Like many of you, after Neural Parasite was nerfed from permanent to 15 measly seconds (not 12 like widely believed)I have rarely bothered to research it. I was having trouble with mass Thor+(misc other terran units), so I thought I would investigate it a little more to find out if Neural Parasite still was a good Thor counter. This is not an article on how ZvT sucks for Zerg and Neural Parasite is the answer by any means.
I am a 450 Diamond player (edit: one month later...now 900, thank you ladder inflation), not a pro, so take my analysis from that viewpoint. The below analysis assumes you are able to protect your infestors somehow for 15 secs, perhaps simply by keeping them behind your wall of roaches and casting Neural Parasite at its max range of 9 on the forward most Thor. I shall separate my findings in two sections, "Theory" and "Evidence".
THEORY
Thor cooldown: 1.28secs damage: 30 damage x 2
Neural Parasite Duration: 15 secs (EDIT: previously had this at 12 game seconds, but it's 15)
This means if you NP a thor it'll usually have time for 11 shots (occassionally 12 shots) before the effect wears off.
11x30(damage)x2(Thor's dbl atk) = 660 damage
Neural Parasite on a Thor deals 660 damage to the terran that would have been dealt to the zerg, for a damage dealt swing of 1320 health. But that is not actually all it does, as a butterfly effect is triggered in the respect that more zerg units are alive because of the decreased damage by the Terran force, so the zerg's firepower is also increased by units that would otherwise be dead. A very rough estimate is that by the Terran losing 660 points of firepower (enough damage to kill 4.5 roaches, factoring in their 1 armor) output over a 15 second period, the zerg will gain about 2 roach worth of additional fire power during that period. Forgive me if that is not perfectly accurate, it could be more like 1 or 1.5, it just depends when the roaches would have died to determine how much firepower is lost.
Bonus "Butterfly Effect" Damage from Zerg units staying alive longer: 2 roach @6shot each = 12 shots x 16 dmg = 192 damage just from units that are alive because the Thor is not shooting at the zerg for 15 seconds.
So the "should be dead" roaches did 192 damage for the Zerg, great, lets review:
Effect of Neural Parasiting a single Thor *660 damage to Terran over 15 secs *660 less damage to Zerg over 15 secs *192 more damage from roaches that aren't killed by NP'ed Thor Total Health Swing: 1512
Well that's pretty awesome, but in truth there is even more to consider.
Terran inflicts X less damage over that 15 secs because of the units lost from suffering 660 (NP'ed Thor) + 192 (the "ghost" roaches) damage. Of course the units are not lost immediately, but lost or damaged at some point over the 15 second span.
It is reasonable to assume that 852 damage over 15 secs might be enough to kill two unsieged tanks (160hp units) by the 6 second mark. That would be 6 seconds of unsieged tank fire "lost" from having the Thor and a few roaches firing on them, that would otherwise not be firing on them, had Neural Parasite not been cast. Since unsieged tanks cooldown is 1.04, they would lose a total of about 14 shots, or against roaches 288 damage (2x7x25 =350 - 14 = 336).
Revised Effect of Neural Parasiting a single Thor *660 damage to Terran over 15 secs *660 less damage to Zerg over 15 secs *192 more damage from Zerg units that aren't killed by NP'ed Thor *336 less damage from Terran units killed by Thor and "should be dead" zerg units. Revised Total Health Swing: 1848
Believe it or not there are additional levels of butterfly effects that could be calculated, but they become very hard to even approximate, and their effects get smaller and smaller. For example, the 336 damage of firepower the Terran lost from the two siege tanks that died at the 6 second mark means even more zerg units are alive to inflict MORE damage for those final 6 seconds.
Conclusion: you can get something like a 1500 - 2000 point health swing for Neural Parasiting a SINGLE Thor.
Possible Conclusion: Neural Parasiting two Thors would lead to greater than a 100% increase over the above analysis, considering the exponential nature of even more Zerg units alive to fire upon the terran units.
Caveat: Some have observed that they've NP'ed 4 Thors at once, killed the terran army, then the NP wears off and the 4 Thors kill the rest of the zerg army. This makes sense intuitively, I'd like to analyze it in a future post as it is an interesting scenario where you might be best served to only NP 50% of the Thors at a time.
EVIDENCE
I ran some tests the other day and was sort of surprised, check this out... (both forces stupidly just attack move each other in the open, with no sieging, and yes I know the roaches will do much better if microed, but I wanted identical situations to measure the effect of ONLY Neural Parasite)
Zerg force 1: 30 roaches (2250m/750g/60sup) Result: 5 Thors and 1 tank survive
Zerg force 2 (yes, I know you'd have fewer roaches if you get Infestors): same as above, add 2 infestors with Neural Parasite, cast NP on front two Thors. Result: 14 roaches, 2 infestors survive Retry Result: 12 roaches, 2 infestors survive
APPENDIX, Damage Output for some other Terran units over 15 seconds
Neural Parasite on an unsieged tank firing on armored units... 14 shots@25 each = 350 damage
On a banshee? 12 shots@12x2 each = 288 damage
Stimmed Marauder? 12 shots@20 each = 300 damage (you can even over-stim it to weaken it) Yep, theoretically not useless to Neural Parasite a Marauder, if it has stim and concussive shells. A 480 health swing can be enough to keep 2 roaches alive and probably kill 2 siege tanks (with the help of the 2 roaches that don't die)
Battle cruiser? (ground)65 shots@10 each = 650 damage (air)65 shots@6 each = 390 damage
Thor still will do a lot more damage to most ground units than a battle cruiser because armor will affect the BC's damage output more. (-65dmg versus armor 1, - 130 dmg vs armor 2, etc..).
REPLAY (EDIT: Added this replay)
This isn't the perfect replay, but certainly I've been beaten by Terran of this skill (and worse) before. There are two battles were NP was used, the most important one to look at is the one at the 12 minute mark on my natural. There we see mis-micro from both sides (common among non-pros such low to mid diamond players), as it appears that the Terran once seeing the roach and infestors approaching, runs his forces back to his tanks in order to dodge, what he thinks will be Fungal Growth. During that move I Neural Parasite the Thor, then before he can direct his MM to snipe the infestor I fungal growth them and surround them with roaches. This allows the Neural Parasite to last the entire 15 secs, and because of lucky timing it actually gets off 10 shots (30x2x10 = 600 damage) instead of the usual 9. He only had 1 unit capable of sniping my infestor, his forward siege tank, which might not have had the sight range he needed once his MM retreated. I made a few mistakes myself, I lose one high energy infestor stupidly by running it up, casting nothing, then promptly dieing. Also, I had emerging hatchery units rallied to my ramp that did not participate in the battle as soon as they could have.
The terran's force was not large, but my force was pathetically small as well, and I lose one infestor for nothing.
(Notice the terran misclicks somehow when my overseer approaches and he kills his own Thor, so much for this replay's credibility, oh well, haha!)
In the later battle he has a solid defense of sieged tanks and a couple of Thor, and I'm still all roaches/infestor basically. I Neural Parasite the closest Thor, and it looks like the Terran tries to snipe the infestor but is surrounded by roaches so cannot get in range. Obviously this battle I had a huge army size edge thanks to the earlier win and my harassment on his natural with infestor/muta.
To be honest, I pretty much suck and I'd like to see some examples of pros successfully using NP against Thors. If they don't exist maybe its a tactic that only works for mid Diamond and below, but you can see it appears that it is possible by using a combination of roach wall plus Fungal Growth to protect your infestors.
Interesting post, I will consider using Infestors more now. But how does NP fare against toss? and Infestors are easily sniped as well, wouldn't 12 infested Terrans prove to be more useful? Especially against sieged tanks that can just snipe infestors?
just curious.. lots of people do complain about the NP not being worth it..
nice #'s Johnny.. the only thing that kind of screws up your research is that infesters often get sniped or FF and the NP can get cut short, the inevitable damage swing is always going to be present though, might not be as large as what you post (your #'s are obviously best case scenario)
does EVERYONE know that if you activate NP and immediately hit "R" for burrow you can NP a unit and have your infester burrowed and protected. (might be a trick not everyone is familiar with)
also... for you zergs out there.. I would really like to hear about someone using shift+click'ed infested terrans to create walls in the same fashion as protoss forcefields to block army mobility. A fully charged infestor can drop some 8-9 infested terrans simultaneously from the protection of burrow
On August 18 2010 00:45 hadoken5 wrote: Interesting post, I will consider using Infestors more now. But how does NP fare against toss? and Infestors are easily sniped as well, wouldn't 12 infested Terrans prove to be more useful? Especially against sieged tanks that can just snipe infestors?
I pre-emptively answered this in my 2nd paragraph with:
The below analysis assumes you are able to protect your infestors somehow for 12 secs, perhaps simply by keeping them behind your wall of roaches and casting Neural Parasite at its max range of 9 on the forward most Thor.
Sure infested terran are nice, I invite you to analyze the effects of 100 energy worth of infested terrans (4) versus a Thor army and we shall compare. My hunch is that Neural Parasite will be much better, so long as you keep the infestor from being sniped. I recommend swarming the Terran with many roaches, and/or using a second infestor for fungal growthing any units that might be able to snipe the Neural Parasiting infestor.
Wow there's a lot of number-crunching theorycraft going on there...probably a bit too much.
I'm nearly 100% certain if you go into a fight with your infestors and each one is able to NP a unit for the full 12 seconds, you'll win that battle. But this never happens, the infestor always dies and the NP is cut short.
The issue that people have isn't with the length of time NP lasts, it's with the fact that the infestor is so vulnerable and never gets to use NP to its fullest extent. Yes you say in your OP that you assume that each and every NP lasts for the full 12 seconds, but I have never once seen this turn the tide of a battle that you would have otherwise lost. If you are able to get off a full 12-second NP, your army was far bigger than your opponents and you would have won without the infestors anyway.
Here's a good example of when NP goes horribly wrong. Note that if every NP that Ciara used during the last big battle had lasted for 12 seconds, the battle probably would have been won:
On August 18 2010 00:48 iNdo-Man wrote: does EVERYONE know that if you activate NP and immediately hit "R" for burrow you can NP a unit and have your infester burrowed and protected. (might be a trick not everyone is familiar with)
I find your math sound, if you make a bunch of assumptions, IE
The thor constantly has something to shoot at for 12 seconds straight (they dont pull back) The infestor dosent get killed halfway through All of your roaches are all constantly firing for the duration of NP Absoloutley none of the damage dealt is overkill damage (eg hitting a zergling (35hp) with a thor shot(60 damage) Nobody has any upgrades
Plus, if you are going to start factoring in the damage gained by roaches not dying, and the damage gained by roaches not dying so they kill more so more roaches dont die, you may aswell make one of those geometric series you should have learnt in high school, (a/1-r) http://en.wikipedia.org/wiki/Geometric_series
Although, for all intents and purposes you could give a number anywhere between 500-2500 and it would be reasonably accurate and convincing enough to make people use NP more
I kinda still wish it was infinite, back then i had terrans calling zerg imba, those were the days.
Actually, my issue with NP is directly tied to the 12 second duration.
The prospect of winning a battle, and now having a Thor forever, makes the 200/200 research of NP worth it.
The prospect of casting NP on a Thor, barely winning the battle, then the Thor breaking free and slaughtering whats left of my army makes 200/200 research laughable.
NP had enough issues with infinite duration, capping it at 12 seconds put a nail in the coffin real quick.
On August 18 2010 00:53 BlasiuS wrote: Wow there's a lot of number-crunching theorycraft going on there...probably a bit too much.
I'm nearly 100% certain if you go into a fight with your infestors and each one is able to NP a unit for the full 12 seconds, you'll win that battle. But this never happens, the infestor always dies and the NP is cut short.
Personally, ive had a game where I NP'd about 3 thors, sent in my whole army, kill everything else, lose pretty much my whole army, then the 3 thors lose the NP, kill my infestors then subsequently kill my natural.
You may also want to factor in the opportunity cost of NP. By choosing to NP, you fail to FG the MM ball alongside the Thor, which when stimmed, probably does more than 540 dmg in the stun time. Also apply the butterfly effect used in your logic, and you realize that it really becomes situational and there is no clear cut choice. Often times, I find FG the bio army produces better results (also because infestor dies when NPing).
On August 18 2010 00:53 BlasiuS wrote: Wow there's a lot of number-crunching theorycraft going on there...probably a bit too much.
I'm nearly 100% certain if you go into a fight with your infestors and each one is able to NP a unit for the full 12 seconds, you'll win that battle. But this never happens, the infestor always dies and the NP is cut short.
Personally, ive had a game where I NP'd about 3 thors, sent in my whole army, kill everything else, lose pretty much my whole army, then the 3 thors lose the NP, kill my infestors then subsequently kill my natural.
that probably had nothing to do with the NP length, and everything to do with how well you handled the battle. A 3-thor swing for 12 seconds should be more than enough to decisively win, let alone not lose your entire army and not lose your nat.
gonna call bullshit on that one, unless of course you can provide a replay.
You've analyzed the best case scenario for a neural parasite, but you haven't compared it to anything. Sure np might pull a 1080 health swing when cast idealy, but this swing is only present when compared to the same army in which the infestor sits back and does absolutely nothing.
Let's say you didn't go infestors with neural parasite, but instead put the infestor's resources into roaches instead. You'd have three or so extra roaches even before you factor in the np research costs. Those extra roaches can now tank an additional 450 damage, and can deal a couple hundred damage themselves.
So your damage swing from np over a very basic army isn't 1080, it's more like 400. Also keep in mind that the np requires additional micro to use, and that there are research costs involved, and that when going for infestors your army will experience a window of increased vulnerability while the infestors build up energy. Also your infestors might be killed before the np can last its full duration.
I'm not trying to imply that it's a useless spell (a potential 400 damage swing is nothing to scoff at), but it's not nearly as good as you make it sound. I think infestors and np could easily return to their former state (unlimited np duration) without causing problems.
On August 18 2010 01:05 Caelestis wrote: You may also want to factor in the opportunity cost of NP. By choosing to NP, you fail to FG the MM ball alongside the Thor, which when stimmed, probably does more than 540 dmg in the stun time. Also apply the butterfly effect used in your logic, and you realize that it really becomes situational and there is no clear cut choice. Often times, I find FG the bio army produces better results (also because infestor dies when NPing).
I added my replay after you posted this, but as you can see, when you rush to infestors relatively quickly you can usually employ both Fungal Growth and Neural Parasite. I think the key is to get a couple of infestors out quick so they can start building up energy. One nice thing about fast Infestors is that the Infestation Pit builds almost twice as quickly as the spire, so you will almost always be able to have fungal growth ready, but perhaps not Neural Parasite. The nice thing about the timing of Neural Parasite research is that it takes a while to get Thors, and longer for them to talk to your base, so you can usually have NP ready too.
Of course much like when rushing to muta, if you rush to Infestors without a decent ground army, there will be a timing window there where the Terran can abuse you. I don't even think I scouted well enough to know I'd be facing Thors, its just that I had decided on fast infestors + Neural Parasite before the start of the game to investigate their potential. About 75% of my games versus Terran, they make Thors, so I figured it was an ok gamble anyway.
On August 18 2010 01:26 archwaykitten wrote: You've analyzed the best case scenario for a neural parasite, but you haven't compared it to anything. Sure np might pull a 1080 health swing when cast idealy, but this swing is only present when compared to the same army in which the infestor sits back and does absolutely nothing.
Agreed, that is beyond the scope of my post. Again, I'm not saying Neural Parasite is the best thing since reavers in an undefended drone line, I'm only posting some analysis regarding a specific situation in which NP could potentially be used, assuming you can protect the infestor for 12 secs.
Often times its valuable to study the best case scenario, as in this circumstance, players can see that if the infestor can be protected then NP is probably worth the cost. Once that is understood players can then work on tactics to protect their Infestors.
I could brainstorm on ways to protect infestors, but that was not the subject of my post.
On August 18 2010 00:59 schiznak wrote: I find your math sound, if you make a bunch of assumptions, IE
The thor constantly has something to shoot at for 12 seconds straight (they dont pull back) The infestor dosent get killed halfway through All of your roaches are all constantly firing for the duration of NP Absoloutley none of the damage dealt is overkill damage (eg hitting a zergling (35hp) with a thor shot(60 damage) Nobody has any upgrades
The assumptions were necessary to limit the complexity of my post. Obviously there are 500 different possibilities given variations on just what you've listed. My post shows one fairly common scenario. Other scenarios could be better or worse for the Zerg. Inferences can be made using knowledge gained from deeply analyzing a single scenario, but nobody has time to analyze them all.
The question I was trying to answer was basically: "Does Neural Parasite suck now that it only lasts 12 secs?" Hopefully my post and my replay show that there are situations where it can be very cost effective. Also I think my post illustrates that Thors are usually the best possible target for NP and the numbers show why this is the case.
One big question mark is if you are able to keep your infestors alive after you win the battle. If you are able to (no easy feat), then you might find yourself in a superior position if the Terran has lost his entire army.
Employing Neural Parasite is an unpredictable tactic which can lead to a big win or a big loss for either side. A lot of it is just going to come down to player skill. Again, maybe its effective use depends on the Terran being below a certain skill threshold, I am not sure, all my analysis does is demonstrate its potential.
Plus, if you are going to start factoring in the damage gained by roaches not dying, and the damage gained by roaches not dying so they kill more so more roaches dont die, you may aswell make one of those geometric series you should have learnt in high school, (a/1-r) http://en.wikipedia.org/wiki/Geometric_series
No thanks, I'll leave that as an exercise for the reader. =p
EPT
