videogameinsightsR

R package for the Video Game Insights API (by Sensor Tower). Retrieve player counts, revenue, units, wishlists, reviews, player insights and company data for games on Steam, Xbox and PlayStation, all returned as tidy tibbles with snake_case column names.

Since 0.3.0 the package uses only the v4 API: multi-platform rows, VGI IDs and slugs alongside Steam App IDs, and cursor paging. Interactive docs: https://app.sensortower.com/vgi/api/v4/api-docs/. See dev/api_coverage_2026-10.md for the endpoint-by-endpoint mapping.

Installation

install.packages("videogameinsightsR")

# development version
devtools::install_github("econosopher/videogameinsightsR")

Quick Start

library(videogameinsightsR)

# Store your API token (once per session, or add to .Renviron)
Sys.setenv(VGI_AUTH_TOKEN = "your_token_here")

# Search and inspect a game
vgi_search_games("Valheim")
vgi_game_metadata(892970)

Core Workflow

The typical workflow is search -> inspect -> pull time series -> analyse.

library(dplyr)

# 1. Find a game
games <- vgi_search_games("Elden Ring")
app_id <- games$steam_app_id[1]

# 2. Pull metadata (one row per game; per-platform prices, dates and store URLs)
meta <- vgi_game_metadata(app_id)
meta[, c("name", "genre", "steam_genres", "release_date_steam")]

# 3. Pull the full daily history for one Steam game
ccu   <- vgi_insights_ccu(app_id)$player_history   # date, avg, median, max, min
rev   <- vgi_insights_revenue(app_id)              # date, revenue_change, revenue_total
units <- vgi_insights_units(app_id)                # date, units_sold_change, units_sold_total
wl    <- vgi_insights_wishlists(app_id)$wishlist_changes

# 4. Or grab every metric at once
hist <- vgi_historical_data(app_id)
names(hist)
#> "steam_app_id" "revenue" "units_sold" "concurrent_players" "active_players"
#> "reviews" "wishlists" "followers" "price_history" "daily"

# 5. One day across many games, every platform, optional country breakdown
snap <- vgi_historical_data_by_date("2025-06-01", steam_app_ids = c(app_id, 730))
by_country <- vgi_historical_data_by_date("2025-06-01", steam_app_ids = app_id,
                                          countries = c("US", "DE"))

Genres

VGI assigns each game one genre and subgenre (genre, subgenre, for example “Shooter” / “Shooter (General)”). Steam’s own store genres, several per game, are kept separately in steam_genres and steam_subgenres. The two can disagree: ARC Raiders is an “Action” game to VGI but a “Shooter” subgenre on Steam. vgi_filter_genre() filters the local cache on either (source = "vgi" or "steam").

Function Families

Game discovery and metadata

Function	Description
`vgi_search_games()`	Search by title (local cache, then slug lookup)
`vgi_game_list()`	Full catalogue: `vgi_id`, `steam_app_id` (NA off Steam), `name`
`vgi_game_metadata()`	Metadata for one Steam game
`vgi_game_metadata_batch()`	Metadata for many Steam games
`vgi_games_metadata()`	Metadata by Steam ID, VGI ID or slug, with cursor paging

Snapshots by date

All accept date and optional steam_app_ids. vgi_historical_data_by_date() returns every metric at once and supports vgi_ids, slugs, countries and regions.

Function	Columns
`vgi_concurrent_players_by_date()`	`peak_concurrent`, `avg_concurrent`
`vgi_active_players_by_date()`	`dau`, `mau`, `dau_mau_ratio`
`vgi_revenue_by_date()`	`revenue`, `daily_revenue`
`vgi_units_sold_by_date()`	`units_sold`, `daily_units`
`vgi_reviews_by_date()`	`positive_reviews`, `negative_reviews`, `positive_ratio`
`vgi_followers_by_date()`	`follower_count`
`vgi_wishlists_by_date()`	`wishlist_count`

Per-game history (Steam)

Function	Returns
`vgi_insights_ccu()`	`$player_history`: date, avg, median, max, min
`vgi_insights_dau_mau()`	`$player_history`: date, dau, mau
`vgi_insights_revenue()`	date, revenue_change, revenue_total
`vgi_insights_units()`	date, units_sold_change, units_sold_total
`vgi_insights_reviews()`	date, positive, negative, total, positive_ratio
`vgi_insights_wishlists()`	`$wishlist_changes`: date, wishlists_total, wishlists_change
`vgi_insights_followers()`	`$followers_change`: date, followers_total, followers_change
`vgi_insights_price_history()`	`$price_changes`: currency, price_initial, price_final, first_date, last_date
`vgi_price_history()`	price periods on every platform, by Steam ID, VGI ID or slug

Player insights

Function	Scope
`vgi_insights_playtime()`, `vgi_top_countries()`, `vgi_top_regions()`, `vgi_top_wishlist_countries()`	one Steam game
`vgi_player_overlap()`	one game, by Steam ID, VGI ID or slug
`vgi_all_games_player_overlap()`	walks the catalogue one game per request
`vgi_all_games_playtime()`, `vgi_all_games_top_countries()`, `vgi_all_games_regions()`, `vgi_all_games_wishlist_countries()`	many games; filter by `steam_app_ids`, `vgi_ids`, `slugs`; playtime also by `countries` / `regions`

Publishers and developers

vgi_publishers_overview(slugs = "free-lives")  # per-platform revenue and genre mix
vgi_publisher_list(search = "valve")           # directory (id, name, slug)
vgi_publisher_info(28663)                      # Steam revenue and game counts, releases per platform
vgi_publisher_games(28663)                     # the publisher's Steam App IDs
vgi_all_publisher_games(vgi_ids = 28663)       # VGI game IDs on every platform
# ...and the matching vgi_developer*() functions

Convenience functions

summary <- vgi_game_summary(
  steam_app_ids = c(892970, 1245620),
  start_date = "2025-01-01",
  end_date   = "2025-01-31"
)
summary$summary_table

yoy <- vgi_game_summary_yoy(
  steam_app_ids = 892970,
  years = c(2024, 2025),
  start_month = "Jan",
  end_month = "Mar"
)
yoy$comparison_table

Pagination

Functions that list many records take limit and cursor and return the next cursor as an attribute; pass all_pages = TRUE to fetch everything.

page1 <- vgi_games_metadata(limit = 500)
page2 <- vgi_games_metadata(limit = 500, cursor = attr(page1, "next_cursor"))
everything <- vgi_publisher_list(all_pages = TRUE)

Configuration

# API root (the /v4 segment is added; default shown)
options(vgi.base_url = "https://vginsights.com/api")

# Timeouts and retries
options(vgi.timeout = 30)
options(vgi.retry_max_tries = 4)

# Request caching (seconds, GET only)
options(vgi.request_cache_ttl = 3600)

# Rate limiting
options(vgi.auto_rate_limit = TRUE)
options(vgi.calls_per_batch = 10)
options(vgi.batch_delay = 1)

# Verbose logging
options(vgi.verbose = TRUE)

Important Notes

Platforms: on Xbox, VGI’s “units owned” includes Game Pass players and other free copies; “paid units” excludes them. Xbox and PlayStation figures are VGI model estimates.
Steam App IDs: pass explicit IDs to the snapshot-by-date functions. Without them the API returns its default handful of games.
API host: requests go to https://vginsights.com/api/v4, which redirects to https://app.sensortower.com/vgi/api/v4.
Return types: all data-returning functions produce tibbles with snake_case columns.

Development

devtools::load_all()
devtools::test()
devtools::check(cran = TRUE)

Tests replay httptest2 fixtures under tests/testthat/api/v4/ (re-record with Rscript dev/record_fixtures.R). Live smoke tests run only when VGI_AUTH_TOKEN is set.

License

MIT