1script - 4:39 pm on Sep 2, 2010 (gmt 0)
Can anyone explain to me how throttling can work? I understand the concept in that goof could be wanting (for reasons of its own) to not give a site too much traffic, but I cannot see how they would acheive this technically without moving a site in and out of the serps during the day. This is something I am not seeing.
Well, that's just the thing - it is harder to move your page in and out of SERPs than to simply take your 2000 formerly ranking KWs, retain 200 and throw away (or at best freeze) the rest. Here you go: your traffic has been effectively throttled down to 1/10th of what it once was.
I admit, throttling down is not an ideal term - it implies that the throttle moves up and down (and there is something moving it) whereas in real life the situation is VERY static, just as you described. I also have KWs that stay on their positions since March. They sometimes move + or -1 position but very rarely. Also, the number of those keywords never changes (I mean within 10% margin - the magnitude of random noise in my case).
So, I personally like dvduval's keyword pie idea. You sliced it - you're done, the sizes of the slices don't change. You may later put two or three slices on someone's plate or you can take someone's last slice away (it's your party, can do whatever you pleased) - I have sites in either category.
It is also very troubling concept. You would expect that the "best search engine in the universe" would actually figure out dynamic throttling based on umm... something - not sure what it would be but it's definitely NOT fairness. Yet they seem to prefer the static (simpler, less CPU intensive?) solution. Why? Too busy re-organizing the worlds information perhaps.