The first thing you should ask yourself about LTV is why do you need to calculate it. If you have a mobile game that is monetizing through free 2 play model, it is safe to assume that lifetime value is going to be a leading KPI for your game. Here is why:
If your game hasn't launched yet, you still don't have any real data, just hypothetical one. From this reason, you would use the "naive method". Simply multiplying the ARPDAU by the expected lifespan of a single user. You can refer to this LTV calculator as an example of how to do it.
For soft launch phases, you would need a different method however. The situation here is different as you already have data points about the retention and monetiaztion of the game. Specifically, you would need the ARPDAU and at least the followign retention ratios: 2-day, 7-day, 14-day and 30 day. Modeling out the retention function is a complex mathematical excersise that requires statistical regression, log functions and integral calculations. These resources can do the trick for you:
As your game is ready for showtime, you will have a new requiment for lifetime value calculations. This time it will be about paid marketing and user acquisition. The goal here is to have a CLV that is higher than the cost of bringing a new user. However, you don't need this condition to existing for all your users but only for specific segments. Once you find a segment like that, you can invest in bringing more users from this segment. Since your game already has a user base and data points for a period of time, the LTV calculation for a new segment can relay on this data. You would already have the lifetime value from the existing user base simply because enough time has passed. The calculation for a new segment requires comparing the ARPU of the first 7-days between the new segment and the existing user base and than applying the same ratio to the LTV you already have. Here is a quicker way to do it with a calculator for new segment LTV