These forums have been archived and are now read-only.

The new forums are live and can be found at https://forums.eveonline.com/

Player Features and Ideas Discussion

 
  • Topic is locked indefinitely.
 

[PROPOSAL] Change to ECM
Author
Previous Topic Next Topic
Loius Woo
Sebiestor Tribe
Minmatar Republic
#1 - 2012-08-25 18:05:07 UTC  |  Edited by: Loius Woo
It is often discussed that the current implementation of ECM is overpowered. I am not taking this as a given, but acknowledge that the way ECM works can be; 1. Frustrating for those affected by it and 2. overshadowing of other EWAR types.

I have given this issue some thought and I think that the core of the problem is that when a jammer is effective, the target becomes ENTIRELY ineffective. While this should not be significantly changed, I feel that there is no graceful degradation of performance for a ship afflicted by ECM, it is all or nothing. So my proposed change would allow for a more graceful degradation on the ship being jammed.

The way it works now:
J=jammer strength
S= Sensor strength
N=number of jammers
p=probability of jammer functioning

So p=J/S
and chance of jammer failing is 1-p
with N number of jammers, the chance that all will fail is (1-p)^N
And the chance that at least one will work is 1-(1-p)^N

When the jammer works, the target ship's max locked targets is modified to 0, meaning that they are unable to use any targeted modules.

Proposed Change:
Instead of modifying the locked targets to 0, reiterate the chance calculation for each locked target.

For reference, in statistics, the chances that two unrelated events will follow one another is equal to the chance of one multiplied by the chance of the next one and so on. For example, the chances of rolling a 6 on a six sided dice is 1/6th or 16.6%. The chances to roll two successive 6's is (1/6)*(1/6) or 2.7%. Three successive 6's would be 0.4%

Simplified, the chances of rolling a 6 N successive times is (1/6)^N. So in general terms, the chances of doing anything multiple times in a row if the chance is the same for each one and the outcome of one does not affect the others (which is true in ECM in Eve) the chances of N successive iterations is (p)^N.

The way the server could tackle this is simple, when a jam is attempted the server checks the jam strength and sensor strength both already, in this idea, it would also check max targets. Then the server performs the probability calculation and then generates a random number (rolls the dice) to see if the jam has succeeded. In this proposal the server would calculate the chances of jam successively for each possible target. So if a ship has 4 max targets, then the server would calculate (j/S), (J/S)^2, (J/S)^3, and (J/S)^4. Then each probability is nested before a random number is generated, if the random number is for example in the range of (J/S)^3, then the ship's max target would be reduced by 3.

EXAMPLE:
A decent fit and piloted Falcon has 7 tech two Gravimetric jammers with a strength of 12 (picking a round number for ease which is in the range of a decent falcon)
He is targeting a decent piloted Basilisk with a single ECCM gravimetric II fitted and his sensor strength is double at 44. The Basilisk has 10 max locked targets.

J/S is 12/44 which equals 27.27%

With the way things work currently, the chances of at least one jam being successful out of 7, 1-(1-.2727)^7=89.24%

WIth this change it would be a little more complicated. The jam probability of each jammer would be:
(12/44)=27.27% for one target
(12/44)^2=7.43% for two targets
(12/44)^3=2.03% for three targets
(12/44)^4=0.55% for four targets
(12/44)^5=0.15% for five targets
...
...
(12/44)^10=0.000227% for ten targets.

This seems like a huge nerf to ECM, but wait.

So the chances that with all seven jammers, you have reduced the basilisk's max targets by at least 1 is 89.24% as it was before to jam him out completely.

But, the chances of reducing the basilisk to 8 targets is more complicated because you need either the one target break to work twice or a jammer to get a two target break once. So you need to add those two probabilities. Each successive target number has multiple ways of getting there and I will spare you the detailed math and just post the chances of max targets being 0-10.


Chances that any single jammer among 7 will reduce max targets to the indicated amount.
Max Targets 9=89.23%
Max Targets 8 =41.78%
Max Targets 7 =13.36%
Max Targets 6 =3.89%
Max Targets 5 =1.05%
Max Targets 4 =0.28%
Max Targets 3 =0.07%
Max Targets 2 =0.021%
Max Targets 1 =0.005%
Max Targets 0=0.001%

The ACTUAL chances of reducing max locks by some amount depends on many perturbations of probability (detailed in the next post).

In this example, a single Falcon with 7 jammers of strength 12 against a Logistics cruiser with sensor strength 44 and 10 max targets, if all 7 jammers are on the Logi, it will reduce the lockable targets by one 89.23% of the time and would almost never prevent complete targeting. This is much less effective than the current system and there is a pretty good chance that anyone being jammed will have SOME ability to target something. As a result, the strength of jammers could be adjusted to balance it out. For reference, I will post some results with various jam strength to sensor strength and with different number of jammers. So that effects of balancing jammer strengths can be seen.
Loius Woo
Sebiestor Tribe
Minmatar Republic
#2 - 2012-08-25 18:05:17 UTC  |  Edited by: Loius Woo
After a bit of work and some tweaking, the spreadsheet I am using to figure out the numbers here is done and I have posted it Here: ECM Proposal Document.

It is an excel file with two tabs.

The first tab is a calculator. If you put in a jammer strength, sensor strength and number of jammers, it will tell you, for up to 10 targets, what the chance of one jammer to jam 1-10 targets and the chance for ANY of the jammers to jam 1-10 targets.

The second tab shows a bunch more data.
For any possible J/S ratio (1% to 100%) it shows the chance of a single jammer to prevent 1-10 targets and the percentage chance that any combination of jammers can totally remove 2-10 targets. The combinations can be pretty complicated. For instance, the percentage that a jamming ship can reduce the max targets by six is a combination of the chance to reduce targets by one six times, by 3 twice, by 2 three times, by 2 and by four, by two twice and by one twice etc. You get the idea how it can be complicated.

A graph which shows the progression of both sets of data is also shown.

From this data, if this idea where to be implemented, then I believe that the best range of jammer/Sensor strength's in the game should be about 20%-80%.

The overall point is two fold, one is to provide for a graceful degradation of capability in the face of jamming such that a devoted jamming ship can still perma-jam you and that a couple of jammers will only remove your ability to target some of your max targets and second is to make the system such that under the best possible circumstances, a jamming ship is still capable of being a force multiplier as it is now.

It is interesting to point out that at a certain J/S ratio (about .7) the chances of preventing multiple target locks becomes more likely than jamming a single target lock. This would be to me the best place to separate jamming specific ships from non jamming specific ships.

I know that ECM is functioning as intended and that it is a good system. It is also the only system in Eve that functions in an all or nothing chance based way. I know this as a pilot that spent many many engagements in a falcon. My intention here is to create a system that has a slow increase in capability (so that ships not meant to be ECM ships are marginally good at it) and a steep increase in capability that tapers out to some maximum (so that ECM specific ships are much better at it, but level out). In addition, this makes it so that ECM drones in groups of 5 are much less effective than groups of 25 or more, so that the effects of ECM drones or un-bonused ships is only good when in large numbers.

Please discuss, but don't do so without actually looking at the data and understanding how ECM works now and how a change like this would effect it.

What I am not doing is saying that ECM is bad or game breaking or OP. What I am saying is that there is a more complex and realistic way to implement ECM that doesn't feel gamey and that is not as frustrating for the victims of it. The overall goal with this is to make the ships that are supposed to be good at it and ships that are not meant to be good at it to have a gradual ramp up and ramp down to effectiveness, a gradual ramping of effectiveness that is currently not there.
A Soporific
Perkone
Caldari State
#3 - 2012-08-25 18:55:34 UTC
In my experience, you only need the one target. So, it would make ECM a thing only in large fleet combat, and render it entirely unusuable as a tank or in small gang combat.

It's a huge nerf, and I'm unconvinced that it fixes things as much as eliminates a play style essentially.
Sigras
Conglomo
#4 - 2012-08-25 19:55:58 UTC
for 90% of ships, you only need one target, now this would really screw up logistics but not do much to any other targets.

It is my opinion that anything that completely disables another person and does not give them the ability to do anything except move is a bad game mechanic.

I think that they should change the active ECCM to provide almost no sensor strength (something like 20%) passively; then when you turn the module on, you're immediately unjammed and able to lock things, even though it would have a 20 second cycle time that would disallow early deactivation like a jammer.

You might even consider buffing ECM in lieu of this change

It would also be interesting to introduce a T3 module that was related to sensors, you could script it to be a sensor booster (range or speed +15%) or an ECCM with no sensor strength modifier, it just unjamms you when you get jammed.

This way, dampeners + ECM would still work, but it would still give the pilot on defense something to do to get out of it.
Kitt JT
True North.
#5 - 2012-08-25 21:23:16 UTC
I believe ECM is currently working as intended.
Loius Woo
Sebiestor Tribe
Minmatar Republic
#6 - 2012-08-26 02:55:31 UTC
I figured out the issues I was having with the math and have now posted a new second post and a link to the excel spreadsheet I used to figure it all out.

Please discuss.
Kitt JT
True North.
#7 - 2012-08-26 04:00:07 UTC
I believe ECM is currently working as intended.
Loius Woo
Sebiestor Tribe
Minmatar Republic
#8 - 2012-08-26 04:02:33 UTC
Kitt JT wrote:
I believe ECM is currently working as intended.


You are repeating yourself.

Also, I stated that I understand that ECM is working as it is intended to work, but that the ramp up for it is not in any way graceful and that it is the only system in eve that is a linear relationship. As such, I believe there is a way to make it more complex, while preserving the effectiveness of ECM dedicated ships.

Now, please read the post, analyze the data, compare it to how ECM works now, and THEN make a comment.
Colonel Xaven
Perkone
Caldari State
#9 - 2012-08-26 09:19:50 UTC
Why touching a running system?

www.facebook.com/RazorAlliance
Caliph Muhammed
Perkone
Caldari State
#10 - 2012-08-26 10:27:47 UTC  |  Edited by: Caliph Muhammed
I propose we change weapons to only do damage occasionally. As you see many complain of weapons tendency to affect them negatively. They argue its far too easy to turn on the module without having to aim and that perhaps an across the board nerf to all weapons systems is in order.

I'd also personally advocate a buff to ECM as it does very little damage. Perhaps a damage modifier for ECM is worthwhile considering.

And can we not all agree that miners harvest rocks far too quickly for the good of EVEs economy? Lets lobby for a 50% cut in the rate of ore, gas and ice harvesting.