Categories and Identifiers:

Although most guidelines in the current collection are for designing visual mappings and representations, we include a high-level categorization based on types of visual analytics papers to enable the continuing expansion of the VGR. These categories are referred to as the primary categories, including:

When each primary category is to be further divided, we introduce secondary categories. The identifier is thus a string, starting with a hash (#), followed by a capitalised letter for the primary category, two letters for the secondary category and a short string for the keywords in a guideline. For example: The guideline proposed by Edward Tufte, "Good visualizations should maximize data-ink ratio", is assigned the identifier #DGeDataInkRatio, where D is for the primary category Design and Ge is for the secondary category general. We use xx to indicate that a guideline is yet to have a secondary category.

The guideline sources are classified as being either
Category Hashtag Description Source

The guidelines in this category are normally for advising how to design, develop, and use algorithms and techniques for visualization and visual analytics systems, The secondary categories are yet to be defined.

Category Hashtag Description Source
Algorithm »
Algorithm#AxxSoundAlgorithmAlgorithm used for visualization must be sound: It shouldn't crash, it should be able to do what it was designed for. It should represent a wide range of datasets. It should withstand variation in data complexity (size, dimensionality, time variance).  

The guidelines in this category are normally for advising how to design visual mappings, visual representations, interactions, and other HCI components that may be used in visualization and visual analytics systems. The secondary categories are ordered alphabetically.

Category Hashtag Description Source
BarCharts(BC) »
Bar Charts (BC)#DBCErrorBarHarmfulError bars considered harmful. (Primary) Correll and Gleicher, "Error Bars Considered Harmful: Exploring Alternate Encodings for Mean and Error." TVCG, 20(12):2142--2151, 2014 
Bar Charts (BC)#DBCHorizontalLabelUse horizontal labels. Avoid steep diagonal or vertical type, as it can be difficult to read.  
Bar Charts (BC)#DBCConsistentColorsUse consistent colors. Use one color for bar charts. You may use an accent color to highlight a significant data point.  
Bar Charts (BC)#DBCNo3DDon’t use 3D effects either, especially in bar graphs. By making the bars look like cubes, the tops become obscured and it is difficult to discern where the top of the data really ends. (Primary)" 
Bar Charts (BC)#DBCBarOrderingOrder data appropriately. Order categories alphabetically, sequentially, or by value.  
Bar Charts (BC)#DBCBarSpacingSpace bars appropriately. Space between bars should be ½ bar width.  
ChartJunk(CJ) »
Chart Junk (CJ)#DCJBad Chartjunks are bad, do not use it.  
ColorsinVisualization(Co) »
Colors in Visualization (Co)#DCoMax6ColorsDon't use more than 6 colours together. Too many colours mean similar hues will appear, like BLUE and GREEN which can be difficult to tell apart. (Primary) 
Colors in Visualization (Co)#DCoNoRedGreenDon't use [RED and GREEN] or [ORANGE and GREEN] to make comparison on the same chart. About 10% of men are color blind with red/green being the most common form of color blindness. (Primary) 
Colors in Visualization (Co)#DCoRainbowHarmfulRainbow color Map is considered harmful. (Primary) D Borland and M T Russell II, Rainbow Color Map (Still) Considered Harmful, CG&A, 27(2), 2007. 
Colors in Visualization (Co)#DCoRecommendationIf in doubt, use colour scales from recommendation tools such as ColorCat ( or ColorBrewer (  (Primary) 
Colors in Visualization (Co)#DCoColorSymbolismUse colours judiciously to indicate relationships and choose colour palettes that facilitate the message conveyed in the figure. (Tertiary) 
Colors in Visualization (Co)#DCoCheck4ColorblindDouble-check your colors for the color blind. You can use tools such as ColorOracle ( or ColorCat ( to simulate the effect of different types of color blindness. (Primary) 
Colors in Visualization (Co)#DCoColorblindFriendlyMake figures colourblind-friendly whenever possible. (Tertiary) /six-guidelines-for-scientific-data-visualization/ 
General(Ge) »
General (Ge)#DGeAvoidVisualEffectAvoid using visual effects in graphs. (Primary) Data Visualization 101: How to design Charts and Graphs 
General (Ge)#DGeCauseEffectVisual displays of information should present both cause and effect. (Tertiary)  
General (Ge)#DGeDataInkRatioGood visualizations should maximize data-ink ratio. (Primary) Edward R Tufte, The Visual Display of Quantitative Information, Graphics Press, 1983. (Secondary) (Secondary) 
General (Ge)#DGeExplainDataExplaining the data helps viewers see the relevance in the information.  (Tertiary) 
General (Ge)#DGeNoVisualMathDon't make users do "visual math".  (Tertiary) 
General (Ge)#DGeConsistencyBe consistent in choosing colours, symbols, and fonts for visual representations. (Tertiary) 
GraphsandNetworkVisualization(GN) »
Graphs and Network Visualization (GN)#DGNChangeLayoutChange the layout for the graphical design to improve the readability of the design. Primary) Introduction to Information Visualization. Ricardo Maaza 
Graphs and Network Visualization (GN)#DGNAllowManipulationImplement interactivity so that the user can manipulate the representation to match their needs.  
Graphs and Network Visualization (GN)#DGNOverallStructure Create an approximation of the overall structure but reduce the complexity so that it may be easier to comprehend. This can be achieved by either reducing the volumes of edges displayed on a graph or by removing relationships entirely if they are not of specific interest.  
Interaction(In) »
Interaction (In)#DInNoBlowApart Do not use blow-apart effects. (Tertiary) 
Interaction (In)#DInDirectInteraction Allow for direct interactions with objects that reveal new insights (e.g., sorting via drag). (Tertiary) 
Interaction (In)#DInSparseInteractionUse interaction in visualization sparsely and cautiously. (Tertiary) /2008/1414/pdf/07291_abstracts_collection.1414.pdf 
Interaction (In)#DInOverviewFirstOverview first, zoom and filter, details on demand. (Primary) B Shneiderman, The Eyes Have It: A Task by Data Type Taxonomy for Information Visualizations. Proc. IEEE Symposium on Visual Languages, pp.336-343, 1996. (Tertiary) 
PieCharts(Pi) »
Pie Charts (Pi)#DPi100PercentMake sure all data adds up to 100%. Verify that values sum up to 100% and that pie slices are sized proportionality to their corresponding value.  
Pie Charts (Pi)#DPi5CategoriesVisualize no more than 5 categories per pie chart.  
Pie Charts (Pi)#DPiAvoidComparisonDo not use multiple pie charts for comparison as slices sizes are very difficult to compare side-by-side.  
Pie Charts (Pi)#DPiPieChartBadPie charts are bad and 3D pie charts are very bad. (Tertiary) (Tertiary) 
Pie Charts (Pi)#DPiPieChartBad Order slices correctly by first placing the largest section at 12 o’clock stretching clockwise, and then placing the remaining sections consistently either clockwise or counter-clockwise in descending order. (Primary) 
Process(Pr) »
Process (Pr)#DPrAskOthers Do ask for the opinion of others after designing your visualizations.  
Process (Pr)#DPrSquintTestVisualization should pass the Squint test. When you squint at your page, so that you cannot read any of the text, you should still 'get' something about the page. (Primary) /blog/2006/09/simplicity_of_d.html 
TextandDocumentVisualization(TD) »
Text and Document Visualization (TD)#DTDColoringTextTagsAll nouns should be encoded in black, verbs in red, adjectives in green, determiners in grey, particles in brown, conjunctions in blue, and Interjection in yellow. (Primary) W Weber. Information Visualization, 2007. IV'07. 11th …, 2007 - 
Time-seriesPlots(TS) »
Time-series Plots (TS)#DTSMax4Lines Don't plot more than 4 lines in one chart. If you need to display more, break them out into separate charts for better comparison. (Primary) 
Time-series Plots (TS)#DTSTwoThirdHeightRuleSet the height of a line chart such that the data in the line chart takes up approximately two-thirds of the y-axis’ maximum scale. (Primary) 
Time-series Plots (TS)#DTSUseLinesUse lines when connecting sequential data in time-series plots. (Primary) Kelleher, C., Wagener, T., 2011. Ten guidelines for effective data visualization in scientific publications. Environmental Modelling and Software 26 (6), 822e827. (Secondary) Strange, N., 2007. Smoke & Mirrors: How to Bend Facts & Figures to Your Advantage. A & C Black Publishers, London, UK 
Time-series Plots (TS)#DTSUseSolidLinesUse solid lines only because dashed and dotted lines can be misleading. (Primary) 
Time-series Plots (TS)#DTSLabelLinesDirectlyLabel the lines directly to enable readers identifying lines quickly using corresponding labels instead of referencing a legend. (Primary) /Data_Visualization_101_How_to_Design_Charts_and_Graphs.pdf 
Time-series Plots (TS)#DTSShowZeroBaseline Always include a Zero Baseline if possible. (Primary) 
Category Hashtag Description Source
Category Hashtag Description Source
Category Hashtag Description Source