Can any hero compete with assassin?

[Mughi]

"TheGreatSanguini" původně napsal:

Unfortunately, at least 80% of that theorizing time has been "how can any character compete with Assassin?"

This has been on my mind for the past 15 years or so. Assassination as a skill is clearly overtuned, not only being the absolute best melee combat skill for damage but also giving 85gold per trigger. I don't think it's possible for any other melee build to contend with them, but I've also tried a variety of other methods like Tinkers using full engineer&towers that can't be assassinated (but just makes them siege fodder), or merchants that attempt to swarm the assassin (which tends to just give the enemy free gold).


I've built a variety of assassins attempting to optimize them at each level cap and for each purpose. My favourites have been:

1) Orc Assassin. Being naturally immune to disease and poison puts them in a place where they can't be easily taken out or slowed down by common status, access to Chaos Magic lets them Drain Mana of any opposing wizard heroes removing their entire threat, and demolition exists for high level builds when you run out of other things to invest in, letting them destroy towers in a few hits.

2) Dark Elf Healer. Technically not an assassin, but gets access to assassination in time for it to be very usable as one by level 40. This build with Elcor's Aura as a skill gets to refill its healthbar over and over for nearly free, and synergizes very well with high tier units that also get topped up for free with 300+50 healing on every unit in the bubble of Heal Group. Cure renders the hero almost as immune to disease/poison as the orc, and invigorate is a nice buff if you find that you have full mana that would otherwise just be wasted. While LifeWard and Ressurection are potentially good spells, I rarely go that high in the healing tree with this one because of mana efficiency.

Due to having assassination, early low tier units are no threat and actively give you resources. Due to heal group, late high tier unit fights are patently unfair since their units die and your don't.

3) High Elf Assassin. When heroes are limited to level 30 max, the high elf assassin still gets to invest plenty into the assassination skill before healing magic becomes available at level 20, needing only 5 points to get the good stuff or 6 if you want invigorate as well. No Elcor's Aura means the healing isn't absolutely broken, but still adequate for self-regeneration and curing the inevitable debuffs.

4) Minotaur Assassin. When heroes are limited to level 15 max, the minotaur assassin's free sheep eating cures status debuffs and tops up the healthbar without requiring skill point investment. Unfortunately, it locks you into playing Minotaurs as an army.
Level 15 is an interesting balance point, because the heroes can still potentially be pursued by some units and threatened by critical hits.

So,
Without the healing available, the assassin can potentially be shut down with status debuffs or eventually running their HP low enough that they have to run to a tower. With those available, I'm not sure what else any other class can bring to the table that assassin can't. What are your favourite builds, and how would you contrive to contest the assassin?

Can you create a build that offers something uniquely powerful, of great enough utility, or bizarrely game breaking to beat them? What is your favourite level cap, and your best hero?

[Fendelphi]

Something a bit different.
Anyone tried a High Elf Priest with focus on Arcane Rune(at least 20 points for +100% health to Elven mages), playing as Dark Elves?
The goal is to get level 5 Sorcerers out in the field(if we go for super high level heroes, then 40 Intelligence and 14 Divination will do the trick), so that they spawn with something like 250 hits and can cast the upgraded spells.
Just a small group of these running together is very hard to kill, simply due to amount of health, the damage they deal and the insane damage their spells can unleash. Even heroes should "beware..." of such a group.
Combine that with Major Healing and Resurrection spells and, well, I think you get the point. ;)

[Mughi]

Fendelphi původně napsal:

Unfortunately, I have seen a lot of people(on different fora for different games) present videos or screenshots as evidence, and then all it shows is a modded or bugged or otherwise inconclusive result. So maybe I am just a bit jaded in that regard. Sorry if I came off as aggressive, was not my intention.

So, there was no misunderstanding or misinterpretation. When provided with a video showing how to do the test yourself, and being encouraged to do the test yourself, you decided it was most reasonable to accuse me of lying instead of doing the test yourself. This is not aggressive, it's indefensible. Trying to capture footage, edit video, and make a youtube account all for the purpose of showing you "don't take my word for it" took me about 3 hours of trial and error, which is 4 hours more thought than you put into your post.

Even more frustrating, you've for a 3rd time dodged all the previous questions posed while I was being very polite and patient with you. You've not asked any questions of your own. Instead of openly answering or honestly communicating, you're still trying to "win" an "argument", regardless of honesty. You're promoting a "level 50 Minotaur Warrior(30 Mighty Blow, 5 Weapon Master, rest in Constitution and Ferocity, 35 Dex, rest in Strength)." in one breath, and saying "assassins are too 1-dimensional" in the next. You're saying to use Gnolls and DarkElfAssassins as a counter despite the conventional units that will inevitably already be on the field directly countering them. You're explaining how towers work, despite being able to see how the Demolition skill works for you, even in singleplayer if you have no friends. For the final time, "discussion with someone who appreciates the same niche game as myself is nice." With you still ignoring my questions, you clearly have no intention of entertaining a discussion.

You do have a grasp of theory in this game, and you've put a lot of hours into singleplayer. After your behaviour those are the kindest words I have for you. I'm done with you now. "Sorry-not-sorry".

Joe the Bartender původně napsal:

Great video! This proves that high combat and crit chance doesn't work. I would ask if a mighty blow warrior build works too but I suspect the result would be similar. If you get around to it, I'd be interested in seeing what happens with the other class builds.

The next test I'd want to do for my own knowledge is your idea of Arcane Magic Destruction + Constitution, but since you're requesting another video I'll try to make time for it. To put it mildly though I'm feeling burnt out now, as doing the tests is easy but trying to capture footage I find more difficult. I'll probably end up uploading Firebreath and Invigorate attack speed videos to my new youtube sooner because I can capture that footage in singleplayer with buildings as the targets, throwing out less footage for desync, and not wasting my friends' time.
All that I actually wanted to achieve with the first video was showing the method by which anyone can do these tests themselves, but since you're requesting these I'll put in the effort and try to have them in the next few weeks.

I'll also be spending much more time in TPC, and probably stop by your forums for questions about hero skill reworks, perks/personalities, why CRTL-hotkeying buildings results in the hotkey switching to units it produced, etc. We're having a blast with it so far. The controls are familiar, but everything I know about build orders and economy is turned upside down in it so it'll take some getting used to.

[Mughi]

Joe the Bartender původně napsal:

I would ask if a mighty blow warrior build works too but I suspect the result would be similar. If you get around to it, I'd be interested in seeing what happens with the other class builds.

After reading this thread, my buddy actually WANTED to do another test video (sick masochist).
Results were statistically pretty different, instead of being the 90% assassin favoured like most builds, it was 70%. Not spending anything in Int and Cha went against our general sensibilities, but made a pretty big difference. Still not a gamble I'd want to take though, and since it's economically inferior you're kinda pushed into making that gamble.
https://www.youtube.com/watch?v=xbCX6Wmbdd4

More to come later, and probably some TPC content once we get used to it.

[Joe the Bartender]

Fendelphi původně napsal:

Are Deathblows possible through a Weaponsmaster crit, or will they only occur when the regular attack "crits" into a deathblow?

The chance for a deathblow is calculated before the chance for a critical hit, so deathblows are possible regardless of investment in weaponmaster.

Fendelphi původně napsal:

As for rushes at high levels, remember that towers scale with level.

I forgot 1.03.24 forces this on by default. 1.03 doesn't have levelling towers.

Fendelphi původně napsal:

I havent tested if Entangle and, say, disease or another speed debuff stacks, or if it only takes the most powerful/recent one.

They should stack.

Fendelphi původně napsal:

As for items, that is another reason why the Archmage is a lot of fun. Once you get to rank 8 Alchemy, you can disable all enemy items for a short while. So those heroes who rely on proc effects(which on the steam version does have a cooldown btw.) suddenly have none for half a minute.

It is a solid counter. Will they have enough mana for disjunction + invisibility + acquire + summon quasit + summon imp? And will they be able to regenerate mana fast enough without the summon mana research, at early stages of the game?

Fendelphi původně napsal:

Also, just to clarify. I am not saying(in the slightest) that Assassins are bad. They are not. But I just dont find them as terrifying as some of the other class combos out there, because they are so 1-dimensional. Somewhat tedious to deal with, sure. And if you dont keep a watch, they will have their way with you. But as long as they dont get the jump on me, I feel like I have always dealt with them fine.

The Great Sanguini původně napsal:

You're promoting a "level 50 Minotaur Warrior(30 Mighty Blow, 5 Weapon Master, rest in Constitution and Ferocity, 35 Dex, rest in Strength)." in one breath, and saying "assassins are too 1-dimensional" in the next.

I'm not sure what the 1-dimensional thing is about since melee classes are all similar. Assassins and similar classes might not be toolbox leaders, but they do what they do better than nearly anything else in the game. In WBC3, the loss of a hero equates to the loss of the game and almost nothing is better at taking out heroes than an assassin.

Fendelphi původně napsal:

And speaking of Assassins, Divination(Archmage access...) has a spell that gives nearby units the ability to steal xp from whatever they kill. Combine that with the Assassins from the Dark Elf army, and their 2 research skills, and you can quickly have high level Assassins running around, getting stronger and stronger(With a higher and higher Assassination chance). So that is another way of dealing with high level heroes. Spam them with high level assassins until you get lucky.

Fendelphi původně napsal:

Anyone tried a High Elf Priest with focus on Arcane Rune(at least 20 points for +100% health to Elven mages), playing as Dark Elves?

The Great Sanguini původně napsal:

]You're saying to use Gnolls and DarkElfAssassins as a counter despite the conventional units that will inevitably already be on the field directly countering them.

This. Early-game classes tend to be the most relevant in the game. If an opponent's hero defeats whatever is fielded against them during the early game, chances are that there won't be a mid or late game. In the case of classes like the assassin or warrior, they immediately start the game at maximum power, whereas other classes don't. This makes the early-game highly important and the mid-late less important or entirely unfeasible.

There are some class combos which are more powerful than an assassin in the early game, but not many at all. Investing 35 points into dexterity as a wizard or support hero allows them to get away from assassins or warriors but it also weakens their abilities. It is possible to use spells like entangle to escape from non-running assassins, though my thoughts are still that the large amount of dex investment weakens the wizard (and support heroes too for that matter). Running away doesn't secure any ground either, so anything will likely be recaptured by early-game classes. Against these classes, it's as if they have the whole map to themselves except for your starting mines. Every time a mid-late game hero ventures outside their towers, they risk losing themselves, but early-game classes don't have the same risk.

The Great Sanguini původně napsal:

I'll also be spending much more time in TPC, and probably stop by your forums for questions about hero skill reworks, perks/personalities, why CRTL-hotkeying buildings results in the hotkey switching to units it produced, etc. We're having a blast with it so far. The controls are familiar, but everything I know about build orders and economy is turned upside down in it so it'll take some getting used to.

Sweet! Good to know it's going well for you.

The Great Sanguini původně napsal:

To put it mildly though I'm feeling burnt out now, as doing the tests is easy but trying to capture footage I find more difficult. I'll probably end up uploading Firebreath and Invigorate attack speed videos to my new youtube sooner because I can capture that footage in singleplayer with buildings as the targets, throwing out less footage for desync, and not wasting my friends' time.
All that I actually wanted to achieve with the first video was showing the method by which anyone can do these tests themselves, but since you're requesting these I'll put in the effort and try to have them in the next few weeks.

Don't worry about it, there's no pressure. I don't mind if a video is made or not, but I'll be interested to see anything you do put out. I haven't got the ability to test for myself at the moment since my DVD drive is broken and I currently don't have any digital copies of WBC3.

The Great Sanguini původně napsal:

After reading this thread, my buddy actually WANTED to do another test video (sick masochist).
Results were statistically pretty different, instead of being the 90% assassin favoured like most builds, it was 70%. Not spending anything in Int and Cha went against our general sensibilities, but made a pretty big difference. Still not a gamble I'd want to take though, and since it's economically inferior you're kinda pushed into making that gamble.

Great vid! This time it was a lot closer. So warriors can potentially go up against assassins, and basically trade the assassination econ with greater anti-building power. I agree, it is true that if the assassin keeps generating gold, then the warrior will end up losing if they don't attack the assassin as early as possible.

[Mughi]

Joe the Bartender původně napsal:

Great vid! This time it was a lot closer. So warriors can potentially go up against assassins, and basically trade the assassination econ with greater anti-building power. I agree, it is true that if the assassin keeps generating gold, then the warrior will end up losing if they don't attack the assassin as early as possible.

Yep. I considered building the fighter as a Chieftain for better racial synergy with the Minotaur, but then the build still needs investment in Ferocity, and misses out entirely on Weaponmaster, leaving it less effective for its one purpose of duelling. Unfortunately with it still losing 70%+ to assassin in the duel, having level1 troops instead of level3 due to Intelligence, having worse Merchant score, and no access to racial spells, it's not very useful except anti-building. Reinvesting the points from Str back into Int and Cha to alleviate these problems results in the build having an even worse win rate in the duel, which is its only purpose.

I had hoped that in this thread, I would learn the counter to my friend's assassin so we wouldn't need our gentleman's agreement to avoid the class. That some new piece of information that had eluded me for 15 years would change my worldview...

...And holy crap, did you ever DELIVER with that Destruction+Constitution gamebreaker.
I ended up trying a few arcane builds to get the feel for it, but settled on Dwarf Ice Mage with a stat spread of 6Str 10Dex 33Int and 20Cha. For skills, I picked +4 Ice Magic, +39Constitution, and +10Arcane Magic. This results in a character with 824HP and 119MP, allowing for the specific combination of Calm(level4 Ice) immediately followed by Destruction(level10 Arcane), which instakills any build except those that invest even more into Constitution, like the Dwarf Paladin from earlier (which might actually make a hard rock>paper>scissors). Even builds with LifeWard must be "frame perfect" with its reapplication, which requries a 100% accurate prediction and a bit of luck with latency. It is truly the ultimate build, only needing the first 30 seconds of the match to regen mana and chug all the potions. Having 20Cha gives it an equal or greater effective range than any of the builds that would contest it, making it safe against getting Drain Mana'd unless you get caught with your pants down.

The only downside to the build is its foot speed, as it beats everything in a duel, AND has greater economy stats than anything that could contest it (since we figure merchants aren't a real thing).

So, thanks for opening our eyes to a whole new world. It's purely on the skill of the player to choose the correct time and place to blow up the world, as choosing right wins the game and choosing wrong loses it. This unfortunately doesn't lead us to believe that WBC3 is in any way a more balanced game than we thought, so if that's the problem we want to solve, TPC seems to be the way to go. However, learning about this absolutely hilarious jank strat that neither of us had considered is a real joy.

Is there any other specific build for Arcane Destruction that would be more effective or add to the variety of potential builds for WBC3? Did I make the build wrong, or overlook some other niche WBC3 combo? Any more info is appreciated. We will probably be shifting our focus entirely into TPC as our rts of choice, and start our own discovery process all over again.

[Joe the Bartender]

The Great Sanguini původně napsal:

Is there any other specific build for Arcane Destruction that would be more effective or add to the variety of potential builds for WBC3?

Dark Dwarves also have access to constitution so they can also do the combo. Given that they have access to chaos magic instead of rune magic, they could have the ability to drain mana, which is pretty nasty. As for classes, pick whichever class/race combo suits the best I guess. The wizard classes with access to arcane magic are: alchemist, defiler, druid, healer, ice mage, illusionist, necromancer, pyromancer and shaman.

Alchemists can go around instantly-capturing things and can be particularly useful for the undead due to transmute & instant unit morphing, though they work fine with any race. Defilers probably aren't worth bothering with due to mediocre poison magic (except call of kargoth) and plaguelord skills (unless you want to play plaguelords). Druids are in the same bucket, they're okay but not great and don't have much to abuse. Healers are good due to being able to instantly heal after destruction. In 1.03.24, life ward is disenchanted after casting destruction, but you could still cast life ward or another healing spell afterwards. Plus, healing is generally just really good, especially for tankier races like knights or dwarves. Invigorate is the key to the build, and is one of the best spells in the game. Ice mage gets gemcutting and freeze, both of which are useful and ice floe can be abused with items. Illusionist has the powerful awe spell among others, and has a combo with minotaurs via mutate. Necromancer pairs well with undead as always. Pyromancer has some powerful combos with minotaurs (firebreath due to research upgrades applying over the top of firebreath damage) and dwarves (berserk and drunkenness speeds up the normally-slow dwarves nicely). Shamans don't really give anything outside a chaos magic synergy though, unless you're going with a dwarf arcane destruction build. Those are my thoughts anyway. I hope this helps a bit.

[Fendelphi]

I think there is some talking past each other here. I dont get why you need that aggressive tone though.
I never accused anyone of lying. It looked like invincibility was turned on, due to a mix of not being able to see anything(not even numbers) and it looking like it was bugged. I already explained I had no intention of claiming you were lying, just that it was a really poor proof of evidence(which unfortunately is very common on steam fora), due to it's bad quality, which you fixed in the next video.

I am not trying to win an argument. I am trying to showcase a point(in which lies questions, which you seem to disregard most of the time) that there are factors that can be a deterrent to the Asssassin. And now, with the recent video, it seems you agree that running into a 1vs1 with a decently built warrior is at least a gamble(dies 30% of the time, and leaving the Assassin close to death on at least 2 other occasions). Which is all I really claimed. If we then factor in towers or supporting units, or other factors that bring the Assassin to less than full health, it gets even closer.
I posted my initial responses of potential counters depending on the situations, and have been answering questions since then("as for ...." in pretty much every post since my first one).
What questions are you refering to? If there is a question I have missed, please tell me where. There is a lot of text here.

Btw, the new video does have some issues in terms of "being representative", since several times the heroes just stand idle or moves around, while the other attacks(fight 2, 3, 4, 6, 7, 10). And there is some inconsistencies with character skills and stats, which leads to the possibility that there have been outtakes or redos.
Again, I am not saying this is intentional, but in terms of "proving a point" it unfortunately cant be counted as conclusive, as there are too many things that are unexplained(why are the skills different, which spells were cast and for what purpose). As I said, steam fora has made me jaded, so I can no longer just take people's word for it, until such inconsistencies have been explained.
Also, to see how close the fights really are, we actually need to see the Assassin's healthpool at all times, because that will determine how close it is(since we cant trust the health bar, we need to see the numbers).
This is important, since the Assassin might not be at full health at all times(especially if diving a base or a hero that has a few bodyguards), while it rarely matters if the warrior is at full health or not, since all it takes is 1 assassin proc(except when max health is above 1000, then their current health will matter ofc.).

Items is an argument for martial classes, because they add damage and can change damage type to something more potent(like Cold or Crush, which is very important for a high Combat Warrior), which is then multiplied on crits/deathblows(thanks Joe for clarifying the Crit/Deathblow thing, btw).
For the Assassin, it matters less, because the damage comes from Assassination and the attackspeed cap is 0.4 no matter what(although cold and crush will help their survivability, if they manage to land a crit). The few extra % assassination chance you can get from items wont help much(which is why they usually look for proc or combat boost). They can do the Paladin's Helm+Attack speed weapon, but so can the Warrior if they want to, and it is just another set of "chance", so they roughly evens out in terms of surviability(especially at even higher levels, since that can put the warrior well above 1000 hp, requiring 4 assassinations).
Similarly, items will help casters getting more mana early on(regeneration) and potentially provide them with some survivability(hitting a mine a few times early on while wearing the Paladin's Helmet, or buffing your army with the Orb of Etheria to better protect them).
In case of the Alchemist spells, Items are a very important factor.

For casters, their stat and skillpoint allocation(and viability in general) will depend on level limit and version of the game, as I said in the first post. Generally speaking, they get exponentially stronger as the level increases, because they can maximize their effects for the same mana cost, while also having more mana to spend. So a mage set in a level 15 limit will probably not be nearly as effective(or even viable) and have to focus on different skills compared to a level 50+ mage.


Moving on. The reason why I think Assassins are more 1-dimensional than other martial heroes is because their damage relies on the assassination skill. As such, they are much worse off in handling buildings, both due to their lower raw damage and their general smaller amount of Hits/Combat compared to most other Martial classes. I think we all agree that this is the case.
My argument is that a Minotuar Warrior(or other martial hero) is less 1-dimensional because it both has killing power and the ability to tank some hits(being a martial class, they start with more hits and combat) and destroy buildings(maybe not the towers, but it will make short work of anything but those). In the video above, the Assassin might be a good hero- and unit killer, but will do terrible against buildings with less than 30 damage, compared to the Mino Warrior with near 100.

1.3.24 is the "standard" Steam version(the one that is available the second you download the game), and towers scale in health and damage here. Meaning it needs to be factored in when talking on a steam forum, especially if we look at high level characters. So from my experience, Assassins, and even some martial heroes, cant rush super hard early game, because they are not able to beat the towers efficiently without the help from their retinue or early units(unless the defender sleeps). There are a few exceptions(when going up against piercing towers and have skeletons with firebreath, for example), but even a level 50 warrior with 700 health will only survive about 10 seconds inside an enemy base(if they try to chase a hero).
With Fog of War on, it is also fairly difficult for an Assassin to locate a good target(hero) on the map for them to assassinate(especially on random generated maps where they dont know the layout), so it is hard to gain map control until you have scouted a large part of the map(again, Summoning sphere has Eye of Oros that helps a ton here).
Even if they do spot an enemy hero, as long as that hero have 35 Dex(or really, just 30), the Assassin will probably not be able to catch them before they get back to their base.
That is part of my point with being 1-dimensional as well. If you play a 4 player map or more, you might be able to deny some of the area for some of the players, but the Assassin cant be everywhere at once and can only really deal with units and heroes and cant, unlike Martial classes and Mages, efficiently raid a base for a few seconds(destroying key buildings) or deny mines quickly(destory them) when the oppotunity arises. If for what ever reason they cant hold/deny map control(and against 3 players, this is hard) and cant find the oppotunity to assassinate an enemy hero, their options are limited.


The Archmage, as I said, only becomes such a powerhouse very, very late(way past level 50). By then, you have several hundred Mana(due to a combination of the skill Lore and Intelligence) and can easily have 30+ Dex and 40+ Intelligence.
The entire point with the class is that it gets more potent as you grow in level, while being very weak at low levels, because you have so many different skills that you can use at different oppotunities(in terms of economy, xp, summons and utility) and no distinct focus(no +3 to any skill).
The same goes for other classes with multiple spell spheres. Just leveling 1 spellsphere will eventually meet a softcap due to casting difficulties(which is part of the reason why all spellcasters have access to 2-3 spell spheres).
As for countering early game rushes, it greatly depends on map size and player counts(if you have any nearby neighbours, and if an assassin needs to be on watch for multiple opponents) as I said in one of previous posts. If the assassin/martial hero starts across the map, you might have time to build your main building, which will give your potions more oomph if needed(simply having a lot of mana will also help).
In terms of map control, with 30+ Dex(cast Phantom Steed for more speed), an Archmage can either do a quick regular cap, or use the Alchemy spell Aquire to take them instantly. And with the Invinsibility spell, the Assassin/Warrior has no way of killing it, unless the Archmage is not paying attention(you have access to Eyes of Oros and Summon Guardian, as well as the option of True Sight, so you should have almost constant vision everywhere). Heroes with Arcane or Divination magic(Dispel, see Invisible) is a counter to this, but exceedingly rare on martial focused heroes.


Assassins and Gnolls can be potential retinue and so can be present and have decent levels right from the start. It only takes 1 match victory to have several high level units with Assassination skill in your retinue(because you can have access to xp boosting spells), making it much harder for your opponents to rush you in following games. And if they get a hero kill, they are instantly level 20(if playing at higher levels, ofc.) with xp leech on.
With each level, their %-chance to assassinate increases(starts 4%+1% per level). The point for them is to be a deterrent for hero rushes. If you have a hero that can aquire them at a high level or turn them into high level from the initial starting units(Divination, Pyromancy), they will always be available at the start of a map. Running into a group of those where each unit has 8% chance(1/12 hits) or more to instant kill your hero is a gamble you need to consider. Especially if the enemy hero has the possibility to escape as well.
If we take the Minotaur hero from the video as an example, he could have the Orb of Etheria, which has a 5% chance to cast level 5 Soul Flame(+30 XP, up to a cap of 60). By hitting a few crystal mines here or there(he has little use of them early, and destroying them gives ressources) and bringing some Gnolls as his starting units, he could have a bunch of level 5 Gnolls(9% assassination) shortly after the game started.


As for Sorcerers in the High Elf Priest/Dark Elf Army setup, they become available at tier 3. If the game is anything but a tier 1 rush(which is rare in 1.03.24), you should be able to hold on until you get to tier 3. Remember, Divination magic can cut research cost by 50%, making tiering up a fair bit easier.
Also, these level 5+ Sorcerers(due to Divination) can have over 250 hits and their Flame Pillar is cast as tier 2, and they have both faster regen due to Summon Mana and a higher max mana due to Sorcery research(so eventually, they can cast twice with a minor cooldown in between). Even without the research bonuses, the extra levels provide a decent amount ot extra stats, including mana regen.
Imagine bringing a few of these in your retinue. They wont get the bonus health(because it only works on produced/summoned units on each map), but the damage from their spells is the same. 3 of these can burst out over 500 damage as level 5+ once they have 40 mana(they start with 30). Over 700, if they are level 10+(XP leech can be very potent).
In a similar vein, a retinue of 3 level 5 Summoners(Daemon army) can eventually summon 3 level 5 Daemons, who in turn can summon Imps. However, Summoners are much more expensive to have in your Retinue, so that kind of limits how many you can bring.
Or if you can get 3 level 10+ Scorpion Priests in the retinue, they can each summon two level 6 Fire Elementals. That is roughly 200 damage per "salvo" from the Elementals, plus whatever the Priests do themselves.

[Mughi]

https://youtu.be/ZqYs5-siU3Y
Here's the vid about attack speed for you, JoeTheBartender. In summary: Yes, Invigorate lets a 35dex character attack faster. Invigorate attack speed buff does stack with Ice Floe. Firebreath attack speed is based on the units' attack animation speed, and is not affected by attack speed buffs or upgrades.

But what a freakin' coincidence that the day of.....

"Fendelphi" původně napsal:

I think there is some talking past each other here. I dont get why you need that aggressive tone though.

Yeah, we are talking past each other. I tried fixing that in the very first response. Unfortunately, the problem is that you're full of ♥♥♥♥, and for that reason we will never see eye-to-eye. Even with you being full of ♥♥♥♥, I'd be politely dancing around the ♥♥♥♥ like my first two responses to you with lightly probing questions if you hadn't shattered the barrier of mutual respect. With that down, I have no patience left to be kind to you.

I'm not trying to prove what is or is not true to the world, but I am proving it to myself. If you're still trying to convince ME, then you don't have a chance after so many false assertions. If you're trying to convince the world, then I hope to have left them a way to do their own homework instead of taking your word for it.

Everything you have suggested that is testable, when tested, proves to me that it's false.

"Fendelphi" původně napsal:

Can you imagine the sheer terror of seeing a horde of Spriggans(+100% attack speed) with Fire Breath+all the upgrades?

Firebreath doesn't stack with the attack speed buffs, all upgrades is difficult to get even in ideal circumstances and a very long game. If Speed upgrades worked with the Spriggan base speed, this would be almost valid.

"Fendelphi" původně napsal:

For the Assassin, it matters less, because the damage comes from Assassination and the attackspeed cap is 0.4 no matter what

The original uploaded video didn't include items, so thanks for the reminder. I've updated the video and this post with it. Items DO take you beyond the natural 0.4 cap, and stacks with Invigorate.

"Fendelphi" původně napsal:

Also, since Assassination is 1000 "true" damage, a high level, high health characters(Dark Dwarf, Dwarf, Swarm, Warrior class, in any combination) with a high combat stat can potentially outlast and kill Assassins. Especially if they have access to Healing.

Gets destroyed and has no means of fighting back.

"Fendelphi" původně napsal:

I did a quick test with a level 50 Minotaur Warrior(30 Mighty Blow, 5 Weapon Master, rest in Constitution and Ferocity, 35 Dex, rest in Strength). He dealt 482 damage(enemy heroes health) in 2 hits. I ran it a few times, and he always killed his target in 2-4 hits. No items.

Loses 70%+ in a short duel game and even worse in an economy one.

When a question gets asked that would lead to a test, you refuse to answer. This attempt to insulate yourself from being wrong is why I said you're only trying to "win the argument":

"Fendelphi" původně napsal:

I posted my initial responses of potential counters depending on the situations, and have been answering questions since then("as for ...." in pretty much every post since my first one).
What questions are you refering to? If there is a question I have missed, please tell me where.

"TheGreatSanguini" původně napsal:

What is your favourite level cap, and your best hero?

"TheGreatSanguini" původně napsal:

I'll be testing that later. What is the skill spread you prefer at level50?

"TheGreatSanguini" původně napsal:

How much do you normally invest in Lore, especially considering the 4 spell trees you're taking?

"TheGreatSanguini" původně napsal:

How deep you dive into each spell tree?

"TheGreatSanguini" původně napsal:

What order do you typically stick your hotkeys? I imagine those F keys are very busy.

"TheGreatSanguini" původně napsal:

What's your preferred skill spread on them by level50? Do you go beyond +8Ice and +6Fire?

"TheGreatSanguini" původně napsal:

How do you deal with mana loss, especially if you either fire off these spells and your opponent runs away for 55 seconds, or you wait too long and get hit with drain mana?

"TheGreatSanguini" původně napsal:

Also, do you mind giving me the specifics of your Elementalist and Archmage build, like stat&skill distributions?

...At which point, you made your accusation. You've still chosen to not answer these, deliberately being as vague as you can with non-answers like:

"Fendelphi" původně napsal:

For casters, their stat and skillpoint allocation(and viability in general) will depend on level limit and version of the game, as I said in the first post. Generally speaking, they get exponentially stronger as the level increases, because they can maximize their effects for the same mana cost, while also having more mana to spend. So a mage set in a level 15 limit will probably not be nearly as effective(or even viable) and have to focus on different skills compared to a level 50+ mage.

I'd want to ask followup questions about this Archmage, but we've already seen how that goes. I've already asked specifically about level cap, and we've seen how that goes. I've asked about spells, combos, tactics, and more. You don't have any. And the more specific you get, like with the Warrior items and Gnoll retinue, the more obvious it is that you don't know what you're talking about.

But the biggest joke so far is:

"Fendelphi" původně napsal:

I am not trying to win an argument. I am trying to showcase a point(in which lies questions, which you seem to disregard most of the time) that there are factors that can be a deterrent to the Asssassin.

I've been literally testing everything that has been presented, showing my findings when asked for them, and encouraging mutual understanding and development by showing the methods. Do you know what "psychological projection" is?

"Fendelphi" původně napsal:

I had no intention of claiming you were lying, just that it was a really poor proof of evidence(which unfortunately is very common on steam fora), due to it's bad quality, which you fixed in the next video.

"Fendelphi" původně napsal:

Btw, the new video does have some issues in terms of "being representative", since several times the heroes just stand idle or moves around, while the other attacks(fight 2, 3, 4, 6, 7, 10). And there is some inconsistencies with character skills and stats, which leads to the possibility that there have been outtakes or redos.

"Fendelphi" původně napsal:

Try without invincibility and it might turn out differently.

"TheGreatSanguini" původně napsal:

When provided with a video showing how to do the test yourself, and being encouraged to do the test yourself, you decided it was most reasonable to accuse me of lying instead of doing the test yourself.

So again, you have failed to do any testing yourself. You still don't understand that multiplayer desync causes some inconsistencies in pathing and updating. You can't even keep straight in your most recent post whether you're making an accusation of the video again. Most bizarre of all, you keep egotistically presuming that I'm presenting the video to prove you wrong rather than to show a method. That video was for JoeTheBartender who asked for it, not to make you to choke on your own ♥♥♥♥♥♥♥♥ (but a fine secondary benefit, if you do). I'll be straight with you: there were three fights that got thrown out. Two, the health bars failed to update, and one the warrior stacked on the assassin, giving the assassin two free hits that killed him. All three the assassin won. Why does this matter? It doesn't. You can TEST IT YOURSELF. Whatever you try to present at this point I won't be able to believe is honest, but if at the very least you can be honest with yourself, you can learn something.

When you're so openly dishonest as to say a MaxStr MightyBlow Warrior is less 1-dimensional than a character that has 20Int, 10Cha, and racial spells, I don't need to deconstruct you point-by-point. I don't need to win the argument. I've figured out that there isn't anything to be learned from you, and even writing out this much in response has been exhausting.

But hey, when you figure out how to turn on invincibility in multiplayer then you can teach me, because I haven't found that option yet.

[Lilykegalie]

My favourite build is Barb Assassin, because running is a good skill once you reach the 35 dex cap. Just 3 points in running lets you get enough free hits on a fleeing hero to proc assassinate even if they're at the same dex. I like playing at lvl35 cap best, but also play at 50. Running + assassinate is just unfair against anything but the other top tier builds. I hadn't thought about it before this thread, but calm from ice magic beating lifeward and other things might be worth the 4 points at lvl50. Fey assassin like joethebartender says is good, but reliant on your scouting to make it shine. If you find out where the enemy hero is early enough its just a free win.

I think the summoner matchup for assassin is pretty easy except in the old version where summons stay forever. Being able to see more of the map lets them keep tabs on where my hero is, and home portal lets them avoid being instakilled. But if I can keep up or make them use home portal more than once I usually win. Home portal is good becuase they can run away, but it's bad because they give up the map and run out of mana, and it gives up whatever troops were running with the hero as free kills.
Destruction builds are stupid. They either win big or lose big, and you need to push them out with troops instead of your hero. If you can juke them or surprise them its over, but otherwise it feels unfair. Fey assassin wins this one almost every time though, they have to be very fast with that trigger finger.

TheGreatSanguini původně napsal:

1) Orc Assassin. Being naturally immune to disease and poison puts them in a place where they can't be easily taken out or slowed down by common status, access to Chaos Magic lets them Drain Mana of any opposing wizard heroes removing their entire threat, and demolition exists for high level builds when you run out of other things to invest in, letting them destroy towers in a few hits.

One thing about drain mana is that it doesn't shut down a lot of casters. It will stop the big things like chaos plague, ressurection, destruction etc. but most other mages can just drink their mana pots.

Speaking of pots though, I notice when you do the test fights you don't use the health potions at all. Using health pots puts the fights wayyy more in the assassin's favour, because the enemy health pots do nothing against assassinate, but assassin's health pots keep them alive longer vs. warriors, especially if the keep is finished building by the time they get used. I like the arena map idea. Im surprised the warrior got 3/10, but I bet with pots it would be 9/10 for the assassin easy.
Thanks for the vids, its cool that other people use test maps too. Can you do a fight club vid of units fighting? I'm curious how much of a difference level makes, because I usually only get 10int but I see you always take 20.

@Fendelphi lmao, you're gonna play the victim? Youre ripping on the videos, why not do the tests like he says? Have you ever even played a multiplayer game? If you have something that actually works agianst assassins why not post that instead of a copy/paste article for campaign using sorceror spam. You keep talking like you have all the stats, all the skills, and all the mana. The more you say the dumber you sound.

[Mughi]

Fendelphi původně napsal:

I think there is some talking past each other here. I dont get why you need that aggressive tone though.

Starting again today with a clear head, I'll be specific about what set me off.
With you still ignoring my questions, you clearly have no intention of entertaining a discussion. If you ignore the questions I ask and ignore the statements I make, then pretend you are in some way responding to me, that is not a discussion.

I am not disregarding everything you have said, but it continues to prove you egregiously ignorant and dishonest. I didn't start this thread for empty banter, but because I'm playing the game and testing these concepts. From gnolls to towers, items to sorcerers, what you say can be tested. It's a shame you've never tried any of it, yet propose it as facts instead of questions. Most of these can even be tried in single player, for someone like you who has no friends to test with.

The videos were not directed as an attack against you. The videos demonstrate a tool that you clearly have never considered using, and they demonstrate the experiments that my understanding has come from, reinforced by in-game experience. The stats are clearly displayed, the spells cast were the Drain Mana that I had said earlier (which you can tell from the results, sprites, and availability), and the pathing behaviour is understood by anyone that's tried actually playing this game in multiplayer.

I don't know why you are so intent on critiquing the videos, because that will never prove whatever point you want to make, and it won't make me disbelieve the results I've generated from my own experiments. Nothing can prove to you that I didn't photoshop the stats, or hack the game, or have it replaced by the CEO of youtube. It doesn't have to. You are free to ignore reality, and never check for yourself.

"Lilykegalie" původně napsal:

One thing about drain mana is that it doesn't shut down a lot of casters. It will stop the big things like chaos plague, ressurection, destruction etc. but most other mages can just drink their mana pots.

True. They are limited though, and If I can Drain Mana a summoner for example, they have to burn though at least two of them just to Home Portal. For me to say that it removes the entire threat of casters was hyperbole, so thanks for keeping me in check. I still think it's a very effective tool against most casters.

|"Lilykegalie" původně napsal:

Speaking of pots though, I notice when you do the test fights you don't use the health potions at all. Using health pots puts the fights wayyy more in the assassin's favour, because the enemy health pots do nothing against assassinate, but assassin's health pots keep them alive longer vs. warriors, especially if the keep is finished building by the time they get used. I like the arena map idea. Im surprised the warrior got 3/10, but I bet with pots it would be 9/10 for the assassin easy.

That's a good point. I had deliberately ignored potions as being something equal on both sides, but that's flawed thinking on my end. I hope your testing goes well. The next time my friend is available without hamachi I'll be doing a video of the Destruction arcane magic classes as outlined by JoeTheBartender, and maybe capturing footage for Ice Floe kiting or other hero duelling things.

"Lilykegalie" původně napsal:

Can you do a fight club vid of units fighting? I'm curious how much of a difference level makes, because I usually only get 10int but I see you always take 20.

It's a bit of a pita to set that up as a custom map, but doable in singleplayer. I favour 20int for better spellcasting, and because it starts troops at level3, or level4 with the Training upgrade. I don't think I'd go for 20int without at least dipping into racial magic to make it worthwhile. I'll try to have that done for you soon, thanks for the interest!

"Lilykegalie" původně napsal:

Destruction builds are stupid. They either win big or lose big, and you need to push them out with troops instead of your hero. If you can juke them or surprise them its over, but otherwise it feels unfair. Fey assassin wins this one almost every time though, they have to be very fast with that trigger finger.

That does seem about right. Can you elaborate on any other details of the matchup, or think that there's antyhing more to it than that? What troops do you feel are best to use against them? I'm just discovering this matchup, having only played a few times against them and with them. I'm playing more of Protectors now than WBC3.

[NekRon99]

How many hit points does an average assassin have compared to an arcane destruction caster of same level? I would think the assassin who favors dex cant have a ton of hits. What about a plaguelord alchemist with 10 ranks in leech and maybe 10 in mana regen?
If he kills 5 peons he has enough mana for a destruction cast. Being an alchemist he can stay full up on heal pots after he blows everything up. This guy could cast destruction after every 5 kills and never let up. I havent played multiplayer before so I know I must sound like an idiot but Im enjoying this game and you guys all seem to be experts.

Does multiplayer always mean 1v1 hero duel? Does it become useless to build a base and train units because they just die too easy? Im guessing at the extreme levels thats the case but whats the turning point where you stop trying and its just hero combat?

I know Im no expert, but is there a point to the higher levels at all? If assassin is so OP I would think most people would just not allow the class for the sake of fair and fun games.

[NekRon99]

By the way I'm currently playing that plaguelord alchy and his name is "Mutation". Im loving him right now. my goal is to keep his intelligence at 10 so he only has 50 casting points. I know there's a magic item out there that adds 40 casting points which will allow him to cast the spell that doubles all his gear stats which I'm assuming will bring him up to 130 casting points, which will give him enough to cast destruction. I still haven't decided whether to go mainly strength (for 1 shot kills) or dex for speed. Plaguelord Alchy's get some great skill choices.

[Lilykegalie]

TheGreatSanguini původně napsal:

It's a bit of a pita to set that up as a custom map, but doable in singleplayer

Yeah lol, that's why I'd rather watch a video of it than do it myself.

TheGreatSanguini původně napsal:

That does seem about right. Can you elaborate on any other details of the matchup, or think that there's antyhing more to it than that? What troops do you feel are best to use against them? I'm just discovering this matchup, having only played a few times against them and with them.

I don't think there really is anything else to the matchup... Basic missile troops do fine at pushing the caster out as long as they don't get run over by his troops. Harpies are good if you can get that far in tech but the game is probably over before that point as he either blows up or gets blown up. Your hero will always be better at fighting units than theirs so any free kills you can get on scouting units or groups away from the mage is a win you need to take. If you can go equal with them for map control then you will win eventually but since they can threaten your hero so bad with destruction thats really hard to do. I don't like destruction builds because unless your playing it it feels like your barely playing the game. Either they hit their spell or they don't and thats all the game is.

NekRon99 původně napsal:

How many hit points does an average assassin have compared to an arcane destruction caster of same level? I would think the assassin who favors dex cant have a ton of hits. What about a plaguelord alchemist with 10 ranks in leech and maybe 10 in mana regen?

Its generally not smart to invest in Str, so most builds end up with around the same amount of HP unless using constitution. Your plaguelord will end up the same.

NekRon99 původně napsal:

Does multiplayer always mean 1v1 hero duel? Does it become useless to build a base and train units because they just die too easy? Im guessing at the extreme levels thats the case but whats the turning point where you stop trying and its just hero combat?

At level 50, it depends on the heros build. Most should be nearly immune to units on their own, but units can bodyblock for the hero, disrupt convertion, and give vision of the map. If you can get a truly massive advantage of units it will eventually wear down any hero though and thats the threat that usually forces the hero duel. Hero duel ability is important for controlling the map because if you can control the whole map then you will always inevitably win by massing more units faster. Ironically if you control the map the enemy has no chance to win except for forcing a duel with you. I prefer 35 because that lets the hero builds use all of their skills but can still be threatened by some tier3 units and above. Also destruction builds don't exist at 35 lol.
Its always worth building a base. The heros do all the heavy lifting but units help and potions scale with keep level. But yeah the higher the level the less relevant units become.

[NekRon99]

Thanks for responding Lily,
So a few thoughts come to mind then. I had to scrub my idea of the double power items spell because when I tried it It did not double the mana bonus of my equipped item which meant I had to go higher on my intelligence stat. Also I don't know if items would be allowed. I'm still pursuing him in single player however just for fun, currently level 21.

You stated the assassin will have roughly the same hp as my plaguelord based off of stat choices. So all I really need to do is whittle him down while keeping away from him right? How about at 50th level my plaguelord trains up his dex like the assassin and keeps his intel around 10, just enough to cast whatever damage spell his class choice allows, and taking 20 ranks in leech (for 40 mana boost a kill). Skip the Arcane magic altogether to allow his leech ability to kill what his aoe spell weakens, thereby replenishing his mana as he runs away from said assassin. Even at lvl 35 he could still have 5 ranks in leech, so more running away but same effect.

I figure the assassin will just charge ahead trying to engage so maybe Im a ranger with entangle to slow him up long enough to drop my aoe spell (lightning?) and chop up enough critters to regain my mana as i press on away from him. He takes some damage every time I aoe cast, and cant catch me while he's entangled.

This way I'm not dependent on items since I don't even know if they are allowed in multi. I could see perhaps if the assassin didn't build a base and hence had no critters for me to kill it would negate the leech at least, but he still couldn't catch me. Thanks for reading.

[Mughi]

Lilykegalie původně napsal:

TheGreatSanguini původně napsal:

It's a bit of a pita to set that up as a custom map, but doable in singleplayer

Yeah lol, that's why I'd rather watch a video of it than do it myself.

https://youtu.be/qI_DwmDmzOE
I had trouble getting the AI to do what I wanted, so just took my best shot at it manually. I'll play around in map editor more and see if I can trigger some event or otherwise manipulate the troops into fighting each other more reliably, but here's what I have for now. Also, thanks for the perspective on the Destruction matchup.

"NekRon99" původně napsal:

How many hit points does an average assassin have compared to an arcane destruction caster of same level?

At level 50, most builds have between 300~500HP. By contrast, the Destruction Ice Mage I built has roughly double the average at 824hp. I believe that JoeTheBartender's Destruction Healer could be the second most effective Destruction build, but would have roughly ~700hp because of the deeper investment in another spell tree and Elcor's Aura. The Destruction Healer build is probably better against a "general" hero, but gets directly countered by LifeWard where the Ice Mage one doesn't.

"Lilykegalie" původně napsal:

Its generally not smart to invest in Str, so most builds end up with around the same amount of HP unless using constitution. Your plaguelord will end up the same.

I think this is kind of true, but a little off the mark. While Str contributes to HP, it is not the sole determining factor. HP also scales off of level-by-class. For example:
Knight Warrior with 7Str @lvl50: 511HP
Knight Assassin with 7Str @lvl50: 382HP
Knight Healer with 7Str @lvl50: 293HP
Minotaur Healer with 7Str @lvl50: 293HP

"NekRon99" původně napsal:

What about a plaguelord alchemist with 10 ranks in leech and maybe 10 in mana regen?
If he kills 5 peons he has enough mana for a destruction cast. Being an alchemist he can stay full up on heal pots after he blows everything up. This guy could cast destruction after every 5 kills and never let up.

Destruction on its own is very very bad. It deals your casters HP-1 as damage to everything in his radius, which includes the caster. This puts your hero at 1HP, on the brink of death, and damages your own units around him as well. I had overlooked the use of Destruction for years because I hadn't ever considered stacking it with Constitution for massive damage (enough to take down anything: towers, keeps, hero). Because the plaguelord doesn't have access to Constitution, he would be incapable of doing the tactical nuclear strike. While the 100mana cost of destruction is pretty steep, the real cost of it is the HP spent for it, and that's also where all of its potency comes from. Since the Leech skill doesn't give mana from killing friendly troops, you'd need to be in melee with enemy troops after burning all your health potions, and as a very poor melee build, it seems very risky.

I'm not sure that the tactical flexibility you'd get from throwing the Destruction spell on a normal Alchemist build like that would be worth it, since Destruction takes 10pts in the Arcane tree to get, and maintaining the mana pool to use it means not using your other useful spells that your hero is built around. It is an interesting build, and it seems like there's a lot of incredibly clutch skilled plays that could be done with it, but that skill and clutch is necessary because of the many weaknesses it would have.

If you 1) maintain your hero at max health 2) reduce the enemy hero's health below yours 3) maintain a minimum 100 mana 4) get within range of the enemy hero without taking any damage or getting Drain Mana'd, then you can still win the game with Destruction on that build, but it doesn't seem nearly as potent as the builds listed above with Constitution and one misstep will cost you the game.

If I understand what you're proposing as a build, I expect that it would end up with 4str, 19dex, 30int, 16cha. For skills, I'd take 7 Summoning for Home Portal, 6 Alchemy for Acquire, 10 Arcane for Destruction, 10 Energy, 9 Leech, and 11 Warding (or more Alch+Ritual). This would end up with 18 Speed, 229HP, and 110Mana. While it has a lot of good spells available, casting any one of them takes Destruction off the table for a while. If you were to use Destruction, it would have to be on a hero that has already been significantly weakened by troops. Also, it doesn't have the speed necessary to chase down or escape enemy melee heroes (relying on Home Portal), which when combined with its command bubble is only barely big enough to outrange an enemy melee hero from casting Drain Mana or some other crippling spell.

I think the build is probably better without Destruction in it. I don't think that shorting yourself on Int would be wise with this build, as then you're losing effectiveness on your Energy, and need to invest skill points in Ritual to compensate and actually cast higher level spells without failure. If you would build much different from this though, please let me know so I can figure out your reasoning.

"NekRon99" původně napsal:

Does multiplayer always mean 1v1 hero duel? Does it become useless to build a base and train units because they just die too easy? Im guessing at the extreme levels thats the case but whats the turning point where you stop trying and its just hero combat?

Above 50, or with items, I'd say that the game is basically nothing but a hero duel. At level 50 without items it's extremely hero-centric but bases still matter. No matter what, a base is good to have for the reasons Lilykegalie mentioned, and the lower the level cap the more potent everything else is in comparison to the hero.

"NekRon99" původně napsal:

I know Im no expert, but is there a point to the higher levels at all? If assassin is so OP I would think most people would just not allow the class for the sake of fair and fun games.

I don't think there's much of a point to higher levels, and level cap I find most enjoyable to play at is probably 30 (though 50 is the most fun to theorize builds for because everything's on the table). My friends ended up agreeing to not use the Assassinate skill for the sake of variety, but this thread has been an interesting learning experience.

"NekRon99" původně napsal:

You stated the assassin will have roughly the same hp as my plaguelord based off of stat choices. So all I really need to do is whittle him down while keeping away from him right? How about at 50th level my plaguelord trains up his dex like the assassin and keeps his intel around 10, just enough to cast whatever damage spell his class choice allows, and taking 20 ranks in leech (for 40 mana boost a kill). Skip the Arcane magic altogether to allow his leech ability to kill what his aoe spell weakens, thereby replenishing his mana as he runs away from said assassin. Even at lvl 35 he could still have 5 ranks in leech, so more running away but same effect.

This is an interesting idea. I had really hoped that some combination of Ice Floe and Freeze would work for kiting the assassin. The Minotaur Elementalist I was trying from earlier in the thread based on Fendelphi's writeup ended up being a disappointment, running out of mana very quickly and having no means of recovering.

If you build a Plaguelord Ranger with stats of 8str, 35dex, 10int, 16cha, and skills of 7 Nature Magic, 7 Summoning, 20 Leech (definitely overkill) and 16 Running, you end up with a build that can naturally outspeed the Assassin like a Theif would. I would normally expect Leech to be of very limited utility because it relies on your opponent feeding you troops, but being super fast+entangle probably gives you the best shot at taking down troops without getting yourself baited and killed.

The downsides I foresee to this are that it only has 40 max mana, and if you're using call lightning for damage then you're killing the units that you want to Leech off of. However, the potential exists to have an Assassin Entangled, critical freezed by an Imp, and diseased all in one, which is hilarious. Very interesting, and merits further testing. Thanks for your contribution!