![]() ![]() update_layout ( yaxis_tickformat = "0%" ) line3. update_yaxes ( range =, tick0 = 0, dtick = 0.1 ) line3. columns = # Pareto Chart linec = make_subplots ( specs =, tick0 = 1, dtick = 1 ) line3. sum () * 100 cumulative_per = cumulative_per. ![]() sort_values ( 'player_id', ascending = False ) # Generate CumSum cumulative_per = countdf. reset_index () # Arrange data according to amount of levels countdf. reset_index () # Count the number amount of players according to amount of levels recorded by player countdf = countdf. This will tell us, from the episode how many levels have each player recorded in the lapse of 7 days.įrom plotly.subplots import make_subplots # Group data of amount of levels recorded by player id countdf = df. Interpretation: Most of the players are casual gamers because 75% of them complete the level and don’t repeat itįirst, let’s examine the number of registries per player.Furthermore, there are players with 0 attempts, so we need to evaluate if this is present at level 1, which can explain a problem in retention rate for that episode Interpretation: The registries are consistent, the interquartile range mention that half of the players try between 1 and 7 time to complete each level.50% of the records are equal to or less than level 9 Interpretation: They’re registered as numbers, but for further analysis will be transformed as factors.Also, the analysis won’t consider this as a lapse per player since the records per player are not continuous, so they will be limited as a timestamp Interpretation: Only includes data from January 1st to January 7th of 2014.Interpretation: Not unique and counts with 6814 distinct values which make sense since there is a player with records of multiple levels.□ Exploratory Analysis & In-game interpretations Summary statisticsĮxcluding the outliers we mentioned before, we got the next conclusions about their distribution and measurement: □ About the data Collection process and structureīefore start let’s import the libraries we’re going to needĭf = df df. Note: To facilitate the understanding of the roles of the development team, I invite you to take a look at this diagram that I designed. Players' community: They expect to have an endurable and great experience with a brief response in case of disconformities.Executive Producer: Besides Candy Crush being an IP with internal producers since it’s developed and published by King, the parent company will expect to have an ROI aligned with their expectations.Gameplay Engineer: They require to start working on the difficulty adjustment patch as soon as they receive the final statement.Mobile Designer & User Retention Expert: This is a game whose main income input is in-game purchases because it’s a F2P, the main source of income is centered in retain the engagement in the game and keeping the consumers on the platform.Level Designers: They work aligned with the rest of the Engineering Team because they still have a backend perspective and their next patch release needs to be aligned with the insights given by the analyst.So they require a Data Analyst to help with this task since the developers are seeing only the backend factors affecting the game, but it’s also critical to consider those external ones that affect the experience for the player and the sustainability of this game for the company. None of the past hypotheses are the main intentions of the developers. $H_1:$ The game is too hard so the players leave it and become frustrated.$H_0:$ The game is too easy so it became boring over time.That’s why our diagnosis will start from 2 potential hypothesis: From the perspective of a customer, there can be several points of view that can emerge and, at the same time, can become unnoticed. We’ll review a game that potentially can lead any developer to many unseen problems, considering the abundance of levels. ![]()
0 Comments
Leave a Reply. |
Details
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |