What the measurements are
Power is serve speed and pop is punch-volley speed, both with a radar gun, in mph. Spin comes from 240 fps slow-motion video of topspin and slice serves, in rpm. Swingweight and twistweight are stock readings on a Briffidi; static weight is in ounces and balance in millimeters. A person swings the paddle, not a machine, so small differences may not hold up on every swing.
Blank, nonnumeric and out-of-range readings stay unavailable (for example a 0.01 twistweight placeholder). Rows from my newer measurement session replaced the older numbers for the same paddles; those pages say so in their source notes.
Which paddles are in the 2026 field guide
The field guide covers 98 profiles marked 2026, including paddles from my latest measurement update. Its main catalog and recommendations focus on 86 foam-core profiles; foam-assisted honeycomb and unconfirmed constructions are kept in a separate section. A generation number is not a core material: explicit construction text identifies most foam entries, and manufacturer descriptions fill in abbreviated ones.
Some paddles appear in the datasheet under two or three names. Those rows are reconciled by explicit name, never by fuzzy matching, and count once. If their readings differ, one selected sample is shown and the range is listed in the page’s source notes; readings are never averaged into a new measurement.
What the paddle-page charts show
Rank strips and scatterplots compare a paddle with measured paddles of the same primary shape. A percentile is the share below a value plus half of equal values. A higher number is not always better: a lower swingweight is quicker, a higher one has more plow-through. Sharing a shape category does not mean identical dimensions or outline.
Percentile context beside the ratings
Small badges beside power, pop and fresh spin compare valid readings with unique 2026 foam-core profiles of the same primary shape. Each badge names the group and its metric-specific count; at least ten readings are required. An older foam paddle can be benchmarked against that group without becoming a 2026 member. Other constructions and unclassified profiles do not receive a foam badge.
Standing = 100 × (count below + half of tied values) / group size, rounded to a whole percentile for display. It is raw rank context, not a smoothed score, percentage performance gain or PPI input. Ratings, PPI, measured values and historical plots remain unchanged. The full rank strips use the broader same-shape measured catalog, so their ranks may differ from these explicitly narrower badges. Example and calculation.
Two axes for feel
Every paddle profile has a four-quadrant feel map: soft on the left and stiff on the right; hollow at the top and dense at the bottom. The four quadrants are Soft + Hollow, Stiff + Hollow, Soft + Dense and Stiff + Dense. These are separate impact descriptions, not a quality ranking or a claim about literal empty space. A soft, hollow paddle can still rebound strongly. The chart adds no new weight to PPI and is not reverse-engineered from its score.
The exact-model review description takes priority on each axis, then the matched datasheet description. Manufacturer descriptions remain labeled as such. Explicit soft/stiff and dense/hollow wording gives the main categories; plush/cushioned, crisp/firm, muted/connected and thonky are labeled as leanings. Dense/hollow mixtures and explicit neutral firmness can sit on an axis. A relative comparison alone does not establish absolute firmness, and weighted-only descriptions are excluded from stock placement. V-Sol Tour uses my later, linked near-neutral break-in observation rather than its fresh-sample stiffness.
The dot occupies a coarse category, not a measured coordinate or a confidence interval. A single known axis gets a band with the other axis left unknown; missing descriptions get no dot, not a false center position. The readout identifies the description behind each axis. No sibling paddle’s feel, foam type, grit tier or presumed break-in change fills a missing axis. Existing historical feel graphics remain unchanged.
One main shape, then the series comparison
Most of my court time usually goes into one main shape, followed by shorter comparison sessions with the other shapes in the series. That helps me separate the platform’s core response from the footprint: elongated, hybrid, widebody or standard; aero or flat top; handle room; swingweight and twistweight. The main review carries the deeper break-in story. Sibling pages bring together written comparisons where recorded, the actual measurements and the shape tradeoffs.
This does not mean equal hours on every shape or that every catalog entry has a full court review. The page links the main review and separates recorded observations from shape-based expectations. Long-term hours, durability results and finished weight setups stay tied to the sample that produced them. Small speed differences stay small, and measured exceptions take priority over the usual pattern. Original/Pro series, grit revisions and known thickness differences are kept separate. Family context does not change measured values, six ratings or PPI.
General geometry context: Selkirk’s paddle-shape guide . Model-specific expectations use my linked review and the recorded samples, not manufacturer marketing scores. If stability needs work, the starting experiment remains 3 g total at 3/9, split evenly; finished SW/TW must be remeasured.
How alternatives are chosen
Nearest alternatives use power, pop, spin, swingweight and twistweight together, within the same shape and construction group, among 2025 and 2026 releases. Each difference is divided by a comparison band: 1 mph power, 1 mph pop, 75 rpm spin, 3 SW and 0.2 TW, and the root mean square of those scaled differences sets the order. The bands are editorial choices, not measured uncertainty. The cheaper list applies the same method to paddles at least $10 cheaper. Feel, durability and affiliate status are never part of the distance.
PPI: four components
PPI combines 50% shape-role playing profile, 14.3% sweet spot, 21.4% stock setup and 14.3% break-in confidence (rounded). Exact weights preserve the previous 35:10:15:10 ratio and divide it by 70. Multiply the weighted 0–10 result by 10 and round. Feel fit and personal game fit are separate review ratings, never PPI inputs. Missing personal-fit ratings do not block PPI.
The fixed measurement guide uses power = 8 + (mph − 60) / 5, pop = mph / 5, and spin = 8 + (rpm − 2000) / 500. Each is clipped to 0–10 and rounded to one decimal. A 2 mph gap is 0.4 points before clipping; a 200 rpm gap is 0.4 points. These anchors are scoring conventions, not quality thresholds. A change in the comparison group does not change a rating. Fresh spin and grit longevity remain separate.
Dwell, control and sweet spot prioritize explicit numeric review ratings, then exact-model review wording. Missing inputs use labeled characteristic guidance: soft/plush versus stiff/firm mainly informs dwell, width and stock TW mainly inform forgiveness, and control also considers measured pop. Dense/hollow earns no performance bonus. Stock never inherits a weighted sample’s forgiveness description. Actual width takes priority over a clearly labeled nominal shape width. These are editorial guide values, not measured contact time, control or sweet-spot area. Existing database proxies remain labeled fallbacks when a replacement lacks inputs. See the exact characteristic anchors and equations.
PPI uses the same ratings shown in the profile, with no extra power, pop or spin calibration. Stock setup uses SW, TW, weight and balance. Break-in confidence uses its recorded input, not an mph inference. There are no automatic break-in bonuses. Missing eligible components are omitted and remaining weights are rescaled. Historical article ratings and plots remain as recorded. See PPI scores and estimates.
Spin and surface life
Fresh RPM and spin retention are different measurements. Reviewed paddles use the Buyer’s Lab surface evidence and its named sources, including John Kew’s linked wear tests. A family result keeps its family scope. A datasheet tier without an exact-model citation is shown as a lead to check, not a verified durability verdict. Surface revisions matter, and I keep updating real-play surface tracking.
Review and reply context
Paddle-page notes come from my Reddit posts, my replies and my supplied review drafts. Every note carries its source, and the build stops if a quoted fragment disappears or if a reader’s comment would be presented as my own play. Exact-model context and related-family context are labeled differently; a related note never creates a personal review, a PPI input or a measurement. Reader counterpoints are anecdotes, not an owner survey. Read the buying lessons.
Setup suggestions
The first side-weight experiment is 3 g total, split evenly at 3 and 9 o’clock. That adds 0.106 oz before grips or edge tape. Added weight alone does not produce a new measured SW or TW, so remeasure the finished setup.
Product photos
Photos are manufacturer images, matched by exact model and labeled when they show a shared series or an option illustration. They are not photographs of my test samples. Image sources.