Every Steam Workshop leaderboard ranks mods by subscriptions — the install. Almost none reports the other number sitting next to it on the same page. Favorites get discussed as an engagement sidebar, a nicety, a "bookmark" that lives in the shadow of the install count. So it is worth stating plainly what a full sweep of the catalogue turns up: nearly half of the Steam Workshop has never been favorited a single time.
Across 6,005,684 active, subscribed Steam Workshop mods, 2,799,554 — 46.6% — carry exactly zero favorites. Every one of them has at least one live subscriber; many have thousands. Not one of them has ever been bookmarked by anyone.
And the split that falls out of that is wider than any listing divide this series has measured. The half of the catalogue that has been favorited reaches 1,000 subscribers 26.97% of the time. The half that has not reaches it 0.118% — a 228-fold gap.
That number is large enough to be suspicious, so this article spends as much time taking it apart as stating it. The pooled gap is real, but it shrinks hard under control, and the shrinking is the actual story: favorites are not a lever a creator pulls, they are a symptom of a kind of mod. Here is the honest cut.
The full ladder
Take the 6,005,684 active subscribed mods and bucket them by how many favorites they hold. The reach-to-1,000-subscriber rate does not step — it falls off a cliff between zero and one.
| Favorites | Mods | Share of catalogue | Reach 1,000 subs | Average subs |
|---|---|---|---|---|
| 0 | 2,799,554 | 46.6% | 0.118% | 23 |
| 1–4 | 1,324,830 | 22.1% | 1.588% | 132 |
| 5–19 | 651,755 | 10.9% | 14.346% | 576 |
| 20–99 | 562,984 | 9.4% | 35.231% | 1,595 |
| 100–999 | 521,341 | 8.7% | 78.095% | 5,929 |
| 1,000+ | 145,220 | 2.4% | 99.625% | 54,588 |
The first two rows are the ones that matter, because they are the bulk of the platform. Nearly seven of every ten live, subscribed Workshop mods have fewer than five favorites, and those seven-in-ten account for almost none of the reach. The zero row alone is 46.6% of the catalogue and its mods average 23 subscriptions each.
At the other end, the band is a wall. Of the 145,220 mods with 1,000 or more favorites, 99.625% are past 1,000 subscribers. Being heavily favorited is essentially a proxy for being installed. It is the bottom of the distribution where favorites stop tracking reach at all — a mod can have 197,839 subscribers and zero favorites.
The zero-favorite population is not small — you have installed some of it
The favorite counter is the closest thing the Workshop has to a "I want this again" signal. It is a deliberate tap that a subscription is not: you can install a mod from the game's own Workshop panel in two clicks, but to favorite it you have to be on the item page and choose to mark it. That is why the zero band reads the way it does — it is not full of bad mods, it is full of mods nobody ever needed to find again.
The clearest example sits at the top of the zero-favorite ranking:
| Mod | Game | Subscriptions | Lifetime views | Favorites |
|---|---|---|---|---|
Custom_Switching20hz |
Aimlabs | 197,839 | 10 | 0 |
Custom_ClickingScope |
Aimlabs | 193,970 | 11 | 0 |
Custom_VTFastTS |
Aimlabs | 66,378 | 19 | 0 |
DayOneServerPack-Chernarus |
DayZ | 16,712 | 281 | 0 |
UPG Clothing |
DayZ | 15,505 | 438 | 0 |
Archi - V2 City Scape (old version) |
SpeedRunners | 13,041 | 903 | 0 |
Large-scale upgrade |
Forts | 12,200 | 221 | 0 |
4-4A-A-十字交叉 |
Forts | 11,773 | 20 | 0 |
Why I do this? |
Dota 2 | 11,521 | 544 | 0 |
These are not obscure files. A 197,839-install Aimlabs config is one of the most-installed items in its game — and it has been viewed ten times. The pattern is the same across the list: Aimlabs share-codes, DayZ server packs, SpeedRunners and Forts map bundles. These are mods players arrive at through a code, a server connect, or a friend's link, install once, and never browse for again. There is no reason to bookmark a dependency.
The reverse case shows the same thing from the other side: 308,667 active Workshop mods have favorites but zero subscriptions — bookmarked and never installed. A favorite is intent, not an install, and the two counters disagree in both directions. That is the honest framing for everything below: the favorite signal is behavioral, and a large, structural slice of the catalogue simply never generates the behavior.
The control that halves it
A 228x gap should immediately raise the composition question, and this series has been burned enough times to run the control before publishing the headline. The concern: maybe creators who earn favorites are simply better or more prolific, and the favorite is riding along with something else.
So split the catalogue the only way that isolates the mod from the author and the ecosystem at once — within the same creator and the same game, compare their mods that have favorites against their mods that have none, on the same account and the same shelf.
| Group | Mods | Mods with 0 favorites | Reach 1,000 subs | Mods with favorites | Reach 1,000 subs |
|---|---|---|---|---|---|
| Paired same-creator + same-game | 245,785 creators | 1,277,465 | 0.203% | 1,153,612 | 11.261% |
Hold the creator and the game fixed and the gap is 55x — down from the pooled 228x, but nowhere near a wash. Same person, same ecosystem, and their un-favorited mods reach a thousand subscribers a fifth of one percent of the time while their favorited ones do it eleven percent of the time.
The number falling by three-quarters is worth being precise about: most of the pooled gap is composition — prolific creators, and ecosystems whose games route installs around the item page, both load the zero band. But the residue is a genuine 55-fold behavioural split, and it survives every conventional control (publish year, catalogue size, ex-outsiders). The favorite is not a vanity metric sitting on top of reach; it is tracking something real about whether a mod gets found by a stranger.
The favorite ladder, per install
The magnitude of the split is legible in one more cut: how many favorites a mod earns per 1,000 subscriptions it holds. If favorites and installs tracked each other, this would be flat. It is the opposite of flat.
| Favorites held | Favorites per 1,000 subscriptions |
|---|---|
| 0 | 0.0 |
| 1–4 | 13.7 |
| 5–19 | 17.3 |
| 20–99 | 29.4 |
| 100–999 | 58.3 |
| 1,000+ | 609.8 |
Small mods earn favorites at a tiny rate and large mods earn them at a rate almost 45 times higher — not because big mods are better, but because a favorite is generated by repeat discovery, and repeat discovery only happens to mods that surface in browsing more than once. The zero band is not a lower rung of the same ladder. It is off the ladder.
Where the zero-favorite floor runs deepest
The zero-favorite share is not evenly spread. Some ecosystems publish almost entirely into the floor, and in most of them the reason is the same — the game's install flow bypasses the item page.
| Game | Mods | Zero favorites | Share with 0 favorites | Reach 1,000 subs among 0-fav |
|---|---|---|---|---|
| Poly Bridge | 26,652 | 26,149 | 98.1% | 0.08% |
| Universe Sandbox | 71,991 | 66,362 | 92.2% | 0.00% |
| SpeedRunners | 29,872 | 27,267 | 91.3% | 0.78% |
| KovaaK's | 21,392 | 19,454 | 90.9% | 0.00% |
| People Playground | 29,464 | 26,501 | 89.9% | 0.00% |
| The Escapists 2 | 31,551 | 26,815 | 85.0% | 0.00% |
| Human Fall Flat | 97,016 | 80,199 | 82.7% | 0.00% |
| Kerbal Space Program | 121,805 | 98,344 | 80.7% | 0.00% |
| Besiege | 177,844 | 142,547 | 80.2% | 0.00% |
| Trailmakers | 333,773 | 210,414 | 63.0% | 0.00% |
Poly Bridge at 98.1% zero-favorite is the extreme. These are level-sharing ecosystems — a player publishes a bridge, another player downloads it from a link, and the item page is a formality. Besiege (80.2%), Trailmakers (63.0%) and KSP (80.7%) are the same machine: a builder game where the Workshop is a swap meet for user levels, not a boutique. KovaaK's and Aimlabs, the aim trainers, are the purest case in the whole dataset — their configs are distributed by share-code, which is exactly why the zero-favorite band has almost no reach in them.
Read against the blank-page problem and the thumbnail divide, this is the third structural escape hatch in the series: the mods that never needed a listing to be found are also the mods that never earn a favorite. A zero-favorite share of 98% is not a quality verdict about Poly Bridge's levels. It is a statement about how that game hands content to its players.
The exceptions worth naming
Two cuts keep the story from collapsing into "zero favorites means dead," because the zero-favorite population contains mods that are very much alive.
First, 3,295 mods with zero favorites are past 1,000 subscribers — 0.38% of every 1,000-subscriber mod on the platform. They are concentrated in the ecosystems where installs flow through server packs and dependency chains:
| Game | Zero-fav mods past 1,000 subs | Subscriptions held |
|---|---|---|
| DayZ | 666 | 1,366,749 |
| Forts | 578 | 1,224,481 |
| SpeedRunners | 214 | 464,901 |
| Duck Game | 196 | 269,268 |
| Paint the Town Red | 175 | 358,548 |
| Planet Coaster | 151 | 271,659 |
| Holdfast: Nations At War | 139 | 301,581 |
| Command & Conquer Remastered Collection | 120 | 213,937 |
| Counter-Strike 2 | 114 | 189,162 |
| Aimlabs | 104 | 771,483 |
DayZ's 666 zero-favorite mods past a thousand subscribers hold 1.37 million subscriptions between them — this is a server-modding ecosystem where a mod is installed because a server runs it, and nobody favorites a mod they were required to download. Forts is the same shape at a smaller scale: 578 live zero-favorite mods, 1.22 million installs.
Second, the ranking by zero-favorite reach — which games actually let an un-bookmarked mod climb — inverts the counts:
| Game | Reach 1,000 subs among zero-favorite mods |
|---|---|
| Paint the Town Red | 3.79% |
| Door Kickers 2: Task Force North | 2.06% |
| Forts | 1.69% |
| Counter-Strike 2 | 1.25% |
| ARK: Survival Evolved | 1.01% |
| DayZ | 0.99% |
A zero-favorite mod in Paint the Town Red reaches a thousand subscribers 3.79% of the time; across the platform that number is 0.118% — a 32-fold spread. Paint the Town Red and Door Kickers 2 are games small enough, and moddable enough, that the Workshop install path is still the main one, so even their un-bookmarked mods get found. In KovaaK's, Universe Sandbox, People Playground and Besiege that rate is 0.00% — the item page is not part of the distribution, so nothing in the floor climbs.
What this is not
A missing favorite is not a missing install. The 2,799,554 zero-favorite mods hold 63,448,390 subscriptions between them. That is a lot of installs in absolute terms — it is simply 0.506% of the platform's 12,530,334,578 total, because the average zero-favorite mod holds 23 subscriptions. The floor is enormous in count and negligible in reach.
Favorites and installs disagree in both directions. As noted above, 308,667 mods are favorited with zero subscriptions. If favorites were a clean install proxy, that population could not exist. They are a return-intent counter, and return intent and installation part ways exactly where you would expect — in the mods people bookmark for later and never get around to, and the mods they install on someone else's recommendation and never need to revisit.
Half of a catalogue is not half of a market. The 46.6% figure is a share of mods, and the mods in it are unusually small. It would be a mistake to read "46.6% of the Workshop has zero favorites" as "46.6% of the Workshop's activity is invisible" — the opposite is closer to true: those 2.8 million mods are collectively half a percent of the install base. The count and the weight are different numbers on purpose, and this series has made that split before.
And the favorite is downstream, not upstream. Nothing here says "get favorites and you will grow." It says the two travel together at 55x once you control for who made the mod and where it lives — which is a statement about what a favorite predicts, not a lever a creator can pull. The lever, if there is one, is the one the one-subscriber floor identified: discovery, not decoration.
Method: ModSignal's live dataset of active Steam Workshop items, filtered to is_active items with at least one current subscription — 6,005,684 mods. "Reach 1,000 subs" is the share of a group whose subscriptions_current is at least 1,000. The paired control groups mods by creator_steam_id and game_id and keeps only creator-and-game combinations holding at least one favorited and one un-favorited mod, covering 245,785 combinations. The reverse-direction count (308,667) covers all is_active items with favorites_current >= 1 and subscriptions_current = 0. Data as of October 3, 2026.