Sunday, June 19, 2005

Sop refined

I've been pondering more on my previous idea of subjectivity. It seems that interest also holds with subjectivity, and a common way that this is done is through stories. For example, it isn't as interesting if in a conversation I just mention that I did that and I did this, but more so if i mention things that happened, my reactions and my interpretations. I don't it even matters how small a triviality it is, as long as its a story or an easily imagined or visualised account of the events.

And from this, I think that it is not the content of what is said, or the events that transpired, but the way in which it is retold that makes it interesting. The content has to be decriptive enough, but concise enough that you don't blabber on, and interest has to be kept up throughout the story. The chain of events have to be told in the right order, with a logical sequence of events, so that the other person can understand why you did what you did. And finally, the punch line or the point of the story has to be given at the end, perhaps with a lead up to this, and maybe this being up to as long as the whole story itself. I think that this is quite difficult to do perfectly, and so, of course, I don't think I really have that many stories to tell. Of the ones that I do, I usually need to think back on them, and figure out what i want to say, the details to include, and the point of it - this takes a lot of thinking, really. And afer an attempt at giving it, I can always see what happened, what might have gone wrong or right, and see where I could have improved on it. An annoying thing that happens sometimes is that others might start a story or mention something, and then pass it on to you. For example, you might be a group setting and you hear someone mentioning something about what you've done or a story you've given, stuff it up by giving the point, not starting it properly, or just not in the way that you would like it, and then pass it on to you to give the rest. This annoys me, because the person may have hyped it up too much, or they may have given away the point, or maybe just because I'm not prepared and gotten my ideas together before I start. But then, I think the main thing is for me that I don't have enough practice in it, and don't feel I can do it well yet. I think the thing I need to do is to have my eyes out for more happenings, analyse it in the way described above, and then try it out (and improve and remember it for next time).

I've recently had occasion to use my SoP system, but have encountered a few problems with it. I'm not sure whether I mentioned it before, but the numbers derived from it wasn't supposed to be correlated literally with each other, or averaged or summed up in some mathematical fashion. The idea was a situation where perhaps you were the decision maker, and then you could hear the respective sops from everyone else, formulate and calculate this in your head, then make a decision based on what others have given you as their indication of preference. And in this capacity, it has worked really well (if I may say so myself), as I think it streamlined the decision making process.

Now, there are 3 problems that I think arises now. Firstly, what happens if you are faced with more than 2 options? You can't really have a linear scale in this case, with 2 polar extremes. I think what needs to be done now is to split up the different choices, and then give a rating scale, perhaps from 1 to 10, for each of these possibilities. For example, you might say A: sop 3, B: sop 6, C: sop 5. The problem then is how to express this as a single sop? Obviously in this example the preferred option is B, but how can we arrive at net sop? Would it be ok for that person to say B: sop 6 without taking into account the preference of the other sops? It wouldn't really work just to minus the other sops from the winning one (you end up with B: sop -2, which doesn't really indicate that you want to do B). What you really need is something that works out always to be a positive sop in the end. One way I've thought that this might work may be to average out all the 'losing' sops (e.g. (5+3)/2= 4) and subtract this from the 'winning' sop (6-4=2) and hence churn out your net sop. Not sure how this will work, but I'll try to test it out.

Secondly, how do you compare sops once you hear them. How do you correlate a 5 from one person, with a 5 from another. What does it mean? So now, I propose that for the decision maker, or at least the person correllating and synthesising the sops needs to have a weighting system to the sops, depending on who it comes from. For example, if say there was someone in the group that usually you didn't agree with or was deemed to make bad judgements, you might assign a factor of say 0.5 (50%) to their sop. In this way, if the person in the example said their sop was 4 for going shopping, you might then apply the x0.5, and hence you come to a weighted sop of 2. I guess in a sense this can be called a person factor, or opinionation (yours of them).

Thirdly, how does the decision maker combine and intergrate all the sops from all the inputs from different people? Can we leave it to a simple summation process to arrive at the decision? Once again, sops are subjective scales of prefence, so it might be hard to compare. I think so far though, that addition of all the sops for a particular option, and see which one gets the highest sop, might be a solution. This, once again, will need to be tested out... but I'm optimistic that this may work.

2 comments:

FLuFFy_BuG said...

hmmm, that'd work well if ur options are orthogonally related to one another... if, for example, ur sops were 6, 5, and 3 - what's to say that maybe the 5 and 3 were very similar options, and that the person would really rather one of those 2 (jus the fact that there were 2 similar options prevented commitment to one or the other, hence the less preferred one wins)?!?

if, on the other hand, each value were independent (like, i'd rather this one than one of the other ones, all differences/similarities between options put aside) - then ur system would be perfectly fine...

i've got a visuo-spatial system that formed in my head abt sops as i read ur blog jus now, in which each option is represented by a point (isn't that interesting that "point O", commonly used to denote a centrepoint in geometry, is an anagram of "option"?!?) - and points that are similar to one another can be mapped close to each other, and those that are vastly different can be mapped further away... then, u draw a point at a weighted centre (minimise the sum total of distances (or squares of distances?!?) b/w the centre and the options), and draw a vector map that traces out the direction from the centre to each option by the magnitude defined by the sop attributed to each option, sum the vectors, and the option found to be closest to the endpoint is your final option!!!

this is good in that it can take into account everyone's opinions, and standardise it by saying that their total sops should add to a fixed number (e.g. 10)...

;p - but this is bad, coz it's jus complicated, and u don't wanna get out a piece of paper, writing implement and measuring instrument every time u want to make a decision!!!

for simplicity's sake, jus make all ur decisions dichotomous ;p

Gal said...

Oh, I guess then an important stipulation in the formulation of an individual choice sop, is that their sops for each choice should not be in any way influenced by the other options presented (i.e. things that add weighting to your sop in choice A shouldn't be related to B... e.g. don't decrease your sop for going to place A cos you sorta want to goto place B... just put in the reasons why you want to go to place A, quantify this, and express as a sop for A). The actual interactions of A and B (and any other options) come after you get your individual sops for the different options, which you then can compile and come up with your overall sop for one thing, which you tell to the decision maker.

Perhaps I need to make up some more vocabulary to distinguish individual choice sops, the calculated sop which is given to the decision maker, and the over sop decision for the whole group?