Ambrosia Garden Archive
    • Jamming in Nova


      A logical explanation

      The Short Version: Let Vk be the type-k vulnerability of the missile, let Sk be the total jamming strength of the target, but not greater than 100, and let Uk = 100 - Sk. Let Lk = Uk / Vk, but not be greater than 1. L = L1 * L2 * L3 * L4 gives the odds the missile with lock on target. Note that if Uk >= Vk then Lk = 1.

      Introduction to Jamming in Nova: There are four types of jamming; every missile is between 0 and 100% vulnerable to each, and every jammer has strength between 0 and 100 in each. The vulnerability of the missile refers to what percent of jammers have the possibility of jamming that missile, and the strength of the jammer refers to the percent of missiles that it has the possibility of jamming. Sound confusing? Just wait until we factor in multiple jammers, duplicate jammers, innate (gövt-based) jamming, and system interference! Luckily there's a simple way to find out how likely a missile is to be jammed in any situation, and even what its behavior will be.

      The Basics: Let V1 be the vulnerability of some arbitrary missile to type-1 jamming. If V1 = 0 then it cannot be jammed by type-1 jamming, so let's assume V1 > 0. Let S1 be the type 1 jamming strength of some arbitrary jammer. If that missile is fired at a ship with only that jammer, then it will have L1 = (100 - S1) / V1 odds of locking onto the target. If L1 > 1 then round L1 down to 1. The missile will have J1 = 1 - L1 odds of being jammed.

      An Example: Let's say the missile is an IR Missile, with 100% vulnerability to type-1 jamming, and the jammer is a Civilian IR Jammer, with 20 strength in type-1 jamming. Now L1 = (100 - 20) / 100 = 80 / 100 = 0.8. The missile will lock on 80% of the time, meaning 20% of the missiles fired will be jammed.

      Multiple Jammers: Earlier I used S1 to represent the type-1 jamming strength of a jammer. Now I am going to extend that definition. S1 is the total type-1 jamming strength of the target, and is the sum of the strengths of all the type-1 jammers and the innate type-1 jamming strength of the target. S1 = S1a + S1b + S1c + ... where a is innate jamming and b, c, d, and so forth are different jammers. Mutiples of the same outfit that add jamming do not stack, so only add the jamming strength once for each different outfit type. If S1 > 100 then round it down to 100.

      From now on I will use U1 = 100 - S1 to represent the unrealized type-1 jamming potential of the target, simplifying the equation for lock-on odds to L1 = U1 / V1. Furthermore, I will use U2 = 100 - S2 for the unrealized type-2 jamming potential of the target, where S2 is the sum total type-2 jamming strength of the target, rounded down to 100 if it's greater than 100. Similarly, U3 = 100 - S3, and U4 = 100 - S4.

      Another Example: A Hellhound Missile, with V2 = 80, is fired at a Rebel ship, S2a = 6, that has a Rebel Radar Jammer, S2b = 30. Now S2 = 36, so U2 = 64 and thus L2 = 64/80, or 0.8, meaning 80% of them will hit, and 20% will be jammed. If the Rebel ship did not have the jammer then L2 would equal 94 / 80, which is greater than 1, meaning 100% of the missiles would lock on.

      Combined Jamming Types: To find the odds a missile with more than one JamVuln type will lock on target, simply find L1, L2, L3 and L4, rounding each down to 1 as necessary, and multiply them together. L = L1 * L2 * L3 * L4

      Yet a Third Example: Nanites, with V3 = 10 and V4 = 50, are fired at a Vell-os ship with the Distract Sensors ability, S3a = 35, S3b = 40, S4a = 20, S4b = 35. S3 = 75 and S4 = 55, so U3 = 25 and U4 = 45.

      Now L3 = 25/10 and L4 = 45/50, which become 1 (rounded from 2.5) and 0.95, respectively. L = 1 * 0.95 = 0.95, so 95% of them will lock on, and only 5% will be jammed. Notice how the type-3 jamming didn't help at all, because U3 > V3.

      Interference: I don't know how interference affects confusable missiles. Even with 100 interference, missiles that can be confused by it sometimes aren't. It appears to be a rate of around 1 in 20 not getting confused. Furthermore, a missile can become confused, then later unconfused, and possible many more times. I don't really know how this works.

      Behaviors: A missile not jammed will inertialessly bear down on its target. If it loses its lock when the target isn't directly head then once the lock is lost the missile behaves like a projectile (flies straight) and cannot regain the lock even if the target becomes directly ahead of the missile later on.

      A jammed missile will behave like a projectile unless it's set to turn away when jammed, in which case it will fly directly away from its target, to the best of its turn-rate's ability. No missiles in the stock Nova scenario have the turns-away behavior.

      A confused missile will fly in circles. A confused missile cannot be jammed. Missiles can become confused and then unconfused again. I don't know how or when the game calculates which missiles are confused based on system interference.

      A missile that's decoyed by asteroids, if not confused or jammed, has some set odds of taking an asteroid within a certain radius of it as a new target. I don't know the odds or radius, but it seems to be about 50% and 100 pixels. I haven't done any testing on the odds of a missile that can double-cross its parent ship when jammed actually doing so.

      Technically I haven't done any testing to see if gövt-induced jamming functions as I said it does, I just presumed it works that way. Given what I know about the Nova engine and its history of functioning in logical ways, however, the smart money says I'm wrong, and it's most likely that innate-jamming strength actually add frog-legs to your ship that twitch on simultaneous contact with two different types of metal. EDIT: I just tested and, at least for the player, gövt-based jamming works as expected.

      All missiles fired will travel for a certain distance (time?) locked onto the target, and then they will suddenly switch to their jammed/confused behavior all at once. A missile only checks to see if it's jammed or confused once. After that its behavior is set, though it could still become decoyed or lose its lock if not facing its target.

      More Information: Negative jamming works. It is added into (subtracted from) the Sk total, but that total cannot go below zero. Sensor Boosts (interference modifiers) do not affect the odds of missiles becoming confused.

      About the Formulae: The Nova engine picks a random integer from 1 to the JamVuln of the missile and adds that to the targets jamming, for each jamming type. If any of those numbers are greater than 100 then the missile is jammed.

      From that knowledge I derived Lk = Uk / Vk, and L = L1 * L2 * L3 * L4 to be the lock-on odds. Testing supports this analysis. If Uk >= Vk then Lk = 1. Remember to round Sk down to 100 if it's greater than that.

      A Final Example: A certain Scarab, inherently of gövt 147, has a Polaris, a Nil'kemorya, a Civilian IR, a Civilian Radar, a Cheap Military IR, a Cheap Military Radar, and a Pirate jammer, and is attacked by IR Missiles, EMP Torpedoes, Polaron Torpedoes, and Gravimetric Missiles.

      S1a = 25, S1b = 40, S1c = 50, S1d = 20, S1e = 0, S1f = 60, S1g = 0, S1h = 30
      S2a = 25, S2b = 40, S2c = 50, S2d = 0, S2e = 15, S2f = 0, S2g = 40, S2h = 20
      S3a = 18, S3b = 20, S3c = 45, S3d = 0, S3e = 0, S3f = 0, S3g = 0, S3h = 0
      S4a = 9, S4b = 10, S4c = 20, S4d = 0, S4e = 0, S4f = 0, S4g = 0, S4h = 0

      V1a = 100, V2a = 0, V3a = 0, V4a = 0
      V1b = 0, V2b = 95, V3b = 0, V4b = 0
      V1c = 0, V2c = 0, V3c = 0, V4c = 90
      V1d = 0, V2d = 0, V3d = 100, V4d = 0

      So S1 = 225, which rounds to 100, making U1 = 0.
      S2 = 190, which also rounds to 100, making U2 = 0.
      S3 = 83, making U3 = 17.
      S4 = 39, making U4 = 61.

      The IR Missiles have 0/100 = 0 odds of locking on: they're all jammed.
      The EMP Torpedoes have 0/95 odds of locking on: all of them not confused by sensor interference are jammed.
      The Polaron Torpedoes have 61/91 odds of locking on: 33% are jammed.
      The Gravimetric Missiles have 17/100 odds of locking on: 83% of those not confused are jammed.

      A Chart of Jam Values in Stock Nova:

      Missiles:

      		Vuln1	Vuln2	Vuln3	Vuln4	Confusable
      IR		100	0	0	0	No
      Radar		0	100	0	0	Yes
      Grav		0	0	100	0	No
      EW		0	0	0	100	Yes
      P Torp		0	0	0	90	No
      EMP		0	95	0	0	Yes
      Nanites		0	0	10	50	No
      Multi-torp	0	0	0	90	No
      P Torp FWC	0	0	0	90	No
      Multi-FWC	0	0	0	90	No
      P Massive T	0	0	0	90	No
      Nanite sub	0	0	10	50	No
      Hellhound	0	80	0	0	Yes
      Radar dis	0	100	0	0	Yes
      

      Jammers:

      		Str1	Str2	Str3	Str4
      Civ IR		20	0	0	0
      Civ Radar	0	15	0	0
      Mil IR		30	0	0	0
      Mil Radar	0	25	0	0
      Aur IR		35	0	0	0
      Aur Radar	0	20	0	0
      Polaris		40	40	20	10
      Nil'k		50	50	45	20
      Reb IR		35	0	0	0
      Reb Radar	0	30	0	0
      Pirate		30	20	0	0
      Vell-os (DS)	45	45	40	35
      Chp IR		60	0	0	0
      Chp Radar	0	40	0	0
      

      Some select gövts:

      		Jam1	Jam2	Jam3	Jam4
      Polaris		20	20	10	2
      Vell-os		50	50	35	20
      Pirate		7	5	0	0
      Rebellion	9	6	1	0
      Nil'kemorya	25	25	18	9
      Krypt		100	100	90	75
      ATMOS		100	100	100	100
      Ambrosia	100	100	100	100
      

      It's important to note that the only playable ships with Ni'kemorya affiliation are the basic Raven and the basic Scarab. Not the Nil'k ones, not the ones with cloaks or fast-jumping, just the run-of-the-mill purchasable versions. All other Polaris ships have Polaris affiliation, except for the basic Zephyr, which is, oddly enough, Rebellion-affiliated.

      The Kestrel (and the Mod 'bridge, too, among several other ships) has no inherent government, and thus gets no jamming bonus.

      Note that to even have a chance at jamming Nanites via type-3 jamming you need to have S3 > 90, which can only be attained by doing the Vell-os string until you have the Distract Sensors ability, and then capturing a Nil'kimas ships that comes with a Nil'kemorya Jammer standard. You then have 10 + 45 + 40 = 95 type-3 jamming, so U3 = 5, giving L3 = 5/10. If you go that route then S4 = 2 + 35 + 20 = 57, so U4 = 43 and L4 = 43/50. 5/10 * 43/50 = 43/100, giving you a 57% chance to jam Nanites. The highest possible.

      A Quick Recap: Where Vk is the type-k vulnerability of a missile, Sk the type-k jamming strength of the target, not greater than 100 or less than 0, and Uk = 100 - Sk, let Lk = Uk / Vk, but not be greater than 1. L = L1 * L2 * L3 * L4 is the odds the missile with lock on target. Note that if Uk >= Vk then Lk = 1. Multiply L by C if the missile gets confused by interference, where C, the odds a missile will be confused, is a function of I, the system interference, the formula for which I don't know.

      If you find any errors in this document please let me know. I hope you have found this helpful and informative. Any and all suggestions on how to improve it are welcome. Good luck, and have fun,

      ~Qaanol

      This post has been edited by Qaanol : 21 January 2008 - 08:42 PM

    • Wow, that's some pretty nice work you've done there. Though I'm still not convinced that Nova's jamming system is the best. See the examples I gave in this topic.

      Qaanol, on Dec 5 2005, 09:43 AM, said:

      All missiles fired will travel for a certain distance (time?) locked onto the target, and then they will suddenly switch to their jammed/confused behavior all at once. A missile only checks to see if it's jammed or confused once. After that its behavior is set, though it could still become decoyed or lose its lock if not facing its target.
      View Post

      I could swear I've seen missiles do a loop or two then chase the target for a sec then do some more loops etc.

    • My knowledge of missiles and jammers was previously as reliable as said jammers.

      We can tell an important consequence: from the viewpoint of the jammer, then missiles not vulnerable enough will NEVER be jammed, and then vulnerable missiles will have the probablity of being jammed increase non-linearly from 0% at the limit, up to the jammer's value for a completely vulnerable missile. From the viewpoint of the missiles, however,if the jammer isn't powerful enough then it will always hit, and if the jammer is powerful enough then the probability of hitting will decrease linearly with the strength of the jammer from 100% at the limit, down to 0% for a perfect jammer.

      EDIT: Guy, it makes sense, especially the limit that, if a jammer is low enough, then no missile can be jammed by it. But indeed, very low (but non-zero) vulnerabilities and nearly 100% jamming are best avoided, I don't think they make much sense anyway.

      And, yes, this is my 3000th post. Go me.

      This post has been edited by Zacha Pedro : 05 December 2005 - 06:22 PM

    • A straightforward explanation from Matt Burch in this topic:
      Well, basically it works like this, for each of the four jamming types:
      Each ship's ECM value (i.e. jamming ability) is its inherent attribute govt (zero for ships with no inherent attribute govt) jamming ability plus all the ECM modifiers granted by the various outfit items on the ship. This value is constrained to the range 0-100.
      Each newly-created guided shot's ECM vulnerability is a random number between 0 and the jamming vulnerability of that weapon type, with a maximum of 100.
      If the target ship's ECM is greater than 100 minus the guided shot's jamming vulnerability, the shot is jammed and will fly off course, go dead, or whatever it's set up to do. So, jamming will happen more often with weapons with high jamming vulnerability values, or ships with high ECM values. All four jamming types are handled equally and independently; there is no "bonus" for any particular jamming type, and different jamming/vulnerability types do not interact.

    • Some corrections: Guy is right. Missiles can be confused and then unconfused. Furthermore, it's not a straight I% of missiles get confused type deal, because I'm seeing about 5% of missiles lock on even with 100 interference in the system. Confusion is apparently recalculated, perhaps every frame or second.

      Next, negative jamming does add into your jamming total. Having +100 and -100 jamming leaves you with net 0 jamming, and all missiles lock on (unless differently jammed, or confused/distracted/whatever.)

    • Negative jamming, eh?

      /me has idea for 'cheat plug' jammer outfit that jams all 4 types -100.

    • Qaanol, on Dec 6 2005, 03:56 PM, said:

      Some corrections: Guy is right. Missiles can be confused and then unconfused. Furthermore, it's not a straight I% of missiles get confused type deal, because I'm seeing about 5% of missiles lock on even with 100 interference in the system. Confusion is apparently recalculated, perhaps every frame or second.

      Next, negative jamming does add into your jamming total. Having +100 and -100 jamming leaves you with net 0 jamming, and all missiles lock on (unless differently jammed, or confused/distracted/whatever.)
      View Post

      So, (for each type of jamming separately of course) the jamming for all outfits and the govt's inherent jamming are added, and then only is the result constrained to the 0-100 range...

    • It appears that, even with 100 jamming, every once in a while a missile will still lock on. Somewhere in the range of 1:100, possibly. And only if the missiles are flying straight at their target at the time they "check" whether to be jammed or not.

    • Qaanol, I am pleased to officially invite you to post this info as a new topic in Bomb's Plug-in Guide; this is complex enough to justify a new guide and not just an addition to an annotated template, especially since the subject covers two (even three) resources. Of course by doing so you accept some responsibilities, such as maintaining it, and (trying to make) making it consistant with the layout of my other plug-in guides. But I think it's best if you do it yourself, so that it appears yours (instead of me adding it there).

    • Thanks, Zacha, I'll try to put something together that lives up to your standards of excellence, but I probably won't be able to finalize and post it until January 2006.

      I have a rough draft made, delineated in a manner that should be readable and useful to plug-in makers, featuring large bold paragraph headers and section markers, but I want to streamline the prose, eliminating all extraneous descriptions, so that only the bare-bones facts remain, and they are explained concisely and completely.

      Actually, I think I'll leave a little entertaining filler as a hook at the start, but it'll be clearly demarcated so anyone who just wants the jamming behavior information can skip over it with no harm done.