The most common skill gap in climbing is not laning. It is closing. Players reach Gold with decent mechanics and above-average decision-making in the first twenty-five minutes. Then the game reaches thirty minutes. The map stops moving. Kills stop happening. Eighteen percent of games end in draws or surrenders. It is not because the game state is unclear. It is because no one knows how to convert an advantage into a win.
The closing problem is mechanical. It is not abstract. Games stall because teams do not have a framework for converting numerical advantage or objective advantage into gold and then into nexus health. They know they are ahead. They have no idea what "ahead" translates to in action items. So they do nothing. They farm. They wait for someone else to mess up. The enemy team waits for the same thing. Twenty minutes later, the team with a fifty-five percent gold share should have ended. Instead, it went to a coin flip teamfight at Baron at fifty-five minutes.
The difference between a team that closes in twenty-eight minutes and a team that stalls until forty-five minutes is not mechanics. Both teams can click. The difference is a framework. One team understands what pressure means. The other does not.
Why Games Stall: Teams Do Not Know How To Convert
A game is stalling when no team is moving toward a win condition. Both teams are farming. Both teams are gathering for objectives. No team has a clear plan for how the next five minutes is going to improve their position. This is the definition of a stalled game.
Stalling happens because players confuse "being ahead" with "being able to win." If your team has a 5k gold lead at thirty minutes, you are ahead. You are not winning yet. Ahead means you have a higher probability of winning. It does not mean the win is guaranteed. You need to convert that 5k into objectives, then objectives into map control, then map control into a win condition.
The conversion mechanism is always the same: use your advantage to take an objective, use the objective to take a second objective, use the second objective to close. The three-step process. Most teams understand step one. They push mid with a numerical advantage and take a tower. Then they stop. They farmed the tower, now they wait for the next objective spawn. The Baron is not up for three minutes. The Elder is not up for six minutes. So they farm. Sixty seconds later, the enemy team has caught up in farm. The advantage evaporates. The objective they took is now irrelevant.
The confusion is that players treat objectives as endpoints instead of tools. They take Baron and think they have accomplished something. They have not. Baron is worthless if you do not immediately use the Baron-empowered minions to threaten a second objective or to force a situation where the enemy has to engage you in a bad position. The Baron is only valuable in the five-minute window after you take it. If you wait and farm, the Baron buff falls off and you have traded fifty seconds of your game to take an objective that did nothing.
Stopping for three minutes to farm after taking a tower is the same mistake as stopping for three minutes to farm after taking Baron. You have created a gap where the enemy team can catch up in resources. The gap closes. The objective you took is no longer an advantage.
The stalling pattern therefore works like this: team A takes an objective, team B catches up in farm, team A gets impatient, team A forces a fight from a bad position, team A loses the fight, team B gets a kill bounty, team B is back in the game. The objective was taken thirty minutes ago. The value from the objective is now zero. The game is back to the win percentage it was at when the objective was initially taken.
Baron as a Tool, Not a Prize
Baron is the object most players misunderstand. They treat Baron as a win condition. "Get Baron, take towers, win." This is the logic. It is incomplete. Baron is a tool for creating a win condition. You need to understand what Baron enables and when Baron is worth trading for.
Baron grants minion empowerment for five minutes. This means your minions do additional damage and gain additional health in a moderate AOE around them. This is only useful if you have a plan to use these empowered minions to threaten an objective or to force a decision on the enemy team. If you take Baron and then do nothing but farm, the empowerment falls off and you have spent fifty seconds on an objective that did not accelerate your win.
Baron is worth taking when you have a clear next step. You take Baron, immediately rotate to midlane, push toward the enemy base with the empowered minions, and use the pressure to either take the midlane inhibitor or to force the enemy team into a bad engagement. If the enemy team wants to defend, they have to fight you in a position where your minions are doing the damage. This is the Baron advantage. Not the objective itself. The tool.
Baron is not worth taking when you have already used your win condition and you are waiting for another win condition to spawn. If Elder Dragon is spawning in ninety seconds, taking Baron forty-five seconds before Elder is lowEV. You will not have time to use the Baron minions effectively. You will be forced to contest Elder immediately, the Baron buff will be half-expired when Elder kills occur, and you will have traded fifty seconds for marginal advantage.
The mental model for Baron is therefore: what is my next objective after Baron. If the answer is "take the inhibitor," Baron is correct. If the answer is "nothing until Elder," do not take Baron. Farm instead. The five minutes of empowerment is only valuable if you have two objectives queued up to use it.
This applies to all objectives in the late game. You take a tower. Your next objective is not "farm for four minutes." Your next objective is "use the tower I just took to access the next tower or to create map pressure that forces a bad decision."
The chain is: take objective one, immediately threaten objective two, take objective two, extend to the base. The wins happen in chains. The stalls happen when you take an objective and then stop moving.
Death Timer Awareness in Late Game
Death timers are the most underestimated mechanical factor in late game. At fifteen minutes, if you die, you are back in game in twenty-five seconds. At thirty-five minutes, a death is a fifty to sixty-second timer depending on your exact level. At forty-five minutes, it is sixty-five to seventy seconds.
This matters because a single death in the late game is frequently game-losing. If your team is at a 5k gold lead at thirty-five minutes and your ADC dies, you are down 5v4 for sixty seconds. During those sixty seconds, the enemy team is scaling their advantage and farming. If they take an objective during those sixty seconds, they are converting their four-player advantage into a concrete resource. By the time your ADC respawns, the gold gap is 3k instead of 5k.
More critically, a death at thirty-five minutes frequently cascades. Your ADC dies. The enemy team, now 5v4, contests the next objective. They kill two of your team members while you are still at base. Now you are 3v5. You cannot contest Baron. The enemy team takes Baron, takes inhibitor, and your 5k gold lead is now a -3k gold deficit within ninety seconds.
This means that in the late game, the decision of whether to fight is not a symmetric decision. If your team is ahead, you are asking the question: "If we fight and someone dies, do we still have enough defensive resources to hold without them." If the answer is no, do not fight. A death at thirty-five minutes plus minutes is not a slight disadvantage. It is a cascading loss.
The five-second decision rule: if a five-man fight would leave you with a player dead at sixty seconds when the enemy is pushing base, that fight is incorrect. You should position in a way where a death does not cascade into a base push. The death timer is long enough that you need a failsafe.
This creates a mental model where late game is the opposite of early game. Early game, you take fights where someone might die because that person comes back before the enemy scales. Late game, you avoid fights where someone might die because that person stays dead long enough for the game to be decided.
When to Force and When to Wait
The forcing decision is where late game strategy lives. You have an advantage. The question is not "can we force" but "should we force." This requires understanding your win condition and whether forcing accelerates it.
You should force when forcing is your best path to completion of your win condition and when you have a plan for what happens if the force goes badly. Example: it is thirty-five minutes. Your win condition is "take Baron and then inhibitor." You have a 5v4 numerical advantage because the enemy support is dead. The enemy team is trying to stall until they scale. You force Baron immediately. You are trading your five-player advantage for a concrete objective. If Baron goes wrong and someone dies, you are still 4v5 but at least you have the Baron objective. If Baron goes right, you take Baron and then immediately push for inhibitor.
You should wait when waiting is your best path to completion of your win condition and when you are not risking the game by waiting. Example: it is thirty-five minutes. Your win condition is "wait for Elder to spawn, win the Elder fight, and close with Elder buff." The enemy team has a scaling win condition that gets better the longer the game goes. You wait until Elder is spawning in sixty seconds. You do not force at thirty-five minutes because forcing is bad EV. You are trying to stall until Elder spawns, win the Elder fight, and then use Elder to close. Forcing now just gives the enemy team time to scale.
The second example is the mistake most players make. They confuse "we are ahead" with "we should force." Being ahead means you should be playing to close. But closing is not always "force immediately." Sometimes closing is "stall and scale slightly slower than the opponent."
The complicating factor is that the longer the game goes, the more you need to know about the enemy team's win condition. If you are at thirty-eight minutes and you have no idea whether the enemy team has Kayle at twelve stacks scaling or whether they have all champions on their final form, your decision-making is blind. You might wait, thinking you are in a "stall until Elder" scenario, when you are actually in a "force now or lose to scaling" scenario. The information gap creates decision errors.
This is where late game macro separates tiers. A Platinum team makes decisions without the information necessary to make them. A Grandmaster team spends the early late game phase gathering information. They check what is the enemy win condition. They understand the exact power curves of enemy champions. They know if they are on a clock or if they have time.
The 5v4 Rule: Never Force When Outnumbered Late Game
This is the single easiest rule to implement and the most commonly violated.
If your team is in a 5v4 state at thirty minutes or later, you do not engage the five-player team. This is non-negotiable. Your advantage is not numerical. Your advantage is positional. You have a player advantage that is meaningless if you engage the enemy team in a position where they can use their positioning to mitigate the disadvantage.
The enemy team's goal when they are down a player is to stall until the player respawns. Your goal is to use the player advantage to take an objective that cannot be defended by four players. You do not get the objective by fighting. You get the objective by using the numbers advantage to force the enemy team into a choice: defend the objective and fight 5v4 (bad) or let the objective fall (bad). Either way, you win.
The moment you engage the 5v4 enemy team, you have eliminated your advantage. You are now in a fight where skill matters more than numbers. A good fight from the 5v4 side can catch a player out of position or use terrain to reduce your effective numbers. You are trading your reliable advantage (one more player) for an unreliable skill check.
This rule breaks down only at one point: if the enemy player respawns before the objective falls. If the enemy support is dead for forty seconds, and it takes sixty seconds to take the inhibitor, the player respawns and defends. You cannot take the objective without engaging the 5v4. In this case, you have to fight, but you are now fighting from a worse state because the player advantage is no longer reliable.
The decision is therefore: before forcing a 5v4, calculate whether the objective you are forcing falls before the dead player respawns. If yes, force. If no, do not force. Set up the objective, farm until the player respawns at full strength, then re-engage.
Wave Setup Before Objectives
This is the mechanic most players never learn. You do not take Baron by walking to Baron. You take Baron by setting up waves, creating pressure that forces a decision, and taking Baron when the enemy team is too far from Baron to defend.
The process is this: if you want to take Baron, the ideal state is that all three lanes are pushing into the enemy. The enemy team is forced to choose which lane to defend. If they defend one lane, the other two take towers. If they come to Baron, you have numbers advantage and you take Baron and then a tower. They lose either way.
To create this wave state, you need ten to fifteen minutes of setup. You place minion waves in all three lanes such that the enemy team's defenses are spread. This is not complex mechanically. It is not high elo. It is just understanding that minion waves exert pressure and that you can orchestrate waves to create space for objectives.
Most teams in the late game do the opposite. They congregate in one place to make sure they do not get caught out. Then they are surprised when Baron gets taken on the opposite side of the map. They had no minion pressure set up. They had no wave exerting space. So the enemy team took the objective for free.
Wave setup is particularly critical before Baron because Baron is the longest objective to take. If you do not have waves pushing in all three lanes, the enemy team respawns from a death and immediately stops your Baron. But if you have minions crashing into turrets in all three lanes, the enemy team respawns and has to make a choice. Defend base or contest Baron. You take Baron when they are split.
The process for wave management is: before you plan to take a major objective, spend two to three minutes setting up waves. Get a slow push going in one lane. Get another slow push in another lane. Make sure a fast push is incoming in the third lane. Then take the objective when these waves hit their walls.
A slow push is when you have more minions on the map than the opponent and you are not attacking, just farming and letting the wave naturally push. This takes ninety seconds to two minutes to translate into actual tower pressure.
A fast push is when you group minions and push hard, forcing the wave to hit the tower quickly.
The coordination of these timings is where the game is actually decided. A team that understands wave mechanics and coordinates them can take objectives against a team with superior teamfight because the superior teamfighting team is split.
Elder Dragon and the Execute Mechanic
Elder Dragon is unique because it grants an execute passive. Any damage you deal to enemies below twenty percent health automatically kills them. This is not a damage buff. This is a teamfight mechanic that changes engagement patterns completely.
When you have Elder, your teamfight win condition is not to outdamage the enemy. Your win condition is to get any enemy below twenty percent health and the execute mechanic kills them automatically. This means the Elder buff teamfight is heavily in favor of the team with Elder, and it is also less dependent on precise execution. You can win teamfights by just doing any damage because any damage eventually triggers the execute.
This makes Elder the ultimate closing buff. You take Elder. You immediately go take a major objective with Elder buff active. You force the enemy to defend. You engage them. Execute mechanics delete a player. You are now 5v4. You convert the 5v4 into base damage. You win.
The critical decision is therefore when to force for Elder. Elder Dragon spawns after one of the two dragon types has been killed. At that point, Elder is available every minute. You should force Elder when your win condition has exhausted other options. If you are in a stalling scenario where you need a teamfight win condition to close, Elder is that win condition. If you already have a win condition through wave pressure and Baron buff, Elder is less critical.
This is subtle, so let me be precise. You are at thirty-five minutes. You are 4k gold up. You have minion pressure in all three lanes from setup. You can take Baron without forcing Elder. You do not need Elder. You take Baron, use the minions, take the inhibitor. Game closed.
Alternatively, you are at thirty-five minutes. You are 4k gold up. The enemy team has better wave clear. Your waves are getting punted back. You cannot take Baron safely with just your numbers. You need a teamfight win to have enough space to take objectives. You force Elder. You win the Elder teamfight with execute mechanics. You are now 5v4. You take Baron with the player advantage. Game closed.
Both paths work. The distinction is whether you have a wave-based win condition or a teamfight-based win condition. Elder works when your win condition is teamfight.
The Final Closing Pattern
If you take all of these mechanics together, the late game pattern is straightforward.
You identify your win condition. It is one of three: wave pressure, Baron, or Elder. You check the enemy team's scaling and their respawn timers. You set up waves in advance so you have lateral pressure. You force your win condition when the enemy team is forced to make a choice. You execute the plan. You move to the second objective. You close.
The closing takes between five and fifteen minutes if you are not wasting time. The game that stalls at forty minutes is a game where step one failed. The team either did not identify a win condition, did not set up waves, or got impatient and forced from a bad state.
The difference between a thirty-minute close and a fifty-minute close is usually five minutes of setup and five minutes of actually having a framework instead of improvising. It sounds small. It is the gap between Gold and Platinum.
The execution is not complicated. The players are not bad. They just do not have a model for how the late game works. They do not know that minion waves exert pressure. They do not know that Baron is only valuable if you have a second objective queued up. They do not know that a death at thirty-five minutes is cascading.
Once you have the model, the execution becomes mechanical. You apply the same principles repeatedly. You set up waves, you take objectives, you close. By the fifteenth time you run this pattern, it is unconscious. Your team closes thirty-minute games routinely because you have turned late game closing into a repeatable process instead of a coin flip.
This is where most teams leave twenty percent of their winrate on the table. They win the early game or they win the mid-game teamfight. They do not know how to convert that into a close. They hand the game to the enemy team by stalling. The skill is not complicated. It is just underemphasized in education content.
Frequently Asked Questions
How do I know when I have reached the closing phase of the game.
The closing phase typically begins at twenty-five to thirty minutes. You know you are in the closing phase when the team with the most gold advantage should be applying that advantage to take objectives. If you are five thousand gold ahead at thirty minutes, you should be closing in the next ten minutes. If you are still playing at forty-five minutes, you are in a stalled state and you have made a closing error.
What is the difference between a macro win condition and a mechanical win condition in late game.
A macro win condition is something you set up before the fight happens. Examples: wave setup that forces the enemy team to be split, Baron buff that empowers minions and forces a choice, Elder buff that gives execute. A mechanical win condition is something you win by outskilling the opponent in a teamfight. Late game should be predominantly macro. If you are trying to win by outskilling the opponent when you are six thousand gold ahead, you have abandoned your advantage and you are trying to win with skill instead of with resources. This is lower EV than macro-based closing.
Why do games stall when my team is ahead.
Games stall because you are taking objectives without having a plan for the next objective. You take a tower. You farm for three minutes. The enemy team catches up. You take Baron. You farm for three minutes. The enemy team catches up. You need to take objective one, immediately threaten objective two, and move to objective two without a farming pause. The pause is where the stall happens.
How do I know if my team should force or wait late game.
Ask two questions. First: what is my win condition. Is it wave pressure, Baron, Elder, or a teamfight. Second: if we wait five minutes, does the enemy team's win condition become better or worse. If your win condition is Elder and the enemy win condition is scaling, you should wait for Elder. The longer the game goes the better your position. If your win condition is Baron and the enemy win condition is also scaling, you should force now because they scale better than you do.
What is the most common closing mistake.
The most common closing mistake is farming after taking an objective. You take a tower at twenty-eight minutes. You are 4k gold ahead. You farm for four minutes waiting for Baron. The enemy team catches up to 2k gold behind. You take Baron and you take inhibitor, but you are now 2k gold ahead instead of 4k. You had the game one and you threw the advantage. The fix is: do not farm between objectives. Move from objective one to objective two without pausing.
Should I always take Baron if it is available.
No. Baron is only worth taking if you have a clear next objective. If Elder is spawning in two minutes, waiting for Elder and skipping Baron is often better. If you just took inhibitor and Elder is not spawning for eight minutes, taking Baron to occupy the next five minutes with the buff is correct. Baron is a tool. It is not always the right tool.