By Martine van Bijlert

As journalists are starting to pack up and go home and observers are formulating their conclusions (some irregularities, need to work on the voter registration) it seems that the real contest is yet to start. The network of governors, district governors, police chiefs and local commanders, that was mobilized in the run up to the elections and that had seemed to play a surprisingly minor role in the process (apart from some campaigning assistance) has kicked in. And has gone overboard in the process.

Consistent and credible reports from the south and the southeast have been coming in for days now: massive and blatant ballot stuffing; the removal or invalidation of votes for rival candidates; complete overhaul of ballot boxes; intimidation of witnesses and IEC staff; systematic removal of the publicly displayed tally sheets.

Local IEC representatives are informally releasing figures that have no relationship to observations on the ground and that are beyond the absurd. See for example the figures provided by Toryalai Ghaznawi from the Kandahar IEC, quoted here, on the high voter turnout in his province: 60 percent -- which is an amazing 480,000 voters -- with a particularly high turnout in Spin Boldak. Observations on the ground, on the other hand, have put voter turnout in Kandahar at an estimated five to fifteen percent (see for instance the observations by Alex Strick van Linschoten), which is much more plausible given the circumstances. And everybody who has been following the elections in the south knows what has happened in Spin Boldak and who was involved.

A man from Kandahar, whom I had known for years and who was a candidate in the provincial council elections, called me on Saturday to give his election day assessment. "It was very good, very transparent." I told him that was not what I had been hearing, but he assured me things had been fine. I asked him if he had any preliminary count results. He didn't, as he had just returned from a district, but he would find them for me.

A few hours later he called, in a rage: "In Spin Boldak, they used 85,000 votes; 75,000 for Karzai! If this is a real election, then send me a delegation of UNAMA, EU and the IEC and we'll see if we can find 10,000 people who hold a voter card with a matching ID! In Zheray they used 30,000 votes! Send me a delegation and we'll see if we can find 5,000 people! In Khakrez they used 50,000! Send me a delegation and we'll see if we can find 10,000 people with a card! Even the provincial council members [those that candidated themselves again] claim to have won 30,000 or 40,000 or 50,000! It is a lie!"

I called someone from Ghazni: "How did the election go in Ghazni? Or how it did not go?" He laughed. "No, no, there was an election. It took place in the governor's guesthouse, and in the compounds of the district governors, and in several houses. It's still ongoing." I asked him what he meant. He explained that it had been decided that Karzai should win with 250,000 votes, out of an imaginary total of 340,000. Not enough boxes had been filled yet to reach that number, so the filling and counting continued. It would be funny, if it were not so worrying.

You wonder if we are really going to be asked to believe that there has been voting in the fourteen Pashtun districts of Ghazni. Voting in those areas was going to very limited under the best of circumstances, given the extremely limited reach of the government, and these were not the best circumstances.

The Taliban in Ghazni had been most articulate in their anti-election threats and they were controlling the major roads well before the elections. People didn't move. Claiming anything other than a low turnout here is an indication of how gullible or indifferent to details we are thought to be.

The first unofficial figures were reported on Saturday evening, based on the counting of 4.5 million votes: Karzai 71 percent, Abdullah 23 percent, Bashardost four percent and Ashraf Ghani one percent. The remaining two million votes are mainly from "remote areas where Karzai was strongly supported," which should set alarm bells ringing. The IEC leadership is now in a difficult position. It will be very difficult not to call fraud when it is so blatant and when international observers are looking over your shoulder, but who will have the stomach to openly challenge the powerbrokers who were involved.

The wait is now for the official release of count figures, which should start on Tuesday on a rolling basis. Things to look out for are implausibly high turnouts in insecure areas; implausibly high numbers of female voters in insecure areas; implausibly high numbers of votes for certain provincial council candidates neatly divided over the districts and polling stations and coinciding with high numbers of votes for a single presidential candidate; and local landslide victories. Any move to release figures without providing the necessary detail -- at least on district level -- should be challenged by the relevant observer organizations. Provinces to particularly watch include Ghazni, Kandahar, Helmand, Wardak, Logar, Uruzgan, Zabul and Kapisa.

While we have been busy with the number of incidents, the total turnout figure, and whether candidates and their supporters will decide to contest the outcome or not, there has been a major and systematic overhaul of the election outcome in the insecure parts of the country. If this is left unchecked the message will be unambiguous: there is no government, there is no law, and the internationals are fine with that. This means there is no real hope for improvement, which is a dangerous message to give in those areas.

Martine van Bijlert is the co-director of the Afghanistan Analysts Network, where this post was originally published.