Suggestions to Balance Twinking

[baddax]

No, it's completely pointless. The result would have no bearing on any suggestion that's been made.

I disagree.
Although not perfect it would give a rough idea as to the effect of the effect of gear vs. experience. Its possible for turbine to affect game play via loot tables and such but not when it comes to player skill level. You could even have them run the solo quests and time them and see how long it takes that way there is no party influence.

[Riorik]

No, it's completely pointless. The result would have no bearing on any suggestion that's been made.

An experiment to compare the effectiveness of player ability plus equipment would strike at the very heart of the matter being discussed. Experimentation is much of the very foundation of every bit of scientific knowledge known to humanity. Saying a result of an experiment "would have no bearing on any suggestion that's been made" is incorrect.

I can only agree with Aeyr. The entire point of his suggestion is empirical evidence that would provide support or withdraw it for the opinion that a mix of player experience within a given party is the primary driver, not equipment (or twinking which is the OP).

It's pointless only if you don't want to know the truth or wish to deny it by burying your head in the sand. Sadly, this seems to be most of both your and Borror0's points. Not to consider any other option with a focus on squashing disagreement by summarily & arbitrarily labeling them as FALSE.

Have you had any success with this approach in real life with anybody that wasn't required by other commitments to accept your decisions verbatim without debate?

[baddax]

No, it's completely pointless. The result would have no bearing on any suggestion that's been made.

Only if said suggestions had no true effect on twinked vs Non twinked game play.

[Borror0]

I disagree.
Although not perfect it would give a rough idea as to the effect of the effect of gear vs. experience. Its possible for turbine to affect game play via loot tables and such but not when it comes to player skill level. You could even have them run the solo quests and time them and see how long it takes that way there is no party influence.

If moving left, right, up or down and clicking right mouse button is hard for you (generic you, not directed at baddax), then you've got serious issues.

[Angelus_dead]

I can only agree with Aeyr. The entire point of his suggestion is empirical evidence that would provide support or withdraw it for the opinion that a mix of player experience within a given party is the primary driver, not equipment (or twinking which is the OP).

That is wrong in multiple separate ways. First, you have failed to keep track of what "suggestion" is actually under discussion.

Second, the pertinent question is "Twinked Vs Nontwinked". If one desires information on how important that is, the correct experiment would be to take an experienced player and see how quickly he can run lowbie quests while twinked, then take away his gear and see if he can do it again. The results of that test would be painfully easy to predict, especially when the character involved is not one with a class ability to restore hitpoint damage.

The proposed experiment was "Twinked unskilled Vs Nontwinked skilled". Any scientist would laugh at the idea of changing two different variables at the same time when the investigation is only interested in one.

And that's not even getting into the utter impracticality of conducting a test like that.

[Riorik]

If moving left, right, up or down and clicking right mouse button is hard for you (generic you, not directed at baddax), then you've got serious issues.

So that's your opinion? That "experience" refers to this?

That is wrong in multiple separate ways. First, you have failed to keep track of what "suggestion" is actually under discussion.

Are you seriously suggesting that it's only permitted to discuss the OP or only points that you choose? Is this somewhere in the forum rules here?
http://forums.ddo.com/announcement.php?f=180&a=4

[Angelus_dead]

Have you had any success with this approach in real life with anybody that wasn't required by other commitments to accept your decisions verbatim without debate?

Forum rules prohibit me from answering that question, just like they also prohibited you from asking it.

[Aeyr]

the pertinent question is "Twinked Vs Nontwinked". If one desires information on how important that is, the correct experiment would be to take an experienced player and see how quickly he can run lowbie quests while twinked, then take away his gear and see if he can do it again. The results of that test would be painfully easy to predict, especially when the character involved is not one with a class ability to restore hitpoint damage.

The proposed experiment was "Twinked unskilled Vs Nontwinked skilled". Any scientist would laugh at the idea of changing two different variables at the same time when the investigation is only interested in one.

And that's not even getting into the utter impracticality of conducting a test like that.

Impractical to conduct, of course. Yes, one could conduct an experiment with only a single variable changed- its the most elementary type of experiment. I actually encourage you to conduct such an experiment that you proposed and learn the results, not to simply "predict" what it will be and leave it at that.

Of course, DDO is a venue with many interrelated variables which are not so easy to separate much as we'd like to believe otherwise. As to a scientist laughing at the idea of changing two different variables at the same time, that is simply untrue unless the scientist has a very limited capacity to conduct the experiment. There is a simple method usable for such multi-variable experiments that utilizes "design of experiments" to allow the statistical evaluation of the direct and interactive effects of variables upon an output. But you don't even have to do that as a simple parametric multi-variable experiment would suffice. So yes, scientists do conduct experiments in which multiple variables are changed, and with a straight face too. We do it all the time. I must admit that laughing at experiment ideas rather than doing them could be more fun though.

[baddax]

If moving left, right, up or down and clicking right mouse button is hard for you (generic you, not directed at baddax), then you've got serious issues.

That is a gross oversimplification.

[Riorik]

Forum rules prohibit me from answering that question, just like they also prohibited you from asking it.

I doubt it. The possibility I could see is the catch-all/discretion of the forum moderators rule #1 - segment about provocation. If it were a violation, it would only prohibit you from talking about it as a violation...which you did...or quoting it...which you did. Mutual assured destruction doesn't excuse either, if it's actually a violation.

Of course, since I've taken so long to write this up while playing, I notice Aeyr's response to your "factual" claim that a scientist would "would laugh at the idea of changing two different variables at the same time when the investigation is only interested in one."

There is a simple method usable for such multi-variable experiments that utilizes "design of experiments" to allow the statistical evaluation of the direct and interactive effects of variables upon an output.

This just reads like something you pulled out of thin air while Aeyr may actually work professionally in a field that does this.

Smack?

[Borror0]

This just reads like something you pulled out of thin air while Aeyr may actually work professionally in a field that does this.

I assume that what Aeyr refers to are observational studies, and if that is the case, presenting it as a refutation of Angelus_dead's argument is extremely dishonest.

[Riorik]

I assume that what Aeyr refers to are observational studies, and if that is the case, presenting it as a refutation of Angelus_dead's argument is extremely dishonest.

/me sits back and watches the other side of the net.

At first glance I would have thought Angelus to be correct. Classes that I'd taken covering it were a LONG time ago and not high level stuff.
Reading this: http://en.wikipedia.org/wiki/Design_of_experiments
and/or this: http://en.wikipedia.org/wiki/Observational_study

The first two sentences of "Design of Experiments" seems to suggest that Observation Studies are a subset not the only option for "Design of Experiments".

This is like, so way beyond the topic of this thread...but, I guess, if you had a matrix of things you like to test - two or more variables...wouldn't you just work through each possible combination? If there were three or four, the grand total of tests would be higher...and that's not counting the total duplication runs where you're confirming behavior repeatedly.

[Borror0]

At first glance I would have thought Angelus to be correct.

He is. Comparing twinked experienced players to untwinked unexperienced players is not a good way to determine the effect of twinking.

It's as dumb as trying to determine the mathematical performances of two ethnicities by taking Asian kids and black adults. If your two sample groups are not comparable, then you'll have a flawed analysis.

[Riorik]

I'll rephrase - discussion of statistical analysis in the context of experimentation ... if you go to his reference, "design of experiments" mentions observational studies as a possible use, but not the only use.

There is an assertion that this somehow makes it dishonest.

I don't think he's really saying what you claim he's saying. I think he's saying...if you have two variables each with two options, you test at least four times...all possible combinations. Even if you think a data point is irrelevant, you still do it - you still test it because your intent is to PROVE that it's something you shouldn't use, not just assume it's something you shouldn't use.

That's a completely different situation.

So, what are you saying? Why is it specifically dishonest? You've picked a topic about which it's really kinda black or white. It either is or it isn't and you've done so with a rather broad brush stroke. On the surface, I think I disagree but I'm open to more information.

This fun...edit is I was reading more.
Found this on wikipedia about parametric statistics -- which is what he mentioned in the next sentence.
Multi-variable, just assembling the phrase into "parametric multi-variable" seems to mean, a parametric statistical analysis using multiple variables.

Dude, this doesn't look like Intro to Statistics, to me.

[Borror0]

I'll rephrase - discussion of statistical analysis in the context of experimentation ... if you go to his reference, "design of experiments" mentions observational studies as a possible use, but not the only use.

There is an assertion that this somehow makes it dishonest.

To reuse the example I gave earlier:
Angelus_dead's argument: You can't compare Asian kids and black adults, if you want to determinate the effect of ethnicities on mathematical performances.
Proposed counterargument: False. You can compare Asian kids, Asian adult, black kids and black adults, to determinate the effect of ethnicities on mathematical performances.

It is dishonest to present the comment as a counterargument since it does not invalidate Angelus_dead's original argument.

[Aeyr]

I assume that what Aeyr refers to are observational studies, and if that is the case, presenting it as a refutation of Angelus_dead's argument is extremely dishonest.

As much as "observational studies" are presented early on the wiki information for the layman, its entirely flawed to either assume a designed experiment must be one or that such an experiment would be "extremely dishonest." The accurate depiction of a real applicable experiment would be to a) think of any and all variables that you think might have an effect on something (for which there is almost always many that do), then b) test out a wide variety of settings for each variable of primary interest, and then c) analyze the outcome of every permutation tested and use mathematical tools if necessary to determine the probability that each variable or combination thereof effects the output. In fact, its ideal to test every single possible combination of every possible variable many times to get the most accurate results, even those variables which the experimenter might be uncertain even has an effect on the output.

Further, nobody is presenting the idea of conducting an experiment as a "refutation of so-and-so's argument." The real message is to run a well-planned experiment, gather data, analyze that data, and then make conclusions based on the findings. That's how you really make a convincing argument. Saying its a dumb idea and that someone should just believe you when there is no real data to support the claim is the questionable path.

[Riorik]

I edited this into the post after reading Aeyr's comments.

Borror0, I think you just stepped in it. I think you just challenged a subject matter expert compared your to wikipedia level gleanings on statistics and my BS detector is off the scale.

I am, however, going to put an end to my involvement in your thread.

[eonfreon]

WOW.

Never mind.

Can't post anything that would make any sense until I catch up that's for sure .

[eonfreon]

Maybe I'll run one of these experiments.

I choose:

A fully Twinked 1st Level in the Shroud with 1 16th Lvl Piker (so I can get in in the first place).

VS.

A Naked 16th Level Toon in WW.

My prediction:
The Naked 16th Level in WW will survive longer than the Twinked 1st Level in the Shroud.
This should prove once and for all that gear isn't as important as a well-built character.

What?
Too many variables?

[Borror0]

I think you just challenged a subject matter expert compared your to wikipedia level gleanings on statistics and my BS detector is off the scale.

LOL! Cute edit.