LETTER TO WILLIAM CASEY FROM GLENN DIAMOND

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP88G00186R000700850018-7
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 11
• Document Creation Date: December 22, 2016
• Document Release Date: June 4, 2010
• Sequence Number: 18
• Publication Date: November 1, 1985
• Content Type: LETTER

Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88GO0186R000700850018-7
TRANSMITTAL SLIP
TO:
ROOM NO.
BUILDING
REMARKS:
EXO/DDA
DDA REGISTRY
0Q b~o e~ d)
D/OIT RECEIVED A COPY.
~NfO Copy To :
D/0/7 ps'
FROM:
ROOM NO.
BUILDING
EXTENSION
1 FEB 56 24 1 WHICH MAY BE USED.
Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88GO0186R000700850018-7
 Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88GO0186R000700850018-7
EXECUTIVE SECRETARIAT
ROUTING SLIP
ACTION
INFO
DATE
INITIAL
1
DCI
X
DDCI
3
EXDIR
X
4
D/ICS
5
DDI
6
DA
X
7
DDO
8
DDS&T
9
Chm/NIC
10
GC
11
IG
12
Compt
13
D/OLL
14
D/PAO
X
15
VC/NIC
16
OIT
X
[
20
21
22
Remarks
TO #14: Please prepare an appropriate
response.
E, cutive Secretary
Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88GO0186R000700850018-7
 Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88G00186R000700850018-7
4167
f ~ea~l`1 VCc,~~
A ~czv- Gso T r -le-1
V\e 4
4c1ecLeg
R
CORPORATION
SS-37,y
Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88G00186R000700850018-7
0
 Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88GO0186R000700850018-7
il< OR
CORPORATION
~~ .-- 5421 OPPORTUNITY COURT
MINNETONKA, MN 55343
[phone]
Co
~kV
t-A
ce FVVI
CQCam'
"CR
J1,ce ve(
XJr,->
CQ^r.,~.~ ~y ~n cal ~ ~ ~~~ ~ ~-~.~,
~.~o t' L ~IGMoh~
~~ o fps
STAT
Cc
Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88GO0186R000700850018-7
 Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88GO0186R000700850018-7
X1:1:\'1' ~.II,M
AGENT 2.0 is a test of your
analytical skills that challenges
ability. You have the
information. Is the crisis real or
imagined? Is the spy ours or
theirs? Is the file fact or fiction?
You must decide the truth of
Operation Kaleidoscope. Agent
2.0 is a counterespionage
and PC DOS tutorial all
combined into first class
software. Created by Michael
de St. Hippolyte.
Requires IBM, PC with DOS 2.0,
128K memory, color/graphics
adapter, and color monitor.
+1 Corporation. I
Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88GO0186R000700850018-7
 Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88GO0186R000700850018-7
wr %I 3mm
Guide to
,uters
? I 1 II
Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88GO0186R000700850018-7 %
 Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88G00186R000700850018-7
LEADING EDGE WITH NFL CHALLENGE SIMULATION
PC WORLD COVER: "JOE MONTANA PLAYS NFL CHALLENGE"
New York, New York, October 28, 1985 -- The National Football
League has joined the personal computer world with the
introduction of its officially-licensed simulation of NFL
football, NFL Challenge, which graces the cover of the November
issue of PC World, the IBM PC world's largest publication.
PC World's cover story, "Joe Montana Plays NFL Challenge", gives
NFL Challenge high praise indeed. Says John Bello, executive vice
-president *of NFL Properties, the licensing arm of the NFL, "We
have been watching the steady growth and rising prominence of
personal computers in our society and felt that the time had come
for the NFL to get involved. And we're really pleased with NFL
.Challenge."
David Bunnell, Publisher of PC World, says, "When IBM got involved
in personal computing, it was big news, but it was expected. The
'.NFL getting into personal computing is unexpected, and it is very
exciting. It-allows PC World to put NFL Challenge and Joe Montana
on the cover."
The November issue of PC World is available on newsstands
nationwide. In-his conclusion, PC World associate editor Eric
Brown writes, "... NFL Challenge is great fun. It's been along
time since a program has consistently kept me up until 3 in the
morning... Sophisticated simulation games such as NFL Challenge
might provide the spark that's been missing from the world of .
PC's."
NFL Challenge is an extremely sophisticated simulation of NFL
football that uses all of the features'of the IBM Personal
Computer and its pure compatibles. It has many sophisticated
features and features state-of-the-art animated grpahics of actual
football plays.
In the feature article, 49ers star quarterback Joe Montana plays a
quarter of NFL Challenge against Brown in a replay of Super Bowl
XIX. Montana, coaching the 49ers, takes the ball down the field to
score on his first drive, highlighted by a big fourth-and-one
14-yard touchdown pass to Dwight Clark. The quarter ended with the
-49ers leading 13-3.
Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88G00186R000700850018-7
T~1
 Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88G00186R000700850018-7
a bead on his on-screen doppelganger. "Oh, man. What's he throwing
to the weak side for?."'Pass incomplete."
Montana's reaction to NFL Challenge? "I think it's great," he was
quoted as saying. "I especially like the idea that you can use the
Undo key to see what would have happened if you had called another
play. You don't get that luxury on the field."
NFL Challenge was developed by Minnesota-based Xor Corporation
under a license from NFL Properties. Says Buddy Diamond, Xor's
founder, "Our crack team of Harvard and Stanford wiz kids was told
to push back the outside of the envelope as far as the IBM PC is
concerned, and they did it. And our package, created by NFL
Properties' Creative Services Division, which includes the
highly informative NFL Illustrated Playbook, is the best in our
industry."
The product is available in hundreds of retail stores nationwide,
including B. Dalton Software Etc., Wherehouse Entertainment,
ComputerCraft, Schaak Electronics, and Lechmere Sales. It costs
$99.95. Said John Dvorak, noted computer columnist for the San
Francisco Examiner, "If there's a computer game worth $99.95, this
is it, believe me."
CORPORATION
Brown describes Montana. reacting to the animated play graphics of
one play: 'I need help," he said as he saw the eight rushers draw
-IN
Brown's comment in his article sums it up: "NFL Challenge is more
'
than a game; it
s a close simulation that can teach you a lot
about football." For the millions of IBM personal computer users
interested in the NFL, NFL Challenge is a dream come true.
GED
Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88G00186R000700850018-7
 Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88GO0186R000700850018-7
7
3 - 4
2 - 7
23:59
Uses full-strength end-19b4 rosters so bestuse of
results is in evaluating season-to-season change.
ool To edit to 1985 rosters-for. customized simulation,`
NFL Ch 1
l
see a
enge User s Guide pages 17,.18 hand 36.
NFL Challenge Simulation Version. Vi. 03!X?e
Bears won 45.0 of 50.0 games-played
C~~Nft'jc
Average Results from 50 games. played
Vikings won 5.0 of 50.0. games played
Team Bears Vikings
First Downs
Rn - Ps -- Pn
3rd Dn Convs
Time of poss
Total Net Yds
Plays - Avg
Net Yds Rush
R Plays - Avg
Net Yds Pass
At - Cp - I nt
Average Pass
Sacks - Yds
Punts - Avg
Return Yds
Penalty - Yds
Fumble - Lost
23 14
11 - 11 - 1 4 - 9
6 - 13 - 46.2% 4 - 12
35:16 24:50
356
6489 5.6
1
40 - .7
167 4
23 - 13 -
7.0
1 - 5
4 - 39.5
119
7
2 - 1
Results from
Bears
First Downs
Rn - Ps - Pn
3rd Dn Convs
Time of pass
Total Net Yds
Plays - Avg
Net Yds Rush
R Plays - Avg
Net Yds Pass,
At - Cp - Int
Average Pass
Sacks - Yds
213
- 4. 1
52 79
22 - 3.6
134
26 - 13 -
4.5
4 - 31
6 - 40.8
127
4 - 34
1 - 1
50 games played
Vikings
0
0
28.6%
1
33.3%
3
1 - 2 - 0
0 8 - 0. 0%
17:50
159
82 3 8
77 3 0
15 - 7 - 0
5.1
0 - 0
91
28 - 3.3
30
11 - 2.7
61
17 - 7 - 0
3.6
0 - 0
Punts - Avg 0 - 0.0 1 - 44.0
Return Yds 29 37
Penalty - Yds 1 - 10 1 - 10
Fumble - Lost 0 - 0 0 - 0
Maximum Results from 50 games played
First Downs
Rn - Ps - Pn
3rd Dn Convs
Time of pons
Team Bears
Points 55
Total Net Yds
Plays - Avg
Net Yds Rush
R Plays - Avg
Net Yds Pass
At - Cp - Int
Average Pass
Sacks - Yds
Punts - Avg
Return Yds
Penalty - Yds
Fumble - Lost
41
18 - 19 - 4
13 - 19 - 68.4%
42:10.
639
92 - 6.9
302
57 -
337 5' 3
32 - 20 -, 3
9.6
3 - 18
258 44.3
111--493
6
Vikings
27
32
12 - 17 - 3
9 - 19 - 47.4%
36:01
402
88 - 4. 6
142
35 - 4. 1
260
44 - 23 - 3
4.9
9 - 68
92-542. 6
8 - 74
4 - 4
Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88GO0186R000700850018-7
 Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88GO0186R000700850018-7
C~_ Mme.
iWelli ce%r,<
A Microcomputer Program for Calibrating Intelligence Analysts
by Glenn E. Diamond, Xor Corporation
"It is likely that most analysts most of the time, are not even aware
of the background level of uncertainty because their job is to bring
reason out of the chaos, not to study the chaos itself."
A. Generally.
-- Fredric S. Feer
Analytical Assessments Corporation
Colloquium on Analysis and Estimates '(1979)
Consortium for the Study of Intelligence
In the intelligence production system, the Analyst could be said to be
a knowledge worker who fashions information from data for action by the
policymakers and decisionmakers of national government.
Analysts, ust like fine machine tools, must be calibrated constantly
to assure that their outputs are within acceptable tolerances demanded
by the end users of the product. Surely, a normal.result of all human
analysis is partial error. To paraphrase Bishop Berkeley: To be is to
be deceived.
In each analysis, the analyst uses critical judgments, assumptions, and
ogic upon the data to create information in
l usable form. Rarely does
the analyst have the time, inclination, or bureaucratic imperative to
do these two things: a) Seek and incorporate into analysis raw
intelligence that has been filtered out of collection output due to it
being9 false erroneous, or based on deception as determined by
counterintelligence, or b) Systematically analyze the critical
judgments assumptions, and logic underlying the analysis from a
counterintelligence perspective.
B. Calibration Important.
An analyst out of calibration is susceptible to preconception,
self-deception, and external deception. Analysts are the target of
strategic deception; we know that. Deception is expanding exponentially
now we know that. Yet we depend on intelligence analyses and estimates
in formulating our national policy.
R properly calibrated analyst will "turn over as many rocks" as
necessary to produce good intelligence product. This is an art; if too
many "rocks" are turned over, the product has decayed past its
half-life, if too few, the product is more opinion than information as
it was fermented from tunnel-vision or myopia or preconception or
premature cognitive commitment (Harvard's E. J. Langer's term).
Calibration of the analyst is an example of integrating
counterintelligence into all aspects of the inteligence production
system. In fact, the modern analyst must have an "on-board
counterintelligence module".
In a state-of-the-art intelligence production system the
counterintelligence staff analyzes the analysts who have already
analyzed themselves. All collected data should pass to analysis
unfiltered, but counterintelligence-rated.
C.The Simulator.
A microcomputer-based (hence portable) program can be developed which,
though abstract in nature can enable an analyst or the analyst's
superiors or counterintelligence to check analyst calibration.
It will present to the analyst a "test pattern" over which the analyst
can lay a specific analysis for calibration:
I. Identifies data universe of analysis.
2. Uses "n-1, n+1" rule to reach contradiction point.
3. Identifies critical judgments and assumptions.
%_ If
X11 (~J
Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88GO0186R000700850018-7
 Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88G00186R000700850018-7
Page Two
4. Uses "disconfirmed disconfirmation" to check logic.
5. Builds "matrix of analyses" off of analysis.
6. Analyst analyzes analysis of analysis with manager.
The result of this activity can produce useful information for the
analyst, intelligence management, and counterintelligence to create
multi-leveled all-source deception insurance. That is what must be
done.
For instance, step one assures that an accejtable universe of data was
analyzed. That implies data of all "vectors and "scalars", including
negative data, uncertain data, false data, deception data, etcetera.
Step two alters that universe in many ways to ascertain at what point
the analysis breaks down or contradicts itself. This rates the
calibration of the analyst in his data acquisition function, and
especially pointing out anchoring biases.
Step three is an important Process as the analyst analyzes the
analysis; dissects it for the additional tests. The quality of this
dissection is an important calibration check too, since if the checking
is sub-standard the analysis will tend to head in that direction.
Step four is the falsification test. An analysis must be falsifiable to
be usable as finished intelligence product. Every disconfirming element
must be reckoned and either confirmed or disconfirmed. Which leads to
step five, wherein all the various possible analyses branchin from the
main analysis are elaborated. Of course, absolute disconfirmation (the
disconfirming of all disconfirmations) is an "A" One thing the analyst
will learn from this program is that A's are seldom given.
In step six, with other staff, the analyst grades the analysis and
re-calibrates. A record should be kept to aid in future evaluations of
the analyst's analyses and estimates and to assist the analyst from
time to time.
The major problem in analyst training and evaluation is the problem of
defining the analyst itself. This program ends up there. The modern
analyst in the modern intelligence production system probably needs to
be evaluated not in academic terms, but as the hybridization of
collector, counterintelligence, and collator. The shoe-box gives way to
Arthur C. Clarke's "monolith". And objective and subjective observation
of the analyst-in-calibration would be high-quality grist for the
counterintelligence mill.
Though Don Quixote said "facts are the great enemy of truth" it is
important to state that when in a near-infinite potential data
environment, the analysis supercedes the data as the fundamental
weakness. A finite set of data analyzed infinitely will yield more
"fungible truth" than an infinite set of data analyzed finitely.
The Analyst Function can be refined and refined. If we analyze it.
The End.
Glenn E. Diamond is the founder of Xor Corporation, an IBM PC software
development house in Minnetonka Minnesota that specializes in advanced
simulation work. Its first simulation for the consumer market the
popular "NFL Challenge",'officially-licensed by the NFL is tie cover
story topic in the November issue of PC World: "Joe Montana plays NFL
Challenge". Mr. Diamond is a recognized technological thinker who has
made contributions in the fields of third world industrial development
("Export Development Functionality with an Import Substitution Form
Factor for an Optimal Industrial Development Velocity per Increment of
Domestic Resource Costs Employed", AAPRD, 1981) and conventional
deterrence ("Robotic Armored Infantry Command for Ultimate Conventional
Deterrence Along the NATO Central Army Group Front", D&FA-ISSA, 1983).
Sanitized Copy Approved for Release 2010/06/07: CIA-RDP88G00186R000700850018-7