The JOOLA Perseus Mod TA-15 held a place on the USA Pickleball approved list for most of a season. It lost that place on 1 July 2025, alongside five other models. The paddle itself was unchanged that day. The test behind the approved list had changed. USA Pickleball applied the new version to models approved under the older one.
That order of events is the part worth planning for, because it repeats. Certification records one result against one version of a rule, and the rule has already moved twice in two years. A brand that treats approval as a permanent property of the product gets surprised at the worst possible moment. The artwork is already printed by then, and the stock is already sitting in a warehouse.
Four separate costs arrive with a removal date, and none of them appear on a factory quote. This article works through the Mod TA-15 case one layer at a time. It starts with the schedule that removed the model, then covers the six models named on the same notice. From there it explains what the power test reads. That reading decides whether a brand faces a redesign or a retest. It closes with two records a brand can keep. With those in place, a rule change turns into a phone call rather than a write-off.
One Perseus passed and lost approval a year later
USA Pickleball added a power test to its manual in the last quarter of 2024. The test fires a ball at the face at a controlled speed and measures how much energy comes back. The industry refers to the result as PBCoR. The notice that introduced the test also carried a schedule, and most brands skim past that part. The schedule is what decides whether a model already on the list survives the change.
The upper limit began at 0.44 on 1 November 2024 and dropped to 0.43 on 1 November 2025. Both figures are tied to the submission date rather than to the ship date. A paddle submitted after 1 November 2024 was tested against 0.44. A paddle submitted after 1 November 2025 is tested against 0.43. Read on its own, that structure looks protective for any brand holding older stock. A model already on the list appears to keep the result it earned, which came from the rule in force at its submission. Most brands described the arrangement that way at the time.
That protection did not reach the Mod TA-15, because the model had never produced a figure at all. It was approved before PBCoR existed in the manual. When the test arrived, the same build read above the new ceiling, and USA Pickleball removed it instead of grandfathering it. The gap between those two choices is where a certification plan usually fails. A result can be honoured later, while a test that was never run leaves nothing to honour.
Date | What took effect | Which paddles it reached |
1 Nov 2024 | PBCoR limit 0.44 | Paddles submitted after this date |
1 Jul 2025 | Six models sunset for sanctioned play | Models already on the list |
1 Nov 2025 | Limit tightens to 0.43 | Paddles submitted after this date |
1 Oct 2026 | Spin cap 2,100 RPM | Paddles submitted after this date |
The second row is the one to pause on. Every other row describes a rule aimed at future submissions. The second row reached back and took models off a list they were already sitting on. No submission date had warned them about that. A brand that reads a certification date as permanent protection will be caught by this kind of notice. The notice usually arrives after the purchase order rather than before it.
The approved list and the notices behind it sit in the approved equipment database. You can look up a model there by name. An entry records what happened rather than what a supplier intends. That is the difference between a paddle described as approved and a paddle that currently is.
The same notice pulled six models and restored one
Six models appeared on the 1 July notice. Two carried the Perseus name, one came from Gearbox, and three came from ProKennex. A seventh name on the same list is more instructive than any of them. ProKennex Black Ace LG was removed on paper, then returned to the list on 3 June 2025 after a second test.
USA Pickleball’s stated reason was that the LG had been built the way it was certified. The model reached its power figure from an 11 mm profile rather than from any change to the construction. The retest produced a reading that matched the original submission. The first measurement had triggered the removal, so the design was never the real problem. The model was off the list for only a short period.
Two paddles can leave the same list for reasons that share nothing. One was genuinely built past the ceiling and cannot come back without a new construction. The other was built under the ceiling and read above it once. That makes the case a measurement question rather than a design question. The distinction decides whether a brand is looking at a redesign or at a retest. The two carry very different lead times and very different bills.
A third route to removal runs through paperwork rather than engineering. USA Pickleball has also removed models when the sample submitted for testing did not match the units in production. This is a records failure, and any factory can prevent it by keeping one sealed sample from each production run. The sample needs the run number and the date it was pulled written on it. A sample without a label proves nothing two years later.
Four lines on the bill that a delisting touches
A removal date costs a brand in four places, and the first one is the most visible. Units already in a warehouse either sell outside sanctioned play or stop selling altogether. Once the delisting is public, the second outcome is far more common.
Line | What the date hits |
Stock | Units on hand sell outside sanctioned play or not at all |
Trade-in | The brand pays to put buyers into the replacement model |
Channel | Distributors send cartons back and ask for credit |
File | The replacement model needs its own submission and its own name |
The trade-in line is usually the one agreed in a hurry. JOOLA ran a swap programme in at least one market for buyers who purchased between November 2024 and March 2025. Production of the Mod TA-15 stopped at the same time. The brand therefore absorbed the loss on every unit it already owned. Units sitting in a distributor’s warehouse had already been invoiced, so they counted as the brand’s loss rather than the distributor’s.
The factory that pressed those units carries none of that bill, and that asymmetry is worth understanding. A supplier is paid on delivery. The rules that decide whether a delivered paddle can be sold are set somewhere else entirely. A supplier can still hand over the records that make a retest or a replacement submission faster. That is a different service from pressing a good shell, and most briefs do not ask for it.
PBCoR reads the whole assembly, not the material list
The power test that caught the Mod TA-15 replaced a static press. The older method compressed the middle of the face and read how far it gave. That produced one number from one point on the paddle. A static press cannot see how the face is bonded into the frame. It also misses how much the perimeter is allowed to move under load. Two paddles can therefore read identically at the centre and behave quite differently in play.
PBCoR fires a ball and reads the speed that comes back. The figure therefore describes an assembly rather than a parts list. A large share of that reading comes from the press that forms the body. A thermoformed shell and a cold-pressed stack can carry the same face material and land on different figures. The difference lives in how the layers are consolidated and how the edge is closed, not in the material name.
This is where a spec sheet stops being enough. A brief that names a face, a core and an edge as three separate items describes three parts. The test scores one object built out of them. A brand that compares two suppliers on material lists alone is comparing inputs rather than finished paddles. Those inputs do not predict the number that decides whether a paddle stays on the approved list.
A spin cap arrives on 1 October 2026
Power was the first number to move and spin is the second. USA Pickleball published a spin test in July 2026 and applies it to paddles submitted from 1 October 2026. The cap is 2,100 RPM, measured on a dual camera rig. The structure repeats the PBCoR schedule. The cap governs new submissions, while paddles certified before that date keep the result they already hold.
Under the same update, surface roughness and friction stop being a pass or fail line and become a measurement of record. That change matters to anyone selling a textured face. A face can clear a roughness figure and still measure past the spin cap. Closing the gap between those two readings is the surface engineer’s job. That work has to happen before tooling is cut. A new texture normally means a new mold and a new sample round.
USAP and UPA-A are not the same door
Two bodies certify paddles in the United States, and they do not run the same test on the same rig. USA Pickleball uses PBCoR with a 0.43 ceiling and governs sanctioned amateur play. UPA-A uses a propulsion figure with a limit of 0.385 for a new paddle and 0.405 after break-in. That body governs the professional tours.
A paddle can clear one and fail the other, and a recreational buyer will never notice. Nothing printed on the packaging states which file exists. A brand selling into both channels keeps two files, and neither one substitutes for the other. The rest of the matrix sits in which market each certificate covers, including the markets that ask for neither.
What a factory can and cannot promise
We cannot promise that a paddle stays approved. The ceiling is set by a body we do not sit on, and it has moved twice in two years. A supplier making that promise is describing something outside its control. The promise is narrower and more useful. It covers the number that goes into the file a certified build keeps. That file holds the construction, the core thickness, the edge treatment and a sample taken off production.
Those four items belong in the brief before tooling starts. With them written down, a brand has something to compare against when a limit moves. A brand that leaves them to the factory default keeps no record. Once the record is missing, nobody can say why a paddle reads the way it does. Nor can anyone say which part of it changed between two runs.
One habit removes most of the confusion. The file should carry the submission date and the rule version beside the test figure. When 0.43 or 2,100 RPM comes up in a meeting, the answer is already on paper instead of in someone’s memory.
The second habit protects an order. When a supplier quotes a test figure, ask which version of the sample produced it and on what date. A supplier who answers both without checking is quoting a memory. A memory you cannot audit is worth less than an empty column in the file.
A test report is the one document worth a second pair of eyes. It can be sent to the desk that reads a test report. The reply comes back within 24 hours, and it names the figures a brand should confirm with its own supplier.
FAQ
Why was the JOOLA Perseus Mod TA-15 removed from the approved list?
It read above the PBCoR ceiling that USA Pickleball introduced in late 2024. The model had been approved before that test existed, so it had no figure of its own to compare. It lost sanctioned play status on 1 July 2025.
Is the Mod TA-15 still legal to play with?
It remains legal in recreational play and in events that do not follow USA Pickleball sanctioning. Individual tournament organisers set their own policy, and JOOLA’s guidance to buyers was to check with the event rather than with the brand.
Can a delisted paddle get back on the list?
Yes, and the ProKennex Black Ace LG is the case to study. It was removed on the same notice and reinstated on 3 June 2025, after a second test showed the build matched its certification. The route back is a retest rather than an appeal.
Does a new colour or a new name need a new certification?
A new name on a different construction does. The manual ties approval to a model name, so a paddle with a different core or surface counts as a separate model that needs its own submission. A colour change on the same build does not move the file.
What is the current PBCoR limit?
0.43 for paddles submitted after 1 November 2025. Paddles submitted between 1 November 2024 and that date were tested against 0.44, and each model keeps whichever figure applied at its own submission.
What changes for a brand launching in 2027?
A new submission faces 0.43 and a 2,100 RPM spin cap, and it may also need a UPA-A file if the model sells into the professional channel. Both figures attach to the submission date rather than the ship date. A slow tooling schedule can push a model across a limit that did not exist when the project started.