The Scholastic Aptitude Test (SAT) is dreaded by high school juniors and seniors
because a single SATurday morning can ruin the rest of their lives!

This article describes my system for getting extraordinary raw scores on this
challenging test.  Rather than fight the test, I use the SAT's difficulty to my
advantage, leveraging down to a new, elite level of distinction:

Answering all questions incorrectly!

On 2003 April 5th, a Saturday, at the age of 33, I threw away my dignity, mocked my Ivy League education, disgraced my Master's degree, and proved, in just over three hours, that humans can do things "The System" didn't anticipate. Things didn't turn out exactly as planned, but it was a crazy experience!


[ 1] Introduction [ 2] Table of Contents [ 3] The Scholastic Aptitude Test (SAT) [ 4] Calculating the Test Score [ 5] Registering for the Test [ 6] Test Preparation [ 7] Test Day! [ 8] My Test-Taking Experience [ 9] My Test Scores [10] Conclusion [11] Disclaimer [12] Reader Feedback [13] Contact Information


The Scholastic Aptitude Test (SAT) is a test of mathematical and verbal
intelligence.  This test was created by the College Entrance Examination Board
(CEEB), and was administered for the first time in June 1926 to 8,040 people.
In the 2000-2001 test year, more than 2.1 million people took the "SAT I:
Reasoning Test" as part of the admission process for more than 1,000 
institutions of higher education, 80 percent of the national (U.S.) total.

All references to "SAT" in this article refer to the "SAT I: Reasoning Test".

What does the SAT stand for? (From College Board literature) "Originally, SAT was an acronym for the Scholastic Aptitude Test. In 1993, the test was renamed the SAT I: Reasoning Test. At the same time, the former Achievement Tests were renamed the SAT II: Subject Tests. SAT has become a simple way of referring to the SAT I: Reasoning Test."

The College Board has a page describing the fascinating history of the SAT: The SAT is given seven times a year at thousands of testing centers throughout the world. Unless there are special circumstances, the test is given on a Saturday, starting at approximately 9:00 A.M. in each time zone.

Cheating on the SAT! In the United States there is a three-hour time zone difference between the East Coast and the West Coast. At one time this created a SAT cheating opportunity: Professional test-takers on the East Coast would steal or memorize test materials, and relay this information to student test-takers on the West Coast. One business apparently flew students to the West Coast to take advantage of this possibility. These days, with Infra-Red ports on calculators and wristwatches, hands-free earpieces and digital video on mobile phones, cheap wireless cameras that can be planted anywhere in a room or on a person, I'm sure there must be someone out there who had a whole support team helping him through the test! Wouldn't it be hilarious to have an Internet chat room or instant messaging group helping one through the SAT! The transcript would be damaging evidence, though. Undoubtedly other people have simply exited the test room with the test booklet and answer sheet, worked on the test with a group of smart friends and a dictionary, and returned discreetly, perhaps with some sort of distraction. It would be ironic for cheaters to get admitted to Harvard and one day learn the school motto: "Veritas" (Truth).

The testing part of the SAT lasts a total of exactly three hours. There are seven timed sections:

(1) 30-minute Verbal (2) 30-minute Verbal (3) 30-minute Math (4) 30-minute Math (5) 30-minute Equating Section (Verbal or Math) (6) 15-minute Verbal (7) 15-minute Math

The 30-minute sections can appear in any order. Test proctors need to indicate when to start and stop for all sections, so it is necessary for all test-takers in a testing room to encounter each of the two 15-minute sections at the same time. For my test group, both 15-minute sections occurred in the final 30 minutes of the three-hour testing period. The following table describes various properties of the sections:

Verbal Sections: ================ * 75 minutes total (30 min + 30 min + 15 min) * 78 questions total: 40 Critical Reading (5-choice) 19 Analogies (5-choice) 19 Sentence Completion (5-choice) Math Sections: ============== * 75 minutes total (30 min + 30 min + 15 min) * 60 questions total: 35 Standard (5-choice) 15 Quantitative Comparisons (4-choice) 10 Student-produced responses (NO-choice) Or, by question type: 44 Arithmetic and Algebraic Reasoning 14 Geometric Reasoning Equating Section: ================= * 30 minutes total * Used to test new questions, and compare editions of the exam * Does NOT count toward final score * Is NOT explicitly identified to test-takers, and looks exactly like an ordinary Verbal or Math section; thus, taken with the same seriousness as "real" sections


Two Divisions: Math and Verbal

The SAT has two divisions: Math and Verbal.  A score report has performance
statistics for the math division, and performance statistics for the verbal

Three Statistics for each Division

Performance on each division of the SAT is described by three statistics:

(1) Raw Score; (2) Scaled Score; (3) Relative Percentage;

Almost all statements regarding "SAT Score" are implicitly referring to the Scaled Score.

(1) Raw Score

Raw Score is an integer (whole number) that is calculated for each
division of the test.

Applying the following rules to the set of answers for a division of a test
will yield a number (potentially non-integer) that will be the basis of the
Raw Score:

(a) Ignore each blank answer (often called an "omitted question"); (b) Ignore each incorrect answer for math questions requiring "student-produced" (non-multiple-choice) responses; (c) Add one point for each correct answer; (d) Subtract 1/4-point for each incorrect answer for multiple-choice questions having 5 choices; (e) Subtract 1/3-point for each incorrect answer for multiple-choice questions having 4 choices;

The Raw Score is the number computed directly above rounded to the "nearest integer".

Mathematical Detour: All Possible Fractional Values The calculation of the integer Raw Score involves rounding a potentially non-integer value to the "nearest integer". It so happens there are twelve (12) cases of the fractional part: FRACTIONAL PART EXAMPLE ROUNDING =================== =============== ============ 0/12 = 0.0 (0-(0/4)-(0/3)) None 1/12 = 0.083333... (1-(1/4)-(2/3)) Down 2/12 = 0.166666... (1-(2/4)-(1/3)) Down 3/12 = 0.25 (1-(3/4)-(0/3)) Down 4/12 = 0.333333... (1-(0/4)-(2/3)) Down 5/12 = 0.416666... (1-(1/4)-(1/3)) Down 6/12 = 0.5 (1-(2/4)-(0/3)) Up <-- Look! 7/12 = 0.583333... (2-(3/4)-(2/3)) Up 8/12 = 0.666666... (1-(0/4)-(1/3)) Up 9/12 = 0.75 (1-(1/4)-(0/3)) Up 10/12 = 0.833333... (2-(2/4)-(2/3)) Up 11/12 = 0.916666... (2-(3/4)-(1/3)) Up If J is an integer, (J + 0.5) is equidistant from two integers: J and (J+1). Obviously the CEEB needs to specify what to do in this special case -- and it must be deterministic (so random rounding and alternate rounding cannot be used). One candidate convention, called "Banker's Rounding", rounds to the nearest EVEN number (e.g., 0.5 --> 0; 1.5 --> 2; -11.5 --> -12). Another convention is to round to the nearest integer, and round UPWARD for the ambiguous case of (J + 0.5). (e.g., 0.5 --> 1; 1.5 --> 2; -11.5 --> -11) This is what the "FIX" function does for 0.5 on scientific calculators ({2nd,FIX,0,0.5,=} --> 1). This is the convention used by the CEEB, as described by this quote: "If the resulting [raw] score is a fraction, it is rounded to the nearest whole number -- 1/2 or more is rounded up; less than 1/2 is rounded down." Keep in mind that rounding up for 1/2 is not the definition of "nearest whole number"; this quote is merely describing the convention used by the CEEB.

Verbal Division Raw Score Range The Verbal Division has three question categories: ======================================================================= Question Type Questions Min Raw Score Max Raw Score ======================================================================= Critical Reading : 40 : -40/4 = -10 --> -10 : +40 Analogies : 19 : -19/4 = -4.75 --> -5 : +19 Sentence Completion : 19 : -19/4 = -4.75 --> -5 : +19 ======================================================================= (NOTE: All questions are five-choice types.) These three raw scores are reported on the printed score report. Note that (-10-5-5) = -20, whereas (-78/4) = -19.5 --> -19. Which method is used to compute the raw score for the verbal division? Who knows! But it could make a significant difference in scaled score. Math Division Raw Score Range The Math Division has two question types: ======================================================================= Question Type Questions Min Raw Score Max Raw Score ======================================================================= Arithmetic and Algebraic Reasoning : 44 : -10 : +44 Geometric Reasoning : 14 : < -1 : +14 ======================================================================= These two raw scores are reported on the printed score report. I am not sure how the 35 five-choice, 15 four-choice, and 10 student-response questions are distributed between these two categories, so I cannot show the calculations. Again, I am not sure if these two raw score values are added to yield an overall division raw score, or if all questions for the entire math division are considered together as in (0-35*(1/4)-15*(1/3)-10*0) = -13.75 --> -14 for computing the scaled score.

(2) Scaled Score

The Scaled Score is intended to be a measurement of ability on an absolute
scale.  In principle, the Scaled Score enables the objective comparison of
test-taker abilities throughout the history and future of the SAT, regardless
of variations in the difficulty of each edition of the test.

A Raw Score is converted to a Scaled Score through a statistical process called
"equating".  Equating may involve correlating performances between different
questions, and to the same questions or similar questions on past editions of
the SAT.

If the equating process works, and Scaled Scores are an absolute measure of
ability, then Scaled Scores for a particular edition of the SAT do not depend
on the relative performances of test-takers for that edition.

The Scaled Score is a multiple of 10 in the range 200 to 800.

FIGURE: An example conversion of Raw Score to Scaled Score
        (based on example data published by the College Board).

According to these example curves, one actually needs to get at least
seven questions wrong (approximately) in each division to have a chance at the
lowest possible scaled scores.

Is it true that you get a 200 on the SAT just for signing your name? (From College Board literature) "Theoretically speaking, if you just sign your name and don't complete the answer sheet, you would get a score of 200. That's because we don't report scores that are lower than 200. In reality, if we received an answer sheet that wasn't filled out, it would be considered an automatic request to cancel scores and no scores would be reported."

I believe that the first sentence in the quote above is not correct, because it is possible for the conversion curves to yield scaled scores higher than 200 on each division for raw scores of zero (as would result from leaving all answer bubbles blank). Although I think the quote is ambiguous, I take it to mean that leaving all answer bubbles blank on the entire answer sheet (not simply all math or all verbal answers) would result in scores not being reported.

(3) Relative Percentage

The Relative Percentage is an integer between 1 and 99 that indicates,
for each test-taker, the percentage of test-takers who earned a lower score.

The percentages are based on: "the most recent scores earned by current-year
college-bound seniors who took the test at any time during high school."

Relative Percentages are given for the nation and state.

Relative Percentage (Number Of Test-Takers with a Lower Score) Fraction = ------------------------------------------ (Total Number of Test-Takers) Relative Percentage = Max( 1, Min( 99, Round(100*Fraction) )) Lowest Percentile People who get the lowest possible scaled score should get a relative percentage of "0", since it is impossible for any other student to have a "lower score". But "1" is the lower limit on the CEEB's relative percentage value. Highest Percentile People who get the highest possible score could, in principle, have a relative percentage anywhere in the range 0 (nobody did worse), to 99.999... (almost everyone did worse). It cannot reach exactly 100.0 percent, given the definition, but rounding could round up to 100. However, the CEEB limits are 1 percent to 99 percent.

Math Score + Verbal Score = Combined Score

Adding the Scaled Score for the math division to the Scaled Score for the
verbal division yields a combined score.

Because each of the two Scaled Scores is a multiple of 10 in the range 200
to 800, the combined score is a multiple of 10 in the range 400 to 1600.

If a person fills out an SAT answer sheet with all required personal and site-
specific information, but doesn't answer a single question (i.e., "omits" all
questions), the person receives a combined score that can be as low as 400,
but is likely to be significantly higher (e.g., 470) based on the raw score to
scaled score conversion curves for that test edition.

It seems inevitable that every person who takes the SAT will contemplate what
it must take to "get a sixteen-hundred" (1600).  For some, it might be the only
chance to get admitted to an elite school.  For most, it's an urban legend.

In 2002, the average scores on the SAT were 504 verbal and 516 math,
a cumulative score of 1020.

FIGURE: The World Almanac 2003; p.240; Table: SAT Mean Verbal and Math Scores
        of College-Bound Seniors, 1975-2002

Some universities specify an "average SAT combined score" in their admission
application materials.  It may not be legal for public colleges and universities
to establish minimum SAT combined score requirements, but private institutions,
like corporations, can set whatever standards they wish.

Score "Recentering": An Attempt to Establish
Standard Scores

Since the SAT was first administered in 1926, there have been over 20 different
sets of scales used to convert raw scores in to scaled scores, all in an effort
to better describe the abilities of people who take the test.

The most dramatic change in score calculation for the SAT in recent decades is
described by the following quote from the College Board:

    "In April 1995, the College Board recentered the score scales
    for all tests in the SAT Program to reflect the contemporary
    test-taking population. Recentering reestablished the average
    score for a study group of 1990 seniors at about 500 -- the 
    midpoint of the 200-to-800 scale -- allowing students, schools,
    and colleges to more easily interpret their scores in relation
    to those of a similar group of college-bound seniors."

One effect of recentering is that scores are apparently "inflated" relative to
scores before 1995.  For example, what would have been a 1490 cumulative (V+M)
score before 1995 is now called a 1600, the maximum possible score.  Put 
another way, a score of 1600 today could correspond to a score as low as 1490
for scales prior to 1995.  All scores, down to the lowest, are inflated
relative to historical counterparts.

The following page has a table that shows the effect of recentering on 
cumulative (V+M) scores:

The following is a fascinating research paper published by the College Board on
the theory and motivation for the recentering process:

sat_recentering.pdf [380 KB]

If you read the research paper mentioned above, one gets a sense of the 
difficulties faced by the College Board in establishing a standard.  One can
find scandal in the paper if one is looking for it, and it is easy to see how
this complicated problem and associated solution could be regarded as more of
an art than a science.

High Intelligence Quotient (IQ) Organizations

Many high-intelligence organizations believe that there is enough correlation between pre-1995 SAT scores and IQ to use those scores alone as a qualification standard for membership. For example, an SAT combined score of 1250 (1974-1994 SAT editions) correlates with a Stanford-Binet IQ of 132, the top 2% of humanity, and thus qualifies a person for Mensa.

In 1994, the SAT exam was greatly modified: antonym questions were eliminated, the verbal section contained a greater emphasis on reading, non-multiple-choice questions were added to the math division, and it became permissible to use electronic caculators for the math sections. In 1995, CEEB started computing and reporting SAT scores according to the "recentering" scale, as described in the previous section. Because of score inflation, and total lack of ability to differentiate performance beyond a moderately high level, SAT scores could no longer be correlated to IQ with sufficient range or precision to serve as a basis for admission to high-intelligence organizations.

Almost no high-intelligence organization accepts any post-1994 SAT combined score as a basis for admission. However, the Graduate Record Exam (GRE) general test is currently accepted by many high- intelligence organizations for admission. Tests for the general population are designed to magnify differences centered around average intelligence, so only a small fraction of the questions can be allocated to differentiate between high levels of intelligence. Thus, specialized IQ tests are required to accurately assess high intelligence. The following are links to Mensa sites: American Mensa : Mensa International : According to Mensa literature, "Mensa meetings are anything but dull!" Perhaps you will meet Marilyn Vos Savant, who is listed in the Guinness Hall of Fame for the world's highest recorded IQ (228). For information about organizations with truly staggering IQ requirements, follow these links: Mensa is at the LOW end of the spectrum. The Giga Society only accepts people in the 99.9999999 percentile of human intelligence ( 1 / Billion, thus "Giga"), which corresponds to an IQ of 196. Using the CIA's "The World Factbook 2002", and extrapolating to today's date (2003/4/28), the world population is roughly 6,297,263,466. For a Binomial probability distribution, the mean is N*p, and the variance is N*p*(1-p). Thus, one would expect the Giga Society to have 6.3 members, give or take 6 members. It now has at least 5 members. I look at various IQ tests from time to time, and I am usually unimpressed with the "intelligence" being tested. Here are sample verbal questions from one "unofficial" IQ test: [1] 9 : 361 :: Tic-Tac-Toe : ____ [2] Roosevelt : New :: Truman : ____ [3] 5,280 : Mile :: 43,560 : ____ [4] 753 : 776 :: Rome : ____ [5] Horse and Donkey : Mule :: Question mark and Exclamation point : ____ Just for fun I decided to measure Google's "IQ"! Acting as Google's proxy for the IQ test, I found the following plausible answers: [1] "Go" : A Go board has 19 x 19 = 361 places to move, and Tic-Tac-Toe has 9 places to move; [2] "Fair" : Truman had a "FAIR Deal", and Roosevelt had a "NEW Deal"; [3] "Acre" : There are 43,560 square feet in an acre, and there are 5,280 feet in a mile; [4] "Olympics" : The Olympic games were founded in 776 BC, and Rome was founded in 753 BC. [5] "Interrobang" : An "interrobang" is a character that is a composite of the question mark and exclamation mark. How silly is that?! A mule is a "composite" of a horse and a donkey. I seriously considered completing an IQ test using only information yielded via Google searching (i.e., ignoring my own knowledge), and paying for it to be scored, and then publicize Google's "IQ". Of course, any online article describing questions missed by Google-brain will eventually be assimilated by Google. It's interesting to think of static information as a given. Suppose all of humanity were wired directly in to Google. Suddenly, the novelty of trivia and languages vanishes. Just as "human calculators" were made obsolete by mechanical calculators, future advances in technology may make other traditionally valued forms of intelligence obsolete. I happen to believe that the future belongs to video games and infotainment. If there weren't so many cheaters and paid "farmers" around, I'd say that one's "level" in EverQuest (EQ) could be correlated to this new prioritized variety of "intelligence". ... Ha! Had you going there! Or am I just creating plausible deniability for a controversial idea? Huh[insert interrobang here]


The SAT is given seven times each year.  Although you must register (enroll) 
with the College Board before a specific deadline to be assured the opportunity
to take the SAT on a specific date, there is a chance that you will be able to
simply show up at a designated test center on the morning of a test and sign up
to take the SAT.  There is no guarantee that the site will have spare test 
booklets and answer sheets, so it's a bit of a gamble.

Here are several fees that may interest a person planning to take the SAT I:

(1) "SAT I: Reasoning Test" : $26 ============================== This is payment for the test. Included in this price is the option to have up to four official score reports mailed to college and university admission programs of one's choice. (There is a charge for each additional score report if more than four are requested. Score reports are only official if mailed directly from the College Board.) (2) "Student Answer Service" : $ 6 ============================== This is an optional report that shows you an enumerated list of questions, and whether your response was: correct (+), wrong (-), or omitted (O). Maybe this product exists to address possible paranoia about the Raw Score calculation. (3) Detailed Answer Report : $15 ============================== This product is only offered on certain test dates. This report shows the actual questions on the exam, your responses, the correct responses, and an analysis of your apparent strengths and weaknesses, which, in principle, could be used as a basis for deciding how to prepare for taking the test again.

There are many other interesting products, like "Rush Reporting" to get your official SAT scores mailed to colleges and universities sooner, "Additional Score Reports" when you are applying to more than four institutions, and a fee for a personal online peek at your scores a full ten days before they are put in the mail, etc. Perhaps the College Board will eventually offer more exotic products:

(1) T-Shirts, bumper stickers, vanity plates, baseball caps, coffee mugs, greeting cards, playing cards, paper weights, bobble heads, flags, and web sites, all depicting your score report in a handsome way that is sure to grab attention. (2) A score report tattooed on your ass or forehead, or burned in to your retina with a laser for a lifelong visual affirmation of your intelligence. (3) Score reports containing "printing errors". The College Board won't come right out and say what the nature of those errors might be, but will hint that such errors are often favorable, in logarithmic relation to your "donation level". (5) Professional photographers will capture one's test-taking experience on four hours of digital video, followed by a helicopter ride to the official "SAT Award Ceremony" at the College Board compound on a remote tropical island paradise. After receiving a trophy, and shaking hands with some of the educational researchers who devised the latest batch of SAT questions, you will tour their secret underground facility, witness giant machines grading thousands of exams per second, and, after a few drinks, you become an honorary employee of the College Board, and have the authority to grade some poor soul's SAT by hand! Of course this situation must be "contained", which means, umm, total erasure of all those involved, nuclear vaporization of all test centers, and installing Xa'Kel-6 as provisional leader of the Earth people.

Online registration for any College Board exam starts at their web site: After navigating a few menus, I started the registration process for the "SAT I: Reasoning Test". FIGURE: Starting the SAT registration process. The first phase is "signing up" with the College Board, which just involves sharing basic information about oneself. When one has finished with this process, the following confirmation appears: FIGURE: "Sign Up" Confirmation As you can see, I have selected the nickname "Tensai", which in Japanese translates to both "Genius" and "Natural Disaster". This is perfect for my test-taking persona. I recently noticed that "Tensai" is an anagram of "SAT ein(s)", or "SAT 1" in German. Coincidence? Or destiny! In the following you can see that I elected to order the "Student Answer Service": FIGURE: Ordering the "Student Answer Service". The registration process includes many questions about one's high school career, and it's hard to resist thinking that the College Board secretly uses this information as part of their "equating" process, but that's how my paranoid, cynical mind works. As you can see, I decided that I wanted to hear from colleges, universities, and government scholarship programs that are interested in "students like [me]". My persona has a terrible cumulative average and class rank: FIGURE: Some questions about my "high school career". FIGURE: That's what they call "making the grade!"

True Story: I received an award for "perfect attendance" when I was in high school!

The questionnaire seems to go on forever. Eventually I encountered questions about my club participation, awards, etc, but the highlight was the list of sports: FIGURE: I like sports that involve weapons and fighting. I selected sports having some connection to medieval warfare! Perhaps, a century from now, they will have sports like: "U.N. Dodging", "Smart Bombing", and "Nation Building". Sometimes my attempts at exquisite contradiction were detected. FIGURE: Sometimes my attempts at exquisite contradiction were detected. After answering six pages of "sign-up" questions, and answering ten pages of a "personal information" questionnaire, and filling out credit card information, I finally had the magic ticket: FIGURE: SAT I Admission Ticket (some parts highlighted and censored) As you can see, I elected to send my score report to four significant universities:

(1) University of California, Irvine (UCI) (Middle 50% of 17,723 Freshman SAT: V 520-610 M 550-670) Where I earned my M.S. in Physics; Go, Anteaters! (2) University of Pennsylvania (Middle 50% of 9,730 Freshman SAT: V 640-730 M 670-760) Where I earned my B.A. in Physics; Go, Quakers! (3) Massachusetts Institute of Technology (M.I.T.) (Middle 50% of 4,213 Freshman SAT: V 680-760 M 730-800) A school for late-night cleaning crew janitors who know combinatorics, and auto mechanics who love talk-radio; (4) Harvard University (Middle 50% of 6,637 Freshman SAT: V 700-790 M 700-780) My "safety school" in case my scores aren't good enough to get admitted to my top school choices. I heard that dropping out of Harvard in the first year is the path to total domination of the computer industry. It is time I started dominating instead of being oppressionated!

Nothing says "success" more than getting admitted to your alma mater a second time. Don't do this if you have defaulted on all of your school loans or if you are on the run from the authorities. I printed the admission ticket. It is required on test day.


The beauty of my special test-taking system is that it requires significantly
less preparation than other test-taking systems.

Perhaps the best way to learn how to fail is to learn how to succeed.
The College Board offers a handy manual to the SAT I:

www_college_board_taking_sat.pdf [538 KB]

However, I have discovered that failure favors the unprepared mind, and so I did
not read this manual before taking the test.


Make sure you have the following items with you before you go to the test

(1) SAT Admission Ticket (Photocopy or Printout okay) (2) Approved Photo ID (e.g., Driver's License or High School ID) (3) #2 Pencils (Mechanical or wood pencils) (4) Sharpener (5) Approved Calculator

FIGURE: SAT Admission Ticket FIGURE: The #2 Pencil (unvarnished wood) I used to take the test. FIGURE: Pencil sharpener and approved calculator that I used for the test. (Pencil shavings are from the test. Bidding starts at $600.) A person taking the test requires a "#2 pencil" ("#2" refers to the hardness of the graphite). A mechanical pencil, or a regular pencil and a sharpener, can be used. An eraser is handy not only for modifications, but also for removing the inevitable graphite smudges that appear when part of the answer sheet is in contact with the table. I haven't used pencils in almost a decade, and I didn't have pencils or a sharpener in my apartment, so I bought pencils and a sharpener at a local 24-hour Rite-Aid store two hours before the test. The SAT materials include a single test booklet with all questions and an answer sheet with a single fold to form a "four page" form. FIGURE: The four panels of an answer sheet. The answer sheet is dominated by little "bubbles" that can be filled in using a pencil, and later optically scanned by computers. For many years, the building selected to serve as the UCI "test center" has been Rowland Hall, formerly known as Physical Sciences I (PS1). FIGURE: UCI, Rowland Hall, site of SAT Exam at the UCI "test center". FIGURE: UCI, Rowland Hall, Room 104 : Test room FIGURE: UCI, Rowland Hall, Room 104 : View from my assigned seat

Fun Fact: Most SAT proctors are teachers from the high schools where the test is administered. Fun Conspiracy Theory: Teachers at prestigious prep schools will get their students in to Harvard, even if it means testing administration (cough) irregularities!

The following is a rough record of how the SAT was administered at the UCI test center on April 5th, 2003. This is just from my memory, so it isn't very precise.

8:00 A.M.
8:15 A.M.

Arrive at the test center and claim a slip of colored
paper that will admit you to the testing room.

8:15 A.M.
~9:00 A.M.

( 1) Walk to the testing room
( 2) Get in line to sign the attendance sheet
( 3) Sign in, and go to an assigned seat
( 4) Administrators handle any same-day registrations
( 5) Final person count
( 6) Administrators carefully read the exact words from a
      booklet, instructing test takers in advance of receiving
      any testing materials
( 7) Put all belongings under desks and chairs
( 8) Sign the little colored slip of paper that admitted you
      to the testing room
( 9) Administrators count test booklets and answer sheets
      several times
(10) Administrators warn test-takers to not open the test book
      or write anything on the answer form until they are 
      instructed to do so.  Then they describe the duration of
      the test, time announcement policy, and breaks.
(11) Test booklets and answer sheets are handed out
(12) The administrators guide everyone through the process
      of filling out personal information on the answer
      form, such as name, Social Security Number, birth date,
      etc, and other data such as the test center name and
      ID number, and unique code numbers that appear on the
      test booklet.


(1) Section #1 : Explanation
(2) Section #1 : Begin work
(3) Section #1 : Work (30 minutes)
(4) Section #1 : Stop work


(1) Section #2 : Explanation
(2) Section #2 : Begin work
(3) Section #2 : Work (30 minutes)
(4) Section #2 : Stop work


(1) Close test booklet and answer form.
(2) Turn in signed colored room admission card.
(3) Break: begins (only stretching or trip to bathroom)
(4) Break: 5 minutes
(5) Break: ends


(1) Section #3 : Explanation
(2) Section #3 : Begin work
(3) Section #3 : Work (30 minutes)
(4) Section #3 : Stop work


(1) Section #4 : Explanation
(2) Section #4 : Begin work
(3) Section #4 : Work (30 minutes)
(4) Section #4 : Stop work


(1) Close test booklet and answer form.
(2) Break: begins (only stretching)
(3) Break: 1 minute
(4) Break: ends


(1) Section #5 : Explanation
(2) Section #5 : Begin work
(3) Section #5 : Work (30 minutes)
(4) Section #5 : Stop work


(1) Section #6 : Explanation
(2) Section #6 : Begin work
(3) Section #6 : Work (15 minutes)
(4) Section #6 : Stop work


(1) Section #7 : Explanation
(2) Section #7 : Begin work
(3) Section #7 : Work (15 minutes)
(4) Section #7 : Stop work


(1) Close test booklet and answer form.
(2) Write out a designated paragraph in script handwriting
      that basically says that you agree to test policies
(3) Fill in another code number from the test booklet
(4) Wait for your testing materials (question booklet and
       answer form) to be collected
(5) Administrators give instructions for test takers to
       "cancel" their scores, either by signing something
       before they leave, or going online within a couple
       of days.
(6) When test materials have been collected from everyone,
       the test booklets and answer sheets are counted
(7) If everything adds up correctly, miscellaneous questions
       from test takers are answered.
(8) Everyone is dismissed at once.
(9) The Administrators erase the blackboard, pack up all
       documents, turn off the room lights, and leave the

~12:00 PM
Have lunch and try to forget the whole thing!


This section describes some of my impressions of this test-taking experience.

The Math Sections were Easy for Me

I thought I solved all math problems correctly on paper (pages of test booklet).
(It turns out I may have made a mistake on one problem.  See conclusion.)

After solving each math problem, and noting that the solution corresponded to
an existing option, I selected the MOST ABSURD alternative answer.  Since this
test date did not offer a detailed question and answer reporting option, I 
cannot share this extra humorous touch with the public.

For the "student-produced" responses, which cannot subtract points from the
raw score, I simply wrote a sequence of numbers that encodes my name:
C O L I N P F --> 3, 15, 12, 9, 14, 16, 6;  I think this will show up in
my "Student Answer Service" report.  (2003 May 26th: I received my "Student
Answer Service" report, and I learned that one of my "random" answers turned
out to be a correct answer!  Note to self: I'm an idiot!)

Sometimes, for particularly tricky math problems, I felt a little pang of
regret that my mission prevented me from proving that I actually knew the
correct answers.

The Calculator was Handy

For several math problems the calculator came in handy.  Sometimes it seemed 
easier to simply add up a bunch of numbers rather than work out a formula that
would yield the answer.

The Verbal Sections were Difficult for Me

The verbal division of the SAT is dominated by "Critical Reading" subsections.
Each Critical Reading subsection has several sets of Critical Reading questions.
Each set of questions begins with a passage having several paragraphs, followed
by questions that refer to the passage.  Many questions refer to specific lines
of the passage using line numbers.

The passages are extracted from existing literary sources, such as articles,
novels, scientific papers, short stories, and journals.  The passages are not
manufactured by the College Board researchers.  The topics of the passages
are very diverse; essentially random.  The College Board points out that the
opinions expressed in the passages do not reflect the opinions of the College

In principle, presenting a test-taker with various types of written material
(scientific, fictional, historical, journalistic, etc) seems like an ideal way
to gauge how the test-taker might perform in an environment that involves
exposure to the same degree of literary diversity.

My gripe with the "Critical Reading" questions is with the questions themselves.
It is my opinion that the questions, in general, involve subjective factors and
implicit assumptions that often make them impossible to answer in a rational

It is possible that people with the same subjective influences and implicit
assumptions do well in college, because college faculty belong to the same 
dominant species of irrationality.  So, the value of the verbal section
of the SAT as a predictor of performance in college might be quite high.
However, the side-effect of optimizing a test to better predict success in 
college is that truly objective minds might be punished for not keeping up
with the latest delusions.

The following questions are typical of the types of questions found in
the Critical Reading subsections:

"Which of the following would be the most appropriate title for this passage?" "The author of this passage would likely agree with which of the following statements:" "Which of the following best describes the author's opinion about XXXX ?" "According to the passage, the American Indian notion of time involves which of the following concepts:" "The author of Passsage #1 would most likely have which of the following opinions regarding statements made by the author of Passage #2:"

While trying to figure out the answers to Critical Reading questions I am convinced that I fully understand the passages, the questions, and the meanings of all multiple-choice options, and yet there are times when there doesn't seem to be a basis for selecting an answer. Maybe I should take the SAT on a date when the detailed "Question and Answer" service is offered. Then I could debate concrete examples with people and decide if there really is a kind of verbal intelligence that helps people answer these questions in a consistent and rational manner. Or maybe I will comprehend the principles of this variety of intelligence and decide that I am better off without it! Thankfully, my test-taking strategy involves choosing the wrong answers, which made some of the trickier verbal questions much easier. Sometimes the best answer was hard to identify, but the absurd answers were very easy to spot. In 2005, multiple-choice questions will require students to identify sentence errors, improve sentences, and improve paragraphs. I'm glad my future doesn't depend on someone's "improvement" aesthetic! A 25-minute written essay section will also be added to the SAT, which means that the arbitrary forces of subjectivity on the overall verbal score may be taken to an entirely new level. It is very difficult for me to imagine grading criteria for these essays that would transcend personal opinions about writing style. I've heard of cases in which a single essay earned radically different grades from different college professors -- and sometimes radically different grades from the SAME college professor, who unwittingly grades the same essay twice at different times. Three Hours of Test-Taking is Exhausting The SAT requires sitting at a desk for exactly three hours of actual problem solving, filling in up to approximately 165 bubbles with a #2 pencil. When one adds in the time spent writing down personal information, filling in other bubbles, and listening to instructions, the whole experience lasts about 3.5 hours. Something about the need to not screw up over a prolonged period like this is really fatiguing. I was grateful for my inspiration to bring a cold bottle of Pepsi and a bottle of orange juice. Still, near the end of the exam I wished I had a third drink, and maybe a chocolate chip cookie! [Make mental note, in case there's a "next time"! :-O] To answer all math questions, you have an average of 75.0 seconds per question to read, compute, and completely fill in the answer bubble. For the verbal sections the pace is a little bit faster, at 57.7 seconds per question. This seems generous, but to me it felt like I was racing the clock for all sections of the test, finishing each section with very little time to spare. So, the pace of the test kept me feeling a little bit harassed. As the test wore on, I became increasingly distracted by the thought of how tedious it was to repeatedly locate bubbles and fill them in. I almost laughed out loud at various times during the test I definitely passed through some kind of mental barrier as a result of this experience. Up to this point, my relationship to exams was based on the strong desire to get every question correct. My educational career lasted 20 years. Assuming an average of 8 distinct courses each year, and two major exams per course, that's 320 exams! When you add in the CTBS, NMSQT, PSAT, SAT (I), AT (SAT II), AP, GRE, GRE subject test, departmental qualifier, etc, that's at least ten more major exams. I have been conditioned to try very hard to determine correct answers. So, in this latest experience, when I worked very hard to determine the correct answer for each question, and then proceeded to pick the exact WRONG answer (in fact, the most RIDICULOUS answer), I had a very strong emotional reaction. For a while I worried that this new peculiar feeling of freedom was in fact insanity; I was finally making the transition to madness. But after a few minutes of settling in to the routine of NOT filling in the correct answers, and beginning to crave this new sense of accomplishment, I started to think of how radically different my mental state was from that of the high school students in the same room, taking the same test. An earlier show of hands indicated that most of the students were taking the SAT for the first time. So, I think it's very likely there were some people in the room who were terrified. For better or for worse they were confronting destiny. Meanwhile, I was confronting silliness. The difference in perspective seemed so extreme that there were moments when I shook with desperately suppressed laughter. I did not laugh aloud or smile conspicuously, but the instinct to laugh was as powerful and as involuntary as a case of hiccups. I admire people who devote themselves to improvement, and I know very well how stressful it can be to have success depend on a relatively short personal demonstration, but humor draws its strength from fear, and the kind of anxiety surrounding the SAT is so familiar, and yet so meaningless, that one can only laugh or become depressed. From the very beginning of the test to the very end I felt euphoric. I was having complex emotions I couldn't describe. This experience was far more rewarding than I had imagined.


The Torture of Waiting for Scores

I was surprised how bothered I was waiting for my score report, especially
considering the fact that I am not a high school student, and I am not facing
deadlines.  Still, I was checking my mailbox daily, and I even figured out when
the postal worker delivered mail (4 PM for me), and I checked at the earliest
possible time each day.  All day long I was distracted by the anticipation of
getting the score report.  I wrote an e-mail to the College Board and learned
that score reports were sent with first-class postage.  (I bet they get that
question all the time!)

But most people don't have to wait nearly as long to find out their scaled
scores.  About ten days after the test date, scores are available online for a
charge of $13.  On the first Monday three weeks after the test date, scores are 
mailed out to test-takers and any specified institutions via first-class mail
(usually 5 full business days to arrive at points in the US).  Starting on the
date the scores are mailed, the online scores for that test date can be
accessed without paying a fee.

FIGURE: Looking at my SAT scores online.

Unfortunately, the online report does not indicate the raw scores.  Still, it
is reassuring that my scaled scores and percentages are at the minimums.
(NOTE: Contrary to popular opinion, one is likely to have to get at least one
question wrong on the test to result in the lowest scaled scores possible.
Blank tests are interpreted as requests to cancel scores, but even if they were
given scaled scores, such scores would likely be significantly greater than the
minimum scaled scores of 200.)

FIGURE: I love how the advice avoids making any significant promises: Tensai
        can look forward to checking his mailbox for big fat envelopes.
        Oh, brother!  

The Envelope...

Finally, on May 6th, eight days after the College Board claims score reports
are mailed out, I received the following:

FIGURE: Envelope containing my destiny.

But check out the date stamped by the postage meter:

FIGURE: April 24th?  I suspected(*) a conspiracy!

(*...Thanks, Christopher™, for "clearing up" a suspected conspiracy!
[See e-mail section]  Or is it "covering up"...?!)

Failing at Failure: Total Humiliation!

Opening the envelope, eager with anticipation, I find the following report:

FIGURE: Printed score report.

The first part is reassuring:

FIGURE: Scaled scores and percentiles.

But, OH NO!  I actually answered two questions correctly on the test!!

FIGURE: Two questions correct, with no question omitted.

Ugh, I have failed to achieve total failure!  That is pathetic.  But somehow 
this seems like a fitting outcome -- and it is a wide-open invitation for

I can't believe I accidentally answered two MATH problems correctly!  I was
preparing myself for the disappointment of getting a few VERBAL questions
wrong, er, I mean correct, but MATH?!...How?  Sabotage! ;-)

"Student Answer Service" Report

A few weeks later I received the "Student Answer Service" report, which told
me a little more about which questions I answered correctly by accident:

FIGURE: "Student Answer Service" report.

The two "+" signs indicate questions I answered correctly:

FIGURE: Two questions answered correctly ("+") in mathematics.

As you can see, one correct answer was in the "Student-Produced Responses"
section.  It's impossible to get negative raw score points for these questions,
so it would have been better to simply leave all 10 answers blank -- but 
instead I encoded my name (COLIN P F... --> 3,15,12,9,14, 16, 6...), and it
appears that "9" was a correct answer!  As for the other geometric reasoning
question, I'm guessing I simply answered it correctly by accident.


I am confident that I am one of the few people in the long history of the SAT
to answer practically every question incorrectly.  Getting a score of 1600 is
more of a challenge, but my goal was novelty!

"It is better to have answered incorrectly than never to have answered at all!"

In 1994, the introduction of the "student-response" math questions made it impossible to lose points on some questions, which represents a blow to the goal of extremely wrong performance. In 2005, the introduction of the written essay section, replacing a section of multiple-choice questions, will deprive test-takers of the current negative point potential. This experiment grew on me as time passed by, and now I am thinking of other funny angles, like asking Princeton Review or Kaplan if they would be interested in being able to make the claim that a person who participated in their SAT preparation course improved his test score by 1200 points! FIGURE: "Could my future get any brighter?!" -- Colin P. Fahey, immediately after taking the SAT I at the UCI test center I want it to be known that I think the SAT is a reasonable test. I also believe that the College Board web site and e-mail notifications are very good resources for high school students on their way to college. Some aspects of the SAT are bizarre, and perhaps competition in the standardized testing business would be good for consumers (students and institutions), but I can't think of any particular improvements that would be big enough to justify entering the market.


Results not typical.  Getting accepted to college after using this system may
indicate a serious problem.  Consult a psychiatrist before starting any 
strenuous insanity program.

If you are a high school student, please, for love of "The Man" and his "System",
DON'T do anything described in this article!  Over the years you will discover
that meaning in life is derived from being very employed and consuming as much
as you can on weekends and your ten paid holidays.  Doing poorly on the SAT
jeopardizes your proper placement as a cog in the machine.  The Man has many
glorious gifts stored up for you in the System, and he desperately wants to 
give them to you, but when you stop conforming, The Man feels great sorrow,
and may even have to punish you...  I shouldn't even be telling you these
things, young Consumer.  Just... No, that's all I'm saying.  Okay, that's...
Just, okay?, just leave.  I don't want any trouble.  I love the System!  All
hail the System! (Breaking in to song:)  Work, work, work, work, work,
consume, consume!      Work, work, work, work, work, consume, consume! 
(Repeat phrase 50 more times, and the System song begins again.)

This article is NOT an official guide to anything.  Do NOT regard the statements
in this article as being official, up-to-date, accurate, or suitable for any
purpose -- except laughing.  Laughing, and sending me money for no reason.


This article was discussed on Slashdot on 2003 May 10th.  Over a period of
exactly 24 hours, 84,692 people read this article.

The "YOU FAIL IT!" is a recurring comment on Slashdot.  It's surprisingly
appropriate for my project.  And sure enough...

I can't believe someone hasn't said this already.. (Score:1)
by Spruce Moose (1857) on Saturday May 10, @06:24AM (#5925089)


"In Soviet Russia..." is a recurring posting theme on Slashdot.

I can't resist (Score:0)
by Anonymous Coward on Saturday May 10, @07:45AM (#5925231)

In Soviet Russia, the SAT FAILS YOU!!!!!!! Muaaahhaaaaahahahahahahahahahahaha

In Soviet Russia... (Score:0)
by Anonymous Coward on Saturday May 10, @01:51PM (#5926423)

In Soviet Russia, we make fun of the SAT by answering all questions correctly!


Although sections, and possibly even questions, are in different orders for
various test booklets, making it impossible to glance at another person's
answer sheet to cheat, the following is a funny scenario:
I feel sorry ... (Score:5, Funny)
by WeeBull (645243) on Saturday May 10, @05:20AM (#5924969)

... for the guys either side of Colin in the examination room
... glancing across ...

"He answered 'D' on Question 26? But, I'm sure it's 'B'!
Shit ... uhm, maybe it IS 'D'. *rubs out, ticks 'D'*.
WHAT? 'A' for Q27?!? SHIT SHIT SHIT! *rubs out*"

Disclaimer: Don't cheat on tests, cheating is bad, mmmmkay?
No, I'm not responsible for the (v funny) Weebl and Bob
cartoons, I just thought I owed them a link-back.

sure 400 seems low (Score:4, Funny)
by Savatte (111615) on Saturday May 10, @04:55PM (#5927333)
( | Last Journal: Friday October 11, @03:07PM)

but it rounds up to like 540 canadian.

Re:Top 2% (Score:4, Interesting)
by Smudgy (144144) on Saturday May 10, @10:43AM (#5925698)

"(Is there a reason that what any sane person would call
a "zero" is a 400 on the SATs?"

The idea is that each section of the SAT is theoretically scored
from 0 to 1000, with a mean of 500 and a standard deviation of
100 points. After calculating the scores, they drop the low and
high outliers and shift them to 200 or 800 respectively, keeping
three standard deviations from the mean.

Re:Standardized tests -- a test of modesty? (Score:0)
by Anonymous Coward on Saturday May 10, @05:19PM (#5927479)

"I think I'm in the top (some low number) percent of the
population as far as intelligence and knowledge goes."

And... out of curiosity, in what (lower) percentile of modesty
do you think you are?

"We're talking about a modest man who has a great deal
to be modest about"

Re:on the positive side... (Score:1)
by hazem (472289) on Saturday May 10, @05:04AM (#5924924)

I don't think I'd ever want to be associated with an organization
that would look at me, then say "well sure you have a masters
degree, you only got a 200 on the SAT. I'm sorry, we can't hire
you/give you a security clearance/let you ride in our airplane/marry
our daughter."

Actually this might be a good sanity filter! Where do I sign up?

With the introduction of the essay section of the SAT in March, 2005, a new 
kind of grading is required.  Apparently essays will be graded by two humans.
However, essays on the GMAT are graded using computer algorithms to assure
uniform standards...But it opens up the possibility of cheating by simply
writing things that maximize the points awarded by the algorithm.  I can 
imagine writing total nonsense that has all the right properties to fool the
grader algorithm in to awarding a high score.  Now THAT would be a good hack!
Too bad it's just the GMAT.  Here's a link to the E-Rater system:

ETS has an automatic rater (Score:2)
by scotay (195240) on Saturday May 10, @06:57PM (#5927912)

ETS uses their �e-rater� system to score essays []
in the GMAT.

I searched College Board literature to find the policy regarding filling in two
or more ovals on the answer sheet for a single question.  I seem to remember 
that it would be considered an omitted question, but I can't find any published
evidence.  However, filling the "E" oval for a five-choice question is
counted as an omitted question (no points subtracted).

In any case, my objective was to answer all questions incorrectly by avoiding
the correct answers, not through a technicality.  I will not announce any other
low scores on this site unless I get a copy of the "Student Answer Service"
report or the "Question And Answer" report.

Does anyone have any evidence supporting or disproving the following claim?

it's easy, just double bubble! (Score:4, Interesting)
by fugu (99277) on Saturday May 10, @04:59AM (#5924911)

dude, if you fill in 2 bubbles you get the question wrong,
all he had to do was scribble in heavy black down the whole
page and he'd be done in 20 seconds

Here is someone who disagrees with the previous post:

Re:it's easy, just double bubble! (Score:2, Informative)
by devnull17 (592326) on Saturday May 10, @06:33AM (#5925101)
( | Last Journal: Tuesday April 08, @03:29PM)

Nope. IIRC, answers with more than one bubble are regarded
as blank. In the SAT's, blank answers are worth a fraction
of a point, so as to discourage random guessing. You'd score
abysmally, but not that abysmally.

(NOTE: "IIRC" = "If I Remember Correctly")

Keep checking that mailbox.... (Score:2, Funny)
by CPgrower (644022) on Saturday May 10, @05:24AM (#5924978)

"This is a key time in the college admission process.
Stay organized, keep your grades up, and before you
know it you'll be checking your mailbox for fat envelopes."

Yeah, the Dell guy's gonna send him mail saying
"Dude, you're going to college!"

(NOTE: Refers to very popular Dell television advertisement (US, 2003))

thought experiment (Score:1)
by WilyKit (68796) on Saturday May 10, @05:51AM (#5925035)

A better project would be to try to provoke ETS into
investigating you for cheating.

Take the test twice; get ~900 the first time, then give
it your best shot for a 1400+ the second time.

Step two: wait for the phone call.

And a funny continuation of the steps...

Re:thought experiment (Score:2)
by angle_slam (623817) on Saturday May 10, @10:30AM (#5925638)

3) ???
4) PROFIT!!!!!!!!

I liked this bit... (Score:5, Funny)
by Unominous Coward (651680) on Saturday May 10, @07:04AM (#5925158)

near the bottom of the page there is an overexposed photo of the
test taker with this caption:

FIGURE: "Could my future get any brighter?!"

Which schools? (Score:5, Interesting)
by ThesQuid (86789)
on Saturday May 10, @08:40AM (#5925334)

I'd love to know what schools have sent him a prospectus on
attending. Who are the bottom feeders?

...and the answer is...  (Ouch!)

Re:Which schools? (Score:0)
by Anonymous Coward on Saturday May 10, @10:18AM (#5925608)

University of Phoenix-Online

Zen (Score:1)
by johnwbyrd (251699) on Saturday May 10, @09:43AM (#5925505)

When we were about to graduate from high school, you and I and
every other Slashdot member were brainwashed into thinking that
our SAT scores were to determine our lifelong social status.

It doesn't. People in the Real World will never ask you your
SAT scores. No one actually cares what your SAT scores were.

This guy has achieved some sort of karmic grace. He has
reduced the SAT to the elitist joke that it truly is. He has
beaten the SAT.

(1470 combined, in 1987)

Hey, that's my calculator! (Score:2)
by FuzzyBad-Mofo (184327) on Saturday May 10, @10:53AM (#5925753)
(Last Journal: Saturday May 03, @11:51AM)

The TI-36X is a decent scientific calculator, but whoever was
responsible for choosing dark blue print (on black) for the
third function labels, should die a slow and painful death.
Maybe that print is readable in a bright classroom, but not
in my darkened lair..

Silly percentiles (Score:2)
by fm6 (162816) [...] (#5927343)
( | Last Journal: Sunday May 11, @08:10AM)

The Relative Percentage is an integer between 1 and 99
Which, as any math person knows, doesn't make any sense.
Intuitively, nobody is entitled to a 100 or 0, because nobody
gets a better score or worse score than everybody. But a little
thought about the definition of rounding tells you that it makes
perfect sense to give a 100 to the top 0.5% of test takers.
And of course a 0 to the bottom 0.5% -- a group that certainly
includes Colin Fahey!

Rowland Hall 104 (Score:0)
by Anonymous Coward on Saturday May 10, @06:26PM (#5927793)

Heh, I have a class in that very room this quarter.
It's almost an honor now.


----- Original Message -----
From: xxxxxxxxxxxxxx
To: "Colin Fahey"
Sent: Sunday, May 11, 2003 8:12 AM
Subject: Re: Wrong again (SAT test)

The most fun fact about "SAT eins" and Germany is that a private
television channel is named like that. They are the only one who show
star trek, also they kinda specialised in soccer coverage. It's name
probably a reverence to the satellite "Sat 1" which covers most of
Europe with lots of television programs. They have a beach ball as sign.


The following e-mail mentions some of the "loose ends" that I was privately 
wondering about when I finally decided to put my article online:


Found out about your website from the blog, What The Heck?!
( I'm just a tad disappointed you beat me to the
same idea.

Couple things I thought I'd point out. For those two "wrong" answers you
have in the Geometric Reasoning, let's not forget that standardized tests
often find out they make mistakes and place multiple correct answers, or
accidentally program the computer to accept a wrong answer instead of the
correct one. Something like that may have happened here, and everyone got
a bonus question or two. You may have utterly failed after all.

What's really interesting, I think, is that in the breakdown of each type
of question, you are listed in being in the fifth percentile of all
students in each category, despite nearly pulling off the anti-perfect
test. Maybe I'm being a tad too incredulous here, but does this mean that
about one in every 20 people taking the SAT are screwing up every question
in a given section? I'd love to see what scores rank in what percentiles.

Also, don't trust any postmark run on red ink - those are personal postal
meters, and you honestly can alter those to say whatever date you want.
Government agencies actually cannot accept one of those as a postmark date
(I know, for I work in one). Of course, the College Board is not a
government agency, but still, you see my point - they are not reliable.

Anyhow, thanks for the illuminating and hilarious take on the SAT.


I was also surprised that your calculator always rounds up.
I tried both the US-manufactured calculators I own and they do the same.
Although, perhaps the fix function is different from rounding. There was no
'round' function on my calculators; only 'fix'.

I checked various programming languages (C, perl, python) and they all round
to nearest even integer. So, hrm. Maybe it's just different in the U.S.
After all, you guys still use the imperial measuring system! =D

The following was sent to me by an actual high school student who took the SAT
on the same date, in the same room, as me!  We must have been seated within 20
feet of each other.  His take on my presence is totally plausible and funny:



I took the SAT at the same center as you at UCI, matter fact I saw you
there lol. Anyways you seem to be a bright guy doing the whole AI
tetris, next October when I take the SAT I again you take the test
with me and Text message me all the answers of the math portion while
I do the verbal. I'll cover all costs. LOL. Funny site I enjoyed
sharing it with friends. Take care

Follow-up e-mail after I requested permission to quote the previous message on
my site:


Haha yah feel free to post my message on your site. Actually when I
saw you I thought you were one of the people from the sat prep
institutions trying to take the SAT so you are up to date on what
they are testing. Like a "cracking" the test if you will. Some
people were planning to sit next to you to copy off you thinking
you're the SAT master, I wonder how their scores turned out, lol.
Anyways seems your getting a lot of hits on the website that is
great. I plan to apply to UCI in fall, and since you seem to know
it well you can show me around sometime. Take care.

The following person corrected my error regarding the "postmark" conspiracy.
This is really great information:


Hi, loved the whole concept... One of those things that you think
of - and find brilliant in theory, but think, "Naw, too silly to
actually do."

Thank God someone has the guts to chase the nightmare.

Anyway, my point in writing: Your postmark problem. That isn't
a "postmark". Note under the eagle, it says meter and is followed
by a serial number. That is the imprint of a Pitney Bowes postage
meter, and April 24 would be the date (probably, see * below) that
the postage was applied to that particular envelope. In all
likelihood, some poor schmuck (or a small team of poor schmucks)
working in a mailroom had to apply these marks to _hundreds of
thousands_ of these envelopes, depending of course on how many
locations process the mailing of scores. Clearly, they would want
to get a head start on prepping envelopes. And can you blame them?

* You can set PB postage meters _ahead_ to any date, but you can't
backdate them.

Hope this helps. Keep up the brilliance. Or, I guess, the
lack thereof.

Follow-up after my request to quote the message above:


Oh, and a clarification (and some more useless information): Pitney
Bowes postage meters can, in theory, be backdated, but you would
need either a) a series of codes available only to certain PB
technicians and administrators, or b) to break open the dater, run
it backwards manually, reassemble it, and hope that it still
functions. To do so, as you can guess, would violate several civil
and criminal statutes, not the least of which would be Mail Fraud.

This SAT experiment was apparently bizarre enough to get me in to the Discordian
Society ( ), the largest dis-organized religion.
You are now to address me as:  "Erisian Avatar or Assimilated, House of Apostle
of Eris for the Eristocracy and Cabalability"...


Historical Document
Do not use as toilet paper If you can read this, you are in range

(X) Official Document to keep
( ) Subversive correspondence
( ) Archive copy

Setting Orange
Day 57 of the Season of Discord
Anno Mung 3169
------------------------------ -><- -------------------------------
From his High Reverence Namu the Scientoabsurd,
Great Pasha,
Church of the Sacred Chaos of Eris and her Apostles,
Episcopate of the Discordian Society,
House of Eris Apostles for Eristocracy and Cabalability

By the powers confered to him by Malaclypse the Younger, KSC
Omnibenevolent Polyfather of the Virginity in Gold, High Priest

In the name of the Discordian Movement:
------------------------------ -><- -------------------------------
To: Colin Fahey

Dear (X) Sir, ( ) Madam, ( ) Possum,

It has come to Our ears that your individual person transported
itself, on the day 52 of the Season of Discord, anno 3169, to the
Rowland Hall, University of California, Irvine, in order to take
the test known as SAT 1 with the intention of answering incorrectly
to all the questions.

Though We could not understand why you did not question the answers
instead of the opposite, We believe you did achieve a notable
performance in the Way of Confusion. Moreover your personal
testimony of the experience suggests that your soul was touched by
our Lady of Discord, as is apparent in this excerpt :
I definitely passed through some kind of mental barrier as a result
of this experience. [...] So, in this latest experience, when
worked very hard to determine the correct answer for each question,
and then proceeded to pick the exact WRONG answer (in fact, the
most RIDICULOUS answer), I had a very strong emotional reaction.
For a while I worried that this new peculiar feeling of freedom was
in fact insanity; I was finally making the transition to madness.

But after a few minutes of settling in to the routine of NOT
filling in the correct answers, and beginning to crave this new
sense of accomplishment, I started to think of how radically
different my mental state was from that of the high school students
in the same room, taking the same test. [...] Meanwhile, I was
confronting silliness. The difference in perspective seemed so
extreme that there were moments when I shook with desperately
suppressed laughter. I did not laugh aloud or smile conspicuously,
but the instinct to laugh was as powerful and as involuntary as a
case of hiccups.
We believe you were personnally contacted by Our Lady Eris through
your pineal gland at this very moment.

Therefore, by Our Official Capacity of Great Pasha, Episcopate of
the Discordian Society, for your Act of Faith in Holy Confusion and
Subsequent Enlightenment, We appoint you as:

Erisian Avatar or Assimilated,
House of Apostle of Eris for the Eristocracy and Cabalability

Arise !

Receive Our Most Complascent and Psychotic Congratulations.

In the name of His High Reverence Malaclypse the Younger
Namu the Scientoabsurd
Great Pasha
Church of the Sacred Chaos of Eris and Her Apostles

"The Words of the Fool and Those of the Wise
Are not far apart in Discordian eyes."
HBT, The Book of Advice, 2:1



After more than 128,000 visitors to this web page I have received only two
e-mail messages claiming that "friends" attempted the stunt described in this
article before my own attempt in April, 2003.

     CLAIM #1: "A high school student took the SAT in the 1960s, 
     got all questions wrong, then took the SAT again and answered all
     questions correctly and was admitted to Berkeley."

     CLAIM #2: "A high school student took the SAT in 1989,
     getting all questions wrong.  This student was admitted to Harvard."

I have made efforts to get in direct contact with the "friends" to eventually
get proof of their accomplishments, but I have not received any responses.
I have changed some of the comments in this article to leave room for the 
possibility of other people having done the same stunt before me, and I 
would actually be a little surprised to learn that I was in fact the first to
get such an extreme negative raw score, but so far I have not received any
direct claims from people, or any supporting evidence.

I will gladly report any past or future accomplishments similar to my own.
However, I will ask for various forms of evidence before I make any reports on
this page.  If you plan to register for the SAT to attempt this stunt, be sure
to order the optional "Student Answer Service" if it is available for the 
particular test date.

Several months after getting almost the worst possible raw score on the SAT,
I was invited to pursue a Ph.D. in Computer Science at UCI.  Ironically, I had
to take the GRE General Test again (since my 1992 scores are too old), and I
scored a 1410 on the computer adaptive test version with only a week of casual
preparation.  Am I a hyprocrite?  Wasn't the thesis of this SAT article 
something like "sticking it to 'The Man'"?!  I'm too stupid to know, but in a
few years I might be Dr. Hypocrite, and perhaps even Professor Hypocrite.
All praise The System and its mysterious ways!


Colin P. Fahey