Concurrency, not sequence
Timeline
Overlap is the point: 277 of these 277 ran concurrently with at least one other, which is the thing a list of dynasties by region cannot show you. Rows are sorted by cited start date, so a neighbour on this axis is a contemporary — and standing at a year says who else was there. A soft bar end means the sources disagree about when it started or stopped.
How many at once
peak 74 in the 1200s
Spans
Searches names and regions
Drag to see who was standing
Why some rows have no peak
The phase vocabulary — formation, expansion, peak, contraction, end, afterlife — is optional, and the rows that do not use all of it are the reason. The Ghurids barely had a golden age before Khwarazm ended them, and the Tahirids never expanded: they were granted a province and kept it. Forcing every polity through the same 6 acts would mean writing something untrue about at least two of them, so the tag is left off and the row shows the absence.
The spine sits beside the name rather than on the chart. Chapters carry no dates of their own, and a mark placed at a horizontal position on an axis labelled in years is a date whether or not it was meant as one — the same class of error as inventing a figure, committed in pixels. The spine says which phases are written. It says nothing about when they were.
One mark is on the axis, and it is cited: where a source dates its own extent figure to a year, that year is drawn. Sixty of these records carry one.