Continual updates have created a dilemma for live-service games with long lifespans.

Each update brings excitement and builds the community, but also steepens the learning curve and list of changes to keep track of.

This holds especially true in lane-pushing games.

How do we design to overcome these challenges? One solution is expanding our use of modular content.

Systems and Levels

Video games have lots of content, and we generally don't want to shove it all in front of the player at once.

To address this, designers often:

• Break the content into levels, which are bite-sized chunks of content.
• Offer a sandbox where the player can discover content at their own pace.
• Use modular systems that set expectations and create a context for change and discovery.

A modular system is one which has many related parts, where the specific parts can be swapped in or out while maintaining a functional whole.

For example, a car has an engine, wheels, and brakes. We can replace the engine, or the wheels, or the brakes, and still have a functioning car.

The car might feel different to drive depending on the modular parts being used — and this is a good thing.

It means we can personalise the experience, adapt to changing circumstances (such as using snow tires), or upgrade the vehicle over time without fully replacing it.

Video games have countless examples. The races in Starcraft, weapons in Quake, or even available skins for a character are all collections of interchangeable content that support the same fundamental gameplay.

A Taxonomy of Pools

In this article we are interested in pools of content and how the player interacts with them.

Pools might be either:

• Static — the full pool of content can be accessed in every match
• Sessional —a subset of the pool is represented in a match, and players don't have full control over which pieces appear

From a live-service perspective, we also consider if content is:

• Stationary — the pool of content is fixed
• Scaling — the pool of content has a strategy to grow over time
• Seasonal — the pool of content fluctuates, typically through rotation or methodical addition and subtraction of what's available to players

Lets dive right in exploring four games:

Case Studies

Eve of the Apocalypse

The first modular lane-pushing game was Eve of the Apocalypse (DarnYak, 2005), which was packed to the brim with content.

The game had 5 different maps to play on, each with unique objectives, layouts, and item shops.

For a Warcraft III mod — it was an enormous undertaking to create and maintain multiple battlegrounds. The largest was a sprawling war across five lanes with capturable bases. The smallest was a single lane with simple objectives made for brawling.

Importantly, these are not different 'modes'. The rules of the game were the same, it's the content players interacted with that changed.

Eve of the Apocalypse also had teams choose one of four different factions to play as — each with their own armies, troops, and hero pools.

This means before choosing a character, your team first agrees on a faction, and then chooses characters from the pool available to that faction.

Consequently, instead of adding new heroes one by one, EotA is set up to add new factions as a Static Scaling content pool.

Note that each faction has its own army, buildings, troop upgrades, and more. They are as nuanced being a lane-pushing game where your team chooses a faction (as in Starcraft).

Heroes of the Storm

Following suit, the first commercial lane-pushing game to embrace modularity was HotS, which launched in 2013 with 9 maps. These offered variety in size, number of lanes, neutral camps, bosses, and teamwide objectives.

That meant HotS had two Scaling pools: maps and heroes.

The particular design of HotS supported this — the game had characters which could be finely tuned to find success on at least some of the available maps.

This was achieved through creative kit designs, and the "talent" system which allowed tweaking character balance at each stage of the game.

For example, if a character is known to under-perform in the late-game, it's possible to adjust its Level 16 talents, or even replace one. The same technique can give more viability on a certain map.

Talents offer granular balancing. Even if it is "forced" to pick a talent on a certain map, that is still supporting the overall goal of viability across our modular content.

We also see that certain maps, like Infernal Shrines, spawn a random boss monster from a pool of 3 options. Players don't control what they get — so that's Sessional content!

Dota 2

In 2019, an update to Dota 2 introduced Neutral Items. From this pool of about 60 items, around 15–25 will randomly drop for each team during a match as they clear neutral camps.

Because the drop is random or semi-random, this is Sessional content. Players don't control what subset of the pool will be available to them.

Observing patch notes, we can see a trend where occasionally a handful of neutral items are removed from the pool, with a few added in return. Sometimes old items make a reappearance as well.

We can see there is a rotation of content over time — even if it's infrequent and not an explicitly stated policy. Still, I would call it a Seasonal pool.

Overall, Dota 2 has prioritised sweeping, large scale changes to the game to create continual novelty, rather than investing in modular systems.

League of Legends

In its early days, League had one Scaling pool (champions). It has now evolved towards modularity in its level design.

League has lots of reasons to preserve the familiarity of Summoner's Rift, the most popular 5v5 level. It enforces a structure onto gameplay, as well as being an icon for esport viewers.

Nonetheless, there is a need for variety and keeping things fresh.

In 2015, League replaced its classic Dragon boss with Elemental Dragons. Occupying the same location, the particular dragon that spawns is now randomly selected from a pool of 4 options with differing rewards.

Even if they all appeared during one match, the lack of player agency around them makes Elemental Dragons a Sessional pool.

In 2019, the Elemental Dragons became a hook for more changes. Over the first ~20 minutes of a match, specific dragons start to alter the battlefield by adding speed zones, new walls, waygates, and powerups.

This is a smart move. A returning esport viewer won't be alienated by the map looking different than they remembered, because at the start of the match it's still familiar and they can learn as they go.

In 2024, the boss objective Baron Nashor was updated with 3 distinct forms, which include changes to its attack patterns and the surrounding terrain. Again, this helps to keep the experience fresh without adding too much learning — since the core idea of the boss remains the same.

Designing for Longevity

Historically, our genre relied on mostly Stationary content, with a single Scaling pool: the 'Heroes'.

This trend comes from the Warcraft III modding era (2003–2009), when:

• All pools were Static
  • Mods didn't support connecting to a live service or custom lobby options, so most developers didn't explore it.
  • Players could just leave a match without consequences if the subset of content they wanted to play with didn't appear. Sessional content is a privilege of having a captive audience.
• DotA Allstars had an outsized impact on early industry trends
  • Most mods from 2003—2009 invested in polishing 20–30 heroes. They were iterating on and refining a Stationary pool.
  • DotA added heroes faster than anyone else, and became an outlier success. Ergo, heroes as a Scalar pool became standard.

In the 2010's, designers learned 'there are diminishing returns on continuing to grow Heroes as a content pool'. It's a steeper learning curve for new players, trickier to balance, and fragments the market for new skins or cosmetics. 

As a result, the commercial games pivoted to introduce a range of Sessional and Seasonal pools. Adding Baron Pits and Neutral Items allows players to get more fun out of the existing content. It's efficient.

Indeed, we can say 'there are diminishing returns on continuing to grow any content pool'. Thus having a diversified strategy is valuable for longevity.

Conclusions

I believe the modern lane-pushing game designer must be closely attuned to the use of modular systems. All content should be examined through the lens of "will this pool be [Static/Sessional] [Stationary/Scaling/Seasonal]?"

In particular, new games have an opportunity to think modular from day 1. This is an untapped possibility space for real innovation.

In future, we'll talk about what makes for good Sessional content, how various games approach it, and the implications on different player segments.