Data Viz for Dummies III
The matchup lab · same questions, sharper substitutions
Try another view of the same question.
Keep the question and try a different chart. A dot can make ranking easier to scan; a violin can reveal a distribution that a box summarizes; a path through a scatterplot can reveal sequence. Compare each substitution with the familiar form and decide whether the extra detail helps the reader.
Substitution matrix · fit over fame
The alternative chart must reveal enough to pay for its extra reading cost.
Each alternative reveals a feature a familiar chart can compress: distribution shape, sequence, ancestry or grouping. Compare the views to decide whether that added detail helps.
Keep the familiar form when the alternative takes more explanation without improving the comparison.
Interactive substitution board
Why the challenger checks in
Every specialist carries a tradeoff. Toggle from the reason to make the switch to the reading burden you accept in return.Dot plot
when the baseline can whisper
Violin plot
when distribution shape matters
Ridgeline plot
when many shapes need silhouettes
Difference chart
when the gap is the protagonist
Connected scatter
when sequence explains the relationship
Tree diagram
when ancestry matters more than size
Circle pack
when nested scale needs visible containers
Dot plot · low-usage sniper
A bar loses its body and the value stops shouting.
Dot plots rank categories with less ink than bars. Position on a common scale still supports accurate comparison, while the empty baseline makes room for negative values, reference lines, and compact dashboards.
Bars can make the size of a difference more immediately apparent. Compare both forms when the gap between players is the main point.
Seven lineups · net rating
Tempo five leads; rookie run crosses below zero
A dot handles positive and negative ratings without bars extending in two directions. The common scale does the comparison work.Violin plot · film-room x-ray
The box plot opens its coat and reveals where the observations actually gather.
A violin plot shows the estimated distribution for each group. Wide sections mark common values. The inner summary shows the median and middle half, so you can compare both shape and spread.
The shape depends on the sample and the smoothing settings. Show the sample size and check how much the outline changes with those settings.
Shot mechanics · release time by attempt
At-rim attempts snap quickly; pull-ups spread and linger
Each violin contains 24 attempts. Its broad sections show common release times; the inner summary locates the middle of the sample.Ridgeline plot · deep rotation
Several distributions line up, overlap politely, and tell time sideways.
Ridgelines stack one-sided densities so a reader can scan how shape shifts across many groups. They work best when sequence already orders the rows—quarters, months, cohorts—and the silhouette itself is the comparison.
Overlap is a design choice, not free real estate. Too much hides baselines; too little wastes the form. Exact group-to-group comparisons remain easier in aligned small multiples.
Shot distance · period by period
The offense drifts outward after halftime
Q1 includes shots near the rim and farther out. Later quarters and overtime have more attempts in the twenties, near the three-point line.Difference chart · matchup scoreboard
Two lines stop competing for attention; the space between them takes the lead.
Difference charts fill the gap between two series and switch color when leadership changes. Use them when above-versus-below, actual-versus-expected, or candidate-A-versus-candidate-B is the analytical sentence.
The fill can make a tiny gap feel enormous if the scale is cropped. Keep the source lines visible and remember that colored area encodes difference, not a third measured series.
Actual score × expected score · twelve games
Rookie City alternates between overperformance and regression
Gold means actual scoring cleared expectation; blue means it fell short. Crossovers are the events, so the chart interpolates them instead of letting blocks collide.Connected scatterplot · transition playmaker
A scatterplot acquires memory, and the relationship becomes a journey.
Connected scatterplots combine two quantitative axes with an ordering variable—usually time. The path reveals loops, reversals, and phases that an ordinary point cloud discards. Color progression helps recover where the journey begins and ends.
Connection implies sequence, not causation. Crossing lines and many observations quickly create a knot, so label anchors, simplify the path, and consider animation or small multiples for longer histories.
Pace × offensive rating · game order
The season wanders before finding fast and efficient
The path starts at game one in the lower-left neighborhood and ends at game twelve in the upper-right, but the route includes two clear retreats.Tree diagram · coaching staff
Hierarchy stretches out its branches and shows exactly who descends from whom.
Trees spend space to make ancestry explicit. Every edge means parent-child; every level means depth. Use them for taxonomies, organizations, decision paths, and playbooks when structure matters more than how much area each leaf occupies.
A tree is not a generic network. Cross-links, shared parents, and loops violate its contract. If the relationships escape ancestry, choose a graph that can admit them.
Offensive playbook · family to action
Clock state branches into sets, then into the final call
The half-court branch has an extra level for sets between the family and its individual actions.Circle pack · roster bubbles
Hierarchy keeps its containers; magnitude rolls into nested circles.
Circle packing encloses children inside parents and sizes leaves by value. Compared with a treemap, it sacrifices space and precision for unmistakable grouping and an organic overview that helps readers browse nested clusters.
Circle area is difficult to compare, and packing leaves unused gaps. If ranking or efficient space usage is the job, choose bars or a treemap.
Roster allocation · 240 player-minutes
Position families contain the rotation they consume
The outer circles group guards, wings and bigs. Inner circle area represents minutes, and labels identify each player.Postgame · substitution review
Be clear about what the new view adds.
Name the feature you want readers to see, then compare how the two charts present it.
Rank with less ink
Dot plotPosition stays precise; magnitude becomes quieter.
Reveal distribution shape
ViolinKeep n and bandwidth in the scouting report.
Scan many distributions
RidgelineOrdered categories make the silhouettes cohere.
Make the gap primary
DifferenceShow both source lines and protect the scale.
Show a two-metric journey
Connected scatterOrder is data; crossings need landmarks.
Expose ancestry
TreeEvery edge must mean parent-child.
Browse nested magnitude
Circle packEnclosure first; exact comparison later.
No feature changes
Keep the starterNovelty has not earned a substitution.
The best substitution earns the reader’s effort by answering the question more clearly.