| Pages: [1] 2 3 4 5 6 7 :: one page |
| Author |
Thread Statistics | Show CCP posts - 0 post(s) |

chrisss0r
|
Posted - 2009.01.16 18:27:00 -
[1]
Edited by: chrisss0r on 16/01/2009 18:29:06 Well my post about calculation bayesian probability in jamming of other ships stirred up some dirt und finally i got interested in how big the effect would actually be. Since i'm still not willing to calculate the probability in a complex model i did it the other way round. Created a situation in which the bayesian calculus should return a result drastically different from ordinary calculation and ran it through a few million times.
i asked a friend to bring up some code with the following specs:
jamchance: 50% (so he could use a coinmodel and i did not have to admit i needed it for an onlinegame -.-) Targets: 2 jammers: 6
The code first flips each coins. Every round has 6 flips (jammers). each coin is flipped till it's heads or till the 6 flips are used up. If one coin is heads in the first turn the other flips (jammers) can used on the other coin till all flips are used up (1 or 2 targets not jammed) or both coins are heads (= both targets jammed)
Well and now to the results:
The classical approach of a falcon jamming 2 targets with 6 jammers of 50% chance
(1-0.5^3)^2 = 76,5625 %
While the program delivered: 10000 runs each: 89,08% 89,11% 88,97% 88,69% 88,87% 89,35% 88,92% 88,91% 89,32% 88,95%
And these are not picked out and are totally repeatable
Tests with 1 million tries each: 89,07% 89,05% 89,08% 89,03% 89,01% 89,05% 89,01% 89,11% 89,05% 89,08% 89,09%
again not picked randomly and are totally reproduceable.
I would be very gratefull if people who have some coding skills could write something with factoring in 3 or more targets od more/less jammers. Or just write the code themselfs on this model to see if they encounter the same estimation
So it's quite safe to say that the real peobablity is around 88-89 instead of 76,5625 to jam both targets in the situation of a falcon having 6 jammers each with a 50% chance jam the target.
This results in: Classical caluculation: both targets permajammed during the first 40 seconds: 0.58618164 During the first 60 seconds: 0.4487 During the first 120 seconds: 0.2013
Using the real probability estimation (estimated by a few million runs..) During the first 40 seconds: 0.7744 during the first 60 seconds: 0.681472 During the first 120 seconds: 0.4644
So the real probability of beeing permajammed by a falcon for 2 minutes in the situation the model is seeted in is more than twice as high as the wrong and simple formula returns.
And this is not even factoring in overlapping jams and relock time
So happy flaming
|

chrisss0r
|
Posted - 2009.01.16 18:37:00 -
[2]
ah yes the code used:
#!/usr/bin/python # -*- coding: utf-8 -*- import random
class moo: def main(self): # initializing of success counter amount_successes = 0
# outer loop for repetition for i in range(1000):
# first flip. 0 is heads, 1 means tails c1 = self.flip() c2 = self.flip()
# loop for remaining throws amount_throws = 0 while amount_throws < 4: if c1 == 0: c1 = self.flip() amount_throws += 1 if c2 == 0: c2 = self.flip() amount_throws += 1
if (c1 == 1) and (c2 == 1): amount_successes += 1 break
print("Amount of successful tries: "+str(amount_successes))
return
def flip(self): return random.randint(0, 1)
if __name__ == "__main__": moo().main()
|

Lilith Velkor
Minmatar DEATH'S LEGION Red Box.
|
Posted - 2009.01.16 18:40:00 -
[3]
I've done a similar approach, I'll prolly release it to the public (its java so runs on your guys PCs) once I get the code cleaned up to a point it doesnt embarass me too much 
I've included locking times in my method, 'real-world' eve stats for the modules/ships, but not the 'waiting a few sec after jam dropped' tactic to get a few more seconds jam time.
# of jammers and ships are theoretically unlimited though, although I dont have a good model for ecm drones yet thats easy to implement.
|

Sidus Isaacs
Gallente
|
Posted - 2009.01.16 18:47:00 -
[4]
Numbers seems ok. |

Wannabehero
Absolutely No Retreat
|
Posted - 2009.01.16 18:50:00 -
[5]
Very interesting, thanks for sharing.
My conclusion... *ECM should be redesigned* |

bldyannoyed
Sebiestor tribe
|
Posted - 2009.01.16 19:22:00 -
[6]
Edited by: bldyannoyed on 16/01/2009 19:26:22 Edited by: bldyannoyed on 16/01/2009 19:23:09 So what you are essentially saying is that the chance of getting permajammed by a Falcon (or indeed any ECM) that is using MORE THAN ONE jammer is significantly higher than the numbers would at first suggest?
I wonder if CCP know this?
EDIT: Also, what does this do to ECCM? Does this insane maths of yours mean that doubling your sensor strength doesn't halve your chance of getting jammed?
EDIT EDIT: If this is true it should go in the same category as the epic SHC thread tht proves that DPS at max Falloff range is actually only 38% and not 50% as everybody seems to claim. |

lecrotta
lecrotta Corp
|
Posted - 2009.01.16 19:31:00 -
[7]
Edited by: lecrotta on 16/01/2009 19:34:48 DOH....
|

lecrotta
lecrotta Corp
|
Posted - 2009.01.16 19:34:00 -
[8]
Edited by: lecrotta on 16/01/2009 19:36:50
Originally by: bldyannoyed
So what you are essentially saying is that the chance of getting permajammed by a Falcon (or indeed any ECM) that is using MORE THAN ONE jammer is significantly higher than the numbers would at first suggest?
The chance of jamming a ship by using multiple jammers of the type that ship just happens to be has always been high or what would be the point of fitting a rack of a single racial instead of a mix?. Although the trade off in doing so limits your effectiveness against other races significantly.
I am not sure what this has proven tbh as the figures are simple and obvious, a 6 rack of a single races jammers can jam a low number (2) of that particular races ships very effectively is hardly a epiphany imho.
|

chrisss0r
|
Posted - 2009.01.16 19:38:00 -
[9]
Edited by: chrisss0r on 16/01/2009 19:38:26
Originally by: lecrotta
Originally by: bldyannoyed
So what you are essentially saying is that the chance of getting permajammed by a Falcon (or indeed any ECM) that is using MORE THAN ONE jammer is significantly higher than the numbers would at first suggest?
The chance of jamming a ship by using multiple jammers of the race that ship just happens to be has always been high or what would be the point of fitting a rack of a single racial instead of a mix?. Although the trade off in doing so limits your effectiveness against other races significantly.
I am not sure what this has proven tbh as the figures are simple and obvious, a 6 rack of a single races jammers can jam a low number (2) of that particular races ships very effectively is hardly a epiphany imho.
It proves that the chance to jam it is higher than believed in the eve community and maybe by ccp. it does not prove that 6 jammers are better than 4 or so but that 6 jammers on 2 targets don't have a 76% chance of jamming both targets but about 88 |

Etho Demerzel
Gallente Holy Clan of the Cone
|
Posted - 2009.01.16 19:40:00 -
[10]
Edited by: Etho Demerzel on 16/01/2009 19:46:27 Ok. Lets define your model accurately before we start to dig into the results.
Your model is one falcon and 2 targets. Your jam chances towards your targets are uniform, meaning you are using a full rack of multspecs. Multspecs maximum strength is 10, so it is reasonable to think your results can be approximated by a situation where a Falcon finds 2 Battleships without ECCM.
Do we agree until this point?
Now to the results:
Quote:
This results in: Classical caluculation: both targets permajammed during the first 40 seconds: 0.58618164 During the first 60 seconds: 0.4487 During the first 120 seconds: 0.2013
Using the real probability estimation (estimated by a few million runs..) During the first 40 seconds: 0.7744 during the first 60 seconds: 0.681472 During the first 120 seconds: 0.4644
So the real probability of beeing permajammed by a falcon for 2 minutes in the situation the model is seeted in is more than twice as high as the wrong and simple formula returns.
This is indeed, an accurate analysis of the impact knowing the results at each activation has over the total jam chance during this interval.
You may notice that the impact is less significant as the chances grow and more significant as the chances diminush, though. In the First 40 seconds the impact is 32% against 130% in the interval of 120s, for example. This means that this kind of activation is not necessarily significant to neraly perma jam scenarios, but very significant if your intention is to disrupt several ships for some time.
Quote:
And this is not even factoring in overlapping jams and relock time
Relocking times are hard to predict when you are jamming, because you don't know if the target has SBs, you don't know what the target is locking, etc. You can ignore a target that just got out of jam for a few extra seconds in an attempt to maximize the global effectiveness of yoru jammers, but if you overdo it you risk losing more of this effectiveness by allowing volleys to go through unimpeded.
Quote:
So happy flaming
This was a very good analysis of the tactic. You won't heave any flame from me. What you will have is the following discussion about what these results mean.
Basically your results imply that by optimizing the use of your multspec jammers, you are able to greatly improve your chances to impair 2 battleships. If that is the sole purpose of your post, then it is proved and there is nothing to say about it but to agree.
On the other hand, if the purpose of your post is to imply that in virtue of this Falcons are overpowered, the previous results are far from being a good proof of it.
Let me elaborate. Lets suppose that you have two gangs. One is composed by 2 battleships, the second is composed by a falcon and one battleship. By your own results in 120 seconds you will be able to jam both ships 46.44% of the time. If you facto lock time from battleship to battleships without any help from modules, we can overestimate this jam time to be, lets say 65%.
This means, that, IF BOTH opposing battleships are using neither ECCM, nor sensor boosters, in the end your opponent will have 70% of a battleship damage against your full Battleship damage(100%). You have a clear advantage, as you should, or else what would be the point in having a balanced gang. This advantage is not huge, though, and you have nothing near a permajam in either opponent.
Now, if both battleships were using ECCMs, your gang (Falcon+Battleship) would be in trouble. You can repeat the calculations for a 0.25 base chance and you will see that in this case, the two battleships will have a clear advantage. Showing that the correct counter DOES WORK.
Even more, in that same situation, we failed to account the effect of unguided weapons as drones, FoF missiles, etc. We also failed to account teh fact that 2 battleships may last more than 120s against the damage of a single one. |

chrisss0r
|
Posted - 2009.01.16 19:48:00 -
[11]
As I stated already numbers alone will never prove if something is overpowered or not. Though my personal opinion is that falcons are overpowered it has nothing to do with this calcs.
i just wanted to proof that you need the bayesian calcs in small gang situations and get the numbers the whole discussion is partly based on straight. As i already said everyone who is able to put more variables into the code is very welcome. Due to my incompetence when it comes to code i can't do it  |

Camilo Cienfuegos
Earned In Blood
|
Posted - 2009.01.16 19:48:00 -
[12]
The important part is to compare the jam chances over two, three and six cycles as per the original post between the commonly used mean average calculation against the more accurate bayesian probability calculation.
Two Cycles (40 seconds) Mean: 59% Bayesian: 77%
Three Cycles (60 seconds) Mean: 45% Bayesian: 68%
Six Cycles (120 seconds) Mean: 20% Bayesian: 46%
This simply demonstrates that the chance to be jammed over a prolonged period of time is markedly higher than previous believed/demonstrated. I hope this helps to clarify things somewhat. |

Etho Demerzel
Gallente Holy Clan of the Cone
|
Posted - 2009.01.16 19:56:00 -
[13]
Originally by: chrisss0r As I stated already numbers alone will never prove if something is overpowered or not. Though my personal opinion is that falcons are overpowered it has nothing to do with this calcs.
i just wanted to proof that you need the bayesian calcs in small gang situations and get the numbers the whole discussion is partly based on straight. As i already said everyone who is able to put more variables into the code is very welcome. Due to my incompetence when it comes to code i can't do it 
As I said, your demonstrations back your claims up that when trying to jam multiple targets, the right use of the knowledge of previous jams will help you to increase your chances considerably.
This technique has some technical issues when you use racial jammers, though, as you will be faced with the dilemma of pairing the wrong racial against the wrong ship (thus blocking its use against the correct ships), or skiping it and saving it for the next cycle against the correct ship, thus decreasing the effect of the method.
Now regarding the Falcon overpoweredness (I don't even know if this word exists but I will use it anyways ), to make such a claim you need objective results to back it up, as you did (and very well), with your previous claim. Otherwise it will be just an irrational and overemotional reaction to something you don't like, which is fitting to Lyria, but not to you. |

Camilo Cienfuegos
Earned In Blood
|
Posted - 2009.01.16 19:57:00 -
[14]
Quote: This means, that, IF BOTH opposing battleships are using neither ECCM, nor sensor boosters, in the end your opponent will have 70% of a battleship damage against your full Battleship damage(100%). You have a clear advantage, as you should, or else what would be the point in having a balanced gang. This advantage is not huge, though, and you have nothing near a permajam in either opponent.
Now, if both battleships were using ECCMs, your gang (Falcon+Battleship) would be in trouble. You can repeat the calculations for a 0.25 base chance and you will see that in this case, the two battleships will have a clear advantage. Showing that the correct counter DOES WORK.
I would be interested to see if this is the case. Looking at the above figures, I have a feeling that if the 2 battleship gang has any advantage, it will be marginal (~5%) at best. I suspect that the Falcon + Battleship combination might well have the upper hand by a percentile or two.
Key word there is suspect of course. |

Cohkka
Celestial Apocalypse
|
Posted - 2009.01.16 19:59:00 -
[15]
Nice post, keep up the good work. |

Camilo Cienfuegos
Earned In Blood
|
Posted - 2009.01.16 20:00:00 -
[16]
Originally by: Cohkka Nice post, keep up the good work.
Oh, and QFT! |

lebrata
Hedion University
|
Posted - 2009.01.16 20:12:00 -
[17]
Edited by: lebrata on 16/01/2009 20:15:32
Originally by: Etho Demerzel
Lets suppose that you have two gangs. One is composed by 2 battleships, the second is composed by a falcon and one battleship. By your own results in 120 seconds you will be able to jam both ships 46.44% of the time. If you facto lock time from battleship to battleships without any help from modules, we can overestimate this jam time to be, lets say 65%.
This means, that, IF BOTH opposing battleships are using neither ECCM, nor sensor boosters, in the end your opponent will have 70% of a battleship damage against your full Battleship damage(100%). You have a clear advantage, as you should, or else what would be the point in having a balanced gang. This advantage is not huge, though, and you have nothing near a permajam in either opponent.
Now, if both battleships were using ECCMs, your gang (Falcon+Battleship) would be in trouble. You can repeat the calculations for a 0.25 base chance and you will see that in this case, the two battleships will have a clear advantage. Showing that the correct counter DOES WORK.
OMG good point.
Still it was a very nice bit of calculus done by the OP.
|

Terianna Eri
Amarr Scrutari
|
Posted - 2009.01.16 20:16:00 -
[18]
Originally by: Etho Demerzel Now, if both battleships were using ECCMs, your gang (Falcon+Battleship) would be in trouble. You can repeat the calculations for a 0.25 base chance and you will see that in this case, the two battleships will have a clear advantage. Showing that the correct counter DOES WORK.
Can we see the calculations for 0.25 jamming chance, 6 available jammers, against 2 targets? __________________________________
Originally by: Arthur Frayn How much to ruin all your holes, luv?
|

Etho Demerzel
Gallente Holy Clan of the Cone
|
Posted - 2009.01.16 20:19:00 -
[19]
Originally by: Camilo Cienfuegos
I would be interested to see if this is the case. Looking at the above figures, I have a feeling that if the 2 battleship gang has any advantage, it will be marginal (~5%) at best. I suspect that the Falcon + Battleship combination might well have the upper hand by a percentile or two.
Key word there is suspect of course.
Even if that was the case and they barely break even in targetable damage. You will end with 100% damage + the damage of non target weapons against 100% damage, and 200% EHP against 100% EHP. An lets not forget that the falcon will be at 120 km tops because it will be using multispecs, and thus reasonably vulnerable to cruise missiles, for example. =====
"If a member of the EVE community finds he or she cannot accept our current level of transparency, we bid you good luck in finding a company that meets your needs." - CCP kieron... |

lebrata
Hedion University
|
Posted - 2009.01.16 20:22:00 -
[20]
Originally by: Terianna Eri
Originally by: Etho Demerzel Now, if both battleships were using ECCMs, your gang (Falcon+Battleship) would be in trouble. You can repeat the calculations for a 0.25 base chance and you will see that in this case, the two battleships will have a clear advantage. Showing that the correct counter DOES WORK.
Can we see the calculations for 0.25 jamming chance, 6 available jammers, against 2 targets?
Another real time scenario like that with 2 BS fitted with ECCM vs falcon/BS team would be interesting tbh as the fight was closer than i thought it would be without ECCM.
I thought the ecm/BS gang would own the 2BS gang.
|

Camilo Cienfuegos
Earned In Blood
|
Posted - 2009.01.16 20:24:00 -
[21]
The code is very straightforward. If you wish to work out the chances for different strengths of jammers, simply modify the return random.randint line. Currently it'll return only 0 or 1, but you could get reasonably accurate results for 25% jam chance by returning 0 to 3 or for 10% chance by returning 0 to 9 and so on. To modify the number of jammers used, it's the while amount_throws line; simply change it to 2 for a single jammer (pointless), 3 for 2 jammers etc, etc.
I've highlighted the appropriate lines.
Originally by: chrisss0r
#!/usr/bin/python # -*- coding: utf-8 -*- import random
class moo: def main(self): # initializing of success counter amount_successes = 0
# outer loop for repetition for i in range(1000):
# first flip. 0 is heads, 1 means tails c1 = self.flip() c2 = self.flip()
# loop for remaining throws amount_throws = 0 while amount_throws < 4: if c1 == 0: c1 = self.flip() amount_throws += 1 if c2 == 0: c2 = self.flip() amount_throws += 1
if (c1 == 1) and (c2 == 1): amount_successes += 1 break
print("Amount of successful tries: "+str(amount_successes))
return
def flip(self): return random.randint(0, 1)
if __name__ == "__main__": moo().main()
|

Camilo Cienfuegos
Earned In Blood
|
Posted - 2009.01.16 20:31:00 -
[22]
Quote: Even if that was the case and they barely break even in targetable damage. You will end with 100% damage + the damage of non target weapons against 100% damage, and 200% EHP against 100% EHP. An lets not forget that the falcon will be at 120 km tops because it will be using multispecs, and thus reasonably vulnerable to cruise missiles, for example.
If we assume (a common scenario) two close range battleships fitted with ECCM versus one close range battleship with a long range support Falcon, the battleships aren't going to be able to touch the Falcon (a mechanic which to my mind is working perfectly). If we're seeing roughly equal amounts of damage going out against both side, provided the single battleship doesn't foolishly decide to attack both his opponents simultaneously, he has every chance of finishing off one of those battleships before he himself goes down. That then leaves Battleship (admittedly almost dead) + Falcon (with three jammers now free). The chances of the now lone battleship winning despite it's full HP are far less as there are now six and not three jammers on him; the almost dead battleship could probably warp out at this point if the need is there, or provided there's good communication might be able to use the time purchased by the Falcon to repair whilst keeping his target pinned.
It's crucial that we can demonstrate the exact figures so that we can see which side this falls on. If it falls in favour of the Falcon + Battleship, it's indisputable that there's a big problem there.
|

Chainsaw Plankton
IDLE GUNS IDLE EMPIRE
|
Posted - 2009.01.16 20:35:00 -
[23]
someone did this in matlab a little while ago, and posted pretty graphs and such, factoring in lock time and whatnot to see what percentage someone was "out of the fight"
|

chrisss0r
|
Posted - 2009.01.16 20:36:00 -
[24]
i've tried to run it with a 25% chance of jamming
each 10000 runs returns: 47,25% 46,97% 46,24% 46,61% 46,69% 46,72% 46,14% 46,98% 46,79%
will base my calcs on 46%
While the simple calc chance would be 33.42%
Resulting in
both targets permajammed 2 cycles (40 secs) mean: 0.1116 bayes: 0.2116
3 cycles (60 secs) mean: 0.0373 bayes: 0.097336
6 cycles (120 secs) mean: 0.0001 bayes: 0.0009
eccm works kinda good:
Ethniel the gap is bigger than you stated because it's only the probability both ships will be permjammed. you miss out a point where one of the opposing bs is jammed while the other is not
|

lebrata
Hedion University
|
Posted - 2009.01.16 20:39:00 -
[25]
Originally by: Camilo Cienfuegos
If we assume (a common scenario) two close range battleships fitted with ECCM versus one close range battleship with a long range support Falcon, the battleships aren't going to be able to touch the Falcon (a mechanic which to my mind is working perfectly). If we're seeing roughly equal amounts of damage going out against both side, provided the single battleship doesn't foolishly decide to attack both his opponents simultaneously, he has every chance of finishing off one of those battleships before he himself goes down. That then leaves Battleship (admittedly almost dead) + Falcon (with three jammers now free). The chances of the now lone battleship winning despite it's full HP are far less as there are now six and not three jammers on him; the almost dead battleship could probably warp out at this point if the need is there, or provided there's good communication might be able to use the time purchased by the Falcon to repair whilst keeping his target pinned.
That is totally subjective as the jammers could fail a lot at the start of the fight getting the BS melted or they could work very very well at the start of the fight giving the single BS a good advantage.
Originally by: Camilo Cienfuegos It's crucial that we can demonstrate the exact figures so that we can see which side this falls on. If it falls in favour of the Falcon + Battleship, it's indisputable that there's a big problem there.
I would disagree with that as i feel a BS + ewar ship VS 2 close range BS should be favorable towards the ewar/BS team as its a much more tactical setup than just tank/gank.
|

Camilo Cienfuegos
Earned In Blood
|
Posted - 2009.01.16 20:50:00 -
[26]
Code for 25% jammer chance, two targets, 3 jammers per ship:
#!/usr/bin/python # -*- coding: utf-8 -*- import random
class moo: def main(self): # initializing of success counter amount_successes = 0
# outer loop for repetition for i in range(1000):
# first flip. 0 is success, 1-3 failure c1 = self.flip() c2 = self.flip()
# loop for remaining throws amount_throws = 0 while amount_throws < 4: if c1 != 0: c1 = self.flip() amount_throws += 1 if c2 != 0: c2 = self.flip() amount_throws += 1
if (c1 == 0) and (c2 == 0): amount_successes += 1 break
print("Amount of successful tries: "+str(amount_successes))
return
def flip(self): return random.randint(0, 3)
if __name__ == "__main__": moo().main()
Thanks to chrisss0r for the original code.
|

Camilo Cienfuegos
Earned In Blood
|
Posted - 2009.01.16 20:51:00 -
[27]
Quote: i feel a BS + ewar ship VS 2 close range BS should be favorable towards the ewar/BS team as its a much more tactical setup than just tank/gank.
It certainly is, which is why if it's mathematically stacked in favour of them as well as tactically, I believe we have a problem.
|

Etho Demerzel
Gallente Holy Clan of the Cone
|
Posted - 2009.01.16 20:54:00 -
[28]
Originally by: Camilo Cienfuegos
If we assume (a common scenario) two close range battleships fitted with ECCM versus one close range battleship with a long range support Falcon, the battleships aren't going to be able to touch the Falcon (a mechanic which to my mind is working perfectly). If we're seeing roughly equal amounts of damage going out against both side, provided the single battleship doesn't foolishly decide to attack both his opponents simultaneously, he has every chance of finishing off one of those battleships before he himself goes down. That then leaves Battleship (admittedly almost dead) + Falcon (with three jammers now free). The chances of the now lone battleship winning despite it's full HP are far less as there are now six and not three jammers on him; the almost dead battleship could probably warp out at this point if the need is there, or provided there's good communication might be able to use the time purchased by the Falcon to repair whilst keeping his target pinned.
It's crucial that we can demonstrate the exact figures so that we can see which side this falls on. If it falls in favour of the Falcon + Battleship, it's indisputable that there's a big problem there.
I agree that the numbers are important. And fortunately chrisss0r just provided them in hsi post above.
Now we have a base chance of 46% against 2 battleships with 1 ECCM each.
Before we start let me list my assumptions:
1) In average the drones account for 25% of a battleship damage 2) The battleships have no other form of untargeted damage as FoF, Smartbombs, etc 3) The relock time increases the effect fromt he ECM from 46% to 60% (look the charts you referred to for validation)
Now using a single battleship damage as 100% the total damage of the two jammed batleships will be 50% (from the drones) + 150% * 0.4 = 60%, or a total of 110% of a battleship damage. As you see the advantage is clearly on the 2 battleship groups, as it should be considering they have ECCM equiped. Note that a single module in each battleship turns the fight in their favor.
Now if you consider what effect 2 tracking disruptors (one in each battleships) used by a curse would have in the same fight you will clearly see why the Falcon is not the best buddy to the single battleship. =====
"If a member of the EVE community finds he or she cannot accept our current level of transparency, we bid you good luck in finding a company that meets your needs." - CCP kieron... |

lebrata
Hedion University
|
Posted - 2009.01.16 20:56:00 -
[29]
Originally by: Camilo Cienfuegos
Quote: i feel a BS + ewar ship VS 2 close range BS should be favorable towards the ewar/BS team as its a much more tactical setup than just tank/gank.
It certainly is, which is why if it's mathematically stacked in favour of them as well as tactically, I believe we have a problem.
That makes no sense at all.
|

chrisss0r
|
Posted - 2009.01.16 21:02:00 -
[30]
Edited by: chrisss0r on 16/01/2009 21:02:15
Originally by: Etho Demerzel
Originally by: Camilo Cienfuegos
If we assume (a common scenario) two close range battleships fitted with ECCM versus one close range battleship with a long range support Falcon, the battleships aren't going to be able to touch the Falcon (a mechanic which to my mind is working perfectly). If we're seeing roughly equal amounts of damage going out against both side, provided the single battleship doesn't foolishly decide to attack both his opponents simultaneously, he has every chance of finishing off one of those battleships before he himself goes down. That then leaves Battleship (admittedly almost dead) + Falcon (with three jammers now free). The chances of the now lone battleship winning despite it's full HP are far less as there are now six and not three jammers on him; the almost dead battleship could probably warp out at this point if the need is there, or provided there's good communication might be able to use the time purchased by the Falcon to repair whilst keeping his target pinned.
It's crucial that we can demonstrate the exact figures so that we can see which side this falls on. If it falls in favour of the Falcon + Battleship, it's indisputable that there's a big problem there.
I agree that the numbers are important. And fortunately chrisss0r just provided them in hsi post above.
Now we have a base chance of 46% against 2 battleships with 1 ECCM each.
Before we start let me list my assumptions:
1) In average the drones account for 25% of a battleship damage 2) The battleships have no other form of untargeted damage as FoF, Smartbombs, etc 3) The relock time increases the effect fromt he ECM from 46% to 60% (look the charts you referred to for validation)
Now using a single battleship damage as 100% the total damage of the two jammed batleships will be 50% (from the drones) + 150% * 0.4 = 60%, or a total of 110% of a battleship damage. As you see the advantage is clearly on the 2 battleship groups, as it should be considering they have ECCM equiped. Note that a single module in each battleship turns the fight in their favor.
Now if you consider what effect 2 tracking disruptors (one in each battleships) used by a curse would have in the same fight you will clearly see why the Falcon is not the best buddy to the single battleship.
that calculation is not complete since it does not factor in the probability that one bs is jammed while the other is not. that is somewhat compensated by factoring relock time in by an estimate. my estimate is that the break even for 2 bs vs 1 bs + 1 falcon is around 1 bs having 1 and 1 bs having 2 eccm fitted.
|
| |
|
| Pages: [1] 2 3 4 5 6 7 :: one page |
| First page | Previous page | Next page | Last page |