would be a nasty little op and yah then it would be far better then night strikes

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •