DOTA2 roles & the early game

Ronin Storm

Administrator
Staff member
After a bit of a losing streak in DOTA2, I've been doing my research into what went wrong and why. It's difficult, with pub matches, to control the outcomes when your team is so fragmented and can range from pro to clueless in the same game. However, I believe that there are themes that we can use to help improve our games and make even the losing ones seem like we were making more of a difference.

For clarity, I'm really only interested in the impact in lower level pub matches. Pros will do it differently, I'm sure, but they've the experience to make that call. For the rest of us...

First theme I'd like to look at is the concept of roles.

When you look at the heroes you can choose from for a match, there's the little icons giving an idea of some of the things they're good at and, if you mouse over the icon, a set of stars (1 to 3, IIRC) that indicate how good at that they are. The icons give an idea of the capabilities of that hero, but how they actually work is down to your choice of role.

I think it was this blog post from Team Dignitas that set me off in this direction. There's plenty of ways to configure for lanes, but this one made sense to me and segued into roles well.

So, what roles does your team need?

You always need a Carry. Why? Because you need someone who is going to get so hard that the become the lynchpin for your team and, in some cases, will be unstoppable even if the majority of the enemy team pile on them in a teamfight. The Carry heroes can handle this well because they scale well to the end game.

Carry heroes come in (at least) two flavours: Hard Carry and Semi Carry. The Hard Carry is the late game powerhouse and properly farmed and levelled they are basically unstoppable (e.g. Faceless Void). The Semi Carry is a lighter weight version of the Carry, but in some cases they reach a good peak earlier in the game, which can be useful for driving a quick victory (e.g. Juggernaut).

Your team only needs one Hard Carry. Why? Because feeding gold and experience to more than one slows the whole team down, potentially allowing the other team to level their Carry faster and thus win.

"But, but, but, surely a team of five Carry heroes will be awesome. I see that happen in pub games all the time and they pwn our carry support characters."

That's taking the wrong view of it. If your team has let five opposing Carry heroes farm to effectiveness then you're doing it wrong. I'll take one fully farmed Carry over five partially farmed Carries any day, because the one with full farm and levels will eat the others alive (with a little handy support).

But, crucially, a mistake there is that your other team roles haven't been succeeding at their jobs if that's happened.

So, what other roles?

Well, every Carry needs a babysitter for the early (and possibly middle) portions of the game. That's where a Support hero comes in.

The job of the Support hero is to ensure that the Carry farms in safety. This can be done in a number of ways and often in some combination. The Support's job includes:

  • Harrassing the heroes opposite the Carry so that they can't interfer with farming.
  • Warding the approaches to the Carry's farm location so it's harder to gank the Carry.
  • Healing or buffing the Carry so they can farm for longer.
  • Denying friendly creeps to ensure that the lane isn't pushed into a dangerous farming area (i.e. nearer the opposing tower).
  • Stacking and/or Pulling neutral creeps into your creep wave to pull the lane and thus create an even safer place for your Carry to farm, while denying farm to the other team.
  • If necessary, dying in place of the Carry so they can continue to farm.

There's probably more than that, even, including things like controlling runes and setting up ganks as part of their harrassing job, but all of it is focused on the singular job of keeping the Carry in lane, last hitting, and picking up experience.

Now, done perfectly, the Carry will be last hitting every enemy creep and the Support will be denying every friendly creep in a sort of synchronous motion that keeps the creep wave near (but probably not quite in range of) your tower for maximum farm potential. Wouldn't this mean that the Support character never gets any farm (aside from the 1 gold per second that everyone gets)?

In short, yes, but that's still successful as for many Support heroes farm isn't entirely necessary to do their jobs. But practically, there's always going to be a small subset of last hits that the Carry can't realistically get and they're fair game for the Support hero that's babysitting them. The rest, though, need to all go to the Carry. Gold equals items, and items equals superpowers, and that's where you need your Carry to be.

So, great, that's the Carry and the Support. What about the other three heroes?

Their jobs are also pretty varied and will change depending on exactly who you're facing, but broadly there's another two major roles to be considered:

The most obvious is the Ganker: a hero with powers to quickly stun or disable opponents and lay in quick, punishing damage with the aim to kill them or, at the very least, nearly kill them. Killing them is best, of course, because they lose gold, and you gain gold and experience, but nearly killing them is okay because it pushes them out of lane, thus stops them farming (which will kill them in the end). An obvious target is the opposing Carry but they're far from the only target and, indeed, any successful gank (one where you get a kill and don't die) is still a win for your team.

The slightly less obvious is what I call the Lane Controller: a hero with powers to help slow or stop the opposing Carry from farming. To understand what needs to be done here, consider what the opposing team is trying to do: they're trying to give farm to their Carry. So what' the Lane Controller's job? To deny that farm. That can be done in a few ways, of course:

  • Deny creeps; this both denies experience and gold, plus it pulls the lane into more dangerous territory for the opposing Carry.
  • Stack and/or pull your own neutral camps; effectively another deny, though it tends to take a bit more work to achieve including cutting down some trees.
  • Harrass the Carry and/or babysitting Support. If the Carry is afraid to come to the line to last hit, you're winning. If the Carry runs out of regen, your hits "stick" and will eventually force the Carry out of lane. Same applies for the Support, but it's more indirect and the ultimate target is still the Carry.
  • Set up ganks on the Carry and/or Support.
  • Block the opposing neutral creep camp to prevent stacking and pulling with observer wards, usually, which often don't actually need to be in the camp, but tactically near the camp.

I'm almost of the opinion that, in the early game, the complex tasks need to be carried out by the Support and Lane Controllers and the Carry has just one job: stay in lane and last hit everything. That feels quite contrary to the more usual "oh, n00bs play support" and I think that may be why pub matches go south so often.

So, how does this actually apply to the early game?

It's about laning, specifically the laning phase.

You know about the safe lane, mid lane, and suicide lane (the earlier Dignitas laning articles uses "short lane" for safe and "long lane" for suicide)? For Radiant, that's bot, mid and top, respectively. For Dire, that's top, mid and bot.

You also know that it's pretty common for lanes to be handled in a 2-1-2 format, right? But why?

It's about splitting the roles you've taken appropriate to the needs of the team. DOTA2 is more akin to Team Fortress 2 or American Football in that all characters or positions are not equal and when played correctly (i.e. to role) in cooperation with the rest of the team then it's a matter of who's better, but when played badly (i.e. uncoordinated, in conflict, not to role) then it's a matter of who screws up least. I think you'd agree it's more fun to lose a game where you always felt you were in with a chance but the other team just had the edge than to win a game where it was just a long sequence of screw ups but somehow you didn't screw the pooch as badly as the other team.

So, 2-1-2 is about roles. Think of it this way:

Safe Lane: 2 heroes, Carry + Support
Mid Lane: 1 hero, Ganker
Suicide Lane: 2 heroes, 2x Lane Controller

The Carry goes to the safe lane to make his farm easier. The Support goes with the Carry.

The Ganker goes to the mid lane to get levels faster so they can be effective in early game ganks.

The Lane Controllers go to the suicide lane to counter the opposing Carry.

There's plenty of other possibilities:

One of the Lane Controllers could be switched for a Semi Carry or Ganker to try to get an early edge by being aggressive.

The mid solo hero could be switched for a Semi Carry, again to try to get an early edge in power.

One of the sucide lane heroes could be taken out completely to be played as a Jungler (note: suicide lane, not safe lane; the Carry pretty much always needs their Support). Technically this'd be a 1-1-2 pattern.

... and doubtless more options than that once you get truly experienced.

But the point is to ensure your Carry is farmed properly while the opposing team's Carry is under farmed.

So, to go back to an earlier thought, why not 5 Carry heroes?

Experience is split evenly between the heroes on your team that are near the kill. That's true whether they're a Carry or Support. However, last hits only go to one hero. Thus, at best, with 2 Carry heroes in one lane they are getting half the gold each. Assuming roughly even last hitting skill, suddenly the opposing single Carry is getting gold at double rate.

Additionally, Carry heroes know they're supposed to be staying alive. With two Carry heroes in one lane, when the gank comes in on one of them, there's no natural hierarchy and in the second or two it takes to figure out who should be saving who one of you can be down, and then the other one shortly follows. With a Carry plus Support scenario, there's no confusion: the Support dies to save the Carry if that's what it takes (fine, you don't want to be feeding, but if it's an either-or choice, always better the Support dies than the Carry).

Finally, real Hard Carry heroes tend to be weaker earlier in the game because they're item and level dependent, where real Support heroes tend to be stronger earlier in the game because they're often independent of those considerations (sure, they may be crunchy, but they can still be effective). Thus a well played Support can clean the clock on an early game Carry purely because the game balances out that way.

So it's about balance, it's about covering the roles the team needs, and then performing them well.

...

So, yeah, wall-o-text. Happy for comments, especially if I've missed something important or got something wrong. Again, bear in mind that I'm not terribly interested in if it works at pro level because, frankly, I'm not playing them and am not likely to be doing for years to come, if ever. But at lower skill pub games, I think a coordinated team of five who cover their roles well will usually beat the snot out of a similar team who are confused on their roles (even if, character by character, they're not terrible players).

I think I'll be writing more of these (when I get time) to cover some other aspects of the game as I get happier with my understanding of them.
 

Ronin Storm

Administrator
Staff member
Got one following thought:

How does this actually affect what we practically do in a Single Draft pub game?

I think that, very quickly, we need to be able to identify the following:

Who has good Support characters?

Who has a Carry they can play?

Pick one Carry and a Support to babysit them. These heroes will be taking the safe lane.

Then:

From the remaining character picks, what roles can be covered?

Lane Controller? Ganker? Jungler? Semi-Carry?

Pick 1-2 Lane Controllers, 1-2 Gankers, no more than 1 Jungler, and no more than 1 Semi-Carry, in that order. If you're taking a Jungler, make sure that the Lane Controller you leave in the suicide lane has a good escape power otherwise they'll just feed. If not, no Jungler.

The Lane Controllers go to the suicide lane. The Ganker or possibly the Semi-Carry goes to the mid lane. And if there's a Jungler, he takes the Jungle next to the Carry + Support.

I agree that this is much harder when playing just a duo in a pub match because usually the other guys on your team either have other ideas or are clueless and this leads to role conflict. However, if we can dominate a team (4 of us, ideally 5) then we take control of our picking and can coordinate properly.
 

Panda with issues...

Well-Known Member
Agree with much of what has been said.

Will add a couple of points that I think make sense, though obviously I've not played as much as you.

I think if we're playing single draft, then the roles on the team are going to be decided partially on who's best (or least worst) at what, but far more strongly by the character selection everyone gets, as you've pointed out.

I think this also feeds in to the skills taken at a basic level.

If you know you're going to be on your own in lane, whether that's because someone is jungling, or because you're in the midlane, I think you need to:

a) take a character with an escape power (I know, I know, this requires being organised in the character select screen, more than we are now. But as we play more, and have a greater roster of comfortable characters, the panic will end and we can be more sensible about this and think about roles based on what we have to choose from, rather than what we chose. A subtle but important distinction I think, and we're beginning to sort of do this in lobby).

b) select that as your first skill. Escape powers have different flavours. The traditional ones are the ones that give you a speed boost, or invisibility, but others can be used too, though aren't necessarily as effective. - Slows and stuns can buy the time for an escape, if you can get them off.

I also consider solo mid to probably require some sort of template ability, which can keep you from being overwhelmed (others probably call this AoE or crowd control or something, but coming from a tabletop background, they're always be template attacks to me), and as far as I've got, it seems a lot easier in the midlane as a ranged hero. The escape isn't QUITE as important in the mid, as the towers are closer, but I think it's still worth thinking about.

I think that we (assuming 5 THNers on TS, an ideal scenario) need to follow a kind of following flowchart:

1. As an individual player, can you play all the 3 characters you have access to? - If so, good, if not, immediately discard the one(s) you can't. This is something you can do personally without bothering the rest of the team. If you can't play any of them, well, that's a bit of a problem, and we can really come back to it later.

2.If anyone only has one character they can play as, they need to speak up, as they are essentially locked in to that role. Similarly, if someone only has a choice of support characters they can play, or carries, or junglers, they also need to say at this point as it will affect the next step.

3. Next, rather than all shouting at once, or going through each person's list, I propose going through by role. It seems sensible to start with the carry (hard if possible).

Someone in the team needs to step up and take control of this kind of discussion really, and just say 'right, what are our carry options?' - We then call them out, and chose the most appropriate. This player is then 'done' and might as well keep quiet. Then we go through by role, presumably finding a support for the carry (preferably ranged if they are melee, or vice versa, though almost all the supports seem to be ranged).

Next we try to ID a capable character for solo midlane. Ideally a decent ganker with some sort of template and an escape. - Obvious ones seem to be things like Queen of Pain, who can blink, and has 2 separate template attacks, Death Prophet, who hasn't really got an escape, but can silence which is kind of helpful in that situation, and also has a template.

Finally we sort out the 'hard' lane, which probably wants a semi carry and another support, or swap out the semi carry for someone that's also good at ganking, or good in a teamfight. Ideally these 2 would have some sort of useful escape abilities or stuns etc.

If someone is going to end up jungling, then different priorities will be required about the hard lane character choice, as they'll be on their own.

4. If someone doesn't know any of the characters they can choose, we have a brief discussion over which is easiest/best for the team.

5. Select skills based on role and lane, prioritising escapes if necessary. (Done in game)




I know we only have 45 seconds or whatever to do this, but I think if we're organised, it's not impossible, or even particularly difficult, especially at the moment where people may only have a real choice of one or 2 characters since the others they don't know.

It may require seceding choice to a sort of pseudo team captain (poss. Ronin) who gets to make a lot of the calls/do the talking. We're playing this as a team game, so should be willing to take a hit for the team in the interests of a better game in the long run. So far, having played with Trax, Ronin and Ghostwolf, people have been very willing to mix it up, and play a less ideal character in order to benefit the team.

Just my thoughts.
 

Ronin Storm

Administrator
Staff member
Okay, great, good reply and I was thinking something somewhat similar in terms of the character choice period. Just to try to get that down to a short, bullet-point list, it might be:

  1. Silently (& quickly) review your characters. Discard ones you don't know.
  2. Those with remaining Carry characters declare, with preference.
  3. Those with remaining Support characters declare, with preference.
  4. Team chooses a Carry and a Support. These will safe lane.
  5. Those with a Ganker or flexible Semi-Carry declare, with preference.
  6. Team chooses a Ganker / Semi-Carry. This will (usually) mid lane.
  7. Team reviews remaining choices and selects based on the following criteria, and these will (usually) suicide lane:
    • Player familiarity
    • Stuns / disables
    • Escape powers
    • Harrass powers
    • Burst damage powers
    • Jungling possibility
    • Counter-picking

The team selection phase is 60 seconds, but I believe you can still pick after that ends and the game is in warm-up (75 additional seconds, I believe).

The final two hero picks aren't quite so template driven. If I were to use an American Football analogy, I think it'd be like deciding how many wide receivers you're having this play, or whether you need a tight end, and I think this as much about play style as it is about figuring the opposing team's play.

As Panda says, after the first step of self-selection, you may end up with no good choices (for you, which is all that matters because you need to be able to play them). I read a good piece of advice that says that if in doubt, pick a character with stuns because you can never have too many stuns.

Also, as Panda says, I think that this gets very much harder when playing with non-THNers. In a 4 plus one random I think it can still be done as we can probably soak one extra Carry or something weird like that. With only 2-3 of us, though, I think it's getting too tricky; we can give it a whirl, but I don't think we should expect that to be game changing.
 

Nanor

Well-Known Member
The points I got to briefly skim over (dinner now) were quite interesting. It'd be good to try and drum up 10 of us for a 5vs5 at some stage.
 

Panda with issues...

Well-Known Member
I think that's a good list. Not exactly the same as mine, but probably phrased better, and less dictatorial in nature, which is likely a good thing.

I wonder how we'll do in actual play with this. It often seems like we fade late in games, but that could be to do with opposing teams often selecting full team of carries. - In this case it would not be surprising that we would struggle. However, this would also suggest that we need to get the game 'won' more efficiently in the laning and midgame phases, which probably requires a few things:

Generally, I think we still need to work more on communication, and planning.

Something I think we really need to do is take better control in the midgame. I think that typically we do quite well in the laning phase, due to being fairly disciplined now, and also since we choose a balanced set of characters. - We should be coming out ahead here, and so need to capitalise upon this.

We generally play very conservatively, and that's ok, we've trained ourselves not to die etc. However, I think that to really take a stranglehold on the game in the midgame, we need to take some risks, risks which can be significantly mitigated by teamwork and communication and discipline, all of which we can almost certainly out compete the other team in.

I think now we have a reasonable grip on when the game is starting to shift from laning to midgame, but we need to capitalise upon this, and the helpless communicationless opponents. I think this is be done in 2 ways, ganking and pushing.

I think that we really need to start talking about ganking. Typically, the most vulnerable player for this is in the midlane, as they are normally on their own. Conversely, they are usually also the most suited to being the 'hammer' in this situation, due to the type of characters which normally inhabit midlane, access to runes, map position and greater experience due to solo laning. People on the flanks need to call if they see an opportunity for a gnk, particularly in the hard lane. I've had a few very effective games where I've moved out of midlane and come up behind the enemy in their hardlane and crushed them, multiple times, and this can be facilitated by greater communication. Simply continuing to hold the lane at this stage of the game isn't going to get us a win overall, we need to take the game by the throat, and I feel we're very capable of doing this. Similarly, if there's an opportunity, the midlaner can be used as an 'anvil' and the enemy midlaner can be ganked by a player coming from one of the sidelanes, whoever is most appropriate to the task. I truly believe we really need to start planning this and trying it, it's not something we really think much about at the moment. If it doesn't work, well, all that's lost is a little gold and experience for the possible ganker.

However, when it does pay off, I think this is the time to organise a significant push. - Bring in two other players for a total of 3 and attempt to take down that first tower - 5 exposes us too much in the beginning of the midgame IMO, and leaves us vulnerable to a counter push elsewhere. Later, when you need to go deeper, the full team may be required, but I certainly believe that 3 players is enough to take down the first towers. - A successful push doesn't HAVE to destroy the tower, simply taking a large chunk of health off it is significant if we are not exposing ourselves, just be disciplined, and wait for another opportunity. - When towers are on lower health, you may be able to just scare off an opponent to make the space, rather than require an actual kill.

It's worth talking about some basics here: If the opposing mid player was killed, then one of our players from each of the side lanes should head down the river and push leaving one on each sidelane to farm and control the lane/provide sight. If it's a sidelane that was ganked, then there should already be 3 players there to push - In this case one of the players from the other sidelane should 'sweep' into the middle and occupy that lane for an experience boost and lane control until the push is over and the traditional midlaner returns. The deeper towers are more dangerous, and probably require teamwipes or multiple deaths to break down, as well as more of us - It is always worth watching who on the opposing team is dead for this sort of an opportunity, as well as for how many players to commit - If it's a full teamwipe, we could easily commit all 5 players, but similarly, a full teamwipe could mean we don't need all 5. The carry could potentially be left to free farm while the rest of the team hits the tower.

Something I've noticed in the late game is we tend to bunch up following in the creep corridor - This has led to numerous situations where we've been hit by ultimates and wiped. I've been trying to discipline myself to stay out of the creep lane in these situations. - I think something we could work on better is 'formations' - At a basic level, we need to spread out - It makes us less vulnerable to template attacks and also reduces the 'noise' within fights and the chances of missclicking. At a more experienced level, we can do better I think about positioning ourselves - Stronger heroes at the front, harassers on wings, ranged support a little behind etc - When we're all attacking a tower, it seems stupid to all stand on top of each other - We should be in either a concave or convex sort of shape abutting the tower depending on how many ranged or melee heroes we have (ideally a mix).
 

Ronin Storm

Administrator
Staff member
I'll reply to Panda's thoughts on tactics a bit later. For now, based on the little icons with star ratings as provided on the character portraits in DOTA itself, here's a quick list of the top characters in each category (in no particular order).

Hard Carry

All these are 3 star Carry heroes.

  • Faceless Void
  • Medusa
  • Morphling
  • Phantom Assassin
  • Spectre
  • Troll Warlord

Support

This is a combination of Lane Support and Support stars.

  • Dazzle (6)
  • Lich (6)
  • Wisp (5)
  • Keeper of the Light (4)
  • Lion (4)
  • Warlock (4)
  • Crystal Maiden (4)
  • Vengeful Spirit (3)
  • Chen (3)

Other Carry

These are heroes who show up on the list for a decent combination of Carry and Durability, but aren't on the Hard Carry list above. Maybe these would be considered to be Semi-Carry?

  • Skeleton King (5)
  • Doom (4)
  • Lifestealer (4)
  • Dragon Knight (4)
  • Slardar (4)
  • Brewmaster (4)
  • Lone Druid (4)

Disablers

This is a short list of 3 star Disablers.

  • Lion
  • Bane
  • Shadow Shaman

Escapers

And 2 and 3 star Escapers.

  • Morphling (3)
  • Anti-Mage (3)
  • Weaver (3)
  • Queen of Pain (3)
  • Slark (3)
  • Storm Spirit (2)
  • Clinkz (2)
  • Phantom Lancer (2)
  • Bounty Hunter (2)

Nukers

Finally, 3 star Nukers.

  • Queen of Pain
  • Lion
  • Leshrac
  • Lina
  • Zeus
  • Tinker

Lists elsewhere are clearly going to contradict this, and perhaps they're right, but this is how DOTA2 itself rates the heroes so may be worth a thought.
 

Panda with issues...

Well-Known Member
I think you've missed Drow Ranger in hard carrys. The slut.

He's doing it based on the game's own ratings (ratings I must admit I haven't seen or worked out where they are).

We can probably put together a sort of THN list.

For example, I'd agree with Nanor that Drow is a hard carry, she does almost nothing other than kill people. Riki too. He offers nothing other than death.
 

Ronin Storm

Administrator
Staff member
He's doing it based on the game's own ratings (ratings I must admit I haven't seen or worked out where they are).

Exactly. (You can find the star scores by hovering over the icons that appear in the top right of each hero's character card.)

I could list all the 2 star Carry heroes; I'll do that another time.
 

Zhar

New Member
Sorry that I couldn't get myself to read those walls of text.
But when it comes to winning and losing streaks it's very simple. The further you get into the game the more blatantly obvious the following becomes.
The matchmaking is balanced around giving you a 50% winrate.
If you're above this, good for you, you're better than most other players, if you're below this, you're not learning as fast as the medium dota player.
Now and then I get on a winning streak of about 6 matches won, after which I will be grouped with the most daft players known to mankind.
The opposite can then be said about the losing streak of 6 I get because after being pitted into hell, I get matched against those exact same people that were making my life miserable by being on my team.
I've been to hell and back, and back to hell…and back.
You either embrace this system and go for the highest as possible win rating you can achieve with the not so sharpest tools of the shed you get to work with, or you give up. The choice is yours.

Also I hope some of you are playing dota tonight and if you are, would you mind if I joined in for some fun?
 

Ki!ler-Mk1

Active Member
Well bugger that if I'm gonna play a game that sets me up to loose half the time. /me goes to casino instead.

[strike]So that is how it groups you up with allies, how does it decide with opponents(?), it would be nice if it had the system from starcraft 2 where IIRC all the players are grouped into 3 subcategories; If you rise above a certain win ratio you move up into the category above. Its almost the same except its for premades only.[/strike]

Having now thought about what I just wrote, I realise that grouping people with inferior allies is the actual problem. I don't see why it cant just keep putting you with equal peers against equal enemies, giving the same theoretically W:L without hurting gameplay.

Im going to assume that after a certain skill/rating level the game starts to run out of players?
 

Ronin Storm

Administrator
Staff member
The matchmaking is balanced around giving you a 50% winrate.

I don't think that's true for DOTA2, based on the following:

http://dev.dota2.com/showthread.php?t=70618

We don't enforce a 50-50 win rate (Maybe we should?) instead we always try to make the best match possible. I'm always happy when I see someone with a win rate close to 50% though since it means that we haven't been skewing one way or the other in giving matches that were too hard or too easy.

It is true for Starcraft 2's matchmaking system, though.

I happen to be around 50-50 just now, but I can honestly attribute many of my more recent loses to bad team composition and/or poor laning skills.
 

Ronin Storm

Administrator
Staff member
Yeah, I do suck.

More last hit drills. :)

Actually, I'm moderately serious. Had a pub game late last night as Faceless Void and my farm was sub par (and my use of Chronosphere, but that's a different issue) so I boomed too late and we'd already been raxed in mid by the time I was ready to lay the smack down. Did a 12 minute practice vs bots later and had my farm up from maybe 2 creeps per minute to more like 4 creeps per minute (personal high for me, though still not super). ~2100 gold in laning isn't bad going.
 

Zhar

New Member
Well, it might not entirely be about forcing you to get a 50% winrate, but it does give you worse teammates the more games you in a row, which does push you closer to the 50% winrate thing. As soon as I get on a 5+ winstreak I get depressed after the 2 matches that follow.
Just pay a little attention to the next you win a lot of matches in a row, you are doomed to lose the following one, two or more.
 

Panda with issues...

Well-Known Member
Well, it might not entirely be about forcing you to get a 50% winrate, but it does give you worse teammates the more games you in a row, which does push you closer to the 50% winrate thing. As soon as I get on a 5+ winstreak I get depressed after the 2 matches that follow.
Just pay a little attention to the next you win a lot of matches in a row, you are doomed to lose the following one, two or more.

Stop playing with randoms?
 

Zhar

New Member
Stop playing with randoms?
I wish to, but when everyone I know who plays the game is only 100-ish wins into it and when I queue with those people they get pitted against the people I usually get matched against, those with 500+ wins and usually in a stack.
 

Ki!ler-Mk1

Active Member
Sorry that I couldn't get myself to read those walls of text.
But when it comes to winning and losing streaks it's very simple. The further you get into the game the more blatantly obvious the following becomes.
The matchmaking is balanced around giving you a 50% winrate.
If you're above this, good for you, you're better than most other players, if you're below this, you're not learning as fast as the medium dota player.
Now and then I get on a winning streak of about 6 matches won, after which I will be grouped with the most daft players known to mankind.
The opposite can then be said about the losing streak of 6 I get because after being pitted into hell, I get matched against those exact same people that were making my life miserable by being on my team.
I've been to hell and back, and back to hell…and back.
You either embrace this system and go for the highest as possible win rating you can achieve with the not so sharpest tools of the shed you get to work with, or you give up. The choice is yours.

Also I hope some of you are playing dota tonight and if you are, would you mind if I joined in for some fun?

I know you spelled it out there, but this means that there is no point in getting to be good at this game(?), and does this still apply to premade 5 man partys, or in party play when you get good you just get pitted against equally nasty partys, and long queues?

Ugh, could have said that better, i am playing another hero battle game, and most losses(and wins) are caused by ppl who have never played before. Thinking of switching games.
 
Top