JWPlatt wrote:

Ok. Victor Laxman departs with Marie Sutherland.

(Actually, not really shocked. I figured as much with a title like "Exodus" which is why I was trying so hard to get the meeting before this episode's end.)

I'm going on faith that Mr. Laxman is a man of his word. We'll meet again.

Hey, JW... did I miss something? Are you saying Laxman and Marie have left also? I have looked, but can't find anything on the various forums that confirms, if that is what your post means.

Dadguy

EDIT: Heheh...just logged on and saw the DRC KI Mail from Laxman. I hadn't seen anything on the "Cavern Today" when I posted, or here in the Mystonline forums. There is a UO post now but it wasn't there a short while ago.

For those few of you that didn't know: Laxman sent out a DRC KI mail tonight that says he and Marie are going to the surface to look for more funding, especially after the Yeesha visit, they are more convinced they must "find a way".

_________________
Member of the Guild of Greeters

JWPlatt wrote:

I'm going on faith that Mr. Laxman is a man of his word. We'll meet again.

Then maybe you'll get a chance to ask him for the back data he has squirreled away. That would make the job of doing a decent analysis a lot easier than starting ab initio and getting only a weeks worth.

I have two big problems with what Laxman was telling you.

The first is this weird fixation on "spikes". (And now "exponential spikes".) The only way for spikes to be a problem is if the points made on the day of a spike did not benefit the lake as much as points made on other days, and if the amount of decrement in this benefit depended on the magnitude of the spike. Adjusted, of course, for the effects of the mysterious D'ni safety mechanisms. Nonsense. OOC, it could be done with enough effort, manually or even programmatically, but I seriously doubt that Cyan is actually putting in that effort. IC, how could Laxman possibly know this without some way of measuring the lake effect? Yet he says the dock dalek is dead and all he is looking at is what is going into the lake. What is the point of this implausible complication?

The second problem is Laxman's continuing insistance that the numbers are "arbitrary". According to my dictionary, the meanings of that word include "based on personal whim", "random", and "assigned no specific value". How do you measure something with "arbitrary" numbers? Yet Laxman claims the numbers are "useful" to him, while we peons would undoubtedly see implications in them that aren't there. More nonsense.

If someone came to my office and told me that they were tracking a process, that they were measuring the inputs to the process with arbitrary numbers, and that they were not able to measure the outputs of the process and correlate it with the inputs, I would be wondering just what they thought they were accomplishing and who turned them loose from the asylum. If they then pronounced themselves satisfied with the level of inputs, I would probably die laughing before calling the nice men in the white coats. But that's the balloon juice we're getting here.

This project could have been a wonderful thing for MOUL, with lots of opportunity to tie the activity into some knowledge of the Cavern ecology, revelations about D'ni history and technology, etc. It just needed a little thought put into making the background at least semi-plausible for an audience that probably averages a bit higher on the IQ scale than the audience for, say, GTA. But I think it has descended into total insanity and is past the point now where anything good can be salvaged from the mess. I'm very, very disappointed.

_________________

*SLMW 1.0* No animals were harmed in the production of this message.

BrettM

I totally agree with you


But (OCC) Unfortunalely Cyan presented the Lake Project to us unprepaired.
With no material on hands to make it convinving and interesting.

To include all the amount of details ( extra backgtound story, artefects, datas, images etc) that people are expecting it will take much more time, ressources and money...

To bad

BrettM wrote:

The second problem is Laxman's continuing insistance that the numbers are "arbitrary". According to my dictionary, the meanings of that word include "based on personal whim", "random", and "assigned no specific value". How do you measure something with "arbitrary" numbers? Yet Laxman claims the numbers are "useful" to him, while we peons would undoubtedly see implications in them that aren't there. More nonsense.

I believe it is Mr. Laxman's lack of scientific brainwashing - er, I mean, scientific training that is causing confusion over the term "arbitrary." What Laxman is probably trying to say is that the units are not meaningful, rather than the "number" - that is, an amount expressed in terms of those units. If Laxman had published a number such as 123,678,987, an endless debate would have erupted over what that number means. For myself, endless debates over large numbers are probably more entertaining than endless debates over mindless slogans like Find a Way, Make a Home, Call if you're going to be Late for Dinner, etc., but to each his or her own.

All of this nonsense could have been avoided if the DRC had decided on using an index approach, which is what it appears they think they were doing by putting 100% on the vertical axis. As long as there is a one-to-one relation between the index number and whatever real number underlies the data, no information is lost and we avoid the endless debate. But by presenting us with two graphs based on essentially two different index numbers - actually, six, taking into account the three different averages in each graph - the DRC has revealed the utter inadequacy of their alleged scientific training.

"The only way for spikes to be a problem is if the points made on the day of a spike did not benefit the lake as much as points made on other days, and if the amount of decrement in this benefit depended on the magnitude of the spike. Adjusted, of course, for the effects of the mysterious D'ni safety mechanisms. Nonsense. OOC, it could be done with enough effort, manually or even programmatically, but I seriously doubt that Cyan is actually putting in that effort."

Why do you doubt it? It's not really harder for them to track "smoothness" than to track a total sum value. It could be as simple as the standard deviation of the daily totals, taken a week at a time. (Those are, after all, the periods that the DRC charts display.)

Obviously there are all sorts of numbers they *could* be tracking. But I am willing to listen to what Cyan says, rather than assuming off the cuff that they're lying to us!

Is it illogical for them to use a "no spikes" standard of progress? No, it is quite plausible. A major problem in any kind of on-line grinding game is trying to guess how much effort players will put in. You can pick a number ("1000 players will do 2 batches day, so let's scale it to finish up in March") and then find that 3500 people are working 4 batches a day. Then you have to either arbitrarily change the rule (bad -- see "Great Zero") or tell your designers to create a new Cavern lighting model by Tuesday (bad -- see "mass office suicide").

Plus, you wind up rewarding players for being online twelve hours a day. (Very bad. See "World of Warcraft.")

"No spikes" is a plausible solution to this problem. By looking for a consistent level of effort, you are measuring organization, rather than grinding per se. If player interest in the project flares up unexpectedly -- or dies down unexpectedly -- it doesn't throw off your estimated timeline. There might be an indirect effect due to the law of large numbers (more contributors => smoother curve, by simple randomness) but it will still primarily be a matter of how steady and reliable people are, rather than how many people are slaving in the salt mines.

(I am not here arguing that Cyan *is* behaving this way, or for this reason. I'm only saying that it is not an absurd design decision.)

_________________
Andrew Plotkin (KI#00102725)
Uru Newcomer's Guide and Answer Sheet
-- http://eblong.com/zarf/uru/newfaq.html

BrettM wrote:

But that's the balloon juice we're getting here.

I was following all the arguments and debate until I hit Brett's comment.

Sorry. I'm really sidetracked now. Nothing else matters. Balloon juice

So many questions now.

_________________
They told me I was gullible, and I believed them.

belford wrote:

(I am not here arguing that Cyan *is* behaving this way, or for this reason. I'm only saying that it is not an absurd design decision.)

I see your point, but it doesn't really comfort me. What such a choice would mean is that not all pellet points are created equal. The points someone made today might have a greater effect or a lesser effect than the same number of points made yesterday, which would render the sorry feedback mechanism we have even more useless, as we would have no idea of how much the points actually counted. If this is the case, even lake-score numbers provided by Laxman would suffer as a feedback mechanism, as they are not adjusted yet for the spikes and we have no clue as to the adjustment method being used.

It also doesn't comfort me because it would mean that the project is on a pre-determined schedule (more or less, anyway), and the results of our actual efforts will be diddled as necessary to fit that schedule, making them irrelevant (or less relevant, at least) to the final outcome. Which is what upset people about the Great Zero calibration.

@Teknobubba: You haven't heard that expression before? Balloon juice is the stuff used to inflate balloons: hot air.

_________________

*SLMW 1.0* No animals were harmed in the production of this message.

Quote:

I see your point, but it doesn't really comfort me. What such a choice would mean is that not all pellet points are created equal.

I don't see that as bad, *if* it is still an interesting challenge to work optimally.

Quote:

It also doesn't comfort me because it would mean that the project is on a pre-determined schedule

That is not what I was suggesting.

To be more precise, I was imagining a system which would end at a particular time *if everyone cooperated perfectly*, regardless of how many players were involved or how many hours they worked. If the pellet output was non-optimal -- spiky -- then the process would take longer.

Or, indeed, it could be a weighted mix of "total output" and "smoothness"; and Laxman is telling us that the smoothness is the bigger problem right now.

_________________
Andrew Plotkin (KI#00102725)
Uru Newcomer's Guide and Answer Sheet
-- http://eblong.com/zarf/uru/newfaq.html

_________________
Grand Master
Guild of Archivists

Mr. Laxman is indeed a man of his word. Despite his departure from the Cavern, he has seen to having the lake effect pellet numbers he promised in our meeting on 11/4/07 forwarded to me for distribution. The numbers for 11/5/07 are now posted on the Pellet Points hood imager. I will arrange the same for other interested hoods.

These numbers should not be seen as a call to increase or decrease your pellet production and contributions to the lake. They are meant to inform and provide for analysis. He is currently satisfied with how we are doing and says, "I'm not interested in trying to get larger numbers to participate artificially. If they aren't interested in contributing for the sake of contributing, then it's only more likely to encourage even larger spikes as they gain and lose interest." He just wants to keep it steady and true.

I ask that the daily data be distributed only through available means within the Cavern at least until the monthly graph is released.

_________________
OpenUru.org: An Uru Project Resource Site : Twitter : Perfect Speed Is Being There.

The numbers for 11/6/07 are now posted in the Pellet Points hood and any other hoods which request it.

I should explain to those who have asked that Dr. Watson is collecting the pellet data in Mr. Laxman's absence.

_________________
OpenUru.org: An Uru Project Resource Site : Twitter : Perfect Speed Is Being There.

JWPlatt wrote:

I should explain to those who have asked that Dr. Watson is collecting the pellet data in Mr. Laxman's absence.

Ahh... I wondered how it was happening. So Watson didn't leave, eh? How come we didn't hear from him? Or did I miss something?

_________________
Nothing to see here, move along.

I looked at the numbers and they don't say much more to me than the graph does.
I'd be happy to hear an explanation/analysis here or IC (since you wanted to keep the numbers out of the forums).

_________________
KI: 59666
http://www.uruobsession.com/

BrettM wrote:

The first is this weird fixation on "spikes". (And now "exponential spikes".)
. . . . . .
The second problem is Laxman's continuing insistance that the numbers are "arbitrary".

It seems pretty simple to me: the magnitude of the pellets are not what's important, it's the rate. There were spikes before that were hindering the process, and recent efforts to push immense numbers of pellets in the lake at once are ironically holding up the process.

So: what we need is a positive and constant d(pellets dropped)/dt. Sure, the pellet scores are arbitrary, but the relationship between them is not. The data clearly has correlation, and if they are not releasing data past a month or so, I would bet that the data past that point is having a negligible effect on the lake; as in old values are discounted toward the overall health of the lake, possibly completely ignored.

If you think about it, this is a much more interesting challenge than "hey, repeat this task a billion times and the lake'll glow". This will require cooperation on a game-wide scale. Surely there's some group - or mayhaps a guild - that could help orchestrate this. . .

_________________
Hover Conversion Industries - Division 17
Relavahn --- KI:06877562

Junee wrote:

I looked at the numbers and they don't say much more to me than the graph does.
I'd be happy to hear an explanation/analysis here or IC (since you wanted to keep the numbers out of the forums).

We're going to need a few more numbers before we can do anything with them at all, since two days isn't much to work with. We're also going to have to make a few assumptions about how the numbers were derived, since we have not been given any information on this, and none of the data that went into the calculations.

For example, it is an assumption that the "daily average" is the total number of points for that day divided by the number of explorers dropping pellets that day. (Before you say "Duh, that's obvious", think about what else it might mean in the mind of someone who thinks the numbers are arbitrary, and that a percentage scale is a good method of indexing the results, and that truncating the vertical axis does not encourage errors of interpretation. ) But, we don't know the number of explorers, so what do we do with this information?

What about the "total average"? Is that the total number of points since the beginning divided by the number of days? If so, what is the time frame? (The "lake points" were apparently introduced on July 26, but Er'cana was released a month before that.) Or is it the average of the daily averages, expressed in per-explorer terms?

This is going to take some head scratching.

_________________

*SLMW 1.0* No animals were harmed in the production of this message.