Communication and Ranked Head: The MMR You Lose Without Playing Badly

This is the only guide on this site that is not about the map, and it exists for a practical reason: in the replay reviews we do, the most common student phrase is not about farming or about item. It is "I lost because the team".

Sometimes it's true. The point is that, true or not, it's the only variable in the match that you have full control over and almost no one works on. There's no motivational speech here: what follows is what changes the result.

Part 1: Communication that serves

The problem of unused information

Most of what is typed and said in a ranked is information that does not become a decision. "Top 2" is information. It doesn't tell anyone what to do with it, and that's why most people don't do anything.

The useful form has two parts: what I saw, and what it allows.

• Instead of "top 2": *"top 2, you can get the jungle from below now"*.
• Instead of "mid miss": *"mid is gone, be careful who is pushing underneath"*.
• Instead of "without vision": *"without a view of the river, it does not cross through the middle"*.

The difference seems small and it is not: the second way transfers a ready-made decision, and a ready-made decision is what an unknown team can execute.

The four calls that are worth more than the rest

If you only make four types of calls per match, make these:

1. What are you going to do? "I'm going to push it down", "I'm going to farm the jungle from above". It prevents the team from fighting in four without knowing it.
2. What you don't have. Ultimate on reload, no TP, no buyback. It's the information that changes the most team decisions and the one that least people give.
3. What's available. Roshan free, open rook, formed field. Opportunity with a deadline.
4. The specific danger. Not "be careful", but "the Lion has an ultimate and disappeared ten seconds ago".

What not to say

• Diagnosis of someone else's mistake in the middle of the game. It doesn't fix anything and worsens the execution of the listener. If the person made a mistake, he will make a better mistake angry.
• Discussion about what has already happened. Every second spent on this is a second in which no one is looking at the map.
• Request without instruction. "Help here" is noise. "Come from TP downstairs, you can kill" is called.

Part 2: The Real Cost of Tilt

Here it is worth being specific, because "don't get angry" is not an executable advice.

What irritation does, mechanically, is change your decision function. You don't slow down or miss more clicks: you start to choose differently. Specifically:

• he buys a fight that he would not buy, because the fight would solve everything at once;
• stop looking at the minimap, because looking brings bad information;
• He plays to prove one point (that the problem was the other), and not to win.

None of this appears as "played badly". It appears as a bad decision, and a bad decision is exactly what separates medal banners.

The two rules that count on MMR

Rule 1: Stop after two consecutive losses in which you played badly. Not after any two losses: losing well played is normal and does not indicate anything.

Rule 2: Don't play ranked right after a game in which you got angry. Play an unranked one, or stop for an hour. The cost of one less game is zero; the cost of three games played in anger is an entire range.

This isn't self-care, it's economics: your MMR is the sum of the decisions you've made, and decisions made in anger have a worse hit rate, whether you're Herald or Divine.

Part 3: What to do with the teammate who is playing poorly

There's only one thing that works, and it's annoying: playing around the problem.

If your carry is dying alone in the jungle, the answer isn't to explain the mistake. It's to put vision where he farms, warn when the map tightens, and plan the team's plays assuming he'll be there. You won't fix the player during the match. You can change the cost of his mistake.

And the question that replaces the discussion: what is still possible to win from here? It is useful because it changes the target. A game with a bad player is still winnable with four good decisions; it ceases to be winnable when the other four decide that it is not.

Part 4: The Mute

Silencing those who are getting in the way is the most efficient tool that the game offers, and it has a consequence that many people forget: you lose the information that this person would give.

The practical criterion: mute the text chat before the voice chat. Text almost never carries a useful call and is where the discussion happens. Voice, even from someone angry, still brings "top 3" at the right time.

And mute early. Muting at minute 5 preserves 35 minutes of decision; muting at minute 35 only registers that you have held out until then.

Part 4.5: Play in a group, and why it changes the problem

It's worth a parenthesis about playing with friends, because almost all communication advice assumes random team.

In a group, the problem is reversed. The noise is no longer an angry stranger and becomes too much conversation: the group talks all the time, about the match and other things, and the useful information is lost in the middle. Teams of friends usually have worse communication, not better, and that's the reason.

Two fixes that work:

• Separate conversation from call. Agree that when someone says the name of a lane or a hero, it is called and everyone stops talking for two seconds. It's silly and solves.
• One person per decision. Five people discussing whether it's worth doing Roshan is the fastest way not to do Roshan. Someone decides, the others execute, and the discussion is left for after the match.

And there is the opposite problem, which is a group of two or three within a ranked of five: the group talks on the outside and the whole team does not receive the information. If you play in pairs, the rule is simple: everything that is called goes to the channel that everyone hears. What remains private is the conversation, never the decision.

Part 5: How to Measure if It's Working

Communication and head are the easiest things to think that you've improved without having improved. Two objective measures:

1. Count your useful calls per match in the replay. A useful call is the one someone has made. Five per match is already more than average.
2. Write down, at the end of each session, how many games you played to win until the end. Not "how many you won": how many you played in the entire game, without giving up in the middle. This number goes up before MMR, and it's what pulls MMR after.

What to hold on to when the chat heats up

• Information without instruction does not become a decision. Say what you saw and what it allows.
• Four calls: what am I going to do, what I don't have, what is available, the specific danger.
• Do not diagnose someone else's mistake in the middle of the match.
• Tilt doesn't make you play slower: it makes you choose differently.
• Stop after two poorly played losses, and never play ranked with fresh anger.
• With a bad companion, play around the problem instead of fixing it.
• Mute text early, voice last.