Presenting Data e.g., bronze, silver, gold ordered e.g., support, - PDF document

3/31/2017 Types of Variables IMGD 2905 • Qualitative (Categorical) variables – Can have states or subclasses • e.g., rank: [platinum, diamond, gold] – Can be ordered or unordered Presenting Data • e.g., bronze, silver, gold  ordered • e.g., support, tank, jungler  unordered • Quantitative (Numeric) variables Chapter 2 – Numeric levels – Discrete or continuous • e.g., gold per minute, deaths, character level • e.g., kills + assists / deaths ratio, win percentage Variables Qualitative Quantitative Ordered Unordered Discrete Continuous 2 Categorical: Bar Chart Outline • Chart containing rectangles (“bars”) where length represents count, amount, or percent • Types of Charts (next) • Better than table for comparing numbers • Guidelines for Charts “Exploring Exer-Walls as a Healthy Alternative to Paywalls in Mobile Games” • Common Mistakes http://www.cs.wpi.edu/~claypool/mqp/paywall/ Demo: imgdpops.xlsx Note: bars could be sideways, too Categorical: Pareto Chart Categorical: Pie Chart • Wedge-shaped areas • Bar chart, arranged (“pie slices”) – most to least represent count, frequent amount or percent of • Line showing each category from whole cumulative percent • Best if few slices since quantifying “size” of • Helps identify most pie difficult common • Comparing pies also Sort. difficult New column for percent [=B2/SUM(B$2:B$12)] “The Effects of Latency and Jitter on a First Demo: imgdpops.xlsx New column for running [=SUM(D$2:D2)] Person Shooter: Team Fortress 2” Note: $ “locks” value in (e.g., B$12 versus B12) Demo: imgdpops.xlsx http://www.cs.wpi.edu/~claypool/iqp/tf2/ Insert combo plot 1

3/31/2017 Numeric: Frequency Distribution Categorical: Cross-Classification Table • Multi-column table that presents count or percent for 2+ • Groups of numeric values • May include percentage categorical variables and frequency – Good for comparison across multi-categorical data • Typically equal size • e.g., Survey of Champion – Sometimes ends are open “skins” bought with RP (for extremes) • Bin size/number variable – 1, 2, 1, 0, 3, 4, 0, 1, 1, 1, 2, Poll class! 2, 3, 2, 3, 2, 1, 4, 0, 0 – Too many and not readable – Cluster into groups – Guide: – Report frequency per group • 100 or less 7-10 • 101-200 11-15 Skins Freq. Percent • 200+ 13-20 0 4 20% Insert Pivot Chart 1 6 30% Select Major through Grade 2 5 25% Demo: grades.xlsx Drag Majors to Axis 3 3 15% Drag Grade to Axis 4 2 10% Drag Grade to Values Histogram Cumulative Distribution • Bar chart for grouped numerical data – No (or small) gaps btwn adjacent bars • Cumulative amount of Ages of professional data with value or less League players • Easy to see min, max, median • Compare shapes of distributions https://www.mathsisfun.com/data/images/bar-chart-vs-histogram.gif Demo: lol-patches.xlsx https://www.reddit.com/r/leagueoflegends/comme Select Banrate data nts/4x5s9m/analysis_of_age_in_league_of_legends/ Sort low to high New column for percent [=ROW()/42] Demo: grades.xlsx Select column  paste down all Select both columns Select GPA data “Nerfs, Buffs and Bugs - Analysis of the Insert  Scatter plot with lines http://www.leaguemath.com/e Insert  Statistics Chart  Histogram arly-vs-late-game-champions/ Impact of Patching on League of Legends” Can adjust bins, overflow/underflow http://www.cs.wpi.edu/~claypool/papers/lol-crawler/ Time Series Plot Stem and Leaf Display • Associate data • “Histogram-lite” for analysis w/out software with date – e.g., exam scores: 34, 81, 75, 51, 82, 96, 55, 66, • Line graph with 95, 87, 82, 88, 99, 50, 85, 72 dates (proportionally 9| 6 5 9 spaced!) 8| 1 2 7 2 8 5 http://www.soundandvision.com/content/violence-and-video-games 7| 5 2 http://www.polygon.com/2014/9/12/6141515/do- violent-video-games-actually-reduce-real-world-crime 6| 6 Demo: majors.xlsx 5| 1 5 0 Sel. year and majors Insert  Line Chart 4|  More Line Charts 3| 4 11 2

3/31/2017 Radar Plot Gold compared to average, LoL NA teams, by role Scatter Plot • Also called • Two numerical variables, one on each axis “star charts” • Reveal patterns in relationship or “kiviat • Setup “right” models (later) plots” • Good for quick visual “Intelligent Simulation of compare, Worldwide Application Distribution especially for OnLive's Server Network” when axes http://www.cs.wpi.edu/~claypool/mqp/onlive/ unequal Demo: lol-rates.xlsx Demo: lol-rates.xlsx Select top line {win, Select two of {win, pick, ban} + 1 row pick, ban} num Insert  scatter plot Insert  Other  Radar scatter plot http://www.thescoreesports.com/lol/news/2561-using-gold-distribution-to-understand-team-dynamic-global-na-lcs-and-lpl 14 Many More Charts! Game Analytics Charts https://en.wikipedia.org/wiki/Chart Gunter Wallner and Simone Kriglstein. “An Introduction to • Bubble • Gantt Gameplay Data Visualization”, Game Research Methods , pages • Waterfall • Nolan 231-250, ETC Press, ISBN: 978-1-312-88473-1, 2015. • Tree • Pert http://dl.acm.org/citation.cfm?id=2812792 • Gap • Smith • Polar • Skyline • Player choices (e.g., build units) • Violin • Vowel • Density of activities (e.g., where spend time on map) • Candlestick • Nomogram • Movement through levels • Kagi • Natal • If common chart effective for message, use • Learn/use other charts as needed Player Choices – Pie-Chart Player Location – Heat Map (1 of 2) (Custom game, comparative study) 3

3/31/2017 Player Location – Heat Map (2 of 2) Movement (1 of 2) Assassin’s Creed Where play testers failed Result: Make red areas easier (game: Infinite Mario , clone of Super Mario Bros.) http://www.gamasutra.com/blogs/JonathanDankoff/20140320/213624 /Game_Telemetry_with_DNA_Tracking_on_Assassins_Creed.php Game: DOGeometry - build road to veterinary house Movement (2 of 2) Player Behavior - Node-link Shows exploration, where stuck Outline Guidelines for Good Charts (1 of 5) • Types of Charts (done) • Require minimum effort from reader • Guidelines for Charts (next) – Perhaps most important metric – Again, “art” not “rules”. Learn with experience. – Given two, can pick one that takes less reader Recognize good/bad when see it. effort • Common Mistakes a a b c b e.g., c Direct Labeling Legend Box https://xkcd.com/833 24 4

3/31/2017 Guidelines for Good Charts (2 of 5) Guidelines for Good Charts (3 of 5) • Minimize ink (1 of 2) • Maximize information – Make self-sufficient – Maximize information-to-ink ratio – Key words in place of – Too much unnecessary ink makes chart cluttered, hard symbols to read • e.g., “Gold IV” and not • e.g., no gridlines unless needed to help read “Player A” • e.g., “Daily Games Played” – Chart that gives easier-to-read for same data is not “Games Played” preferred – Axis labels as informative as possible • e.g., “Game Time (seconds)” .1 1 not “Game Time” • Same data – Help by using captions (or • Downtime = 1 – uptime http://www.phplot.com/phplotdocs/conc-labels.html title, if stand-alone) • Right “better” • e.g., “Game time in seconds versus player skill in total hours played” Uptime Downtime 25 26 Guidelines for Good Charts (3 of 5) Guidelines for Good Charts (4 of 5) • Minimize ink (2 of 2) • Use commonly accepted practices – Present what people expect – e.g., origin at (0,0) vs. – e.g., independent (cause) on x-axis, dependent (effect) on y-axis – e.g., x-axis scale is linear – e.g., increase left to right, bottom to top – e.g., scale divisions equal • Departures are permitted, but require extra effort from reader  so use sparingly! 28 Checklist for Good Charts Guidelines for Good Charts (5 of 5) • • Axes Scale – Are both axes labeled? – Are units increasing left to right (x- axis) and bottom to top (y-axis)? – Are the axis labels self-explanatory • Avoid ambiguity – Do all charts use the same scale? and concise? – Are the scale and divisions shown on – Are the scales contiguous? – Show coordinate axes both axes? – Is bar chart order systematic? – Are the min and max ranges • at right angles – Are bars appropriate width, spacing? appropriate? • Overall – Show origin – Are the units indicated? – Does the whole chart add information • Lines/Curves/Points vs. • usually at (0,0) to reader? – Is the number of lines/curves – Are there no curves/symbols/text – Identify individual curves reasonably small? that can be removed and still have – Are curves labeled? the same information? and bars – Are all symbols clearly distinguishable? – Does the chart have a title or caption • With key/legend or label – Is a concise, clear legend provided? (not both)? – Is the chart self-explanatory and http://www.carltonassociatesinc.com/images/confusion-new.jpg – Does the legend obscure any data? – Do not plot multiple concise? • Information – Do the variables plotted give more variables on same chart – If the y-axis is variable, is an indication information than alternatives? of spread (error bars) shown? • Single y-axis – Is chart referenced and discussed in – Are grid lines required to read data (if any accompanying report? not, then remove)? 29 5