Methodology

A TTDNA rating runs from 0 to 100 and means the same thing everywhere on this site. 80 on a Butterfly sheet is 80 on a DHS sheet, and the same scale is used when blades are rated. There is no second set of numbers and no per-brand conversion table.

A rating is only useful if it can be compared. Manufacturers each define their own range, their own zero point and their own idea of what a 10 in control means, so their numbers cannot be compared across brands. TTDNA ratings are produced in one place under one set of rules.

A TTDNA rating is a standardized comparison value. It is not presented as a laboratory measurement performed by TTDNA, and no claim is made that every characteristic on this site was physically tested in one place under one protocol.

Every rating starts from source material, weighted by how strong that source is. Stronger evidence takes precedence; supporting evidence never overrides it on its own.

Primary sources

• Manufacturer product pages and official specifications
• Official technical documentation
• Published construction and material information

Independent data

• Standardized equipment tests where they exist
• Reputable technical databases
• Published measurements such as sponge hardness, weight and thickness

Supporting evidence

• Retailer specification listings
• Experienced player reports and community data
• Product reviews and observed playing behaviour

Where sources conflict, the stronger tier decides. Where a primary source contradicts a retailer listing, the primary source is used. Where two strong sources genuinely disagree, the conflict lowers confidence rather than being averaged away silently.

Manufacturers do not share one rating standard. Butterfly’s “Speed 13” cannot be treated as equivalent to another brand’s “Speed 13”: the ranges, reference products and test conditions differ. Brand figures are also positioned against that brand’s own back catalogue and move when a new flagship needs room.

TTDNA therefore converts source information onto its own standardized internal scale so products from different manufacturers can be placed side by side. Normalization improves cross-brand comparison. It does not convert a manufacturer rating into a laboratory measurement, and this site does not present it as one.

Every rating traces back to six core inputs:

Speed · Spin · Control · Arc · Hardness · Tackiness

Those inputs are established from the physical properties of the sheet — sponge hardness, topsheet grip, surface type, published construction data — from published and independent specifications, and from the behaviour that surface family reliably shows in play. Everything else on a product page is calculated from them. The same input produces the same output on every page load, and every rating can be reproduced from the six numbers above.

When no directly comparable information exists for a specific model, TTDNA infers an expected range from the evidence that does exist: sponge hardness, rubber type, topsheet construction, blade composition, ply structure, fibre type, thickness, product category, related models within the same line, known characteristics of similar equipment, manufacturer positioning, and available testing or player evidence.

These values are labelled as estimates on the product pages. They exist to make products comparable when standardized measurements are unavailable, and should be read as informed, evidence-based estimates rather than measurements.

Scores that summarise several characteristics — an overall score, a style match, a suitability figure — are composite TTDNA metrics. They are calculated from multiple relevant characteristics and are designed as a quick comparison aid.

A composite is not a separately measurable physical property and is not presented as one. Because playing styles weight characteristics differently, a composite cannot replace the individual ratings or personal fit; read it alongside the characteristics it comes from.

Confidence describes how much of a rating rests on direct evidence.

• Higher — several strong sources agree, direct specifications or published measurements exist, and little inference is required.
• Medium — good information exists, but some normalization or inference is needed to place it on the TTDNA scale.
• Lower — limited public information exists and a larger part of the rating is estimated.

Lower confidence does not mean poor data. It means wider uncertainty around the value, and it is stated so the number can be weighted accordingly.

Individual data blocks carry a compact status marker.

• Verified — confirmed through reliable published sources.
• Published — taken directly from manufacturer or technical documentation.
• Normalized — derived from available specifications and evidence and converted to the TTDNA comparison scale.
• Estimated — no directly comparable standardized measurement exists; the value is an evidence-based TTDNA estimate.

For blades the same principle applies to construction. Where the exact materials and layer order are publicly confirmed, they are shown as verified. Where they are not, the stack is shown as the best available construction estimate and unpublished plies are stated as such rather than filled in.

These are the exact derivations. Hardness is sponge hardness on the normalised degree reference.

Power             = 0.70 x Speed   + 0.30 x Hardness
Grip              = 0.75 x Spin    + 0.25 x Tackiness
Catapult          = 0.55 x Speed   + 0.25 x (100 - Hardness)
Throw angle       = 0.60 x Arc     + 0.25 x Spin      + 0.15 x (100 - Hardness)
Elastic           = 0.50 x Speed   + 0.50 x Catapult
Spin sensitivity  = 0.60 x Grip    + 0.40 x Spin      - 4 if Tackiness <= 50
Short game        = 0.70 x Control + 0.30 x (100 - Speed)
Blocking          = 0.60 x Control + 0.20 x (100 - Hardness) + 12
Backhand flip     = 0.50 x Spin    + 0.30 x Control   + 10
Counterloop       = 0.50 x Speed   + 0.35 x Spin      + 0.15 x Control
Opening loop      = 0.55 x Spin    + 0.25 x Arc       + 0.20 x Control
Chop              = 0.60 x Control + 0.40 x (100 - Speed)
Serve             = 0.60 x Spin    + 0.25 x Control   + 0.15 x Tackiness
Service receive   = 0.55 x Control + 0.45 x Short game
Difficulty        = 0.45 x Speed   + 0.30 x Hardness  + 0.25 x (100 - Control)
Beginner friendly = 100 - 0.85 x Difficulty
Durability        = 70 + 0.06 x (100 - Tackiness) - 0.15 x (Speed - 70)

Six inputs feed everything: Speed, Spin, Control, Arc, Hardness, Tackiness.
Arc is the shape of the ball path. Throw angle is how high the ball leaves the
rubber. They are separate inputs and are never treated as the same number.
Catapult and Short game are chained into Elastic and Service receive before
rounding, so no drift builds up. Every result is rounded and clamped to 0-100.

When a core number cannot be established for a specific model, the surface family supplies it. These are the only baseline values in the system, and they are the reason an obscure long-pips sheet still gets a full rating set instead of a page of blanks.

Thickness options and OX are variants of one model, not separate products. A variant starts from the base sheet and is shifted by fixed rules, so a variant comparison shows the shifted rating rather than the base sheet.

Inverted     1.5 mm and under   Speed -5   Spin -2   Control +4   Power -6   Short game +4
Inverted     1.6-1.8 mm         Speed -3   Spin -1   Control +2   Power -4   Short game +2
Inverted     1.9-2.0 mm         reference sheet, no shift
Inverted     2.1-2.2 mm         Speed +2   Spin +1   Control -1   Power +3
Inverted     Max                Speed +3   Spin +2   Control -2   Power +4
Long pips    OX                 Speed -6   Power -7  Blocking +4  Short game +4
Anti-spin    OX                 Speed -5   Power -6  Blocking +4  Short game +3

Short pips and medium pips follow the same thickness pattern as inverted, with the shift applied to blocking instead of spin. OX exists only where it exists in reality: long pips and anti-spin. A thickness rule never moves a rating by more than 8 points.

Two products are compared on the full rating set, every time. Nothing is dropped because one side is thinner on data.

• Differences below the tie tolerance are reported as similar, not as a win for either side.
• Some ratings are neutral — tackiness, arc, catapult, throw angle, elasticity, spin sensitivity, difficulty. Higher is not better there, only different.
• Player-profile similarity weights the ratings differently per playing style. What comes out is a match for that style, not an overall quality score.

If a rating looks wrong, report it with the model name and what you are basing it on. The fix goes into the inputs so it propagates everywhere, rather than being patched on one page.