4 million searches in a week: what Minecraft players actually want.
In the seven days to 2 August our catalogue handled 3,983,440 searches across 874,178 distinct strings. That is a fairly direct picture of what people want from Minecraft right now — but only if you know what the numbers count. Read straight off the top of the list, this data tells you two things that are flatly untrue. So here is the picture, and here is how we nearly got it wrong.
Trap one: half the top results are not searches
The raw list of most-searched strings opens like this: verity, horror, ver, the, gun, mo, ve.
ver was searched 12,572 times. ve, 9,560. veri, 9,095. verit, 4,171.
Nobody wants a mod called ve. Our search runs as you type, so every partial word on the way to a real query lands in the log as its own row. The apparent popularity of two-letter strings is thousands of people spelling verity one keystroke at a time.
This matters more than it sounds. A demand report built on the raw list would conclude that the most wanted things in Minecraft are ver, the and mo, and somebody would go and do something about it. Everything below is filtered to strings of four characters or more with the obvious typing fragments removed, and adult search terms stripped out entirely - they are a real and large share of the log, and they are not what this article is about.
What people actually search for
| Search | Times | Matching mods |
|---|---|---|
| verity | 37,714 | 56 |
| horror | 15,489 | 1,576 |
| action | 8,581 | 706 |
| wither storm | 8,262 | 113 |
| fnaf | 8,079 | 256 |
| ben 10 | 7,149 | 61 |
| guns | 6,743 | 429 |
| stalker | 6,598 | 159 |
| create | 6,194 | 7,161 |
| shader | 6,097 | 1,932 |
| one block | 5,883 | 1,491 |
| sonic | 5,368 | 149 |
| tacz | 4,609 | 62 |
| naruto | 4,443 | 90 |
| backrooms | 4,119 | 131 |
| poppy playtime | 3,735 | 174 |
Two clean groups fall out of this, and they behave completely differently.
Named things beat categories, by a lot
The biggest single term in Minecraft modding right now is not a genre. It is one horror mod, searched 37,714 times in a week - more than twice the entire horror category and six times shader. People do not arrive wanting "a horror mod". They arrive having watched a specific thing on a specific channel, and they type its name.
That pattern repeats all the way down: Wither Storm, FNAF, Ben 10, Sonic, Naruto, Backrooms, Poppy Playtime. Franchises and named mods, not descriptions. Generic words like better, more, realistic and real do appear, but they sit well below and they behave like browsing rather than searching - the person does not know what they want yet.
Where demand runs past supply
The third column is the interesting one, because it turns searches into a ratio. Create is searched 6,194 times and has 7,161 matching mods - more supply than demand, and every result is a real option. Minecraft as a bare word returns 22,495, which is another way of saying it returns nothing useful.
Then there is the other end:
- Ben 10 - 7,149 searches, 61 mods. 117 searches per available mod.
- TACZ - 4,609 searches, 62 mods. 74 per mod.
- Wither Storm - 8,262 searches, 113 mods. 73 per mod.
- Naruto - 4,443 searches, 90 mods. 49 per mod.
- Sonic - 5,368 searches, 149 mods. 36 per mod.
- Backrooms - 4,119 searches, 131 mods. 31 per mod.
Compare that with horror at 10 searches per mod, or create at less than one. A franchise search concentrates thousands of people onto a few dozen files, which has a practical consequence: those few files carry enormous traffic, so when one of them breaks it is not a small problem. It is the reason our hourly re-check targets the most-opened mods rather than sweeping the catalogue evenly.
Trap two: the biggest source of empty searches is the most popular term
8.21% of searches returned nothing. The obvious next question is which searches, and the obvious answer is wrong.
Top of the empty-handed list is verity - 361 searches that found nothing. Which looks alarming until you put it next to the 37,714 searches for the same word that found plenty. The failure rate for the most popular term in the catalogue is 0.96%, and those 361 are typos, odd spacing, and searches fired while a filter was narrowing the results.
Any term with enormous volume has an enormous tail. Sorting empty searches by absolute count just re-sorts the catalogue by popularity and calls it a gap. The genuinely missing things live further down and are much smaller: avaritia at 238, falsity at 191, twixxel at 188 - names from videos and other games with little or no Bedrock presence yet. Those are worth chasing. Verity is not.
The rule we now apply to every number here: before acting on a count, ask what one row in that table represents. One keystroke or one intention? One person or one page of scrolling? An absolute or a share? Every one of those questions has caught us out at least once, and each time the wrong answer pointed at real work that did not need doing.
What we do with it
Search demand steers three things. It decides what our scrapers prioritise, because a franchise with 117 searches per available mod is worth going and finding more of. It decides which mods get re-checked most often, since concentrated demand means concentrated damage when a file dies. And it feeds the synonym handling - people search in their own language and their own spelling, and matching only exact English strings would strand a large share of the catalogue.
If you want to see the results rather than the analysis, the Bedrock catalogue and Java catalogue are both fully searchable, and our breakdown of what actually installs covers what happens after you find something.