Two poster sizes, one catalogue item
Counting what a web page offers
One saved SeedMivo homepage contained 194 image candidates inside 97 responsive image elements. Counting the advertised images would double-count each placement. Of those placements, 96 linked to a title page. They pointed to 76 different catalogue paths because some titles appeared more than once. The remaining placement was the large hero backdrop.
Ninety-six elements offered 185-pixel and 342-pixel widths. The hero backdrop offered 780-pixel and 1280-pixel widths. Each pair shared one image filename. Across the whole sample there were 77 distinct filenames.
Count the units separately
| Unit counted | Count |
|---|---|
| Responsive image elements | 97 |
| Advertised image candidates | 194 |
| Paired candidates with the same filename | 97 |
| Distinct image filenames | 77 |
| Elements linked to catalogue paths | 96 |
| Distinct linked catalogue paths | 76 |
Show the R calculation and output
candidates <- read.csv("candidates.csv", stringsAsFactors = FALSE, na.strings = character())
placements <- split(candidates, candidates$element)
same_file <- vapply(placements, function(p) length(unique(p$asset)) == 1L, logical(1))
width_sets <- vapply(placements, function(p) paste(sort(p$width), collapse = " + "), character(1))
first_candidate <- candidates[!duplicated(candidates$element), ]
linked <- first_candidate[nzchar(first_candidate$linked_title_path), ]
stopifnot(all(same_file), all(table(candidates$element) == 2L))
cat("Responsive image elements:", length(placements), "\n")
cat("Advertised image candidates:", nrow(candidates), "\n")
cat("Paired candidates with the same filename:", sum(same_file), "\n")
cat("Distinct image filenames:", length(unique(candidates$asset)), "\n")
cat("Elements linked to catalogue paths:", nrow(linked), "\n")
cat("Distinct linked catalogue paths:", length(unique(linked$linked_title_path)), "\n")
print(table(width_sets))Output
Responsive image elements: 97
Advertised image candidates: 194
Paired candidates with the same filename: 97
Distinct image filenames: 77
Elements linked to catalogue paths: 96
Distinct linked catalogue paths: 76
width_sets
185 + 342 780 + 1280
96 1 R version 4.6.0 (2026-04-24)
What the browser chooses
An HTML width descriptor gives the candidate image's intrinsic width. The browser can use that width, the page's declared display size, the device pixel ratio and other conditions to choose a resource. Measuring actual downloads would require recording network requests in a specified browser session. TMDB builds its image URLs from a base, a size and a file path. For example, changing a size segment from w185 to w342 can request another rendition of the same image.
Reproduce the comparison
The capture was made on 19 November 2026 at 20:17 UTC. We selected each image with a srcset attribute inside the homepage's main element, split its width-described TMDB URLs into individual candidates and recorded the nearest enclosing link. For this capture, all candidates used ordinary comma-separated URLs with no embedded commas. We checked that each width descriptor matched the size segment of its TMDB URL. The supplied CSV replaces each distinct filename and linked path with a stable identifier within this snapshot. An empty path marks an element outside a link.
Save the CSV beside the R Markdown file, then run its code chunk in R. The linked-path count applies to responsive placements on this one homepage. It cannot establish the total number of films or series in the catalogue.
Download the 194-row CSV · Download the R Markdown source