SlashStrike’s Guide on Skill-Builds


Hi guys, SlashStrike here (not the same person as slahser), and this time I’m bringing you a guide on skill-builds. I haven’t stopped writing guides on heroes, this is just a different general guide. In fact, here’s a strawpoll ( through which you can help decide on what hero my next guide is going to be on. Since I’m not making any money by writing this, I’m shamelessly plugging my stream over at – it’s got follower and donation notifications, a moobot, cool music and of course some high 6k – low 7k gameplay! If you enjoy this or my other guides, make sure to tune in. Back to skill-builds.

I believe this is a significant aspect of DotA that is simply not discussed enough. Casters do point out skill-builds from time to time, but often provide uninformed analysis, trying to reach the conclusion of whether a build is ‘good’ or ‘bad’, or whether they agree or disagree with it. A skill-build can hardly be disagreed with (as long as there’s at least 1 level in each spell) – it is simply an indication of the playstyle that the player (and team) has chosen to go for. Now, that playstyle – that is something that the caster can choose to agree or disagree with.

A topical example would be Fear’s Spectre versus NiP the other day ( He took stats at levels 5, 7, 8, 10, and 12, while maxing Spectral Dagger with only 1 point in his other abilities. The caster said he disagreed with it, which I think is an uninformed statement to make (don’t get me wrong, I’m not looking to call out the caster here). This is because Fear had also gone for the Phase-Aquila-MoM build, which translating to his playstyle meant that he was mostly hitting lane creeps, neutral creeps and ancients in the early game, until he finished his Radiance at 20 minutes. Now, how useful are Desolate and Dispersion for farming creeps? They aren’t, at all. And it is evident that this is exactly what Fear had in mind, because as soon as he was about to finish his Radiance (20 minutes in), he also got to level 13 and started levelling his spells instead of stats.

Nowadays, skill-builds in DotA are a lot more flexible than they used to be. There are fewer heroes that follow the same skill-build every game, more heroes that can adapt to the game, and of course more carries that incorporate early stats as part of their skill-build. But over-all, with teams becoming more efficient at getting levels on even their lowest-priority supports, players feel more comfortable experimenting with their skill-build in the early levels.

As for heroes that do take a lot of farm, they are subsequently expected to level up a lot faster, and therefore can afford to level up skills according to the immediate state of the game. Your skill-build dictates how you can most effectively play your game, and will probably decide how well you do on your lane, how quickly you can farm afterwards, and whether you have enough damage and disable for that kill or not.

Table of Contents

  1. What makes a skill-build good/bad
  2. Playstyle
  3. Skill Scaling
  4. What does a skill offer?
  5. Unique Effects
  6. Closing Words

1. What makes a skill-build good?

A good skill-build is one that allows you to play your hero to the fullest potential. There should be no case in which you wish you had a higher level of another skill. If you find yourself not making good use of a skill soon after you have put a point in it, then your skill-build is not as effective as it could be. Most of all, however, it depends on the playstyle you will go for in your given game. There are far more good skill-builds than there are bad skill-builds.

Ok then, what makes a skill-build bad?

A bad skill-build is quite apparent most of the time. Were you unable to set up a kill as a support Earthshaker because the 1.0 second stun on your level 1 Fissure wasn’t long enough? That’s an indication that maxing Enchant Totem first was a terrible idea and your skill-build is bad. Did you pick Dazzle to sustain the push, but are then unable to do so because Dazzle is maxing Poison Touch and Shallow Grave with no / too little points in Shadow Wave? Again, an example of a bad skill-build, because it limits the hero’s ability to bring to the table that which he is meant to offer – in this case, a spammable heal.

2. Playstyle

As everyone knows, there are many heroes that can do well in different roles. Often, their skill build is adapted to this.

For example, take a Naga Siren that is played in the typical farming hard-carry style – Riptide and Mirror Images are maxed out because they speed up farming, while only a single point or often none at all are put into Ensnare, as the Naga is not looking to fight any heroes early on. While this could be a regrettable decision if the Naga misses a kill on a target that just teleports away, that situation is unlikely to arise because a Naga that has gone for that skill-build will be more focused on chasing creeps, and not heroes. A support Naga Siren, however, is more likely to level up the Ensnare since it’s her only disable apart from the ultimate, and gives her ganking potential.

Another case would be Lina – when played mid, players often max out Dragon Slave and then Fiery Soul, leaving only one point in Light Strike Array and then maxing that last. This is because thanks to a mid-Lina’s level advantage, her right-clicks are more significant in the early game (making the AS from Fiery Soul more valuable), and it also allows her to pressure the enemy tier 1 mid tower. A support Lina on the other hand often only takes one or no points of Fiery Soul, and chooses to max out Light Strike Array, because the two nukes are going to be her main contribution to the team, not her physical attacks.

However, the reason this section is called playstyle and not role, is because there can also be different playstyles for the same hero fulfilling the same role.

For instance, Axe is nearly always played as a position 3 initiator that takes some jungle farm to secure his blink. Because farming the blink is so important, Counter Helix is maxed out first in 95% of games. It makes sense since levelling Counter Helix improves his kill potential and damage by roughly the same amount that levelling the other spells would, but it is the only spell that helps him farm.

In some cases you will see the Axe skip Battle Hunger entirely and just max out his Berserker’s Call alongside the Counter Helix – this is often when he is zoned from the lane and unable to assert any pressure until after farming his Blink Dagger, at which point having a level 4 Berserker’s Call will be much better than having Battle Hunger.

Yet let’s say the enemy team went for an aggressive tri-lane and Axe is left to face the enemy offlaner in a 1on1, or he’s simply up against a rather weak dual lane that cannot zone him effectively – in this case, we often see Axes pick up at least one point of Battle Hunger because it gives them very early kill potential.

Most skills grant all their effects starting at level 1, meaning that a 1-1-1 build at level 3 is generally good on most heroes. There are still many exceptions to this, of course.

One of these exceptions would be a Witch Doctor that is supporting. His skill build has decent variety – Paralyzing Cask is generally maxed out first, but then some max Voodoo Restoration along-side it, while some skip / leave it at level 1 and choose to max out Maledict. The difference between the two is fairly simple – does the team need more healing or more damage early on? It is also important to note whether the enemy offlaner can effectively be zoned and/or killed without Maledict. However, a level 1 Maledict is one of the weakest abilities in the game because it provides literally nothing other than damage, and the damage is negligible until higher levels, meaning it is better to either max it or not skill it at all – unlike with many other spells, getting just 1 level in it gives you very little benefit.

3. Skill Scaling

I briefly touched upon this concept in my previously released Ember Spirit guide, so those of you who read it may remember these examples from there.

A term that is very important to define and understand, is the so-called ‘value point’. The value point of a skill is the level at which you get the largest increase in the spells potency. Most skills with an obvious value point have it at level 1, and they are also known as ‘1-point-wonders’.

For instance, take Yurnero, the Juggernaut – Often his Bladefury is left at level 1 because it grants you the full 5 second magic immunity, and the total damage only increases by 100 each level, starting at 400. This means you get a 25%, 20% and 16% damage increase per level, and a slight cooldown reduction. A more extreme example of a 1 point wonder would be DK’s stun, starting at 2.5 seconds and only increasing by 0.25 seconds as you level it, meaning 10%, 9% and 8% per level.

While fewer, there are also skills that have a clear value point that is not at level 1.

For example, Sniper’s Headshot – you get the full slow from level one, but the damage starts off at 15 and increases by 25 each level, meaning you get a 166%, 62.5% and 38.5% damage increase, which is why level 2 can also be considered the ‘value point’. Another example of this would be Bounty Hunter’s Shuriken Toss – with the damage increasing by 100%, 25% and then 30%, level 2 is also obviously the level giving most value for your skill point.

This is once again a very important concept to understand when deciding how to level your skills. A regular skill with regular scaling may not have any clear value points, but it is evident that every next skill-point put into it gives you less in return.

Case in point, Phantom Lancer’s Spirit Lance – 100/150/200/250 damage, 10/20/30/40% slow, 2/4/6/8 second illusion duration. Important to note that the slow duration is a constant 3.25 seconds. Very simple stuff. However, let’s take a look at the percentage increase with each level.

  • 50% damage, 100% slow and 100% duration increase from level 1 -> 2.
  • 33% damage, 50% slow and 50% duration increase from level 2 -> 3.
  • 25% damage, 33% slow and 33% duration increase from level 3 -> 4.

Clearly, the skill improves massively from level 1 to 2, then not so much from level 2 to 3, and then only a little bit from level 3 to 4. On a hero such as Lancer that relies on right-clicks for most of his damage, and being an illusion-based carry that likes stats so much, it makes sense to keep Spirit Lance at level 2 or 3 and take stats in favor of maxing it out. Does this mean maxing it out is bad and is going to lose you the game? Of course not. Intricate as DotA may be, a 10% difference in amount of slow, or the 38 hp, 26 mana and 2 damage you do or do not get from a point in stats is hardly going to dictate the course of the game. Nonetheless, at the end of the day every little bit counts towards the end result, which is how you use your hero and what you can accomplish with it.

Never forget to look carefully at how a certain skill scales, because there are many that would surprise you. For example, ZeusArc Lightning. It deals 85/100/115/145 damage, bouncing 5/7/9/15 times. Notice that little jump at the end?

  • 17.7% damage, 28.6% bounce increase from level 1 -> 2.
  • 15.0% damage, 22.2% bounce increase from level 2 -> 3.
  • 26.1% damage, 66.6% bounce increase from level 3 -> 4.

Clearly, levelling from 3 to 4 is a great boost compared to the other points. This is why you either keep it at level 1, or max it all the way once you start putting more skill points into it. It also explains the different variants we have been seeing lately, of some Zeus players leaving one point in Arc Lightning and maxing out Static Field, and other Zeus players leaving one point in Static Field and maxing out Arc Lightning.

Lastly, it is important to note that many big AoE ultimates do not scale well enough to warrant levelling them over your other spells.

For example, the benefit of having a Puck‘s level 2 Dream Coil over a level 1 Dream Coil is quite small compared to getting that level 4 Phase Shift and reaching un-killable status an entire level earlier. “But what if you don’t get the kill because Dream Coil’s stun didn’t last long enough?” By the time you’re level 11, a 0.75 second stun should never make the difference between getting a kill or not. If it does, it means your initiation and gameplan is far too shaky and unreliable.

Same goes for Magnus’ Reverse Polarity – the stun only increases by 0.75 per level, and while that is certainly valuable and worth taking over stats, it’s not worth taking over an extra level in Empower which is what’s going to boost your carry’s farmrate and damage in teamfights.

Another mention goes to Enigma’s Black Hole. Not only does the duration stay the same, the manacost is increased massively with each level, essentially making it detrimental to level it past level 1 until you really have more than enough mana.

4. What does a skill offer?

This is an important aspect when deciding how many skill-points to put into it, and when. There’s a difference between skill effects, i.e. what the skill does (damage, heal, buff, vision) and skill utilities, i.e. what the skill can be used for (farm, push, scout). Of course, most skills are made up of several effects. Often some skills, or some heroes’ skill-sets that are considered very strong, are made up of many effects, i.e. they ‘do a lot’.

Take for example the skillset of Troll Warlord versus that of Antimage.

Troll’s skills buff his armor, HP, MS, AS, BAT, and damage, offer him two nukes (help him farm), a slow, a miss chance debuff, stun potential, a buff to his whole team making it easier to take objectives, and give him global combo presence with heroes such as Legion Commander, Spirit Breaker, Clockwerk, Void, etc.

Antimage’s skills buff his magic resistance, boost his damage versus heroes (but not if they BKB), give him mobility (helps him farm), and potential for a big AoE nuke + ministun.

Just from the difference in length between the two sentences, it is obvious why Troll is more versatile and better in 95% of cases.

The term ‘nuke‘ refers to a spell that simply deals damage in one burst (often magical). Nukes are generally the most versatile. They help you farm, they are useful in any engagement, and they help you wave-clear when pushing/split-pushing/defending. Spells that are pure nukes (i.e. do nothing but deal damage) are almost always maxed out first, because they are most effective in the early game, do not scale well, but allow the hero to farm. Examples – Magnus’ Shockwave, Shadow Fiend’s Shadow Raze, Lina’s Dragon Slave, Death Prophet’s Crypt Swarm, etc.

For pretty much any core, maxing out the spell that helps you farm first is your safest bet. You may be sacrificing a little bit of your offensive capabilities, but in the event that there are no kill opportunities you can always fall back on farming.

Skipping a spell entirely is generally ill-advised, because whatever effect the spell gives you can be made decent use of with just 1 point in it, whereas you are otherwise entirely skipping out on it. The exception to this rule clearly is when the spell gives you nothing but negligible damage and/or debuff, which as mentioned earlier is the case with Witch Doctor’s Maledict. Further examples of one-dimensional skills that can be skipped entirely are Bane Elemental’s Enfeeble, Medusa’s Mystic Snake, Silencer’s Curse of the Silent, and Ancient Apparition’s Cold Feet.

5. Unique Effects

A special mention to some skills that have unique effects that cannot be achieved with any other skill. If the enemy bans these heroes, it becomes rather evident what kind of heroes they are looking to pick up. There are probably some more, these are just some of the most prominent ones.

Ancient Apparition’s Ice Blast – prevents healing, incredibly powerful versus heroes that rely on heals/regeneration (Wisp, Slark, Necro, Chen, etc.)

Treant Protector’s Living Armor – heals structures, essentially rendering slow sieges useless and forcing the enemy to fully commit if they want to take a tower.

Wraith King’s Reincarnation – the ideal carry versus long-cooldown skills, a strong counter to lineups relying on skills such as Black Hole, Chronosphere, Doom, Ravage, etc.

Silencer’s Global Silence – the only global disable in the game. Incredibly hard to fight against, and games usually end up revolving around its cooldown.

Night Stalker’s Darkness – the only skill that reduces all enemy vision, including that of wards. Powerful in chaotic fights and ideal versus line-ups that like a clean 5 on 5 teamfight and rely on none of their heroes getting picked off.

6. Closing Words

That’s it for the guide! I hope you found it informative – if you did and wish to see more from me in the future you can support me by following me on this blog, twitterfacebook, youtube and of course twitch.

If you have any requests for the future guides feel free to send me a message on any platform.

Good luck and remember – balance, in all things.


14 comments on “SlashStrike’s Guide on Skill-Builds

  1. Mayouay says:

    Thx, it was good to read, ill try to adapt more on skill builds so 🙂 Keep the good work !


  2. DumpsterPlayer says:

    >Another case would be Lina – when played mid, players often max out Dragon Slave and then Fiery Soul, leaving only one point in Light Strike Array and then maxing that last.

    Hmm.. I think most players take a value point in her passive and max LSA second.


  3. V says:

    Great read Slahs.. Slashstrike! Good job as usual.


  4. Kav says:

    For the future, if you want to prove your theory use some examples of heroes that aren’t generally played or considered the current trend. Every example you gave is basically flavour of the month, almost as if you’ve watched some pro games and wrote a story about them using their skill builds. Not flaming, but your article would be 100% better if you did this.


    • slashstrike says:

      I used current trends as examples because they can be seen working right now. What’s the point of talking about Oracle or Enchantress’ skill-builds when barely anyone plays them? Jugger is a great example because everybody used to go a certain skill-build on him, and then the most common build changed to the current stat-heavy one as a result of his buffs.


      • SpiritB says:

        In line with your Jugger example, I’ve noticed Burning almost always do a 1-1-1 then stats build, taking up ulti on 6/11/16 when farming solo safelane or aggressive trilane hardlane. I guess the reasoning behind of getting only 1 point in crit rather than maxing it out asap like most payers do is to get consistent last hits rather than getting sudden burst of crits.

        Other Jugg build I see is maxing crit and getting 1 or 2 points in healing ward, coupled up with phase boots and mask of madness, can do a solo roshan at level 8.

        So all in all, your playstyle dictates your skill build. GLHF!


  5. H47 says:

    I don’t really agree with your reasoning with the Puck ultimate part. Wouldn’t your plan be equally shaky if you mess stuff up because you don’t have level 2 Phase Shift instead of level 1 at level 11 or because you got too greedy and ended up in a situation where you have to Phase Blink out of danger when enemy support are still squishy enough to die to you alone, assuming you are of course hailing from a solo lane and in most cases, mid lane? Your initiation and game plan would still be shaky if you fuck up in another way because of your skill build choices. Just think your reasoning is way too subjective for that and doesn’t cover it in a way to have it down to science, but closer to covering it at the cuz u suk lol level and not taking into account that the enemy doesn’t die in 1.5 seconds, but would need 2.25 seconds isn’t any more of a blunder than not taking into account that your Phase Shift is 2.25 seconds and not 3 seconds. The level 2 of Phase Shift doesn’t give you the window to Blink away from any situation, but to get from level 11 to level 12 requires 2200 exp, the same amount you’d need to get both levels 13 and 14 together. Therefore I think that the lvl 2 Phase Shift isn’t really worth 2200 exp. You’re stuck on level 11 the longest time out of all levels in comparison to how much exp is available on the map and if you spend the point on something that doesn’t directly help you then, but build towards something, you might miss a lot of chances with just being stuck in the middle of your late value point project. Even if you’d get the Phase Shift level 4 earlier, you don’t really need to struggle to get it anyway and don’t really need it faster besides in some borderline scenarii, as getting levels isn’t particularly hard around middle levels of the spectrum, but upgrading from lvl 11 to 12 is a Sisyphean labor and you should take into equation that the skill point you’re going to use at level 11 will be in play for a while now and if it is a practically worthless skill point, then it stunts your growth more than a skill point that is an overkill. Additional levels of Coil also increase your nuke potential and give you better solo burst nuke pickoffs, which is something Puck is pretty good at early, but doesn’t do nearly as well come 20 min mark. If we really went for the min-maxing for utility path, why not just ditch skilling Illusory Orb past level 1 or 2 and start going for Phase Shift even earlier? In short, I just think that when discussing skill builds, you should also think what levels come easy and which ones don’t when it comes to exp requirements. I can see where you’re coming from with the faster Phase Shift, but I don’t think it’s worth the risk of taking a weak skill point worth 2200 exp at level 11. I’ve sometimes messed up and skilled additional Boulder Roll over Magnetize level 2 at level 11, accidentally (no point in skipping ulti at level 11 on him) and that felt like the longest level 11 ever and it truly is. As for the Zeus builds, I’ve seen a lot of pro Zeus players get 2 points in Chain early instead of maxing bolt by level 7 on Zeus, while there’s no real value point in early levels. Thoughts on that?


  6. .339 says:

    i didnt know about the stuff u wrote about blackhole,will take it into consideration next time thanks!


  7. gameorg says:

    You didnt metion cooldown decrease when talking about RP and Blackhole. You basicly never see pro players skip the point at level 11, exactly due to this. On those cooldowns, every second matters – more than a point in another skill.


    • SpiritB says:

      I’ve seen iceiceice get only 1 point in Blackhole countless of times, most recent I remember is when the enemy picked a Silencer. So, don’t say you never see pro players skip a point at level 11, because he clearly skipped level 16 Blackhole too. Don’t believe me? search for it and you’ll see.


  8. da says:

    Make a guide about microing, controling groups & stuff like that


  9. […] gosto de falar como Ucranianos e Russos, me julguem), hoje trarei pra vocês a tradução do ótimo guia do SlashStrike sobre Skill-Builds. Pode parecer algo pequeno mas, conforme a leitura for encaminhando aprenderão vários ótimos […]


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s