DOTAFLIX
Play
Matches
Learn
Community
PGL WALLACHIA
LIVE CLASS
TOWER DEFENSE
AIM
CAREER
BATTLE PASS
PlayMatchesLearnCommunityPGL WALLACHIALIVE CLASSTOWER DEFENSEAIMCAREERBATTLE PASSMore
Sign in

DOTAFLIX

Dota 2 lessons, guides and replay analysis with three-time champion coach Filipe “Astini” Ribeiro.

Follow us

GuidesAboutContactTerms of UsePrivacy PolicyLegal Notice

DOTAFLIX is not affiliated with or endorsed by Valve Corporation. Dota 2 and its game assets are property of Valve.

HomeLIVEDEFENSEPASS
Dota 2 Guides›Macrogame

Big jungle and farm cycle: how to choose the moment to invade

Taking the opponent's big jungle is worth nothing if the fields haven't formed. The timing criterion, why the total volume of farm is deceiving, and how the Roshan decides which side of the map matters.

By Filipe "Astini", three-time tier-1 champion coach·August 17, 2026·6 min read·Intermediate

In this guide

  1. Territory without resources is not territory
  2. The invader at the right time buys an option
  3. Why Jungle Farm Total Is Deceiving
  4. The Roshan hierarchizes the two sides of the map
  5. The question that replaces "should I fight?"
  6. The three cycle errors, by position
  7. Checklist for the next ten matches
  8. Where it connects

Ask ten Dota 2 players where the most valuable part of the map is. Nine will point out a lane. The right answer, in modern Dota, is the large jungle: the region of the ancient and large fields, on either side of the river.

It is there that the farm per minute of a core is decided after the 15th minute. It is there that the fights that change the match take place. And it is the only territory whose value changes from minute to minute, which makes it the most interesting decision in the game.

Astini's lesson on this

Stacks - 03Stacks - 03Free17 minWatch the lesson

Territory without resources is not territory

The mistake that this guide is there to fix is subtle, and most players who have ever understood "big jungle matters" keep making:

"LGD takes the opponent's big jungle and has no field formed: it resumes an empty big jungle, an empty cycle. The timing that Nigma enters LGD's big jungle, the field has already formed once, forms a second one. It has to be thought of in a farming cycle."

Occupying the opponent's large jungle with empty fields is occupying ground. You spent displacement, exposed the team, maybe bought a fight, and didn't take any resources. The same invade, done 40 seconds later, would have brought two field cycles.

It changes what you observe before you decide. It's not "am I strong?" or "are they there?". É:

  • Did the fields on that side form?
  • Will they form again within the window where I will stay there?
  • Do I arrive before or after respawn?

The invader at the right time buys an option

Here is the benefit that is almost never mentioned, and which is the most valuable: those who enter the right cycle can refuse the fight.

"Even if she decides not to take the fight, she says 'gee, this fight is at a bad angle', and she just walks out."

If you entered because the field formed, you have already taken what you came for. The fight becomes optional: you look at the angle, think it is bad and leave, without losing anything. Those who entered the empty cycle are in the opposite situation: they need the fight to happen and work, because it is the only way to justify having gone. This is how a timing decision becomes a risky decision.

Why Jungle Farm Total Is Deceiving

This is the most instructive data from the match that consolidated the doctrine: LGD × Nigma, The International 2026:

Big jungle (ancient + large), entire gameClean fieldsEstimated Gold
LGD (lost)263~11.060
Nigma (won)234~10.034

The team that farmed the most large jungle lost: about 10% more fields. If you measure volume, you conclude the wrong thing. What separated the two teams was not how much, it was when: each invasion of the Nigma happened with the cycle formed; most of the LGD did not.

It is worth noting the honest limit of this measurement: the time window trimming does not exist in the public Dota APIs. The list of cleared fields does not come with a time. You can count the total, not to know the timing of each one. That's why the evidence of timing is visual, in the replay, and it's one of the reasons why analyzing replay is still irreplaceable.

The Roshan hierarchizes the two sides of the map

The two large jungles seem equivalent. They are not, and the criterion that decides which one matters is simple:

"They took the big jungle from the top. The Roshan is at the top. Is this the time to fight to retake the big jungle from the bot? No.."

The position of the Roshan defines where the next big goal will be disputed. Territory on the Rosh side is worth more, because it is ground that will be used. To resume the opposite side is to spend a fight on ground that no one will step on in the next ten minutes.

Rule of thumb: before deciding which jungle to contest, look at where the Roshan is and how far it is from respawning. This resolves most doubts without any calculations.

The question that replaces "should I fight?"

The theme of big jungle is not just a tactic: it is what gives content to the central reformulation of Astinism:

"Stop thinking about whether you want to fight or not, whether you're strong or not. The game isn't about that. It's about which areas I'm going to contest, which fields, which waves. And 80% of the time this is going to be a big jungle theme."

"Should I fight?" has no objective answer, so it becomes momentum. "Which field do I want?" has an answer by looking at the map. The fight comes later, as a consequence, and on the terms you chose, not the ones the opponent offered.

The three cycle errors, by position

The concept is the same for everyone, but it appears differently in each function, and recognizing your version of the error is what makes reading a habit.

In carry, the mistake is to farm by fixed lane. The player memorizes a circuit (bottom field, big field, wave, back) and executes it regardless of what the map offers. The circuit is comfortable because it doesn't require a decision, and that's precisely why he produces the carry that farmed a lot and was nowhere to be found when he imported. The fix is to change the fixed lane for a question every two minutes: *what's the next piece of map where I have safe farming for the next 60 seconds?*

In the offlaner, the mistake is to enter the opponent's big jungle by threat and not by resource. He invades because he is strong, not because the field has formed. When it works, he kills someone; when it goes wrong, he has lost the territory he was holding. In both cases, the gold did not enter.

In support, the mistake is not looking at the clock for the team. The 4 is usually the one who has the information (he's on the map, he put the vision) and it's almost never the one who says what it allows. The phrase "you can get their jungle now, the field has formed" changes more matches than a gank.

There is also an error that crosses the three positions: counting the jungle farm by what appeared at the end of the game. The score of clean fields does not distinguish between a field taken in the right cycle and a clean field in empty territory, and these two things are worth very different amounts.

Checklist for the next ten matches

  • Before any invades, look if the field has formed. If it has not, wait or go elsewhere.
  • After winning a fight, the question is not "shall we go for Rosh?" but "what courses and waves are available now?".
  • When you lose territory, do not take back the side opposite the Roshan. Take back the side you want to use.
  • As a support, stacking your big jungle field is playing this theme. It's the cheapest macro item there is for position 5.
  • As a carry, your farm after minute 15 lives there. If you're farming lane because the jungle is contested, the problem is with the team, not your farming pattern.

Where it connects

Big jungle is the second of the three rules of Astinism. The first one (never defend your safe lane) exists largely so you can play this theme: whoever is anchored in the corner of the map defending tower has no way to contest any field.

Filipe "Astini" Ribeiro

Written by

Filipe "Astini" Ribeiro

Coach profissional de Dota 2 e fundador do DOTAFLIX

  • Campeão da ESL One Bangkok 2024 como treinador
  • Campeão da ESL One Raleigh 2025 como treinador
  • Campeão da DreamLeague Season 26 como treinador
See profile and credentials

Read next

  • Astinism: the three macrogame rules that decide the match

  • Never defend your safe lane: the mistake that costs the most MMR in modern Dota

  • Play as if the enemy has Aegis: the stance that resolves the fight decision