I am sorely disappointed at the poor quality of BK Ong's analysis in his latest letter to malaysiakini.

While I agree that the recent delineation exercise has helped the BN candidates in the parliamentary seats of Jerai, Pokok Sena, Baling and to a lesser extent, Kuala Kedah, the shoddy, inaccurate and loose use of terminology and facts in BK's analysis is not helpful at all in trying to understand the impact of the aforementioned exercise.

Let me raise a few general points before revealing the poor use of facts one by one.

Firstly and most importantly, BK fails to differentiate between gerrymandering or delineation

where the boundaries of seats are redrawn thereby causing voters to increase or decrease as opposed to the increase in voters in a constituency because of newly registered voters and not because of a change in boundary lines.

He mentions that in the Sik and Padang Terap parliamentary constituencies, "PAS lost their seats

following the same gerrymandering pattern as above" (meaning the Jerai example). If he had bothered to check the polling districts of these two parliamentary seats, he would have found that they were not affected at all by the recent delineation exercise.

All the polling districts in these constituencies remained as they were in 1999 and no additional polling districts were added. In other words, the boundaries of these two parliamentary seats stayed exactly the same as 1999.

BK also mentioned: "The only unsuccessful case of gerrymandering was Pendang where PAS clung on to the seat against all odds." Again, if he had bothered to check, he would have realised that the electoral boundaries for Pendang were exactly the same as 1999.

Secondly, BK uses the term 'voter import' in a very misleading way. By using this term, he gives the impression that voters have somehow been illegally 'brought in' to a particular constituency.

He doesn't seem to take into account in his analysis the possibility that there could be a natural increase in the number of voters in a particular constituency. After all, shouldn't the number of registered voters in a constituency, one that has not been affected by the delineation exercise, increase over time?

Again I point to the Pendang example. BK says that: "The most significant exception to the gerrymandering success of BN was Pendang where BA retained the parliamentary seat by a majority of 333 despite a voter import of 3,892. BA gains 9% in votes." As mentioned above, Pendang was not affected by the delimitation exercise and hence was not 'gerrymandered'. (The final majority was actually 50 and not 333.)

The 'voter import' was the increase in the number of newly registered voters in Pendang since the 1999 general election. The 3,892 translates into a 7.2% increase since 1999 or a 2.4% annual increase averaged out over three years (1999 year end to 2003 year end). It is also close to the total increase in the number of voters in Kedah from 1999 to 2004 which BK calculated at 7.4% and which I've verified. Not an extraordinary annual increase, is it?

Thirdly, many of the numbers which BK uses also seem to be wrongly calculated. BK said: "Overall BA's votes actually surged 4.7% within a context where voters increased by 7.4%. Relative vote loss by BA is only 1%. They obtained 33% of the votes in the 1999 general election compared to 32% on Sunday."

By the BA of 1999, I hope that BK is referring to the combined vote of PAS, Keadilan and DAP at the parliamentary level. With the use of a little bit of common sense, one could immediately guess that the 33% figure which BK gives is erroneous given the fact that PAS alone won eight out of the 15 parliamentary seats contested in Kedah in 1999.

If the 33% for 1999 is correct, it means that the BN won 66% of the popular vote in Kedah in 1999 and still failed to win half the parliamentary seats there.

My own calculations, which I think are slightly more accurate than BK's, show that the popular vote won by the BA (which includes DAP) in 1999 was 44.3% (with DAP winning 2.3% in its sole seat in Alor Setar).

This percentage dropped to 40.2% in 2004. (DAP is not a factor since they didn't contest any parliamentary seat in Kedah). The share of the popular vote won by the BN increased by about 4% from 55.7% in 1999 to 59.8% in 2004 (rounding-up to one decimal place!)

Hence, not only are BK's numbers of the BA's share of popular vote wrong, his spurious claim that their relative loss is only 1% is off the mark as well. Note that if he had used the correct numbers, he would be able to say that the BA still commands 40% support among Kedahans, a higher figure that the 32% figure which he uses in his letter.

BK also said: "The other side of gerrymandering are areas where voters are exported. Even in these BN strongholds - Kubang Pasu, Alor Setar - BA support surged significantly."

Now, even I'm getting confused. When BK talks about areas where voters are exported, does he mean areas where voters have been taken out or areas where voters have been brought in?

Kubang Pasu was not affected by the delineation exercise (but BK probably didn't check, did he?) and has seen its number of voters increase by 3,669 from 45,925 in 1999 to 49,594 in 2004.

Alor Setar, which was affected by the exercise, has seen its number of voters decrease by 6,365 from 63,678 in 1999 to 57,313 in 2004. If voters have been exported (to or from depending on BK's definition), why have voters increased in one constituency but decreased in the other?

I'm also confused by BK's definition of "BA support surged significantly." While the number and percentage of votes going to the BA candidate in Kubang Pasu increased by 684 or 5.6% from 1999, the number and percentage of votes going to the BN candidate in the same constituency increased by 4,258 or 19%!

The percentage majority of victory increased from 29.2% in 1999 to 34.6% in 2004! Saying that the BA support surged here is akin to saying that we're happy as a football team when we get hammed seven to three instead of four to one! At least we can say that we managed to score two more goals and we scored 30% of the goals instead of 20%!

At the start of his letter, BK said that: "In Jerai, Barisan Nasional (BN) through EC's delineation

exercise moved in 21,607 new voters and obtained a 10,587 majority to win the seat. The same voters in Jerai actually gave extra support to the PAS candidate by a whopping 23% but unfortunately for them they were overwhelmed by a flood of new voters."

Firstly, the delineation exercise moved in the state seat of Gurun from nearby Merbok to Jerai. Gurun has 21,081 voters in 2004 and not 21,607. The difference of 526 is actually the nett increase in the number of voters in the state seats of Sungai Limau (formerly Sala) and Guar Chempedak, the original two state seats in the former parliamentary seat of Yan.

Secondly, I'm wondering where BK got his 'whopping' 23% figure from. Consider the following facts. Nasharuddin Mat Isa won the parliamentary seat of Yan in 1999 with 16,041 votes or a majority of 182 votes. The combined vote for the PAS candidates in the state seats of Sungai Limau and Guar Chempedak was 15,865 in 2004, a decrease of 176 votes from 1999. How a

decrease of 176 votes translates into an 'extra support by a whopping 23%' mystifies even the most flexible of statisticians.

(Technically speaking, one should calculate the votes garnered by the PAS Jerai parliamentary candidate using the parliamentary streams of Sungai Limau and Guar Chempedak but because I don't have these figures, I've used the state returns as proxies. This is acceptable because vote splitting was not significant in the Jerai parliament seat)

I don't have the space or time to elaborate further on how BK misleads by using only the increase in percentage of votes by the BA candidate and not comparing this to the increase in the percentage of votes for the BN candidate. Not forgetting the failure to take into consideration a higher voter turnout and its impact on his analysis.

But I think I've shown in my examples here that his analysis has been poor and inaccurate, at best, and misleading and malicious, at worst.

If this is any indication of his work in general, then it would call into question other aspects of his analysis on his website .

BK, while I admire your enthusiasm and sense of justice, I hope that you will do your homework before coming out with your next installment!