Experiment 34
The first experiment that produces full candidates. At every period the cribs allow (27-29, and 53-95 sampled every third) under the 24 conventions, the forced key slots are fixed and the free slots are chosen to maximise the English score of the other 73 letters. The question is how English the best text is, against the same procedure on shuffled K4.
What the method does with the 24 known letters.
For a repeating secret wordthe secret word The word or string a cipher needs besides the method. of length 27, 28 or 29, the 24 known lettersthe known letters Sanborn has confirmed that positions 22 to 34 of the message read EASTNORTHEAST and positions 64 to 74 read BERLINCLOCK. fix 24 letters of the word and leave a few unknown. We fill the unknown ones in every way and read the messagethe message The 97 letters of English that Jim Sanborn wrote and enciphered. that comes out, keeping the one that reads most like Englishthe English score A number saying how much a stretch of letters looks like English, computed from how common each run of four letters is in English text.. If the cipher really were of this kind, one filling would read like a real message. We also run the same game on scrambled versions of the carved textthe carved letters The 97 letters cut into the copper at the CIA, shown at the top of every experiment page., to see how English-looking a text can be made by filling alone.
Words in the margin are defined on the glossary page.
The carved letters go in at the top; the reading comes out at the bottom.
A secret word of 74 letters repeats under the message, one letter per place, so place 75 uses the first letter again. On the sculpture's table, a known letter over its carved letter fixes the key letter in its slot: 24 slots are fixed that way.
The other 50 slots are free. The experiment tried all 26 letters in each, again and again, keeping whatever made the reading look most like English. That is a lot of freedom for 73 letters.
The reading scores -3.70 on the English scale, where real English is about -3.6, and the same climb on a shuffled copper scores better, -3.23, so under the rules of 2026-10-15 it fails the baseline gate. Its key takes 356 bits to write down, almost all of its loss of 5.03.
Laid out as the copper carries K4: OBKR ends the line that finishes K3, then three lines of 31. Left, as carved, the 24 known places in accent. Right, the candidate: a known place it gets right is an accent cell, one it misses is outlined. This is the lowest-loss reading that passed the original gates.
in: the copper, always the same
OBKR ends the line that finishes K3, then three lines of 31
out: the candidate
checked: the known letters in place, and the reading turns back into the copper
The loss of the best candidate, part by part. Zero is solved.
| Known letters missed | 0 of 24, ten points each | 0.00 |
| English | the 73 free letters score -6.71; English text scores -3.6 | 3.11 |
| Key | 144 bits to write it down (key words are cheap, arbitrary letters 4.7 bits each); each 60 beyond 60 costs a point | 1.40 |
| Loss | of 240, the worst possible | 4.51 1.9% |
All 18 candidates that passed every gate, lowest loss first.
| rank | missed | English | key | loss | share | |
|---|---|---|---|---|---|---|
| 1 | 0 | 3.11 | 1.40 | 4.51 | 1.9% | best |
| 2 | 0 | 3.21 | 1.33 | 4.53 | 1.9% | |
| 3 | 0 | 1.18 | 3.52 | 4.70 | 2.0% | |
| 4 | 0 | 1.51 | 3.28 | 4.79 | 2.0% | |
| 5 | 0 | 0.44 | 4.46 | 4.90 | 2.0% | |
| 6 | 0 | 3.66 | 1.25 | 4.90 | 2.0% | |
| 7 | 0 | 0.71 | 4.22 | 4.94 | 2.1% | |
| 8 | 0 | 0.10 | 4.93 | 5.03 | 2.1% | |
| 9 | 0 | 0.41 | 4.69 | 5.10 | 2.1% | |
| 10 | 0 | 0.01 | 5.16 | 5.18 | 2.2% | |
| 11 | 0 | 1.43 | 3.75 | 5.19 | 2.2% | |
| 12 | 0 | 1.30 | 3.99 | 5.29 | 2.2% | |
| 13 | 0 | 0.00 | 5.40 | 5.40 | 2.2% | |
| 14 | 0 | 0.00 | 5.63 | 5.63 | 2.3% | |
| 15 | 0 | 0.00 | 5.77 | 5.77 | 2.4% | |
| 16 | 0 | 0.00 | 6.10 | 6.10 | 2.5% | |
| 17 | 0 | 0.00 | 6.34 | 6.34 | 2.6% | |
| 18 | 0 | 0.00 | 6.57 | 6.57 | 2.7% |
Redo the method on a few letters with pencil and paper.
The same steps run by the code on the real letters.
period 27, KRYPTOS alphabet throughout, Vigenere; 24 slots forced, 3 free, climbed by English (3 restarts)
key YWOYNKYUYELYOIECBAQKMRDUMRI
text PEUZYOVDEPTIPLQBAOHBOEASTNORTHEASTDYFGMRRCWXTSZAVWCXZYNGDOTILLSBERLINCLOCKOFKOAWIHFOGQKNNNVTCUQVM
English of the free letters -7.790 (English about -3.6, noise about -6.5)
Family, key space, the constraint under test, and the clues it reads.
0005 left periods 27-29 and 53-97 alive under every convention, and 0009 showed the forced key stretches are noise; but noise in the key is compatible with a real periodic cipher if the key is not a word. The remaining test is the plaintext: fix the 24 forced slots, choose the 3 to 73 free slots by coordinate descent on the quadgram score of the 73 non-crib letters, with twenty random restarts, and record the best score per period and convention. A true periodic cipher at one of these periods would give a text scoring near English, about -3.6, at that period and convention; the same climb on shuffled K4 calibrates what overfitting achieves, which rises with the number of free slots. Candidates are submitted; they pass the crib gate by construction, so the English deficit and the key cost carry the verdict, and a 27-letter random key costs about 67 points of loss on its own. Bounds: coordinate descent is not exhaustive above period 27; period 27 is close to exhaustive with twenty restarts over 3 free slots.
Written before the run; the result decides it.
At periods 27-29 the best English score is below -5.0 per quadgram under every convention, the same as on shuffled K4, far from the -3.6 of English. At periods 53-97 the climb reaches -4.5 or better on both K4 and shuffled K4 because of the many free slots; the gap between them is under 0.2. Candidates pass the gates but carry a loss above 60 from the key cost; none reads as English.
The harness's record: candidates, gates, loss, and the negative result this experiment leaves behind.
All 18 candidates, lowest loss first.
period 29 key HYHMKBNENTXBQRMAEROOOXYHANZEF (5 free slots climbed); kry/kry/std/beau
period 28 key NZVDMORFNOEEADIAGKSSSUELRXLO (4 free slots climbed); kry/std/std/vig
period 56 key BFFCTASVNMWWAXSAUQSJPGWPJDPMNBFXOMFSIWDMFVJTLJFMOEMVEAHI (32 free slots climbed); kry/std/std/varbeau
period 53 key PRPUBCROXCHYSSRSHDNLQCECHPTSDQDDDEHDKTQUHQMEEHLMPTBKH (29 free slots climbed); std/std/kry/beau
period 68 key SDHLGKTHUVFNRRDEIXWHPREZYPYXXSYDKZSSVSGKUTHKTWVLCAFONAADRVOJCODFQXDE (44 free slots climbed); std/std/kry/vig
period 27 key EFHYHMKUKBNENTXBQRMAEXYHANZ (3 free slots climbed); kry/kry/std/beau
period 65 key TGNWRJLFKIYNHDJRTJHPXWXAKGZTOAXAFDJDFWXQRUGDKOMSJLBTEWZZQKPTPWCRG (41 free slots climbed); kry/kry/kry/beau
period 74 key LGZVHKFBYMIMUHPSIHZMLQTELWEHGZUPFHUIBBFTXDHEKZNUABBJSOGDXEPUSRFOGHTTKPBKSD (50 free slots climbed); kry/std/kry/vig
period 71 key KSDEGWSDMKXVMAYZNVJUIQTELWEHGZUPFHUFMTEZYWBPEPUHJVTJPMDVTVBWFQUOGHTTKPB (47 free slots climbed); kry/std/kry/vig
period 77 key HSUPMCUYLTXJVEGYTUVDPMCLWNQHMHLKFGFKQKXYGDAGNLNRQWVRTYNCMWQEKMCBYHITCDFYZDURP (53 free slots climbed); std/kry/kry/varbeau
period 59 key IYIFUCJFBFUJDXDKGVFJYJMWZCUJURLLXVUFSBCVWRNZBZMFKHTIHKGANRY (35 free slots climbed); kry/std/kry/beau
period 62 key FOGHTTKPBKSDJZDKAFFBLQTELWEHGZUPFHDOEJCLOOLEEKMSFEKOVKEXIVEASR (38 free slots climbed); kry/std/kry/vig
period 80 key XXZVLYUUDKQKHVOGLYXCTBKVSBPCXFCTACADMYQSJQCSAZOSTDQFWDKDOWQWJXPLRCFPLJIHUATZZZOS (56 free slots climbed); kry/kry/std/vig
period 83 key JSRDFXJVYXNXFNCNCSAFWSJRXTUOSORAMNCAZEZBOGLYRONFQMDORNNDEXUPZOFICOPEJKMCZKGEUSUCPSE (59 free slots climbed); std/kry/std/varbeau
period 86 key IZHXAZWDYMNTQMMIWYHMXZPBYXYCCUYQABQYIZDWIOIAPUWFVEZRDSOKLHILUZROGCQPUQMJNAXUQJIHJFLQZY (62 free slots climbed); std/std/std/varbeau
period 89 key NACPMJYYGVZKPKXKWQZMXQRTPHIUUARAYZSHZSHZXYXKDEQHKALUSKZBPUATKMVUGUHMRUKAFKAEONSFIHNKGNTZF (65 free slots climbed); std/kry/std/beau
period 92 key KVUKDZQJQDDANGRPTOAWTRDUMRIYWOYNKYRJFXJMOHYMKZHSNVPDNXEUVQJVCWIELYOIECBAQKDASNSHKYXCTQVDGIKQ (68 free slots climbed); kry/kry/kry/vig
period 95 key FALPWPOZKOFGDGYGYMDMJMCLWNQHMHLKFGXKQHZRHUEXABHQDWGHKKCPTKAYEAZBYHITCDFYZDATQMCOKGZPLSACOFWVNKB (71 free slots climbed); std/kry/kry/varbeau
Best English at 27-29: [-7.257, -6.808, -6.709]; shuffled K4 at 27-29: [-6.831, -6.886, -6.51]. Best overall -3.277 at period 95 vs shuffled best -3.233920610278189.
What was searched, as the harness recorded it.
ZWAWFAFIWGMGOGMDROTMAEASTNORTHEASTISTCDAPPUDSFROKLPDGBCUMLDWLBYBERLINCLOCKADHARIUHVWCQFISIQMNGDJKRYIHBKSCNNJARISEYUCCTEASTNORTHEASTWHCOFHCGDMXKAATSIXNHCIURRORLDBERLINCLOCKTHEJUFYFLIUAILUSCILVXVLYJAEIVIZKOYDBEACIOSSLEASTNORTHEASTVCACZLEVESITMILANBACMIIXDCHKYBERLINCLOCKATTHIZQLLPDLOFDFOIJGUSMPATEOVERSCURFONVGFSQYEASTNORTHEASTARELEDFEATHESTHEPRESHANINTSTIBERLINCLOCKNFGEEHNBSHCEBUDGEARESSUISINGHICNNJARISEYUNDEEASTNORTHEASTNETINTHEREINTEDTOTHERECORDSONBERLINCLOCKHIMJUFYFLIUAILUSOFFSOFVRROURUNXYOROCSVIOLEISEASTNORTHEASTANDINTHERATINGTHESTINTERSAIDEBERLINCLOCKTANYLOWCUVRISRQWLTMEDSLYVWPJDWEXAHRENTERNINOEASTNORTHEASTTHEPARTERANDANDINTHECONTERATIBERLINCLOCKSANDMEARECTUNWVSGUCAJFAZONVSENAGRAYSTEDITSUREASTNORTHEASTESINTHERATIONANDTHESTIONSIDERBERLINCLOCKMICHEMWHATHEWPNOSTNOXLSIRWAHOTALLISAREATHENTEASTNORTHEASTINANDSOMESANDINTHERENTHERSEARBERLINCLOCKSSIONESISSTINGETPVCHACTHUKSALLTHAVICEFFICESAEASTNORTHEASTNATIONANDINATESTINANDATTHERESBERLINCLOCKENSTIONESERTERETOODSLCQWHEUSHESTIONSTEONLYCOEASTNORTHEASTNTHETHESTATESTIMENTSANDINGATIBERLINCLOCKONTHEWARKENHISTHESOMESGTOVENOTHESTHERSTTHESAEASTNORTHEASTRERESIDENTERSAREARETOTHERATERBERLINCLOCKSONOFTHISFREETERSINEEDLBERSDATEDTOARELETHESTEASTNORTHEASTORESTHATTHEREATTHETHEAREANDINBERLINCLOCKTHERESANDITISREELETOURJOAREACHALLIONTHEREINTEASTNORTHEASTHEWITHEARECOMENTHEREINTERESANBERLINCLOCKDRESTATINTHEDALECTINGMOWAROUNTRETHEREATEDINTEASTNORTHEASTERESTOREASTEVERESANDINTHETHESBERLINCLOCKTATTHECONTHESTORATANAMEISCOREATTHETHESTHEREAEASTNORTHEASTNDCONTHERESERINTHESTERANDERANBERLINCLOCKDRENTHEREARATERSTHEMARXOFORESTINTEDINTHEARENEASTNORTHEASTTHESTHESTHERENTERSANDINGANDINBERLINCLOCKGSTANDTHERENTERSARETHALRPRESIDENTHESTHERESANEASTNORTHEASTDERSTHEPRESINTHEREARENTHERETHBERLINCLOCKERESIDEDINGTHETHEREARTIHow this experiment sits among the others is on the lineage page; its family's open hypotheses are on the families page.