Posted by Lindenblüte on October 20, 2006, at 14:24:37
Dr. Bob and psycho-babble community,
[*note- I kindly request all general comments regarding the necessity of blocks, or blocks to specific posters, emotional and psychological sequelae of blocks etc. to another thread.]
I would like this thread to focus exclusively on the mathematics of the current block length system.
In addressing the problems of the exponential growth of block lengths, Dr. Bob instituted a new algorithm of block length determination.
I would like Dr. Bob to explain his reasoning on why it is necessary to round DOWN when determining the reduction in block length that is given for civil behavior since the previous block.
As I am currently prohibited from linking you to a post which contains incivilities, I am unable to link you to posts by Dr. Bob where he metes out the length of the block. (Isn't that droll?)
Instead, you will have to take for granted my honest copying and pasting from the block posts I will use as examples to illustrate a bias in how block lengths are determined.
**************
Example 1.previous block: 1 week
period of time since previous block: 51 weeks
uncivil toward a particular individual or group: yes
particularly uncivil: yes
different type of incivility: no
clearly didn't understand PBC and made effort to reply: no
provoked: no
uncivil in multiple posts at same time: no
already archived: noIf we take 51 weeks, divide by 10, and round down, that's a reduction of 5 weeks. If we apply that to her previous block, that takes her back to 0. And if we go from there, that's 1 week.
**********
Example 2.previous block: 8 weeks
period of time since previous block: 65 weeks
uncivil toward a particular individual or group: yes
particularly uncivil: no
different type of incivility: no
clearly didn't understand PBC and made effort to reply: no
provoked: no
uncivil in multiple posts at same time: no
already archived: noIf we take 65 weeks, divide by 10, and round down, that's a reduction of 6 weeks. If we apply that to your previous block, that's 8 - 6 = 2 weeks. And if we triple that, that's 6 weeks.
*************
Example 3.previous block: 1 week
period of time since previous block: 12 weeks
uncivil toward a particular individual or group: yes
particularly uncivil: no
different type of incivility: no
clearly didn't understand PBC and made effort to reply: no
provoked: no
uncivil in multiple posts at same time: no
already archived: noIf we take 12 weeks, divide by 10, and round down, that's a reduction of 1 week. If we apply that to your previous block, that takes you back to 0. And if we go from there, that's 1 week.
***********
Example 4.previous block: 4 weeks
period of time since previous block: 18 weeks
uncivil toward a particular individual or group: yes
particularly uncivil: no
different type of incivility: no
clearly didn't understand PBC and made effort to reply: no
provoked: no
uncivil in multiple posts at same time: no
already archived: noIf we take 18 weeks, divide by 10, and round down, that's a reduction of 1 week. If we apply that to your previous block, that's 4 - 1 = 3 weeks. And if we triple that, that's 9 weeks.
***************************
I believe that in the spirit of a supportive website, posters should be given as much credit as possible for their civil periods following a block. Of particular concern is when the previous block occured during the previous "exponential growth" block length algorithm period.
In example 1, a year has elapsed since a block. This is enough time to give the poster a "clean slate", or a zero, from which the current block will be added to.
In example 2, well over a year has passed since a block. However, 65 divided by 10 will be rounded DOWN, to be a reduction in 6 weeks. Since the post from over a year ago (which must have occured during the exponential growth phase) was 8 weeks, there is now a positive, non-zero number to be multiplied by the factor of three, yielding a lengthy block of six weeks. [one can only imagine how hard it would be not to be able to take one's medicine, talk with one's friends, or see one's therapist for six weeks!]
I feel it's important to recognize the impressive accomplishment of civil behavior for 65 weeks, and in rounding DOWN, Dr. Bob essentially says that five weeks of being civil is of no consequence.
In example 3, a poster's slate is wiped clean after 12 weeks of civil behavior. Interestingly, the previous block was only a week in length, and occured during the phase since the new algorithm had been implemented.
In example 4, a poster has been behaving civilly for 18 weeks. However, because the algorithm rounds DOWN, eight entire weeks of civil behavior are ignored (nearly half of the time elapsed since the previous block!)
I feel it's important to recognize the impressive accomplishment of civil behavior for 18 weeks, and in rounding DOWN, Dr. Bob essentially says that eight weeks of being civil is of no consequence.
Because this rounding DOWN occurs before the multiplication factor of three, Dr. Bob's algorithm suggests that 8ignored weeks x 3 = 24 weeks of being civil are of no consequence.
*****************I would like to call for review of these particular aspects of the new block length algorithm.
1) why round DOWN?
2) why penalize posters whose previous blocks were determined by the previous system (exponential block length increases)? Why should these posters be treated differently than posters who have blocks under the current system?
3) why is the reduction factor applied before multiplication?
4) why not issue a set of guidelines by which one's "record" can be wiped clean?
My modest proposal is that a block of 1 week be expunged from the records after minimum 5 weeks of civil behavior. 2 weeks = 10 weeks and so on.
Futhermore, I believe that a reduction in block length should be taken into account without rounding, and after multiplication.
Let's take Example 4. One possible algorithm for determining block length is as follows.
If we take a block length of 2 weeks (for being uncivil towards a particular individual or group) and add 4 weeks (the length of the previous block), we have 6 weeks.
Because your previous block occured 18 weeks ago, if we divide that by 10, that is a reduction of 1.8 weeks.
your block will expire in 6.0 - 1.2 = 3.2 weeks, or 22.4 days
Only at this final stage of calculation is it appropriate to round to the nearest significant unit, in this case days. (Dr. Bob, surely you haven't forgotten one of the basic tenets of the natural sciences? I wouldn't want my pdoc to round my weight to the nearest ton, then figure out the appropriate dose at that point!)
The block length would be
22 days.
What do posters and Dr. Bob think about this?
Thanks for your input
-Li
poster:Lindenblüte
thread:696312
URL: http://www.dr-bob.org/babble/admin/20061018/msgs/696312.html