If you want to be serious, once you have chosen a basis for your analysis you must respect the numbers you get.

Trying to arbitrarily remove some entries is not correct first, and next you won't know for sure if people have quit before the next day where, miracle!, they will be absent of the 30-day active crunchers anyway.

It would already be a mistake on a 30-day period, so imagine on one week or one one day! Particularly during the summer holiday season.

On June 30 the top cruncher of our team has quit as you say, and I know that at least one teammate has believed so. Personally I was not too worried because I know that his results fall when he has several days of vacation. He is in charge of a pool of servers and he probably does not want that anybody could attribute a possible problem to WCG. And guess what: end of last week he has joined again as you would say, and his results are already back to their usual level.

People having 4 weeks vacation at once are much less common in France than it used to be several years ago, but everybody may have one week, even in the US.

Once you have chosen a period of time you must stick to it and not play with numbers. And if you are analysing numbers to determine an "average cruncher" profile, you cannot work on too short periods in a process which is well known as being cyclic.

Cheers. Jean.

I had not gotten to that conclusion yet, but that makes sense, since most crunchers have a single computer and dual core is as small as you can get.

So the Average WCG Cruncher has a C2D or AMD X2 desktop CPU running 24x7, which would be faster than having 2 single core computers, and anything more than that puts you on the road toward being a monster cruncher.

Over the next few years, that average will increase significantly as new users come online with faster (and multiple) computers and 'Old Reliable' breaks down and gets a much deserved upgrade (possibly to a Quad).