Usage Stats: what the ladder really runs
Every logged Colosseum battle, turned into a live picture of the ladder — usage, sets, cores, matchups and, for the first time, what actually gets switched in on you.
Every rated Colosseum battle already gets recorded for replays. Now every one of those records also feeds a live picture of the ladder: which Pokémon trainers actually bring, what they run on them, who they pair them with, what gets switched in against them, and how all of it goes. It lives at pbf.gg/usage, and as a Usage Stats tab in your dashboard under Game tools.
Nothing on it is a survey or a guess. The counts come straight from the battle logs, are refreshed every half hour, and there is no minimum rating to be counted. If you battled this month, your team is in these numbers. Right now, that is:
This month’s ladder, live
This is not a screenshot. It is the same data the usage page reads, fetched as you opened this post. Switch formats, or pull the rating cutoff up to see what the top of the ladder looks like. The percentage is the share of teams carrying that Pokémon in the last 30 days, the chip is the change against the 30 days before, and the trace is its day-by-day usage. Click any row for the full breakdown.
The rating cutoffs are the server’s own PvP roles, belts included. A team counts toward a cutoff when its trainer’s rating going into the battle met it, so “PVP Expert” is what Experts and above bring, not what everyone brings against them. Click a role to open the ladder at that cutoff.
What moved
Every row on the ladder carries its change against the previous window, so the meta’s drift is one sort away. These are the Pokémon whose share of Singles teams changed the most this month among the ones that are actually used, live from the same numbers as the ladder above.
Every Pokémon, fully opened up
Click a Pokémon and you get its whole competitive profile for the slice you are looking at. Every panel changes with the format, window, rating and rules you pick, so the Garchomp that Experts run and the Garchomp that everyone runs are two different pages.
Every move seen on the Pokémon, with how many sets ran it.
Held items, abilities and how often the hidden ability shows up.
The actual EV spreads people use, not a textbook recommendation.
Who it is brought with, and how much more often than chance.
The opponents it wins and loses against over a whole battle.
How the teams carrying it did, and how often it is still standing at the end.
The most common set, exported in one click.
The newest public replays with it, and a preview card for Discord.
Every species page has its own preview card, so dropping usage.pbf.gg/garchomp into Discord shows the numbers right there, and the share button on the page gives you that short link.
Team cores
Usage tells you what is popular. Cores tell you what is popular together. The format overview lists the most common two-, three- and four-Pokémon cores, each with the share of teams running it and the win rate of those teams. It is counted from a random sample of teams, with a cap on how many times one trainer’s identical team can count, so a single grinder cannot fill the list with their six.
Switch-ins: what actually comes in on you
This is the one we are proudest of. Matchup tables tell you who beats whom over a whole battle. Switch-ins tell you something sharper: when a Pokémon comes in, what does the opponent immediately bring in against it, and does that answer work?
The rule is strict on purpose. Only the opponent’s very next voluntary switch counts, and only if it happens by the following turn. Lead Pokémon are ignored, and so is anything that comes in just because the previous Pokémon fainted. What is left is the deliberate answer, read straight out of the battle log, turn by turn. Both directions are shown: what comes in on a Pokémon, and what that Pokémon is brought in on.
The win rate next to each answer belongs to the side that made the switch, so a red number means that answer tends to lose the game, however sensible it looks on paper.
Likes
Every Pokémon page has a heart. One like per trainer per format, and the format overview keeps a Community favourites list of the most-liked Pokémon. It is a small thing, but a Pokémon that is barely used and heavily liked is a story worth telling, and you get to write it. The hearts below are real, and they count.
The fine print
Everything on the usage pages is anonymous totals. A replay you kept private stays private and never appears in the recent-replays list; it still contributes to the counts, like every other battle. Cosmetic formes fold into their base species, while formes with their own dex entry, such as Rotom-Wash or Landorus-Therian, are counted on their own. Little Cup and 2v2 Doubles have their own ladders, and 2v2 has no rating cutoffs because those battles are unrated.
The page is marked Beta because we already know what we want next: which Pokémon actually lead, which ones were brought to the battle versus benched, and who gets the KOs. Those need a little more from the game’s side, and they are on the list.
Quick quiz, free LP
Three questions, all answerable from the Usage Stats page, and a one-time LP reward for every eligible account that gets them right. You need a linked Discord and Roblox account, and there is one reward per person.
Go see what the ladder is really running
And if you disagree with the numbers, there is only one way to change them.
Enjoyed this post?
Be the first to like it — it tells the team what to write more of.