Why Forcing Compositions Works in Teamfight Tactics
It’s all about the combat.
Advice for gaining rank in TFT follows a common refrain:
Memorize a team composition, then spam it over and over again.
A team composition typically consists of some 6–8 Champions with some specific items to prioritize.
Although the best compositions tend to cycle in and out, each patch typically has one or two dominant builds that simply destroy any of the other compositions.
At higher ranks, team compositions become deadlier and more optimized.
It’s to the point where optimized compositions beat lesser compositions pretty much every time. Meme builds with double Warmogs just can’t hold a candle to metagame compositions.
And that’s because a deliberate choice made around how combat works in TFT.
The choice to make combat mostly deterministic.
High in Complexity and Determinism
The claim that TFT’s combat has predictable outcomes might be surprising, especially given how chaotic battles can get.
When Players fight each other, they get grouped into pairs and square off against each other. Their units spring to life, and a melee ensues.
Battles can be difficult to keep track of. Each unit has their own abilities and mana pools. Units also have unique passives that can confer additional traits and bonuses. They can even hold powerful items to augment their strength.
Late game fights can resemble large scale battles, complete with the chaos of multiple overlapping abilities and attacks. They also play out in real time, meaning that complex calculations are run for every instance of damage.
However, for all their spectacle and complexity, the fights hide a fundamental truth.
The fights are mostly scripted. They are complicated Rube Goldberg machines balancing hundreds of tiny variables.
Granted, there is an innate 25% critical strike chance for every auto attack. And certain abilities and items (Taliyah stun, Jeweled Gauntlet) add some variability to the mix.
But if you replayed the same fight against each other — rewinding the fight to the beginning — you would get the same result 90% of the time. Even for those late game fights, you would get the same chaos playing out in practically the same way.
These fights have high determinism.
Which is just another way of saying they have low randomness.
Sliding Scale of Randomness vs. Determinism
What determines how combat resolves?
Do your small decisions about positioning and itemization matter? Or does combat mostly revolve around coin-flip abilities that target random characters?
The answer depends on where combat falls on the Sliding Scale of Randomness Versus Determinism.
- A largely random combat system relies on randomized targeting for spells, randomized aggro patterns, and randomized damage calculations. The same fight won’t always resolve the same way.
- A mostly deterministic combat system relies more on predictable rules for how units are targeted for spells and how units are aggroed. Battles give more consistent results.
Let’s first talk about the advantages of randomness. It adds variability and can generate excitement even when the same battle is playing out for the umpteenth time. Combat is less ‘solved’.
Randomness is also the great equalizer. It can allow less optimized builds to beat more optimized ones. Higher levels of randomness tend to give more casual friendly experiences.
However, high amounts of randomness devalue individual decision-making. You can create the ideal composition, with every unit positioned perfectly, making sure your Zephyr hits your opponent’s stacked carry — only to have the battle slip through your fingers because of a coincidental Taliyah stun.
This can lead to moments of frustration. Players can feel like they’re losing even when the opponent ‘deserved’ the loss.
It’s a matter of tradeoffs. What do you want your combat to emphasize?
Teamfight Tactics Chose Determinism
For Teamfight Tactics, the developers decided to tip the scales towards Determinism.
That means that your choices matter.
The placement of units, the choices in team composition, the determination of how to distribute your items — each decision can significantly increase the odds of victory.
After hitting Rank 1 during Set 2, Mismatched Socks talked about all of the tiny optimizations he made to push combat towards his favor in this Reddit post.
Here’s how he manipulated Singed’s AI targeting:
“The key is that Singed always runs to the furthest unit, and when he gets there, he looks for the furthest unit and runs to that unit and repeats the process.”
Consider the following hypothetical situation:
In this example, the two Singed would run towards each other. After reaching their destinations, they would repeat this process again.
If your Singed was less stacked than the opponent’s, you could use this positioning to counter the enemy Singed while also keeping your Twitch safe in an unobtrusive corner.
“I think if that even if you only have a tiny edge over the other TFT players, over a large amount of games, that will translate to a huge amount of LP gain.”
Why Teamfight Tactics is Hard to Balance
So, choices matter in combat. Even seemingly minute decisions on positioning can have a large impact on how battles are played out. High determinism, low randomness.
However, this comes with tradeoffs.
As combat is mostly deterministic, the power level of compositions are sensitive to seemingly small changes. Changing the value of a couple of variables can change the entire combat equation.
New patches can introduce entirely new equilibriums.
Butterfly Effects
In Patch 10.10, Ziggs was buffed so that his spell only cost 40 mana to cast instead of 45 mana. That meant that he was able to cast his spell more often.
At first glance, this doesn’t seem too significant. Maybe the intent was to just slightly buff Ziggs because they noticed that his win rate was a bit low.
This change was added to the patch notes and thrown into the ocean of complexity that is TFT.
The result was significant. Ziggs suddenly became S+ tier and the most contested unit. Ziggs based compositions were now outcompeting every other composition.
Riot Mort, the Lead Designer of Teamfight Tactics, had this to say about the situation:
Let’s try to parse what happened.
First, what’s the difference between 40 mana and 45 mana? Intuitively, it feels like a change of five mana should be relatively minor.
At a micro level, this change alters how often Ziggs can cast his spell. Requiring less mana to cast Ziggs’s spell means that his overall damage output is higher.
But wait, there’s more! You can also put items on Ziggs to amplify his damage. Mana generating items on Ziggs can add even more compounding value to his damage output.
At a macro level, Ziggs has traits that synergies with other units. Units that share passives and traits with Ziggs are suddenly more valued.
Changing a small scaling factor can jettison a previously weaker composition into the forefront. Small changes can have outsized effects because they can have ripple effects on the rest of the system.
An entirely new ecosystem can emerge, with its own patterns of balance.
Balancing TFT is a Nightmare
Like we established earlier, Teamfight Tactics has highly deterministic combat. Choices matter. They are the levers that Players use to influence the outcome of the game.
Combat is balanced around hundreds of tiny variables. But combat with low randomness can start to create problems.
The balance changes from earlier have very little noise to them. No variability. That’s why the sensitivity becomes so off the charts.
Balancing the game becomes like changing the inputs of an enormously complicated system. A Black Box with mechanisms too intricate to reliably predict.
What’s more, TFT is an ever evolving game. Features are added at an aggressive cadence, making balancing the game even more of a Sisyphean task.
Until combat introduces more controlled randomness into the equation, every new patch asks the same question:
What is the new ‘most powerful’ composition?
Thanks for reading.
Catch me live at: https://www.youtube.com/channel/UC12qwsLyBQP01lmC22C-l1w
Socials: https://linktr.ee/deltagonist
