Categorised | Columns

Sweeping police percentages!

SHAH ALAM, 20 July 2009: The investigation into the death of political aide Teoh Beng Hock, 30, is 60% completed, said Inspector-General of Police
Tan Sri Musa Hassan.

FOLLOWING reactions to the terrible tragedy of political aide Teoh on 16 July 2009, I was drawn to the figure put forth by police chief Musa: 60%.

Odd, I thought. That figure is surprisingly accurate. Nowhere has Musa been reported to have noted that the number was an estimate. Our investigations are 60% done, he said. No more, no less. Not 60.35%. A round 60.

So I call Malaysian Statisticians for Verisimilitude Union (MSVU) president Datuk Ong Chin Teng. I ask him whether it would be accurate, or even possible, for the IGP to affix a percentage to a police investigation. I am assuming that such an investigation, by definition, could not be quantified in the same way one may quantify — say — the download progress of a pirated .avi copy of Harry Potter and the Half-Blood Prince.

To my surprise, Ong answers that, yes, it is possible to do so.

“There are many different tasks that the police would undertake in the course of this inquest: analysing glass fragments, measuring the height of the building, checking the timeline of events,” Ong says.

“If we contrast the tasks already performed with the total tasks on schedule, we can arrive at a number.”

But surely there are some elements to detective work, such as new evidence, or witnesses who come forward, that would mean such a convenient number would be thrown off?

“It is standard practice for statisticians to allocate about 10% of the final estimation to ‘surprise factors’,” Ong explains. He adds that this number remains constant, as studies of criminal investigation trends have shown that the probability of “surprise factors” usually arises in predictable quantities.


                                                              (Pic by cobrasoft / sxc.hu)
“The probability of the 15% error margin being accurate is 94.2%,” Ong says.

“It’s very human to think that there are things we can’t put numbers to,” Ong continues. “But the truth is, you can apply a figure to everything. Just look at economics!”

The Fendy-Kevorkian theory revisited

Ong reveals that according to the Fendy-Kevorkian theory — and philosophy — of “social tally accumulation”, almost everything can be counted and calculated for: from life expectancy, to the grief felt by any one individual at a funeral, to abstract concepts such as equality and moral good.

This paradigm of looking at the world was new to me. The last percentage number that directly affected my life was 64% — the result of my SPM-level Math exam.

Intrigued by Ong’s sharing, I decide to do more reading about Jack Fendy and Richessa Kevorkian. With their “social tally accumulation” principle, these two brilliant actuarial scientists from Gallaher College, Vermont, had discovered an equation that was capable of accounting for… well, practically anything!

P(k) = AccnT (12 + z)-1

Where “P” is total precipitate;

“k” is the ratio of deferrable expenses to the estimated gross results;

“AccnT” is the number of certified accountants involved in the tally, times the duration of the tally, averaged daily;

and “z” is the amount of unknown unknowns involved.

While “social tally accumulation” was devised in 1965, its methodologies wandered in the wilderness for decades, and developed its reputation as a fringe science. Statisticians now put this down to the fact that the technologies of the time were just not able to cope with the amount of number-crunching required to employ the Fendy-Kevorkian equation effectively. This rendered it impractical for everyday application.

Today, however, with fairly capable arithmetic machines in our pockets, the Fendy-Kevorkian equation is available for any person to use. Surfing online, I find numerous iPhone applications that require users to merely fill in a questionnaire of quotidian details. Here is one to calculate satisfaction in one’s romantic life.

“Social tally accumulation” has revolutionised many aspects of society, including politics. Political scientist Roy Arundhati, in his landmark essay The Calculus of Unbridled Equality (not to be mistaken for Indian writer Arundhati Roy’s The Algebra of Infinite Justice), concludes:

“Through application of the life-quantifying solutions provided by F-K theory, politicians in burgeoning democracies have been able to accurately read the moods of their electorate. Political parties across the globe are now using focus groups and statistical science to tailor their platforms to respond to voters, and alter their perceptions.”

Arundhati argues that this means politicians are better able to “push through popular partisan agendas”, thereby satisfying their respective “citizen-adherents”.

Far-reaching effects

Ong agrees that the Fendy-Kevorkian theory had far-reaching effects. He points out that statisticians like himself were being employed by Malaysian parties — on both sides of political spectrum — to determine campaign and administration strategy.

“I just drew up a paper on the Teoh issue for a Member of Parliament (MP): in real, hard numbers, what the death means for the state of Selangor, and how these circumstances may benefit or hinder laying out their strategies in the next few months,” Ong says. He declines to reveal the identity of the MP in question, due to a confidentiality agreement.

“The data suggests that Teoh’s death was actually good for both the ruling and opposition coalitions,” Ong quips — about 60%, with a 5% error margin, for either.


                                                         (Source: humboldt.edu)
My mind 75.4% blown, I decide to ask Ong about the amount of hope Malaysians currently have for positive political change.

“That’s about 37.6%,” Ong says. “But I think there’s a 94% chance that figure will change in the next six months!”

I also ask Ong to calculate my probability of a romantic encounter in the next month. He gives me an answer — but some data, naturally, belongs in the private sphere.


Zedeck Siew apologises to statisticians, accountants, and emphatic readers.

Post to Twitter Post to Google Buzz Post to Delicious Post to Digg Post to Facebook Post to StumbleUpon

Tags: , , , , , ,

10 Responses to “Sweeping police percentages!”

  1. Pratamad says:

    It’s not about whether or not the “60%” is accurate. It is about whose mouth it was uttered from. And the character of that one person has more meaning than the actual number uttered. And we all know who Musa Hassan is.

  2. roastpork says:

    Selangor Police Chief Khalid Abu Bakar said the investigation is ‘more or less’ done: http://malaysiakini.com/news/109433

    What percentage does ‘more or less’ translate to?

  3. James says:

    The central question is not about the accuracy of the quantitative aspect of the issue and investigations. What is the truth behind the heinous deed? Only the MACC (and probably PDRM too) have the truth and they being the bad guys are not going to disclose of course.

  4. armstrong says:

    If police can be trusted, Mahathir would be a woman.

  5. Nicholas Aw says:

    A well written analysis Zedeck. Your article seems to have come out of an episode of the TV series ‘Numbers’ where we all know the mathematical ability of Charlie Epps.

    I must admit that I’m really bad at numbers. But I’m okay with ‘figures’ (of a different kind). I have been making regular comments at The Nut Graph so once in a while this comment is made in lighter vein. Keep up the good work all you people at The Nut Graph.

  6. Rajesh Taluar says:

    Since the MSVU and the theories you put forth state we could put a number to anything, I’m willing to bet the percentage of idiocy in Musa and those involved in the investigations would result in some pretty high numbers. Come on, Musa, Teoh’s grieving family deserves real answers, not pointless percentages! What was he trying to say when he gave that figure? That the police is actually doing some work for a change?

  7. Is it not time that the Public stopped taking notice of any statements or rhetoric made by the PDRM, MACC, AG, politicians before we are sent to Tg Rambutan for listening to them, and the poor newscasters over TV who are made to repeat these rhetoric to a helpless mass of TV viewers and radio listeners. The newspapers have no choice because they have to survive – they are the workers in the press. So they continue dishing out rhetoric after rhetoric to a public that is already satiated with nonsense.

  8. yeo kien kiong says:

    Was wondering what percentage will Altantuya’s murder case be solved… rationally…

  9. Katherine says:

    I found this an entertaining read, which speaks volumes for the author’s skill in deftly weaving in humor into the dry subject matter that is quantitative methods. I wish textbooks could do the same. Well done!

  10. kclim says:

    Percentages can be used as a tool to deceive people. We need to know what procedure or constraints or variable is taken as a whole. Leaving the variables that do not meet our assertions will portray the false information that we wish to portray in our crooked mind. Don’t believe in [everything] the figures portray for they never tell us what yardstick they are using. Only if all variables are disclosed can we only believe in the figures given.


Most Read (Past 3 Months)

Most Comments (Past 3 Months)

  • None found

Advertisement


<

Advertisement


<
  • The Nut Graph

 

Switch to our mobile site