Friday, June 28, 2013

Let's Play (Part 4) Mirana


Every week or two I generally try to play a new hero to discover what it’s like playing it, and in this series of blogs I will make a post on my experiences with it.

The Celestial Starlight skin, created by Oni and Zaphk

Mirana (Princess of the Moon)

Mirana, the Princess of the Moon, is based off the character Tyrande Whisperwind in Warcraft 3, the High Priestess of the goddess Elune, and thus also leader of the Night Elves. Favoured by Elune and touched with the powers of the goddess herself, Tyrande is both a capable leader and formidable fighter. She organized the defense of Kalimdor against the initial waves of undead, and played a key role in the defense of Mount Hyjal, sacrificing the Ancestral Guardians and the World Tree to defeat Archimonde and the Burning Legion.

In DOTA2, Mirana is a cunning ranged hero that excels at helping her team set up ambushes and surprise attacks. Her high mobility and invisibility skill make her very difficult to take down.

Mirana has always had a decently good win rate and pick frequency, currently trending in the top 30% win rate and pick rate on DOTABUFF. She is difficult to master and she is often underestimated and seen as a low impact hero.


Moonlight Shadow is one of the best team buff skills in the game, giving your side a large advantage in teamfights by giving them persistent invisibility.

Her high mobility and low farm dependance make her one of the most versatile heroes in the game, able to play nearly any role in any lane.


Falls off in strength by late-game due to poor ability scaling.

A poor dueller - she relies on ambushing her opponent and winning through strength of numbers, if challenged head on in an even fight she likely has to flee.

Sacred Arrow

This skill inflicts longer stun and heavier damage the further it travels. It will do its maximum stun and damage at a range of 1500, and has a maximum range of 3000. For reference, the size of your DOTA2 viewscreen on a 16:9 monitor is very roughly 1250 tall and 2250 wide.

I'll take a moment to talk about skill design (and game design in general). This is an example of a really good skill mechanic - it takes a long time to master for the user, but playing against it is simple from the point of view of the victim (avoid the arrow!). The mechanic is concise - it can be summed up in a single line - but it has multiple uses.

1) Ambush (initiation) - firing it from out of vision range at long distance is a low probability but high reward and low risk play. If it hits you might have a kill, if it misses you haven't lost much.

2) Teamfight (extension) - in a larger clash you cannot afford to miss this skill, as both teams are committed to fighting, so you will use this as a follow up stun by firing it at an enemy that cannot evade it.

3) Backstab (assassination) - if you can sneak up behind an enemy safely under the cover of invisibility you can get into point-blank range and use Sacred Arrow and Starstorm for high burst damage. (360+525=885 damage)

A hypothetical skill with all the opposite qualities to this would be terrible design. This hypothetical bad skill mechanic would be so complicated it would take several paragraphs to explain, yet it only has one use, and is very difficult for the victim to play against because there are three diferent ways to escape from the skill and only one is viable at any particular time.


Her leap is instant cast and will disjoint projectiles heading towards her, which will allow you to dodge a variety of nasty spells. (In most cases, if a projectile is en-route to you and you leap, that projectile will miss). A large part of playing Mirana is being aware of what skills you can disjoint, and deliberately baiting the enemy into using them on you. As always bear mind that you can not disjoint Sven's Storm Bolt, Alchemist's Unstable Concoction and Phantom Lancer's Spirit Lance, but it's worth using the leap anyway - even though those projectiles follow you through your leap, it will stun / slow you at a distant, hopefully safer location.

Moonlight Shadow

If you are using this to initiate fights, make sure that no one on your team can be seen during the fade time - if someone is spotted in the fade animation, it's likely the entire enemy team will fall back to a safer position until its over, wasting your ult.

It's also a good ability to just use in a teamfight - it will allow your supports and initiators to do their jobs more effectively under the cover of invisibility (they can channel their skills while invisible), and will allow stunned or injured allies to escape. At level 3 with a fade time of only 1.5 seconds it's very difficult for the enemy to target anyone as they're fading into and out of invisibility every few seconds.


Mirana is a good pick when no one else on your team has natural invisibility or has Shadow Blade as a core item - this will force the enemies to use gold buying detection, unless they want to fight every teamfight at a big disadvantage.

Heroes with long stuns and disables will allow you to land your arrow more reliably. Some good examples are Bane's Nightmare, Nature's Prophet's Sprout, etc


None, really. Mirana's skill set is so large and versatile that there's no single aspect of her play that you would want to specifically counter.

Item build

Mirana is quite commonly built with physical damage items even though she doesn't have innate damage scaling. How big is the gap between her and other traditional carries who do have passive scaling, though?

I've put some numbers into a spreadsheet to give us an idea of how they compare - DPS output at level 25 with no items, Mirana has 153, Sniper has 213 and Drow has 267. With a late-game item loadout costing about the same - Mirana with (Phase / Midas / Mjollnir / Daedalus / Yasha / Drums) has 896 DPS, Sniper with (Treads / Midas / Shadow Blade / Mjollnir / Daedalus) has 992 DPS, Drow with (Treads / Shadow Blade / Dominator / Daedalus / Butterfly) has 1264 DPS.

While the game doesn't boil down to numbers, it's nice to know what they are. Sniper's long range and ministun can't be factored into this - neither can Drow's 60% movespeed slow and large AOE silence, or the fact that she loses her Marksmanship bonus if there's enemies close to her, or that Mirana has much better mobility and utility than either of those heroes. If you're played in the #1 farm priority position, you may as well go this route.

My personal build for Mirana however, is as follows.

Phase Boots, Aquila and Urn give her the much needed mobility and sustain she needs in her ganker role. I generally prefer to go the magic burst route with Mirana, rather than physical damage - she's simply too fragile to go up head-to-head with enemy heroes, and she doesn't have any innate scaling on physical DPS anyway, so there's no real advantage to building physical DPS on her. Building a Dagon will augment her already potent burst damage, hopefully killing fragile enemy heroes before they get a chance to use their spells. Once the Dagon is fully upgraded, the natural item progression leads to an Ethereal Blade, which helpfully also provides a good physical damage boost in the form of the the 40 agi stat which will be useful when pushing the enemy base.

Justification for Dagon - some really arbitrary modelling, assuming magic resist = physical resist at level 16. Assume a level 16 Mirana with Phase, Aquila and Urn - she has 106 damage attacking 0.93 times per second for a total of 99 DPS. If she had a Maelstrom, that would boost her DPS to 172 including the chain lightning procs. If she had a Dagon, it would boost her DPS to 112 but she would also have a 400 damage energy burst upfront, which would take 7 seconds of attacking to break even (60 DPS gap between the two builds). It's not often you get to attack the same target uninterrupted for 7 seconds, especially so for heroes that have escape mechanisms. Physical damage scales better than magic damage into the late game, but as they say, if you win the early game, they won't get to play their late game. It all depends what sort of heroes you're facing on the enemy team. Even so, even a Mjollnir compared to a Dagon 3 (for equal cost) requires 5 seconds of uninterrupted auto-attacks to break even with the 600 energy burst. (also, all other physical DPS items are inferior to Maelstrom / Mjollnir as a first item)

An early Midas doesn't go astray either, as Mirana can farm one reasonably safely while still being a contributor in teamfights. She doesn't have any particular synergies so you're free to build whatever you need at the time - Desolator if your team has other minus armor effects, MKB if the enemy has evasion, straight up Skadi if the enemy has easily kiteable melee carries (like Naix), Deadalus for pure damage, Butterfly if the enemy threat is primarily physical damage right click.

Wednesday, June 26, 2013

Short Story Design


Today's blog will be a break from game design and instead talk about short story design.

A cinematic short, or a trailer, is an art form of its own. Two minutes is the standard length, like how songs are around three minutes - any longer, or any shorter, and it feels awkward.

There is a truism that short stories are harder to write than novels, because you have to accomplish the same thing in 10% of the words. The same is true of cinematic shorts- you have to tell an entire story - have a beginning, middle, climax, and resolution - in a two minute sequence.

I will use the Starcraft 2 Trailer to talk about the structure of trailers, and later compare the trailers from DOTA2, LOL and The Dark Knight Rises.

Tools of the trade

Film-makers have two very strong tools to manipulate the emotional state of the audience - the music track, and speed of the cuts (scene transitions). Slow and calm music puts the audience in a relaxed and pliable state - fast and dynamic music builds a sense of excitement. Slower cuts and transitions, versus faster harder cuts, achieve the same effect.

I will describe the four parts of a story or narrative, and how they're each given completely different music tracks and cuts to achieve this.

1) The Setup

Starcraft 2 trailer setup: Jim in depression over Kerrigan

We always start by introducing the characters and world to the audience. The tone and pace is relaxed - you want to ease them in. You need some sense of normalcy in order for the audience to connect emotionally - it's a narrative tool called anchoring - where you connect the audience to the starting point of the movie, and then use that starting to bring them to wherever you need the story to go. This is not necessarily a happy or peaceful setting - it's just setting up the initial state of the world, it could be a dystopia or a person in depression.

You use simple, raw emotions, common situations everyone can relate to. This isn't where you get artsy, this isn't where you puzzle the audience by making them ask questions, or reveal some mystery. If the audience is confused, they can't "sync" with the story and build that emotional attachment. If the audience isn't "synced" whatever conflict or tension you introduce later - danger, zombies, disasters - will have no effect, since they were never on board with you in the first place.

The music here is slow paced, maybe piano tracks and haunting strings. Cuts are slow and low energy.

2) The Conflict

Starcraft 2 trailer conflict: Jim joins the rebellion to get revenge on Mengsk

Now that you've built up the sense of normalcy and invited the viewer to share in it, it's time to introduce the conflict, sometimes with a bang, sometimes with a line of foreboding dialogue. This is where you disrupt the feeling of normalcy you just built in the first 30 seconds. You should reveal conflict, but keep it ambiguous. Hint at larger schemes within schemes. Mystery keeps the viewer interested, hungry for more. This sequence is kicked off here when Jim resolves to join the rebellion and get his revenge on Mengsk - he picks up his badge and holsters his weapons.

The music here changes quite clearly. It's common to use faster tempo strings to build tension and anticipation.

3) The Climax 

Starcraft 2 trailer climax: Jim fights open war with Mengsk and the Zerg

At its heart, a story takes the audience on an emotional journey. You've shown them normalcy, and you've broken it by introducing conflict and creating tension. Now you escalate the tension into the full blown exhilaration of some kind of hinted resolution - usually a massive showdown or war. (In romantic trailers this is where the couples are running into each others arms or kissing). This is where your most awesome scenes and set pieces are revealed - characters at their most powerful, emotions at their highest, and the music built to a strong climax. In the Starcraft 2 trailer, the climax sequence starts from the point Mengsk tells Jim "You're in way over your head," and Jim replies "We'll see about that."

The music here changes quite clearly. It's common to add choral elements (very exciting!) and increase the amount of drums and cymbals. The cuts get even faster and more frenetic as the music draws to a climax - see how fast the cuts are right at the end, barely half a second each.

4) The Stinger

Starcraft 2 trailer stinger: Jim reloading his sniper gun after sniping a hydralisk

After the climax and fade to black, it's common to add an "extra" moment or scene as a resolution. The stinger here is Jim reloading his sniper gun after killing a Hydralisk and letting the heavy bullet casing slam into the ground. In a subtle way it signifies the "end" of the fight, mirroring the feeling of the "end" of the trailer. Having a stinger as a book-end makes the short cinematic feel complete - it's strange to end the cinematic on the climax - there needs to be a resolution or release to the scene, either with humour or finality.

These are all completely normal narrative techniques, common to all media - novels, short stories, songs, even games - just compressed down to a 2 minute cinematic short.

Again, DOTA2 vs LOL...

Now that you've gotten an overview of short story composition, watch the latest cinematics of both games, and compare them. Neither game has any kind of story element whatsoever - how will their makers design a two minute cinematic short?

DOTA2 BETA launch cinematic "What does a hero truly need?"

League of Legends cinematic: "A Twist of Fate".

DOTA2 BETA Launch cinematic

The setup - you see a blacksmith of some sort, in his shop, working on his wares, and he's talking directly to you. You've presumably asked him for advice, and he's saying, ah, I get this question all the time... and clearly he's trying to sell you something. He talks about different heroes having different needs. Immediately this establishes a whole bunch of things.

1) The player - you - are a hero. And this is an RPG, where choice matters.
2) The game involves buying items to become stronger.
3) There are three types of heroes - those that rely on Speed, on Strength, and on Wisdom.

The conflict - the music immediately changes (you hear the faster tempo strings?). The first scene shows the Drow Ranger - a hero - possibly you! - fighting some creeps. The tension is increased in the second scene, where Axe fights the Bloodseeker. This establishes that there is going to be PVP.

The shopkeeper's narrative keeps you firmly anchored in the present (in his shop) throughout the trailer, making it clear the montages of battles you're being shown are possible futures, or maybe tales of past valor. The cinematic climax sequence starts when the camera zooms out and you see the entire field of battle, the music ramps up further with choral elements, you see armies battling across a river crossing. The knight jumps off a cliff, transforms into a dragon and soars into the sky, locking claws with the green drake and they fall to earth - boom! - then fade to black.

The shopkeeper drags you back to the present, and then delivers the stinger - "and that, is for you to decide..." before ending on the title screen.

The trailer is made entirely from in-game art assets - all the hero models are just eye level shots of the actual DOTA2 characters, so by watching the trailer, you can get a good idea of the art style and values the game will have. I guess it would have cost very little to make. The graphics are pretty mediocre (you can see how inferior it is compared to the proper CGI trailers before and after this even in these tiny screenshots) but it gets the job done.

LOL Cinematic - A Twist of Fate

A big caveat here - Riot have explicitly not called this cinematic a trailer. It's a marketing cinematic, perhaps, and a thank you to their fans, made to bring the game to life. But yet the fans - and also people who have never played LOL before - haven't had the most positive reaction to it - and they should, given the sheer quality of the CG and animation - and I've been trying to analyze why.

There is no anchoring scene at the start. Immediately you see Twisted Fate with his playing cards, waiting at a grave yard. He looks at a city in the distance, then fights a supernatural scarecrow (Fiddlesticks). Very cool and amazing fight scene.

This establishes very little, and only raises questions and confuses the audience, exactly what you don't want to do at the start. Did they arrange for a duel? Or was he ambushed? What is that city? Is he trying to protect it, or attack it? Wait, it doesn't matter, because we don't see Twisted Fate, or Fiddlesticks, or that city again in this cinematic.

The cut to the second fight scene - and the third - is extremely jarring. There is no sense of anchoring or continuity at all - not in location (totally different environment) not in time (it goes from night to day instantly) and not in faction (they don't seem like they're allied or connected in any way). There's no narrative to tie them together - it's just sequence of fight scenes devoid of context or meaning. By the third fight scene most people are getting bored, feeling - when is something going to happen? When is the story going to start?

There is no gradual build up of tension or pacing - the story doesn't move the audience emotionally from one place to another. All fight scenes are equally fast paced - arguably, the first fight scene was the strongest and most dynamic, and the last fight scene the least (Annie, summoning the fiery teddy bear). There isn't even a climax sequence - Baron Nashor gets revealed at the end, but they don't actually fight it. He roars at the champions and the screen fades to black. The ending felt very weak for multiple reasons - lack of proper climax sequence, music score was weak (it does build up, but not as effectively as the other examples on this article), at least partially because the opening was so action packed that it didn't allow any possibility of "build up" and also, because it lacked a stinger - there was no feeling of resolution.

This is akin to a song, which has no soft, or loud parts. No verse, chorus, or bridge, to create contrast. No real beginning or ending. It's just maximum volume all the way.

This cinematic is more than twice as long as the DOTA2 trailer, yet by the end of it, someone new to the game still isn't quite sure what sort of game this is. It could be an MMO - a raid boss appears at the end - this is quite a common format for an MMO cinematic. It could be something like Street Fighter or Soul Calibur - we see a LOT of individual 1v1 duels, but nothing indicating this is a full 5v5 team combat game. The cinematic is fully CGI with no resemblance to the art assets or even art style within the game. Dodging falling rocks isn't part of the game in any way (Ryze, lol).

The LOL cinematic displayed some spectacular fight choreography and great graphics, but the design and narrative aspect of it was very disappointing. It also breaks the rule on cinematic length - people get fidgety and start feeling "it's going on too long" at about exactly the two minute mark. Even if it's designed "only for the fans" and not for newcomers, breaking virtually every rule of storytelling and narrative is a bad idea, and it really shows. I'm not actually sure fans (such as me) watching it are getting a much better experience than newcomers.

Movie Trailers

Movie trailers also follow the same rules. The Dark Knight Rises trailer is a good one.

This trailer makes exaggerated use of soundtrack to achieve the transition changes.

The setup (1) is drawn out, a shot of the city and a dispirited looking Bruce Wayne, with a wistful piano track. The tension is hinted at in the words "A storm is coming, Mr. Wayne." The trailer continues with scenes of destruction and hopelessness. Then (2) the coming conflict is hinted at - the kid asks "Do you think he's coming back?" and gets the reply "I don't know." The question is significant, and the music track changes here to include strings (building tension) and the trailer moves into the territory of Batman rising again. The trailer moves into the climax sequence (3) when Catwoman says "You've given them everything." and Bruce replies "Not everything. Not yet." The music score now incorporates choral elements, and the cuts get faster and faster. After the climax and title screen, the movie delivers the stinger (4) where Catwoman says "My mother warned me not to get into cars with strange men," and Batman replies "This isn't a car."

Tuesday, June 25, 2013

Playing Better (Part 1) Dual Purpose Ward Spots


This will be a series of articles covering a very, very broad topic in DOTA2 - how to be a better player. I didn't really want to talk about this topic (as there are a million and one guides online) but I now see this as more of a "brain dumping" exercise where I can store some basic knowledge off-site in the cloud for new and intermediate players, so I don't have to explain myself to people and just link them to these posts.

Dual Purpose Ward Spots

It's a simple concept. You put a ward in the place you want to get vision. Yet we could probably write pages and pages of articles on warding strategies, best spots, how to counter ward, how to avoid being counter warded, etc

I'm going to keep this simple - here are the top 12 most important ward spots - 10 of them are notable because they perform dual functions very early in the game, and the last two are the well known "super ward" locations that give huge 360 degree vision useful in mid to late game.

Oh, this guide is presented by "Princess and the Frog".

1. Block Dire Pull + Tower Vision

The Radiant top lane will use this ward spot to block the Dire Pull. This ward also gives you vision behind their tower. This is an aggressive ward that will allow you to dive their tower more safely.

A common mistake is to place the ward too low (right inside the camp) so its vision gets blocked by trees and doesn't give useful vision of the back of the tower or either side of the camp.

2. Block Dire Pull + River Gank Vision

The Radiant top lane will use this ward spot to block the Dire Pull. This ward also gives you vision of the gank path coming from the river. This is a defensive ward that will allow you lane more safely, and also makes it safer to pull the hard camp to your lane creeps if you are gaining the advantage in lane. Additionally this is less likely to be de-warded, as not many players know how to place this ward consistently.

The ward placement on this has to be very precise so I have included two pictures. In the first picture you can see a patch of grass between two trees (to the left and slightly higher relative to the burning torch). You want to play the ward as far to the right as possible on that grass patch. You'll start by trying to place the ward right up against the tree on the right, getting a bunch of "invalid ward placement errors" and moving your mouse slowly to the left and continuing to try place it until it allows you to drop it.

The common mistake is in thinking you're trying to place the ward as far to the top, or top right as possible. You're actually trying to place the ward as far to the right as possible. I can do this consistently now and it may be worth your time setting up a bunch of custom lobbies to experiment getting it right.

3. Block Radiant Pull + Block Radiant Hard Camp

The Dire bot lane will use this spot to block the Radiant Pull (bottom right). Because it can be a bit tricky to place, I've included two pictures here - before and after. In the before picture, you can see that there is a bush next to the trees, in the after picture, you can see I've planted a ward directly on it.

This ward spot will also block the Radiant Hard Camp (top left) and provide huge useful vision.

Some players block the Radiant pull by putting the ward directly in front of the Radiant Pull camp (bottom right), similar to the Dire Pull camp ward I talked about earlier. However the tree placement here means that putting the ward in the camp doesn't give vision behind the tower, making it significantly less useful - but it can still give some vision of the jungle area around the pull camp if that's what you're after and the enemy has no jungler.

4. Bottom Rune + Block Radiant Medium Camp

The Dire bot lane uses this ward spot to get vision of the rune and gank path. This ward spot also blocks the Radiant Medium Camp.

5. Bottom Rune + Gank Protection

Both sides can use this spot to get vision of the rune. This ward spot also reveals the entire area of the map beside the lane, protecting you from ganks.

6. Bottom Rune + Entry to Radiant Jungle

Both sides can use this spot to get vision of the rune. This ward spot also reveals the two key entrances into the Radiant jungle, and may detect junglers moving from the hard camp below it to the easy camp off to the left.

7. Bottom Rune + Block Radiant Hard Camp

The Dire bot lane uses this ward spot to get vision of the rune and gank path. This ward spot also blocks the Radiant Hard Camp. This also gives vision of Radiant jungler movement and possible path for Radiant mid moving bot through the forest instead of river to gank. Not the most useful spot but it is also less likely to be de-warded as well.

8. Bottom Rune + Roshan + Radiant Medium Camp

Both sides can use this spot to get vision of the rune. This will also give vision of both the Roshan entry point and the Radiant Medium Camp. A common mistake is to place the ward too high - placing it slightly lower, as shown here, will give vision of the medium camp as well without compromising vision of the Roshan entrance.

It's also common for players to place the rune watching ward on the high ground to the left of the rune. The 4 other bottom rune ward spots I have detailed here have more utility.

Note that the Dire can use this single ward spot to protect themselves as they kill Roshan, but the Radiant will need to spend two wards to protect themselves - there are two separate paths from the Dire side of the map that lead to Roshan.

9. Top Rune + Radiant ancient camp

Both sides can use this spot to get vision of the rune. This will also give vision of the Radiant Ancient Camp.

10. Top Rune + Dire Medium Camp

Both sides can use this spot to get vision of the rune. This will also give vision of the Dire Medium Camp.

A common mistake is to place the ward too low - placing it higher, as shown here, will also give you vision of the approach to the medium camp as well. This ward will also reveal movement between the medium camp and the left-most hard camp, exposing the enemy jungler to possible ganks.

11. Dire Super Ward Spot

Both sides use this spot to give them a commanding view of the Dire top lane approach and jungle, usually in mid to late game.

I wouldn't really call it a mistake, but players generally place the ward right in the center of the painted symbol - by placing it on the extreme left - as shown - the ward will also give you vision of the easy camp. It's a minor improvement, to be sure, but it costs you almost nothing to do.

12. Radiant Super Ward Spot

Both sides use this spot to give them a commanding view of the Radiant bottom lane approach and jungle, usually in mid to late game.

Asymmetrical map - Radiant or Dire advantage?

As you will have noticed, the map is very different on both sides. There are many apparent advantages for the Dire from a warding perspective - their ancient camp is difficult to ward so it's safer for them to stack and farm, it's safer for them to kill Roshan and they can block the Radiant jungle camps much more effectively. In fact, Dire can even secretly access Roshan without revealing themselves to river wards by using mobility skills and items to hop over the cliff on their side of the map. In competitive games, Dire consistently scores 50% more Roshan kills than the Radiant, netting them a large advantage.

Yet on average both Dire and Radiant have almost the same win rate. To compensate for the Dire advantages, Radiant have other laning advantages - their jungle layout allows their jungler to gank more aggressively, their pull camp also allows a secondary pull into the medium camp, and their jungle, midlane and safelane are considered more secure. But this will be the subject of another blog post.

Warding in a nutshell

Don't blindly place wards just because it's a "good spot". Wards are precious - always place them to achieve a specific goal.

Place wards where you anticipate the fight is going to be. If your team is being aggressive - a gank or push oriented lineup - then placing aggressive wards is a good idea - putting them in the enemy jungle, behind their towers. This will greatly assist your team's pushes and ganks by giving you critical vision over the battleground and preventing counter-ganks. If your team is being defensive - a turtle line up, with late game hard carries - then you need to place wards defensively, to create safe spaces for your carries to farm.

If you're not aware of the state of the game, you could waste wards providing vision in areas no one intends to visit! For example if your team was playing defensively, having a ward in the enemy jungle doesn't do any good - in the first place, the enemy wouldn't be spending time jungling, they would be trying to push your towers and gank you instead - and even if your ward spotted an enemy, your team is presumably too weak to even consider going across and invading their jungle.

Besides these 12 spots, there are about 40 more "common" ward spots - they basically boil down to deciding that you need vision at a particular spot, and planting a ward there to achieve it, making sure it's out of range of enemy tower detection. They're all very single minded in use, and not particularly interesting to talk about, but important nonetheless. Don't be afraid to buy wards, regardless of what hero or role you're playing - they cost very little, provide a lot of benefit, and as the person who places them, it allows you to personally control the flow the game, creating safe zones for yourself or exposing enemy territory to attack.

Monday, June 24, 2013

Competitive Roundup (Part 3)


There are so many tournaments going on at the same time that it can get overwhelming trying to keep up with them. In this series of blog posts, I will share matches which I have found notable and worth watching in the past week or two.

Dreamhack Finals "Perfect Game"


Tournament:  Dreamhack Finals Game 3
Teams: Quantic versus Alliance
Stream: ???

I almost could not obtain this match for you guys in time for Monday - best game of the whole of last week, I watched it live, but again there's no official uploader. Thanks to Hexor who must have bought the ticket to the game, recorded and uploaded it to his youtube channel.

Quantic are probably the best drafters in DOTA2, and their lineup looked a lot better than the one Alliance picked. In fact, Alliance's heroes were confusing - they picked Puck, Nature's Prophet, Chen, Alchemist and Wisp - a team with very little damage capability, which should in theory lose in early game, mid game, and late game to the lineup that Quantic picked - Undying, Warlock, Batrider, Shadow Demon, Gyrocopter. I mean just look at that teamfight potential, AOE, and lockdown.

Despite the casters feeling the game should be a certain loss for Alliance, and despite losing in CS across every lane, Alliance show that they're in the DH Finals for a reason. A couple of flawless teamfights later, they close out the game 22-0. As Ayasee said, we may never see a game like this again.

Quantic draft Naga support

Tournament: Dreamhack Group D
Teams: Alliance versus Quantic
Stream: LuminousInverse

This match happened two weeks ago but the Dreamhack group stage was not casted or uploaded as far as I know. This is a delayed replay cast, with Luminous and EternalEnvy.

Alliance is one of the strongest teams, winning both G1 LAN and Dreamhack. Quantic Gaming, also known as DD, came runners up in the TI3 Western Qualifier, and are generally regarded as the the team with the best drafting, coming up with innovative team compositions and many times winning on the strength of their draft alone. This game is one of them - Quantic run a very interesting pushing team with a Naga Siren support - a hero previously used as a carry - and overrun Alliance's mid rax by 20 minutes.

Alliance baffles Chinese teams


Tournament: G1 LAN Finals Game 2
Teams: LGD vs Alliance
Stream: Beyond The Summit

This happened quite a long time ago, but I wanted to highlight it anyway.

LGD is considered one of the top teams in the world, with a very strong showing at the TI2 tournament last year. An interesting fact (to me at least) is that LGD is sponsored by a chili sauce company of the same name, Guizhou Laogandie Food. Generally the Western teams get sponsored by companies at least tangentially related to gaming in some way...

The Chinese teams in the G1-LAN were unprepared for the playstyle that Alliance brought with them. Chuan, leader of IG, said in an interview that he knew they were in trouble when he was watching the Alliance draft and did not understand their strategy at all. Their playstyle leans towards "Tri-Core" where they split their farm up among 3 heroes who all get strong, while Chinese playstyle leans towards "4 protect 1" where their entire team and laning setup is designed to guarantee farm on one hero. Well, and part of the problem was that they didn't respect Admiral Bulldog's Lone Druid enough to ban it - he is well known for having an unhealthy obsession with playing the Lone Druid, which he is admittedly very good at.

LGD's strategies were ineffective, they try every trick in the book but they never really get any traction in the game and ultimately go down to Alliance. Nice casting by Luminous and LD, many memorable lines in this one. "If your Gyro is 0-5-2, is that not feed?" "He's... had a lot of pressure. Let's be fair."

Friday, June 21, 2013

Let's Play (Part 3) - Juggernaut


Every week or two I generally try to play a new hero to discover what it’s like playing it, and in this series of blogs I will make a post on my experiences with it.

Release artwork for the Juggernaut


The Juggernaut is based off the character Grom Hellscream in Warcraft 3, a legendary Orc Blade Master and Chieftain of the Warsong Clan. Grom is remembered for both his failings and his valor - he was the first one to drink of the demon Mannaroth's blood, allowing himself and his clan to fall under the Burning Legion's control. Yet Grom ultimately redeemed himself by wresting his sanity back through sheer force of will, attacking and killing Mannaroth at the cost of his life, freeing the Orcs from the Burning Legion once and for all.

In DOTA2, the Juggernaut is a wily melee combatant with the ability to deal massive amounts of damage while avoiding retaliation. He can also heal his team, allowing them to mount a sustained assault on the enemy base or weather a prolonged siege.

Juggernaut is a classic hero that hasn't seen much change over the years. He currently has a 49% win ratio, yet is the 8th most picked hero (out of 101). He's generally seen as a fun to play and popular hero, yet well balanced.

This is probably my most played hero, currently standing at 33 wins and 17 losses, one of the best win rates I have on this account.


Highly lethal early and mid game. He is the quintessential snowball hero, capitalizing on his ability to score early kills to build an unstoppable advantage by mid-game.

Difficult to pin down and kill, with Bladefury giving him magic immunity and Omnislash giving him invulnerability nearly the entire fight.


Falls off in strength by lategame because his spells don't scale as well as other carries.

Most of his power is in his Omnislash, which can be counter-picked against or neutralized lategame with the correct items.


An interesting quirk of this skill is that the Juggernaut can land normal autoattacks during Bladefury - on targets that don't get affected by it. You can auto-attack things like buildings (towers, barracks) and siege engines while spinning next to them. You can't auto-attack magic immune heroes, though, since the heroes are still considered a valid target for Bladefury, it's just the damage they take is reduced to zero.

Healing Ward

Many people have wondered, why does an assassin type hero have a team healing ability? Isn't this completely counter to the entire concept of the Juggernaut?

I'll make a slight detour here to overall team strategy:

A common game mechanic between asymmetrical teams is that one will be the "Aggressor" and one will be the "Defender". The Defender has the better late game, they try to stall the game until they reach the point they can win. The Aggressor has the better early game, they need to try to win early. These roles are mutable and are only defined in relation to each other - a team with 1 hard carry might be the "Defender" when playing against a team with no hard carries, yet they might be the "Aggressor" when they play against a team with 2 hard carries.

Having a Juggernaut tends to put your team in the "Aggressor" role since the Juggernaut takes up farm yet isn't a hard carry himself - his power falls off late. If you carry out a successful gank and don't manage to take a tower, you may actually be putting yourselves further behind - since you've commited extra heroes to that gank, while the enemy hard carry gets free-farm in another lane, and the game isn't any closer to ending. This is the reason why the Juggernaut has the Healing Ward as a skill - it allows your gank team to heal up and immediately push the tower in the lane they just ganked.

As you can see the Healing Ward is a key ability on the Juggernaut and should be maxed after Bladefury. Practice placing the ward out of reach of enemies during teamfights, either by anticipating their movement or by hiding it in the trees or on high ground near the team clash. If you know the enemy has big initiators (Enigma, Magnus etc with a Blink Dagger) you can preemptively place it a bit further back then move it up so it heals your team through the initiation stuns.


This is a very simple skill, but a little forethought can go a long way towards maximizing its use.

1) There's nothing worse than seeing the enemy alone, starting your Omnislash, then having enemy creeps arrive from the fog and absorb your attacks. If you're ever not sure if there are enemy creeps approaching from the fog, check your own creeps in lane - their movement exactly mirrors the enemy creeps, so you can always know where the enemy creeps are.

2) You need to keep track of what skills your enemies have which avoid your Omnislash. Get into the habit of also checking the enemy inventory before you gank them - smart opponents will buy items such as Shadow Blade, Ghost Scepter, Force Staff, or Manta Style which can nullify your Omnislash. Bait them into using those items prematurely, then kill them.


Juggernaut synergizes well with heroes with disables and AOE nukes, as they allow him to land his Bladefury and Omnislash. He typically does well in an aggressive push oriented lineup.


There are many heroes which have abilities that negate your Bladefury and Omnislash. The full list is extensive, but generally involve abilities that grant invisibility, mobility, illusions / summons or invulnerability. Playing against them isn't impossible, but is definitely more difficult.

Item build

As a snowballing ganker, you want move speed and regeneration so you can gank continuously, and level advantage so you can continue killing your opponents. He is reliant on his Bladefury and Omnislash combination to kill. This is why it's vital he doesn't fall behind - if the enemy ekes out enough of a level and item advantage that they no longer die to his combo, the Juggernaut basically becomes useless for the rest of the game.

1) Movement and stats - Juggernaut needs the mobility provided by Phase Boots to position himself to use his Bladefury and Omnislash. He typically needs some Ironwood Branches at the start, because if he does not take any points in stats, he won't have enough mana pool at 6 to use Bladefury + Omnislash combo. Having some branches at the start allows him to take one level of Healing Ward instead of having to put two points in stats, and you can convert the branches into a Magic Wand, or just sell them. Key point on Phase Boots - the Phase movespeed effect cancels if you use a skill, so if you use Phase and then Bladefury it will dispel the Phase. The correct sequence is to use Bladefury first, then Phase, so you can get the advantage of extra movespeed and no collision, but that will not always be possible.


2) Lane sustain - As the Juggernaut is a snowballing carry, he needs regeneration in order to continuously threaten and kill his opponents. In fact you should build the regeneration portions of the items here even before finishing your Phase Boots. There are two possible choices, either a Battlefury or a Vladmir's Offering paired with an Urn of Shadows. The Battlefury is the more passive option, as it takes much longer to farm, and doesn't help his ganking or killing power - it primarily helps accelerate his farming. This is a dubious choice, as he finds it difficult to compete with true late-game carries anyway. The Vladmir's Offering and Urn combination is more aggressive, as it comes online earlier, and the Urn offers an extra source of nuke damage to finish opponents off if your combo doesn't kill them. Soul Ring and Medallion of Courage are notable alternatives, as the Juggernaut generally doesn't fear retaliation, but their following build up can be awkward.

3) By mid-game, realistically, your game should be more or less decided. Either you've secured a large advantage, or you haven't. From this point your hero power peak ends. The Juggernaut doesn't have the mobility or disables to dish out his substantial click damage effectively. As hero health pools and autoattack damage increase, Bladefury becomes obsolete, becoming a net DPS loss for you when used, leaving your Omnislash and Healing Wward as your main teamfight contribution. With that in mind the Aghanim's Scepter is a critical item for the Juggernaut, giving a big boost to your ultimate, adding 50% to its damage at level 2 (6 to 9 hits) as well as reducing its cooldown, and giving you a much bigger health and mana pool to freely use your Healing Ward. It depends on the pace of the game and enemy hero composition, but sometime it may be better to skip the Aghanim's and go straight for late game items.

 Realistically it doesn't matter too much what items you buy now, but in general, here are some good builds.

4) Vladmir's Offering and Urn into a "Push" build - the idea is that you build a Desolator and Assault Cuirass to stack armor debuff effects in order to win the game early with stronger pushes. You get the DPS trifecta in this build (Lifesteal from Vladmirs, Damage from Desolator, Attack speed from Assault Cuirass)

5) Battlefury into a "Dueller" build - with the accelerated farm you build a Butterfly, Helm of the Dominator and Skull Basher. You also get the DPS trifecta in this build (Lifesteal from Helm, Damage from Battlefury, Attack Speed from Butterfly) and you further get a hero lockdown in the form of the Skull Basher, which extends into an Abyssal Blade.