Throne and Liberty price trends and market timing

Which way each region's market is moving today, which weekday gear is historically cheapest, and the live counters behind both. The three auction houses have diverged sharply — read them separately, not as one market.

Free for everyone — no account needed

What’s moving today

One line per region, each measuring its own market against the same base day, then the biggest movers in each direction.

Biggest movers of the week

Only items on the snipe list — the ones this tracker actually watches — are ranked here. Each kept a real order book on all seven days; thin and spiky books are dropped before ranking.

loading…

Rising

Largest gains over the week. A window to sell into — open an item to check its live floor first.

Item7d change

    Nothing climbed in this region this week. Try another region.

    Falling

    Largest drops over the same week. A window to buy into — open an item to check its live floor first.

    Item7d change

      Nothing dropped in this region this week. Try another region.

      Most restless

      Ranked by churn: every day's price move added up, both directions. Big churn means the price rarely sat still.

      Item7d churn

        No prices moved enough to rank in this region. Try another region.

        When gear is cheapest

        Which weekday auction-house prices are historically lowest. Every tracked item is compared with its own typical price, so a 500,000 lucent weapon and a 2,000 lucent material count the same.

        Best day to buy
         
        Thursday
        Typical saving
        versus the dearest day
        0.0%
        Worst day to buy
         
        Sunday
        Items analysed
         
        0
        Observations
         
        0

        The week

        Each day against the weekly norm, after the market's own rise or fall over that week has been removed. Below zero is cheaper than usual.

        Across the week, hour by hour

        Weekday and hour

        Cell colour is the price against normal — the key below the grid says which end is which. Hover a cell for the detail behind it. Hour of day on its own does not hold up — only combined with the weekday does it carry anything, which is why nothing here ranks hours by themselves.

        CheaperDearer not enough data

        By category

        The same measurement run separately for each kind of item. A narrower slice carries fewer observations, so its shape is noisier than the market-wide one — the spread column says how much day-to-day movement there is to work with at all.

        CategoryCheapest dayDearest daySpreadItems

        How this is calculated

        Every item is measured against its own typical price rather than in lucent, so expensive gear cannot drown out cheap materials. Then the market's overall move during that week is subtracted — without that step a market that simply drifted downwards would show up as "Fridays are cheap", which is a trend, not a weekly rhythm.

        What is left is the part of the price that depends on the day of the week. The spread is real but modest: a few percent, not a sale.

        The exact method, and what it is not
        • Each observation is divided by that item's own median price in its own region, then clipped to the 0.5–1.5 range so a typo listing cannot move the result.
        • The daily market level is removed with a centred seven-day average, which cancels any smooth trend exactly.
        • Cells are combined with a mean, not a median. Roughly three quarters of item-hours are unchanged listings repeating the same price, which pins a median to exactly the item's own typical value and hides the effect entirely.
        • Split-half validation on alternating weeks gives r = 0.87 for the weekday shape. Measured over the first six weeks of the record: Thursday came out below the weekly norm in six weeks of six, Friday in five of five.
        • Hour of day on its own does not replicate (r = −0.70) and is never presented as a ranking here.
        • Thursday carries a small caveat: two Thursdays in the record hold about a third fewer observations after collection outages. The effect holds in the weeks either side.
        • This is historical statistics, not a promise about any listing on screen right now.

        Counted, not claimed

        Every figure below is a live count from the database that runs TL Tracker — nothing sampled, nothing estimated, and each one says what it counts.

        Reading the counters…

        Counting…

        The counters did not answer. Nothing is broken on the trading side — this page just could not read them. Use Refresh above to try again.

        What it reads

        What it flags

        Readings per hour · last 72 h

        Alerts per day · last 30 d

        What it flagged, and whether it cleared

        loading…

        Figures the ledger's own endpoint does not hold, read from the analysis service instead. Where the two endpoints measure the same idea — readings stored, days on record, alerts this week — the ledger above is the one that prints it, so no number on this page is stated twice from two sources.

        Pulling the latest numbers…

        These four are warming up. The ledger above is unaffected — check back shortly.