The 70/30 Rule: Balancing Game Time and Speaking Time in ESL

The problem with most gamified ESL lessons is one of ratios. A teacher commits to "making speaking more fun", reaches for a quiz tool, runs an engaging lesson, and at the end of the period realises their class produced about three minutes of English in total. The energy was high, the students enjoyed it, and almost no speaking happened.

The instinct to gamify is correct. The execution usually isn't. What's missing is a session-design rule that tells you when the game has crowded out the learning. The 70/30 rule is that rule.

YapYapGo is a classroom speaking practice tool for ESL teachers, designed so the speaking is the activity and the game layer runs on top. This post walks through the 70/30 rule, how to apply it to any ESL speaking lesson, and why most existing gamified tools cannot satisfy it.

The rule, stated simply

In any gamified ESL speaking lesson, the share of time spent on these two things should land roughly here:

• Speaking: 70 percent. Students producing English in pairs, groups, or full-class formats. This includes preparation talk where students discuss what they're about to say in English, but does not include silent thinking time.
• Game layer: 30 percent. Setup of teams, score animation, celebration of points, transitions, and the visible game state. This is the time the activity looks like a game.

A 50-minute lesson with these proportions delivers around 35 minutes of speaking and 15 minutes of game-layer time. Across a 12-week term that's about 7 hours of additional speaking practice compared to a lesson that flips the ratio.

The percentages are not magic numbers. They're a heuristic for noticing when you've drifted. Below 60 percent speaking, you're running a quiz. Above 80 percent, you've removed the gamification entirely and the energy will start to flag in a class that needs it.

How a YapYapGo session hits 70/30

A typical 25-minute YapYapGo speaking round looks something like this:

1. Minute 0-2: Team Maker pairs the class, topic is selected, Fun Mode toggled on. Game layer.
2. Minute 2-22: Pair work runs continuously. Every pair is speaking simultaneously. The Classroom Timer caps the round. Fun Run runners advance or Bubble Blast shots fire as each pair completes a question. Speaking.
3. Minute 22-25: Final scores resolve, animation plays, celebration. Game layer.

That's 20 minutes of continuous speaking out of 25, or 80 percent. The ratio sits comfortably inside the 70/30 band. Adjust round length and you can target the exact ratio you want.

The key structural feature: the speaking and the game run in parallel, not in sequence. Pairs don't pause to watch the game advance. The game advances because pairs are speaking. We have written elsewhere about how Fun Mode is engineered so the speaking drives the game, not the other way around.

How a Kahoot session typically lands

For comparison, run the same exercise on a typical Kahoot session in a class of 30:

1. Minute 0-5: Setup, game pin entry, students join, lobby. Game layer.
2. Minute 5-25: 20 questions, ~30 seconds each. One student "calls out" the answer per question in maybe a third of cases. Total student-minutes of English: roughly 2 to 3.
3. Minute 25-30: Final leaderboard, winner reveal, transition out. Game layer.

The Kahoot session produces 2-3 minutes of English from a 30-minute slot. That's 7-10 percent speaking, 90+ percent game. It violates the 70/30 rule by an order of magnitude. (Our head-to-head comparison with Kahoot walks through the maths in more detail.) Students enjoy it. They also barely speak.

This is not a Kahoot-specific problem - it's structural for any one-at-a-time quiz format. Blooket and Gimkit run into the same maths.

Why the rule depends on parallel speaking

The reason the 70/30 rule is achievable for YapYapGo and unachievable for quiz tools comes down to one design choice: parallel speaking vs serial speaking.

In a serial-speaking activity, one student speaks at a time while the other 29 listen. Even if you ran a "who can answer this English question fastest" game with perfect efficiency, your absolute ceiling is 1 student-minute of English per minute of clock time. A 50-minute lesson tops out at 50 student-minutes across the whole class. That's well under two minutes per student.

In a parallel-speaking activity, every pair is talking at once. A class of 30 split into 15 pairs produces 15 student-minutes per minute of clock time. A 50-minute lesson with even 20 minutes of pair work produces 300 student-minutes. The gap is roughly 6x.

The 70/30 rule is what it is because parallel speaking is what makes the ratio achievable. Without it, gamification is a fixed-budget trade-off: every extra minute of game is a minute of speaking you don't get. With parallel speaking, the game and the speaking aren't competing for the same minutes - they're stacked.

You can run the Team Maker and the Activity Timer to set up parallel pair work in any classroom, including for activities outside YapYapGo. The pattern works wherever you import it.

A quick test for your next lesson

After your next gamified ESL lesson, check the ratio. You don't need a stopwatch - rough percentages are fine. Ask:

• How many of my students were speaking English at the same time on average?
• For how many minutes was the speaking actually happening?
• For how many minutes was the class watching the game without producing English?

If parallel speaking was happening for at least two-thirds of the lesson, you're in 70/30 territory. If most of the lesson was spent watching one student or the screen, the game has become the activity and the rule is broken.

The bottom line

The 70/30 rule isn't a strict numerical target. It's a tool for diagnosing whether the gamification in your lesson is helping or substituting. If you can hit roughly 70 percent speaking, the game is doing its job - adding visible energy without stealing the speaking. If the speaking share drops to 10 percent the way it does in most quiz-tool sessions, the game has eaten the lesson.

Build the speaking first, layer the game on top, and use parallel-speaking formats so the maths actually works. That's the whole rule.

---

Sources:

• Student-time calculations follow the parallel-vs-serial speaking model in Nation and Newton (2009), Teaching ESL/EFL Listening and Speaking.
• 7-10 percent figure for Kahoot speaking share derived from typical classroom observations and YapYapGo internal analysis of comparable session designs.