Summary
In service-game design the Session is the fundamental unit of success: one complete cycle of a player choosing to open the game, playing, and leaving with a reason to return. Where product games design primarily for Playtime (quality and length of a single sitting), service games must engineer the entire session cycle.
Nicholas Lovell breaks the Session into four components (Lovell, The Pyramid of Game Design, see source-pyramid-of-game-design):
- On-Ramp — making it easy for the player to choose to open the game right now
- Playtime — delivering fun, rewarding or meaningful gameplay
- Off-Ramp — signalling that the session has reached a natural close
- Return Hook — giving the player a concrete reason to come back
This framework applies across platforms, though its weighting varies: the On-Ramp is most critical on mobile, where the game competes directly with social media for attention; the Return Hook matters most in any game seeking daily or near-daily engagement.
Key Ideas
The Attention Problem
Herbert Simon (1971) predicted that “a wealth of information creates a poverty of attention.” In the 21st century, every service game competes not just with other games but with Facebook, Instagram, Netflix, YouTube and every other application on the player’s device. On mobile in particular, the first battle a service game must win is the battle to be opened — not the battle to be enjoyed.
Implications:
- A game with a weak On-Ramp can have excellent Playtime and still fail commercially, because players never get in.
- Mobile games compete with primetime television (comScore 2013: 64% of UK mobile gaming happens at home, primarily in evenings, not on commutes).
The On-Ramp
The On-Ramp is distinct from onboarding (helping a new player understand the game for the first time). The On-Ramp is the friction cost of every session — the barrier a player must clear each time they decide whether to open the game or do something else.
The Starbucks Test (popularised by Torsten Reil, CEO of NaturalMotion): “Have I got time for a meaningful interaction in the time it takes the barista to make my macchiato?” This question decomposes into two parts:
- Have I got time for this game right now?
- Will the game provide sufficient fun or progress for the time and effort I must invest to start playing?
Factors that help the On-Ramp:
- Short load times
- Instant rewards visible immediately on loading
- “Not tidy your room” — opening the game should feel like arriving at a welcoming world, not receiving a list of chores
- “One more go” gameplay — short, repeatable Base Layer units
Factors that hurt the On-Ramp:
- Splash screens and developer logos eating load time
- High-end 3D engines (slower to load — a real trade-off, not an absolute problem)
- Requiring online connectivity and matchmaking before gameplay begins
- Interrupting menus (e.g. asking “which save file?” when the player has only one)
- No suspend behaviour — games that crash or log out when placed in the background punish players who need to be interruptible
“Regardless of what platform it’s on, clicking on a game icon should make me feel good. Games are about fun and escapism. When I come back to your game, my first feeling should be ‘ah, I’m glad I opened this game!‘” (Lovell, The Pyramid of Game Design, Ch. 5)
Activation energy (CCP’s framing): different platforms demand different levels of commitment before play begins. A mobile idle game has very low activation energy; VR has very high activation energy. Design decisions appropriate for one level of activation energy do not transfer automatically to another.
The Session-Length Paradox
If you shorten the promise of session length, the actual session length may increase.
The promise of session length is the player’s expectation of the minimum time investment required before the game delivers value. If a player believes they need two uninterrupted hours to make meaningful progress, they will not open the game unless they have two hours available. If the game promises “you can get something done in five minutes,” they may open it in a five-minute gap — and stay for an hour.
Classic example: PopCap’s Bejeweled Blitz promises exactly 60 seconds of gameplay per round. At its height, average session length was 43 minutes. Supercell’s Clash Royale offers a 30-second On-Ramp (open a chest, start a new countdown) and three-minute matches; average sessions regularly extend far longer.
This paradox also applies to Retention metrics: addressing On-Ramp problems often raises the number of sessions per day more than it changes session duration.
Playtime
Playtime is governed primarily by the Base Layer (see pyramid-of-game-design). The key additional concept here is pottering: the low-intensity, relaxed activity that allows players to extend sessions voluntarily without burning out.
Pottering takes many forms:
- Enjoyable make-work — rearranging inventory, optimising factory layouts, decorating a farm for aesthetic reasons
- Theory-crafting and world exploration — reading lore, experimenting with builds, browsing tech trees (appeals to Bartle Explorers and Quantic Foundry’s Immersion motivation)
- Self-expression — customising avatars, arranging spaces; more monetisable when visible to other players
Pottering solves the chocolate cake problem: consuming too much enjoyment at once leads to satiation and eventual aversion. Service games need to cater both to players with five minutes and players with two hours — pottering gives extended-session players something satisfying to do after the main tasks are complete.
The Off-Ramp
Service games are not trying to maximise session length per session. They are trying to maximise the number of sessions a player has over weeks and months. Exhausting a player in one long session can reduce the total play relationship.
The Off-Ramp signals to players that the session’s main tasks are done and that it is acceptable to leave. Key principles:
- Signal completion, do not force it
- Let players stay if they want (pottering)
- Plant the Return Hook before they go
Common Off-Ramp mechanisms:
- Energy / life systems — the player runs out of resources needed to continue (e.g. Candy Crush Saga’s lives, Marvel Contest of Champions’ energy). Controversial but effective. Lovell’s view: treat energy as a retention mechanic first, not a monetisation tool. Energy signals “you have done the efficient thing — it is OK to leave now.”
- Natural task completion — all crops planted, all machines running, all daily quests finished (e.g. Hay Day)
- Match structure — PvP games end sessions naturally at match boundaries (Hearthstone, Clash Royale)
- Energy as encouragement towards challenge — in CSR Racing, scarce fuel pushes players towards attempting the next boss rather than grinding easy races, creating a meaningful final-decision moment that closes the session on a high note
“A player told me she would not play Candy Crush Saga if it did not have the life system. ‘I know that when I play Candy Crush, I will eventually run out of lives and have to stop. I don’t have the willpower to make myself stop, so if it weren’t for the lives, I would never play the game at all.‘” (Lovell, Ch. 6)
Return Hooks
Return Hooks are the tactical counterpart to the Retention Layer’s strategic design. Where the Retention Layer provides structural reasons to keep playing across many sessions, Return Hooks are the specific triggers that prompt the player to start their next session.
Lovell identifies eight types:
| Hook | Description | Notes |
|---|---|---|
| Welcome Back | Reward players for returning — show what they earned while away (Adventure Capitalist) | Sets the emotional tone; “returning emperor” effect |
| Rewards | Fixed-timer free items (e.g. Marvel Contest of Champions crystal every 4h / 24h) | Builds regular return habit |
| Accumulator Rewards | Daily login bonuses that increase with consecutive days; reset on a miss | Loss aversion drives return; cap at 5–7 days typical |
| Appointment Mechanics | Player-chosen countdown obligations (harvest crops, open chest) | Agency distinguishes this from a Tamagotchi demand |
| Decay | Punishment for not returning (crops wither, base gets raided) | Fallen out of favour in casual games; still used in competitive genres where it fits the metaphor |
| Live Events | Seasonal / limited-time content and competitions | Only appropriate once Core Loop is proven; not a substitute for a weak Core Loop |
| Social | Turn-based obligations (Words with Friends), guild commitments (World of Warcraft raids) | Most powerful long-term hook; most expensive to build |
| Nudge | Local notifications / push alerts | A nudge amplifying other hooks, not a hook in its own right; opt-in rates typically low |
Critical caveat on Return Hooks: Many are easy to implement and can mask a weak Core Loop. Lovell recommends leaving most Return Hooks — daily bonuses, live events, social features — out of soft launch, so that the game’s intrinsic retention can be measured cleanly. Implementing them too early creates a false positive: retention looks adequate, but when events are removed or the novelty fades, the underlying Core Loop is exposed as weak.
“You should not use events as your primary Return Hook until you have proven that you can get high quality retention without them.” (Lovell, Ch. 6)
In Practice
Design the On-Ramp as a separate problem from onboarding. The first-time user experience (FTUE) matters, but every subsequent session also has an On-Ramp cost. Audit the time from app icon to first meaningful action for returning players, not just new ones.
Leave Return Hooks out of soft launch. This is counter-intuitive but important. A clean soft launch — without daily bonuses, live events, or social sharing features — reveals the true retention quality of the Core Loop. Add Return Hooks once you are confident the game works without them.
Design the Off-Ramp deliberately. Most studios design the On-Ramp and ignore the Off-Ramp. A game that players feel traps them, or that they can only stop playing by frustration or exhaustion, will not sustain long-term daily engagement.
Treat energy as a design choice, not a default. Energy systems solve the Off-Ramp problem and signal session completion, but they are clunky and can be perceived as exploitative. Explore other Off-Ramp solutions — task completion, match boundaries, narrative chapter breaks — before defaulting to energy.
Match the Return Hook to the metaphor. A wither mechanic in a farming game feels punitive because it frames absence as neglect. A base raid in Clash of Clans feels natural because competitive threat is part of the genre’s metaphor.
Evidence
“The Session-Length Paradox: if you shorten the Session-length promise, the Session length may get longer.” (Lovell, The Pyramid of Game Design, Ch. 5)
“The key message is that successful service games are not focused on extending the Session length for their players. They are trying to keep players playing for the long haul.” (Lovell, Ch. 6)
“Local notifications are a useful tool to make your Return Hooks more powerful. They are not a Return Hook in their own right.” (Lovell, Ch. 6)
Implications
- Session design is a first-class design concern, not an afterthought added after the core game is built. The Off-Ramp and Return Hook in particular are frequently overlooked in studios with a product-game heritage.
- The competition is not other games — it is everything on the device. On mobile especially, designing for retention means designing against social media habits, not just against rival games.
- Retention hooks added too early obscure fundamental design weaknesses. This is one of the most practically actionable insights from the Pyramid framework for student developers building prototype service games.
Open Questions
- How does the On-Ramp / Session model transfer to PC and console live-service games, where load times and platform boot sequences are longer but players expect longer sessions?
- Do Return Hooks have diminishing returns as players become habituated to daily login bonuses across multiple games simultaneously?
- The chocolate cake metaphor implies an optimal session length for each player — can this be measured and used to personalise Off-Ramp signals?
Related
- pyramid-of-game-design — the Pyramid framework; the Session is its tactical layer
- game-loops — nested loops within Playtime
- reward-systems — reinforcement schedules underpinning Return Hooks
- f2p-monetisation-design — monetisation during the Off-Ramp (energy purchase) and Return Hook (chest openings)
- dark-patterns — energy mechanics, decay, and the line between persuasion and manipulation
- games-as-a-service-development — Dev Agency / Live Agency and when to implement Session features
- post-launch-and-live-ops — live events as Return Hooks in the live phase
- source-pyramid-of-game-design