Every four months, Microsoft hands out awards for people who they believe excel at helping the various Microsoft communities. It is not a certification that can be studied for and earned.
Initially, it was one person who thought that those who were helping out on the Microsoft forums deserved some recognition. It was just an thank you post. This idea was taken up by others at Microsoft, who on their own time, helped expand the list. Eventually some of the recognized were invited out to Redmond to meet the product teams they supported. This was the first MVP summit, a small group of volunteers who enjoyed helping their online communities getting recognition from Microsoft.
This was a grass roots movement and luckily, someone terminated the programme in 1999. Someone thought that the access the MVPs were getting was a potential problem. Over the weekend, those who were helped in the forums were quite vocal with Microsoft upper management and let their feelings be known. The programme was quickly reinstated, but now they had the backing of upper management who remain a staple of the MVP summits as the keynote speakers. The initial MVP staff ran the MVP programme in addition to their regular work. With the reinstatement of the programme, they now had backing and support from upper management (and a budget).
That was a long time ago and the programme has grown. Now, every three months, individuals are singled out for their support of the various communities over the previous year. As part of the Microsoft structure and has its’ own budget and as the programme has grown, MVP perks have unfortunately had to be cut. Luckily, one of the key perks, contact with the product team, is still very much alive and encouraged. Having the summits in Redmond meant most MVPs had direct access to the product teams.
Initially it was based around a few products and the leads were based with those products, now it covers most of Microsoft products. (I am still waiting to meet the MS Bob MVP). Regional MVP leads have been added to address the needs of those MVPs around the world. Only the US leads are product based, so the regional MVP leads need to understand the full suite of Microsoft offerings. So even though they may not know the nuiances of a certain switch in your product, they understand your passion and will try and get someone to help with the answer.
So, why did so and so not get an award? Like the Academy awards, only so many can be recognized. This does not mean that those not receiving an Academy award are terrible actors, just that the Academy saw something to recommend one of the others. Also, the programme has been expanded world wide and it is an extremely difficult task for the MVP leads to evaluate all the recommendations they receive. As far as I know, the MVP leads do not have a quota for having x number of Product Y MVPs in their area, but there are constraints on how large the programme can grow within their budget. So there are many reasons why someone was not awarded.
So why is there not a shopping list for MVP criteria? Back in the old days, the selection was based on participation in the CompuServe forums and a few other lists. Now selection is opened up to a far wider range of activities. So saying 25% of A and 33% of B will get you an MVP award is not possible. So if I answer 500 or 1,00 questions, do I qualify? What is the quality of the answers? Is it a yes/no or does it have a detailed explanation or recommendation. Obviously quantity alone can not be a criteria. Who is to say that talking to a user group of 25 is less important than talking to one of 3oo. Since they are making money, should authors be excluded? A number of authors actually loose money writing about technical products, some actually do get rich. Again, this is another criteria that is difficult to quantify. If the criteria is quantified like other certifications. Then we will have people going for the award rather than helping the communities. One year, several Office MVPs were not awarded because other individuals had higher answer counts. On further examination it was found that some of the answers were not useful and were just there to boost the counts. So I hope that Microsoft keeps the award criteria a mystery. Yes, the MVP leads are human and do make mistakes, but the selection process is not easy.
People who did not honour their NDAs have been dismissed. (Other than joining the mother ship, this is the only way to lose an MVP award).
The award is not something you can study for. Some of the old time MVPs prefer to keep helping the community whether they get another MVP award or not. Their focus is on community.
D’Arcy Lussier has an excellent post on what new MVPs can do http://geekswithblogs.net/dlussier/archive/2012/01/05/148246.aspx
John Marshall… Visio MVP Visio.MVPs.org
john Visio MVP 
In other words, if your motivation and passion are directed properly to help others in your field, you go over and above to help (AND knowing more than the average Joe doesn’t hurt), and people goto you for knowledge/help/guidance, then you are good candidate!
I’ve talked to many MVPs in our area, and they are FANTASTIC at rejuvenating energy levels, getting people psyched up again to continue learning rather than just stagnating and falling by the way-side.
Great blob John (MVP)! Keep it up!