Skip to main content
CrystalMark
PricingStart — $49.99/mo

Live exampleOne real examiner, fully unlocked — the exact profile subscribers see on every active examiner. Every figure is computed from public USPTO records.

Open every examiner →
Law Office M70·Active·Examining since 1991

Doritt Carroll

21,402 applications analyzed·last OA Jul 29, 2026

16.8%−6.1pp vs fleetSubstantive refusalLast 2 yrs · n=3,329 examinedFull record: 28.7% · n=20,916Share of examined apps that drew a substantive refusal (fightable grounds only). Approvals with no refusal count in the base. Last 2 years, vs the fleet's same window.40.2%−0.8pp vs fleetEscalates to Finaln≈6,001 refused appsOf refusals resolved on the merits (Final or approved), the share pushed to Final. Unfought abandons sit out. The n is derived, hence the ≈.63.7%−2.9pp vs fleetRecovery after refusal · when foughtn=3,053 fought refused appsOf refusals the applicant fought, the share that still reached publication. Higher is better. Fighting then abandoning counts as a loss; unfought refusals sit out.
DifficultyHow hard this examiner is to overcome, ranked 1–10 against active examiners. Weighted toward post-refusal behavior; refusal frequency counts least.
8/10Hard

More difficult than 70% of active examiners.

Why 8/10? ▾

Winnability vs the active fleet — how hard to overcome once refused.

Holds once you fight · 35% of score

Once refused, how often they push it to a Final rather than fold.

near fleet (−0.1σ)

Won't withdraw post-Final · 25% of score

When applicants press a Final §2(d), how rarely the refusal comes off.

above fleet (+0.7σ)

Refusal rate (basis-adjusted) · 20% of score

Refuses more than expected for their filing-basis mix.

near fleet (+0.1σ)

Survives TTAB appeal · 12% of score

When an applicant presses a TTAB appeal, how often the refusal survives.

well above fleet (+1.5σ)

Grinds extra rounds · 8% of score

More office-action rounds to reach resolution.

near fleet (−0.2σ)

Weights renormalized over available signals.
Decile across active examiners, n≥30.

Refusal grounds
Of apps with classified refusal grounds · independent rates — don't sum to 100%
66%§2(d) confusionfleet 64% · +2pp
24%Descriptive §2(e)(1)fleet 23% · +1pp
4%Surname §2(e)(4)fleet 3% · +0pp
3%Geographic §2(e)(2–3)fleet 4% · −1pp
10%Other substantivefleet 14% · −4pp

Cites two or more grounds at once on 6% of ground-classified apps.

Mechanical requirements excluded — compliance, not judgment.

Filing → first office action
Median · last 2 yrs · apps they examined from round 1
162 days≈ 5.3 months · fleet 166 days (same window)

Full record: 107 days

Mostly queue time, not examiner temperament.

Examiner's amendments
Share of interventions where an amendment could work · last 2 yrs
23%fleet 25% (same window)

250 of 1,105 amendment-eligible interventions — no substantive refusal in play

An examiner’s amendment fixes the application by phone — no formal refusal to answer, no response cycle. Only interventions an amendment could actually resolve count here (a §2(d) can’t be phoned away). A high share is the cooperative signature.

Stats by filing basis
Apps · refusal (of fetched apps) · publication (of resolved apps)
BasisAppsRefusalPublication
§1(b) intent12,78326.2%79.9%
§1(a) use8,13931.5%68.4%
Foreign · §44 + §661,34132%78.8%
Unknown13843.5%43.5%

The foreign bucket merges §44 and §66(a) — no §66(a)-only rate exists. §66(a) is separable as a count: 860 apps (4% of the docket).

Refusal rate by class
Colored vs the fleet in the same class · a multi-class app counts in each
ClassRefusalShare of examined apps in this class that drew a substantive refusal. Colored against the fleet's rate in the same class.§2(d)Of ground-classified refused apps in this class, the share citing §2(d). Shading marks distance from the typical examiner in this class, ▲ above and ▼ below — a composition read, not a good or bad one. Independent rates; they don't sum to 100%.§2(e)(1)Of ground-classified refused apps in this class, the share citing §2(e)(1). Shading marks distance from the typical examiner in this class, ▲ above and ▼ below — a composition read, not a good or bad one. Independent rates; they don't sum to 100%.SurnameOf ground-classified refused apps in this class, the share citing Surname. Shading marks distance from the typical examiner in this class, ▲ above and ▼ below — a composition read, not a good or bad one. Independent rates; they don't sum to 100%.GeographicOf ground-classified refused apps in this class, the share citing Geographic. Shading marks distance from the typical examiner in this class, ▲ above and ▼ below — a composition read, not a good or bad one. Independent rates; they don't sum to 100%.OtherOf ground-classified refused apps in this class, the share citing Other. Shading marks distance from the typical examiner in this class, ▲ above and ▼ below — a composition read, not a good or bad one. Independent rates; they don't sum to 100%.
Electronics & software · 009 · n=2,96228%71%25%2%1%6%
Advertising & business services · 035 · n=2,70731%67%27%5%4%4%
Education & entertainment · 041 · n=2,61529%65%30%2%5%6%
Clothing & footwear · 025 · n=2,15936%67%5%4%2%32%
Science & technology (software) · 042 · n=1,78927%67%34%2%1%4%
Insurance & financial services · 036 · n=1,06033%67%28%4%5%2%
Cosmetics & cleaning · 003 · n=95329%79%17%2%1%5%
Paper & printed matter · 016 · n=94328%54%24%3%2%23%
Pharmaceuticals · 005 · n=87427%76%24%2%1%5%
Toys, games & sporting goods · 028 · n=82622%78%14%5%0%7%
All 48 classes ▾
Household utensils & glassware · 021 · n=69722%67%18%3%1%14%
Staple foods & bakery · 030 · n=68533%64%26%7%4%8%
Medical & beauty services · 044 · n=67330%68%33%0%4%3%
Food & drink services · 043 · n=65730%71%22%3%4%4%
Medical apparatus · 010 · n=54124%76%25%2%0%3%
Appliances (lighting/heating) · 011 · n=52924%74%18%5%0%7%
Construction & repair · 037 · n=50531%72%17%8%8%3%
Legal & security services · 045 · n=49625%64%26%4%6%6%
Furniture · 020 · n=47424%68%23%4%4%4%
Meats, dairy & processed foods · 029 · n=41234%62%29%9%5%11%
Machines & machine tools · 007 · n=40023%75%26%7%1%6%
Leather goods & luggage · 018 · n=37729%72%9%6%2%16%
Beers & non-alcoholic drinks · 032 · n=37623%81%12%2%6%10%
Wines & spirits · 033 · n=36133%65%13%8%5%16%
Vehicles · 012 · n=36027%74%15%11%0%5%
Telecommunications · 038 · n=36028%75%27%2%0%2%
Jewelry & watches · 014 · n=35228%65%13%4%3%23%
Chemicals · 001 · n=28826%60%33%3%0%10%
Transport & storage · 039 · n=27934%67%34%2%5%4%
Treatment of materials · 040 · n=26832%70%27%5%4%6%
Agricultural & natural products · 031 · n=26335%55%22%5%10%27%
Metal goods & hardware · 006 · n=21126%73%18%4%2%4%
Fabrics & textiles · 024 · n=20725%75%18%4%2%10%
Hand tools · 008 · n=19124%73%11%4%0%13%
Non-metal building materials · 019 · n=17628%54%31%13%4%8%
Tobacco & smokers' articles · 034 · n=13929%76%24%0%0%0%
Lubricants & fuels · 004 · n=11824%75%18%7%0%14%
Lace, ribbons & notions · 026 · n=11416%72%17%6%0%11%
Rubber & plastics · 017 · n=11022%61%26%13%0%9%
Paints & coatings · 002 · n=8025%55%25%10%0%10%
Carpets & floor coverings · 027 · n=7220%50%36%14%0%7%
Cordage & raw textiles · 022 · n=5618%90%10%0%0%0%
Firearms & fireworks · 013 · n=5040%67%50%6%0%6%
Musical instruments · 015 · n=3726%—————
Unclassified · B · n=3231%40%80%10%0%0%
Unclassified · A · n=16——————
Yarns & threads · 023 · n=15——————
Class 200 · 200 · n=9——————

The read — their sharpest class gap is firearms & fireworks (class 013), 40% against a fleet 29% there, on 50 examined apps.

Refusal shading is green below the fleet and red above it. Ground shading is neutral — depth marks distance from the typical examiner in that class, ▲ above and ▼ below. Rates publish at 25+ examined apps in the class; ground shares at 10+ classified refusals. Cells below that show —.

Refusal rate by year
Substantive refusals · per-year n

The read — easing — down about 10 points from 2022 to 2026.

10%22%35%2022n=12742023n=14232024n=14822025n=16102026n=1161fleet 28.4% · since 2007, drawn flat

Data as of Jul 30, 2026 · Built on the public USPTO record — coverage begins Jan 2007, when the USPTO standardized outcome recording; careers predating that are covered from 2007 · n=21,402 applications · Fleet baseline = unweighted mean across 943 active examiners

How every figure is computed →

Your examiner has a record like this one.

One subscription opens the full profile — Pre-OA, §2(d), the Full record, and the Playbook — on every active examiner. Search yours the moment you're in.

Start — $49.99/mo →