Does a smaller supply make a Roblox limited worth more?

Every trader has heard that rare items are worth more. It sounds obvious: if only a few dozen copies of a hat exist, the people who want one have to compete for them. We wanted to know how far that holds on real numbers, so we set the number of copies of each limited against its value for every classic limited our tracker prices. The data is our limiteds file, measured on the October 30, 2026 snapshot: 172 classic limiteds, each with the copy count Roblox's catalog lists for it and our estimate of its value.

The short answer is yes, and strongly, as long as the difference in supply is large. Pick any two of the 172 items with different supplies and different values, and the one with fewer copies is worth more in 12,923 of 14,697 pairs, or 87.9%. When one item has at least ten times the copies of the other, the rarer one is worth more 98.3% of the time. But when the two supplies are within a factor of two of each other, the rarer item is worth more in only 55.3% of pairs, barely better than picking one at random. Supply sets the scale of a price. It does a poor job of ranking items that are close.

Where the numbers come from

The copy count is the total quantity Roblox's catalog reports for each item, which our tracker reads from catalog.roblox.com every time it refreshes. We treat it as the number of copies that exist. The catalog reports no units left to buy new on 171 of the 172 items, so for practical purposes the count is fixed. The counts run from 23 copies of Dominus Empyreus to 6,467,908 copies of Goldrow.

The value is the estimate our limiteds market page is built on, read from Roblox's current resale data: 0.6 times the median of the five most recent sale days in the last 180 days, plus 0.4 times RAP. An item with no sale in 180 days has no recent price, so its estimate is its RAP alone. Fifteen of the 172 are in that position. The tracker shows no estimate for an item with no sale in 90 days, but the file still holds one, and we use it here so that every item is on the chart. Further down we drop the unsold items and check that the answer survives.

The data is from October 30, 2026, a week before this post. Prices move from day to day, so read each figure as that day's.

Ten times rarer, about ten times the value

On a chart with both axes on log scales, the 172 items fall close to a straight line running from the top left to the bottom right. The least-squares line through them has a slope of -1.01: each time the number of copies goes up tenfold, the value goes down by a factor of about ten. The line accounts for 85% of the variation in log value, which is a lot for one fact about an item.

The two ends of the list show the size of it. Dominus Empyreus, with 23 copies, carries an estimate of 13,577,229 Robux. Goldrow, with 6,467,908 copies, is valued at 172. Goldrow has roughly 281,000 times as many copies, and the Dominus is valued at roughly 79,000 times as much.

The table splits the list into five groups by number of copies.

The median estimate falls from 4,336,949 Robux in the rarest group to 168 in the most common. The last column multiplies each item's estimate by its number of copies and adds the results up within the group. It is arithmetic, not a price anyone could get, since nobody could sell every copy at once. But it shows the pattern another way: the median copy counts of the first and last groups differ by a factor of more than 14,000, while the five totals all land between 15.4 and 39.5 billion Robux. When the price falls about as fast as the supply rises, the total stays roughly flat.

Close supplies barely sort

The second chart shows the same items in their groups, one dot per item. The medians step down from group to group, but inside each group the dots spread wide. In the 10,000 to 99,999 group the dearest item is worth 1,803 times the cheapest. In the 1,000 to 9,999 group it is 599 times.

Inside a group, fewer copies still helps a little. Taking every pair of items from the same group, the one with fewer copies is worth more in 65.3% of pairs under 1,000 copies, 62.9% at 1,000 to 9,999, 73.6% at 10,000 to 99,999 and 58.1% at 100,000 to 999,999. In the group with a million copies or more it reverses: the item with fewer copies is worth more in only 44.2% of pairs. Those 27 items are all valued between 145 and 669 Robux, and 24 of them at 196 or less. With millions of copies of each, the price sits near the bottom of the market whatever the exact count.

The exceptions are not only at the cheap end. The Void Star has 2,112 copies and an estimate of 3,476,715 Robux, more than six items with at least ten times fewer copies, among them Dominus Infernus with 27 and Countess of the Federation with 51. Across the list, 161 of the 9,369 pairs that are ten or more times apart in supply break the rule, and they involve 51 different rarer items. Something besides the copy count decides a good part of the price, and this data cannot say what.

Is it the supply, or the launch price?

A fair objection is that the rarest limiteds were also the most expensive when Roblox sold them, and that today's values partly carry that over. Our file records a launch price for 50 of the 172 items. One was free, which leaves 49. Among those 49, rarity and launch price do go together: the rank correlation between copies and launch price is -0.79, where -1 would mean the item with the fewest copies always had the highest launch price.

So we fitted value against both at once. Holding the launch price fixed, ten times the copies still went with about a tenth of the value (a slope of -1.06). Holding the copy count fixed, a launch price ten times higher went with only about 26% more value. On these 49 items, supply carries nearly all of the pattern and launch price adds little on top. Every one of them is now valued above what it launched at, with a median of 37.9 times its launch price. Forty-nine items is a small sample, and the 122 without a launch price on file might behave differently, so read this as a check, not a rule.

The rarest values are the least tested

The rare end has a catch. The 29 items with under 1,000 copies hold 89.0% of the summed estimates of all 172 items. They also sold 31 copies between them in 90 days, about one per item, and only 11 of the 29 sold even once. Fifteen have no sale in 180 days, and 5 have no usable sale on record at all, so their value is their RAP alone, a number no recent buyer has confirmed. We looked at the same problem from the RAP side in how close RAP is to recent sales, where the items with no sale in 180 days turned out to hold most of the RAP on the list.

That raises a fair question: is the pattern real, or a product of unsold items carrying high RAPs? We checked it two ways. First, we dropped the 15 items with no sale in 180 days and used each remaining item's recent sale price instead of our estimate. On those 157 items the fitted slope was -1.04, as steep as before. Second, we kept only the 124 items that sold 50 or more copies in 90 days, where prices rest on many recent sales. The pattern weakens but holds: the slope is -0.84, so ten times fewer copies went with about seven times the value. Using RAP instead of our estimate on all 172 gives a slope of -1.00, the same picture.

So the relationship does not depend on the dormant items. What does depend on them is the very top: the multi-million estimates on the rarest items come with almost no trading behind them. How often items change hands is its own question, covered in which limiteds actually sell.

What this means when you weigh a trade

If you are comparing two items whose supplies are far apart, the copy count is a good first guide: the rarer one is almost always worth more. If the supplies are close, the count tells you very little, and you need the sale history instead. The limiteds trade checker compares items on our estimates, and the limiteds page lists how many copies of each sold in 90 days and when the last one went. For the rarest items, check that last sale date before trusting any value: a price set by one sale long ago is a guess, not a market.

What these numbers cannot tell you

This is a pattern across 172 items, not proof that supply causes value. Rarity comes bundled with age, fame and launch price, and we could only separate one of those, for 49 items.

The 172 are the classic limiteds our tracker found through Roblox's catalog search and could price from Roblox's resale data. Roblox has many more limiteds, and modern collectibles and community-made limiteds are not included. The copy count also includes copies sitting in accounts that no longer trade. We cannot see how many of those there are, so the number of copies anyone could actually buy is smaller than the count, by an amount nobody outside Roblox knows. Serial numbers are a separate question that only applies to Limited U items; what a low serial number is worth covers it.

The file every number here came from is saved with the site at /data/snapshots/2026-10-30/limiteds.json, with the copy count, estimate, RAP, sales and launch price of each item, so each figure can be checked against the data it used.

The short version

Across the classic limiteds we track, fewer copies goes with a higher value, and the relationship is steep: ten times rarer, about ten times the value. It is a rule about orders of magnitude. Between two items with similar supplies it says little, and the rarest items, which hold most of the list's value, are the ones that almost never sell.