22:17:13    OCT 25, 1983
     Phoenix              Awit
REAL mins:     154               72
CPU  secs:     1              2851
             EXPERIMENTAL     Amdahl 5860/2  

D4          0B*E 30 ( 1) Ng8-f6      0B*E  
BG5          R E 40 ( 2) Nf6-e4     .19=E 25 
BH4        109=E 23 ( 3) d5         .29=E 41 
F3         .19=E 26 ( 4) Ne4-f6     .19=E 41 
ND2        .89=E282 ( 5) g6         .39=E 44 
E4          24=E 81 ( 6) Nb8-d7     .19=E 88 
E5          19=E181 ( 7) Nf6-h5     .19=E 45 
C4         .79=E202 ( 8) Bh6        .48=E 71 
PXD5        19=0221 ( 9) Nd7-b6     .19=O 25 
G4         .29=0129 (10) Nh5-f4     .19=O107 
G5         .19=0219 (11) Bg7        .19=O 71 
NE4        .19-0280 (12) Nb6xd5     .29=O 73 
BG3        .36=0 8O (13) h6         .21=O 76 
QD2         19-0321 (14) h6xg5      .17=O 54 
NXG5       .19-0321 (15) Nf4-h5     .28=O 63 
NG1-H3     #09 0235 (16) f6         .19=O 48 
PXF6       .19-M205 (17) e7xf6      .19+M 47 
NE4        .49<M194 (18) Qe7        .29=M116 
QE2        #79 M433 (19) Bxh3       .19>M 91 
BXH3        13<M150 (20) Nh5-f4     .19>M 84 
NXF6+      .29<M161 (21) Bxf6        15>M 15 
QXE7       .19<M103 (22) Kxe7        13>M  9 
BF1        .39<M334 (23) Bxd4       .19>M 22 
O-O-O      .19*M 91 (24) Be3+       .18>E 47 
KB1         13<M311 (25) Ra8-d8     .19>M 26 
BD3        122*M255 (26) c6          29>M 37 
RH1-E1      19<M111 (27) Kf6         19>E 41 
BE4         69<M144 (28) g5         .39>E 58 
RH1        #29 M437 (29) Rh8-h3     .29>M 44 
RH1-E1      99*M152 (30) Nf4-h5     .55>E 48 
BXD5        49*E114 (31) Rd8xd5     .15+E 67 
RXD5        19-E158 (32) c6xd5      .19+L 34 
RXE3       .19-E158 (33) Nxg3       .16+L 59 
PXG3       .19<L155 (34) Rxg3       .18+L 53 
RB3        109<L171 (35) g4         .18$L 47 
RXB7       115*L 85 (36) Rxf3       .13$L 76 
RXA7       .29*L121 (37) g3         .28>L 60 
RA6+       .19<L277 (38) Kg5         16>L 40 
RA8        .39*L153 (39) Rf1+       .19+L 63 
KC2         1F<L 28 (40) Rf2+        13+L 86 
KC3         16-L 86 (41) d4+        .39+L122 
KXD4        15-L130 (42) Rxb2       .19+L 80 
KC3        .69-L140 (43) Rb6        .19+L 68 
RD8        .39-L131 (44) Kf5        .39+L124 
RD2         79-L199 (45) Ke4        .19+L 72 
A4         .17-L214 (46) Rc6+       .19+L 61 
KB4         13-L134 (47) Kf3        .19+L 59 
RD3+       .49-L291 (48) Kf4         16+L 42 
RD4+       .78-L168 (49) Ke3         15+L 30 
RG4        .17-L401 (50) Kf2        .19+L 50 
RF4+       .95-L108 (51) Ke2         45+L 96
RF7        127<L197 (52) Rc8         12+L 56
RE7+         R L 32 (53) Kf2         16+L 34
RF7+       .19-L205 (54) Kg1         15+L 37
A5         129<L279 (55) Rh8        .18+L 66
A6         .59-L202 (56) g2         .18+L 55
A7         199*L164 (57) Kh2        .16>L 49
PA8=Q      .19*L189 (58)
At this point Pheonix resigned.  The new Queen is lost and
it is impossible to prevent Black's promotion without loss of
White's remaining piece.  That said, the ending is not easy for a computer.
The ending was played out in January 1984 and the moves appear
at the end of this file.

 Ow ::    ::    ::    Rb
 ::    ::    :: Rw ::
    ::    ::    ::    ::
 ::    ::    ::    ::
    KW    ::    ::    ::
 ::    ::    ::    ::
    ::    ::    :: Pb Kb
 ::    ::    ::    ::
OCT 25, OC2.4 BLACK'S MOVE 58.
NODES 53370 TERMINAL 29011 DYNAMIC 2734 CAPTURE 10319
SALVAGE 5061 DUPLIC 4401 RECLAIM 2217 BRANCH 1124539
DUP_TREE 8327 REC_TREE 7987 BRAN_SCORED 882217

Commentary, 29 November 2009.
The above listing was produced using ReadIris 7.0 on a scanned image (a TIFF
file) from research notebook #31 (2004-58-33) period May 1983 - October 1983
There were many errors in the extracted text file and these were corrected by
hand by comparison with the scanned image (the *.tif file).

From the above listing we assume that EXPERIMENTAL mode means maxtree = 600 
(workspace of 600 positions per move), search width (MAX_WIDTH) of 19 moves
and assumed search depth of up to 7-ply.

In this game Awit thinks it is slightly ahead for the whole game.  Only 3 of White's
moves were not examined, they are moves 16, 19 and 29:
NG1-H3         #09 0235 (16)    
QE2            #79 M433 (19) and 
RH1            #29 M437 (29)
as indicated by the '#' in the statistics.  Pheonix took 235, 433 and 437 second for those
moves!

At move 53 we see a statistic of 'R' as in 
RE7+             R L 32 (53)
suggesting that there may have been a telecommunications failure at that point, necessitating 
a restart by Awit.  Black's move 52 of Rc8 is questionable, since Rg6 seems better.

Other questionable moves by Awit are (30) Nf4-h5, since Bb6 seems better and 
(31) Rd8xd5, since Nxg3 looks more promising.  (41) d4+ is questioned as is (50) Kf2.

Outcome: Awit 1 Ostrich 0 in Round 5 of the 4th World Computer Chess Championship,
New York 22-25 October 1983.

The ending was replayed in January 1984 as follows:
15:25:50  JAN 28. 1984
          Phoenix             Awit
REAL mins:  42                  18
CPU  secs:   0                   623
                 EXPERIMENTAL      Amdahl 586O/2
A8=Q          R L 38 (58) Rxa8       .19>E 24  
RH7+        .19*1086 (59) Kg1         22>L 20 
KC5          89*L 53 (6O) Kf2        .16>L 47 
RF7+        .f9*L 42 (61) Kg3         15>L 22 
RG7+        .19*L 43 (62) Kf2         16>L 31 
RF7+        .19"L 25 (63) Ke2         15>L 21 
RE7+        .29*L162 (64) Kf3         16>L 27 
RF7+        .19<L 97 (65) Kg4         15>L 23 
RG7+        .39<L 85 (66) Kf3         16+L 28 
RF7+         12<L159 (67) Ke3         15+L 20 
RE7+        .19<L 70 (68) Kf4         15>L 20 
RF7+        .19<L127 (69) Ke4         16+L 22 
RE7+        .19<L 17 (70) Kf5         14+L 23 
RG7         .19<L 42 (71) Ra2        .19+L 51 
KD5         .59<L156 (72) Kf4        .18>L 51 
RF7+        .79*L 50 (73) Kg3         14>L 17 
RG7+        .19*L 58 (74) Kf2         26>L 39 
RF7+        .19<L 17 (75) Kg1         15>l 32 
KD4         .29*L139 (76) Ra3        .44>L 44 
KE4          99*L 55 (77) Kh2        .16$L 40 
RF2         .19*L 27 (78) Kg3        .18$L 24
and White resigns, since cannot prevent the promotion without
sacrifcing the Rook.
    ::    ::    ::    ::
 ::    ::    ::    ::
    ::    ::    ::    ::
 ::    ::    ::    ::
    ::    :: KW ::    ::
 Rb    ::    ::    Kb
    ::    ::    Rw Pb ::
 ::    ::    ::    ::
JAN 28, OC2.6 WHITE'S MOVE 79.     
NODES 18639 TERMINAL 12813 DYNAMIC 628 CAPTURE 2204  
SALVAGE 3516 DUPLIC 4040 RECLAIM 832 BRANCH 412959  
DUP_TREE 7044 REC_TREE 7925 BRAN_SCORED 209864

Right from the beginning 
RH7+        .19*1086 (59) White
recognizes the magnitude of the problem and spends 1086 seconds
on its first move.  Still Awit does not know how to proceed,
                        (6O) Kf2        .16>L 47 
no human would waste time
and would recognize the need to create a block with Rf8.
Still the actual play shows that the draw by repetition
mechanism was working.

There were only two changes between version OC2.4 and OC2.6
but have no idea what they were (they would be in 2004-58-33).