q value

Understanding p value can be confusing. People who get it have an edge over those who don’t.

To make it easier, use q-value.

q = 1 – p. It tells us the probablity (95% or above) that the experiment worked.

Suppose you run an experiment and get p = 0.04. That means q = 0.96. So, there is a 96% chance that your experiment is successful.

Every experiment starts with a hypothesis. For example: “Adding Feature X will improve LTV.”

To test this, we create a null hypothesis, the opposite. So: “Adding Feature X does not improve LTV.”

The p value tells us how likely the null hypothesis is true.

If p is low (less than 5%), we say the null is false. That means our idea is likely right. If p is high, we keep the null. Our idea might not work.

Example:

You want to test if changing the button color will increase clicks.

Null hypothesis: “Changing the color of the button will not increase clicks.”

You run the experiment and get the p-value.

If p is high (say, above 0.05), it means your change didn’t help. If q is high (say, 95% or more), your change worked.

Think of it like this:

High p → Experiment likely failed

High q (or Low p) → Experiment likely succeeded

Use q to simplify decisions. It it more intuitive and helps your team move faster.

Making Decisions

To Take a Decision Faster

1. Do you have enough information? Note that the marginal gain from having additional information is rarely worth the effort.

2. Is it high impact? If yes, go ahead. If not, ask whether it’s worth it.

3. Is it expensive in terms of energy and price? If it’s low, go ahead. If high, reconsider relevance.

4. Is it reversible? If so, make the decision more quickly. Hiring someone is reversible; letting someone go is not.

Writing

Scott Adoms on Good Writing

1. Business writing = clarity + persuasion. Keep it simple.

2. Short sentences work best. Cut extra words.
   For example, “He was very happy” → “He was happy”.

3. First sentence should grab the attention of the reader. Rewrite until it is perfect.

4. Brain imagines the object before the action. So write in an active voice.
   For example, “The boy hit the ball” is better than “The ball was hit by the boy”.

D1 Retention

If your game’s D1 retention is below 30%, stop adding features. There is a deeper issue. The core loop is not fun. Scaling a game with a broken loop will not help it grow.

D1 retention gives an early signal about the fun. Sometimes, onboarding is broken. Users do not reach the gameplay screen. This reduces Day 0 gameplay time, which lowers D1. If users enjoy the game and play for a few minutes on Day 0, they are more likely to return. That is why D1 is closely tied to D0 gameplay.

Increasing Retention

Retention is the percentage of users who come back to the game again after installation after a given time period. For example, if D1 is 30%, 30% of the users come back again on the next day of installation. I believe retention is a primary KPI in LTV as we can figure out how to make money from retaining users. If there is no retained user, there is no money.

How to increase Retention?

Increasing it depends following three things

1. Users should like the game
2. Users should think of coming back again to the game
3. Users should return to the game