بسم الله الرحمن الرحيم
المصادر
86 Classical Works
From Sahih al-Bukhari to Tanzih al-Shari'ah — the entire hadith tradition, structured and graded.
The Nine Books
Major Collections
| المجموعةCollection | المصنفCompiler | الوفاةd. AH | عدد الأحاديثHadith (approx.) | نسبة الصحيح والحسنSahih + Hasan Rate |
|---|---|---|---|---|
| صحيح البخاري Sahih al-Bukhari | al-Bukhari · d. 256 AH | 256 | 7,580 | 95.9% sahīh + hasan |
| صحيح مسلم Sahih Muslim | Muslim ibn al-Hajjaj · d. 261 AH | 261 | 7,360 | 97.7% sahīh + hasan |
| سنن النسائي Sunan al-Nasa'i | al-Nasa'i · d. 303 AH | 303 | 5,679 | 97.2% sahīh + hasan |
| سنن أبي داود Sunan Abi Dawud | Abu Dawud · d. 275 AH | 275 | 5,272 | 93.8% sahīh + hasan |
| سنن ابن ماجه Sunan Ibn Majah | Ibn Majah · d. 273 AH | 273 | 4,338 | 91.4% sahīh + hasan |
| جامع الترمذي Jami' al-Tirmidhi | al-Tirmidhi · d. 279 AH | 279 | 3,924 | 90.0% sahīh + hasan |
| سنن الدارمي Sunan al-Darimi | al-Darimi · d. 255 AH | 255 | 3,513 | 93.5% sahīh + hasan |
| مسند أحمد Musnad Ahmad | Ahmad ibn Hanbal · d. 241 AH | 241 | 26,635 | 90.5% sahīh + hasan |
| موطأ مالك Muwatta' Malik | Malik ibn Anas · d. 179 AH | 179 | 1,829 | 89.9% sahīh + hasan |
Total across 86 works: 449,285 hadith (v3.4 deployed). 350,646 independently graded (78.0%). v3.26 staged adds 8,241 records from 6 more works (Ibn al-Sunnī ʿAmal, Hannād Zuhd, al-Quḍāʿī Shihāb, al-ʿUqaylī Ḍuʿafāʾ, Ibn al-Mubārak Jihād, Ibn ʿAdī Kāmil) bringing the total to 457,526 across 92 works. See v3.13–v3.26 changelog.
ملخص توزيع الدرجات Grade distribution summary
Bukhari (95.9% post-v3.8 honest de-inflation), Muslim (97.7%), and Nāsaʾī (97.2%) lead the sahih+hasan rankings — independently computed, not tuned to match. All nine major collections are above 89%. Every rate is the output of the IPSC grading engine applied uniformly. See v3.8 cluster rebuild for why Bukhari moved from 97.3% to 95.9%.
جدول التقارب الكامل Full Convergence Table
54 collections with 50+ graded records. Every rate independently computed — no grades imported from printed books.
Tier 1 — Sahih Collections
Compilers who applied the strictest narrator selection criteria. These works drew almost exclusively from T1–T3 narrators with verified chain continuity.
| Collection | Total | Graded | S+H Rate |
|---|---|---|---|
| Sahih al-Bukhari | 7,580 | 7,557 | 95.9% |
| Sahih Muslim | 7,360 | 7,359 | 97.7% |
| Sunan al-Nasa'i (Kubra) | 11,701 | 11,685 | 95.3% |
| Musnad al-Sarraj | 1,656 | 1,640 | 96.2% |
| Riyadh al-Salihin (al-Nawawi) | 1,891 | 813 | 95.7% |
| Ibn Jarud, al-Muntaqa | 1,098 | 1,098 | 94.5% |
| Sahih Ibn Khuzayma | 3,092 | 3,040 | 94.3% |
| Musnad al-Shafi'i | 1,768 | 1,768 | 94.1% |
Bukhari (95.9%) — Al-Bukhari required confirmed meeting (liqa') between each teacher and student, not just temporal overlap. His chains through known mudalliseen use tasrih (explicit hearing), not 'an'anah. The 4.1% divergence breaks down to: disambiguation errors on common names, anonymous-narrator penalties, mu'allaqat (suspended citations included as supporting evidence), and legitimate weak chains included for topical completeness. Note on the 95.9% figure: earlier IPSC releases reported 97.3% — corrected in v3.8 after the cluster rebuild surfaced inflated supportingChain counts. The 95.9% number is the honest post-de-inflation figure.
Muslim (97.7%) — Nearly identical criteria to Bukhari but accepted mu'asarah (temporal overlap) instead of requiring confirmed meeting. Includes more maqtu' (783 scholar statements) and mawquf (305 Companion statements) than Bukhari, some with weaker chains since they are non-prophetic content.
al-Nasa'i (95.3%) — Known for narrator criticism some scholars considered equal to Bukhari's. The 4.7% divergence includes chains with T4–T5 narrators included for fiqh relevance, plus 276 mursal chains where the mursal cap applies.
Riyadh al-Salihin (95.7%) — Only 813 of 1,891 are graded because al-Nawawi cited hadith by reference to source collections, not by independent chains. The graded records are the ones with embedded isnads.
Tier 2 — Major Sunan and Musnad Works
Strong transmission networks but broader inclusion. These works intentionally contain mawquf (Companion statements), maqtu' (scholar statements), and sometimes weak chains for scholarly discussion. The 88–94% range reflects this deliberate breadth.
| Collection | Total | Graded | S+H Rate |
|---|---|---|---|
| Sunan Abu Dawud | 5,272 | 5,272 | 93.8% |
| Musannaf Ibn Abi Shayba | 37,943 | 37,942 | 93.7% |
| Musannaf 'Abd al-Razzaq | 21,040 | 21,003 | 93.7% |
| Sunan al-Darimi | 3,513 | 3,510 | 93.5% |
| al-Adab al-Mufrad (al-Bukhari) | 1,359 | 1,349 | 93.5% |
| Musnad Ibn al-Ja'd | 3,446 | 3,346 | 92.9% |
| Musnad al-Tayalisi | 2,767 | 2,766 | 92.4% |
| al-Jami' al-Kabir (al-Suyuti) | 45,609 | 29,010 | 92.4% |
| Musnad Abu Ya'la | 7,553 | 7,427 | 92.2% |
| Musnad Ibn Rahawayh | 2,396 | 2,390 | 92.1% |
| al-Ahadith al-Mukhtara (al-Diya') | 4,240 | 4,212 | 91.9% |
| Sahih Ibn Hibban | 6,673 | 6,650 | 91.6% |
| Sunan Ibn Majah | 4,338 | 4,338 | 91.4% |
| Bulugh al-Maram | 1,572 | 494 | 91.1% |
| al-Targhib wa-l-Tarhib | 5,555 | 2,951 | 90.7% |
| Musnad Ahmad | 26,635 | 26,369 | 90.5% |
| Tafsir al-Tabari | 20,396 | 20,364 | 90.5% |
| Sharh Mushkil al-Athar (al-Tahawi) | 2,930 | 2,910 | 90.1% |
| Jami' al-Tirmidhi | 3,924 | 3,920 | 90.0% |
| al-Sunan al-Kubra (al-Bayhaqi) | 21,597 | 21,575 | 90.0% |
| Muwatta' Malik | 1,829 | 1,803 | 89.9% |
| Sharh Ma'ani al-Athar (al-Tahawi) | 6,661 | 6,661 | 89.8% |
| Musnad al-Bazzar | 9,079 | 8,914 | 88.3% |
Abu Dawud (93.8%) — Abu Dawud stated he included the strongest hadith on each topic, plus weaker ones where no stronger version exists. He explicitly included some da'if hadith and noted their weakness. 269 mursal chains, 112 very-weak records — by the compiler's own design.
Ibn Abi Shayba (93.7%) — A massive musannaf with 16,572 maqtu' and 8,244 mawquf — deliberately non-prophetic content with shorter chains. The 6.3% da'if rate is expected for a work of this scope drawing from major Kufan and Basran networks.
Ahmad (90.5%) — Organized by Companion, not by quality. Ahmad included multiple chains for the same hadith, including weaker ones. 953 mursal chains, 798 anonymous narrator penalties, 1,228 maqtu' statements. The 9.5% da'if rate reflects the musnad methodology — collect everything a Companion narrated.
Tirmidhi (90.0%) — Al-Tirmidhi explicitly included and discussed weak hadith for scholarly analysis. He would cite a hadith, grade it himself (sometimes hasan gharib), and explain why scholars differ. 142 mursal chains. 10% da'if is consistent with his stated methodology.
Tafsir al-Tabari (90.5%) — The graded records are those with parseable isnads. Contains 3,772 mawquf and 9,069 maqtu' reports — scholar opinions on Quran interpretation, correctly graded lower because they are not prophetic hadith.
al-Jami' al-Kabir (92.4%) — Only 29,010 of 45,609 are graded because al-Suyuti compiled from across the tradition. 16,599 records have no parseable isnad (reference entries). The high rate on graded records reflects the strength of his source collections.
Tier 3 — Broader Inclusion / Specialist Works
Collections focused on comprehensiveness, rare chains, or thematic coverage rather than quality filtering. These works preserve narrations the major collections chose not to include.
| Collection | Total | Graded | S+H Rate |
|---|---|---|---|
| al-Mu'jam al-Kabir (al-Tabarani) | 21,586 | 21,516 | 87.9% |
| Mishkat al-Masabih | 6,368 | 1,038 | 87.6% |
| Kanz al-'Ummal | 46,650 | 4,246 | 87.5% |
| Musnad al-Ruyani | 1,565 | 1,498 | 87.4% |
| Jami' al-Usul (Ibn al-Athir) | 9,524 | 807 | 87.1% |
| al-Mustadrak (al-Hakim) | 8,831 | 8,828 | 86.9% |
| Lubab al-Nuqul (Asbab) | 330 | 291 | 85.6% |
| Shu'ab al-Iman | 11,398 | 11,153 | 84.4% |
| Sunan al-Daraqutni | 4,733 | 4,733 | 84.5% |
| al-Mu'jam al-Saghir | 1,195 | 1,188 | 83.6% |
| al-Mu'jam al-Awsat | 9,479 | 9,399 | 82.1% |
| Ibn al-'Arabi, Mu'jam | 2,070 | 2,062 | 80.4% |
| Majma' al-Zawa'id | 10,254 | 5,165 | 79.7% |
| al-Matalib al-'Aliyah (Ibn Hajar) | 4,730 | 4,584 | 77.1% |
| Ibn Hisham, al-Sira | 172 | 171 | 74.3% |
al-Mustadrak (86.9%) — Al-Hakim claimed these hadith meet Bukhari/Muslim criteria but were excluded. Classical scholars (especially al-Dhahabi) widely criticized his leniency — he accepted narrators Bukhari and Muslim rejected. The 13.1% da'if independently confirms what al-Dhahabi said.
al-Mu'jam al-Awsat (82.1%) — Lower than al-Kabir because al-Awsat specifically collects rare and unusual chains (gharib narrations) — by definition chains with fewer parallel transmissions and often weaker narrators. 17.9% da'if reflects this "unusual chains" focus.
Majma' al-Zawa'id (79.7%) — Al-Haythami's index of hadith found in musnad/mu'jam collections but NOT in the Six Books. By definition these are the hadith that Bukhari, Muslim, Abu Dawud, etc. chose not to include. 20.3% da'if is expected — these are the narrations left behind by the major collections' selection process. 859 mursal chains.
al-Matalib al-'Aliyah (77.1%) — Ibn Hajar collected hadith from musnad works not in the major collections. Even more selective toward unusual and weak chains than al-Haythami. 22.9% da'if, 516 very-weak records.
Kanz al-'Ummal (87.5%) — A compilation index referencing hadith in other collections. Only 4,246 of 46,650 are graded — the rest cite by reference without independent chains. The 87.5% rate applies only to records with embedded isnads.
Ibn Hisham, al-Sira (74.3%) — Prophetic biography, not a hadith collection. Includes historical narratives with informal chains, storyteller reports, and poetry. Muhammad ibn Ishaq (the original author, T4) was known for tadlis. 37 very-weak records, 27 anonymous narrator penalties. The 25.7% divergence correctly reflects that sira literature applies different evidentiary standards.
Tier 4 — Fabrication and Specialized Collections
Fabrication catalogs and specialized genres with different evidentiary standards. The rates here test the engine's ability to handle edge cases.
| Collection | Total | Graded | S+H Rate |
|---|---|---|---|
| Tanzih al-Shari'ah (Ibn 'Iraq) | 3,919 | 1,482 | 87.4% |
| al-La'ali al-Masnu'ah (al-Suyuti) | 1,846 | 1,603 | 71.6% |
| al-Mawdu'at (Ibn al-Jawzi) | 1,057 | 817 | 67.3% |
| Kitab al-Zuhd (Ibn al-Mubarak) | 2,039 | 2,025 | 43.4% |
Tanzih al-Shari'ah (87.4%) — Surprisingly high for a fabrication catalog. Ibn 'Iraq wrote Tanzih specifically to respond to al-Suyuti's al-La'ali — arguing many of al-Suyuti's "fabricated" hadith were authentic. The high rate reflects this polemical purpose: he included hadith WITH strong chains to prove they are not fabricated.
al-La'ali al-Masnu'ah (71.6%) — Al-Suyuti's fabrication catalog. 226 very-weak records. The 71.6% sahih+hasan rate exists because many hadith in fabrication catalogs have authentic parallel chains in other collections. The specific chain al-Suyuti cited may be fabricated, but taqwiyah from legitimate parallel chains raises the grade. This is methodologically correct.
al-Mawdu'at (67.3%) — Ibn al-Jawzi's fabrication catalog. 113 very-weak records. Same taqwiyah dynamic as al-La'ali. Ibn al-Jawzi was also criticized by later scholars for being too aggressive — some hadith he declared mawdu' were later accepted as hasan or sahih by other critics.
Kitab al-Zuhd (43.4%) — The lowest rate in the corpus. An asceticism collection from the 2nd century AH: 607 maqtu' statements (30%), 380 mawquf (19%), 117 mursal chains, 67 anonymous narrators. Ibn al-Mubarak preserved a devotional tradition, not a hadith canon. 56.6% da'if is the correct output for a collection that prioritizes ascetic wisdom over chain authentication.
Divergence Causes
Every divergence from 100% has a reason. The IPSC does not claim that every hadith in Bukhari is sahih or that every hadith in al-Mawdu'at is fabricated. It applies the same methodology uniformly and reports what the chains say.
| Cause | Description | Affected Collections |
|---|---|---|
| Compiler intent | Collection deliberately includes weak chains for scholarly discussion or comprehensive coverage | Tirmidhi, Abu Dawud, al-Daraqutni, al-Bayhaqi, al-Hakim |
| Non-prophetic content | Mawquf and maqtu' statements with shorter, sometimes weaker chains | Ibn Abi Shayba, 'Abd al-Razzaq, Tafsir al-Tabari, Kitab al-Zuhd |
| Rare chain focus | Collection preserves unusual or uncommon narrations | al-Mu'jam al-Awsat, Majma' al-Zawa'id, al-Matalib al-'Aliyah |
| Fabrication catalog | Fabricated chains + authentic parallel chains = mixed grades via taqwiyah | al-La'ali, al-Mawdu'at, Tanzih al-Shari'ah |
| Non-hadith genre | Sira, tafsir, or devotional literature with different evidentiary standards | Ibn Hisham, Kitab al-Zuhd, Lubab al-Nuqul |
| Reference compilation | Most records cite by reference, not by independent chain | Kanz al-'Ummal, Mishkat, Jami' al-Usul, Riyadh al-Salihin |
| Disambiguation error | Wrong PID assigned to common narrator name | ~2–3% of da'if grades across all collections |
| Mursal chains | Chain doesn't reach a Companion — grade capped at da'if | Present in all collections, concentrated in early works |
The grading engine independently reproduces the quality hierarchy classical scholars established over centuries — Sahih collections at the top, fabrication catalogs at the bottom — without importing any grades from printed books.
77 Additional Works
Supplementary Collections
Musannaf works
Ibn Abi Shayba, 'Abd al-Razzaq
Mu'jam works
al-Tabarani (Kabir, Awsat)
Mustadrak & supplementary
al-Hakim, Ibn Hibban, al-Bayhaqi
Musnad collections
al-Bazzar, Abu Ya'la, al-Tayalisi
Compilation indices
Kanz al-'Ummal, al-Jami' al-Kabir
Tafsir
al-Tabari
Fabrication catalogs
Tanzih al-Shari'ah, al-La'ali al-Masnu'ah
86 classical works. 449,285 hadith deployed. Every chain graded.
v3.26 staged: 92 works, 457,526 records. Changelog