Optimizatioin graph

[Sir Swindle89]

I'm sure most of us have seen this image before
http://images.community.wizards.com/community.wizards.com/user/malkonnen/d4fcf0760c825a7e0d539bf93822f34b.png?v=201600
well you have now at least.

Has any one tried to actually make a graph like this?
If so what metric did you use(or if not would you use) to measure "Optimization" and "Power"?

[OrganicGolem]

I could feasibly make a graph that compares two classes, MAYBE 3. However once you have too many on the graph it gets a bit too impossible to compare class abilities that do opposite things.

[Coidzor]

Has any one tried to actually make a graph like this?
If so what metric did you use(or if not would you use) to measure "Optimization" and "Power"?

No, not really. It seems like something that was a waste of time.

[Doc Roc]

That graph is... I'm hoping it isn't how you imagine CO.

[Greenish]

That's a neat-looking graph, nice flowing lines and soothing colours, but I have no idea what it's supposed to tell anyone.

[Coidzor]

That graph is... I'm hoping it isn't how you imagine CO.

That's how some people view it, yeah. :smallannoyed:

[Tvtyrant]

And how is a low-op Warlock worse then a low-op rogue? It still has rays and flying to its name without any op-fu.

[Sir Swindle89]

And how is a low-op Warlock worse then a low-op rogue? It still has rays and flying to its name without any op-fu.

I assume all of the lines are fairly arbitrary.

The location of the box is that person's view on "optimizers". Which from an outside looking in perspective would appear to hold true.

[Greenish]

The location of the box is that person's view on "optimizers". Which from an outside looking in perspective would appear to hold true.So the point of that graph is to say that… optimizers optimize?

And claim that they do it as much as possible, all the time, it seems. :smallamused:

[Sir Swindle89]

So the point of that graph is to say that… optimizers optimize?

And claim that they do it as much as possible, all the time, it seems. :smallamused:

well i think it's more that they (we really) don't care how strong somthing is at less than it's fully abused level. granted that can't actually be true because of the generally understood arrangment not to involve pun pun in any serious discussion.


No, not really. It seems like something that was a waste of time. involvment in the hobby as a whole could be considered a waste of time. so don't be hate'in

[Kurald Galain]

And how is a low-op Warlock worse then a low-op rogue?
The graph depicts 4E classes.

Also, it assumes that "power" is defined by "damage per round", and it's obviously not based on solid data (since by definition you can't measure "optimization" to put it on a gradual axis like that).


So the point of that graph is to say that… optimizers optimize?
The point of the graph is to state that a certain class is weak compared to another if neither is optimized, but that the two are comparable in strength if both are optimized.

[Greenish]

The point of the graph is to state that a certain class is weak compared to another if neither is optimized, but that the two are comparable in strength if both are optimized.Well, that's a point well missed by posting this on the 3.5 section of the forum. :smallamused:

[ericgrau]

Well optimizers can also care about the left side. His point was that only optimizers care about the right box. I get a bit peeved when people get peeved at people who get peeved at optimization. They're obviously talking about excessive optimization, which means going above and beyond normal players (via intentional rules abuse or simply a much higher power level than their particular group) and nothing else. Stop trying to change the subject into something else (like normal optimization or trying your best) so you can argue about it. It can go too far, oh well, get over it.

Other than that the curves seem like fairly arbitrary lines drawn with some art program. So as long as you remember they're somebody's opinion and not 100% certain it could spark a discussion I suppose.

[Tyndmyr]

And how is a low-op Warlock worse then a low-op rogue? It still has rays and flying to its name without any op-fu.

And it has UMD as a class skill without any shenanigans. It would be fairly low op to look at random loot including divine scrolls/wands and think "hey, I can use that".

I don't understand this concept of "excessive optimization". The correct amount of optimization is that required to make the character fit the job. Some situations require more, some less.

[Sir Swindle89]

I don't understand this concept of "excessive optimization". The correct amount of optimization is that required to make the character fit the job. Some situations require more, some less.

But "Fit the job" is very vague, and how well you fit matters. a fighter can stand there a deal single target damage just fine, but a wizard can kill it faster. They both accomplish the same goal just one is more optimal than the other.

[Greenish]

But "Fit the job" is very vague, and how well you fit matters. a fighter can stand there a deal single target damage just fine, but a wizard can kill it faster. They both accomplish the same goal just one is more optimal than the other.Fighters are more efficient at single target damage, if given the opportunity.

Besides, max DPR ≠ optimal, in most any group.

[Yora]

As someone who actually took a couple on classes on statistics, I actually like that graph a lot.
Lable the axises A, B, and C instead and you have a very nice representation that neatly shows a percieved situation that is a bit difficult to explain in words. If the person you're explaining it to knows a bit about graphs.
But it's vital to leave the axises without units, or else it doesn't make any sense.

[Tyndmyr]

But "Fit the job" is very vague, and how well you fit matters. a fighter can stand there a deal single target damage just fine, but a wizard can kill it faster. They both accomplish the same goal just one is more optimal than the other.

Tier and "kill it faster" are not the same thing at all.

A wizard is absolutely a higher tier than a fighter. A fighter can still kill most things ridiculously fast.

No, the main reason to optimize a fighter is not "kill things faster"(though a bit of that is nice), it's to do things like give him flight and such. Or a dip into an initiator class for juicy maneuvers. Flexibility.

Fit the job entirely depends on group. If the group is capable of handling the encounters presented, and all players present have a niche that comes up routinely, so they share spotlight time, you're generally doing alright. If someones falling behind, add optimization until fixed. Overall level of optimization varies wildly by group, and there is no correct level.

[XianTheCoder]

As someone who actually took a couple on classes on statistics, I actually like that graph a lot.
Lable the axises A, B, and C instead and you have a very nice representation that neatly shows a percieved situation that is a bit difficult to explain in words. If the person you're explaining it to knows a bit about graphs.
But it's vital to leave the axises without units, or else it doesn't make any sense.

Can I ask where you took those classes?

As someone who took a LOT of Math/Stats/Programming/Engineering classes and works in a very quantitative Wallstreet job dealing with statistical models daily, I look at that graph and see absolute garbage art, not an actual graph meant to accurately model optimization.

For arguements sake.... what are you labeling A, B, C... and why on earth would you remove units and metrics from each axis? Since it's a 2 dimensional graph you basically have A and B (no C), and without units the person viewing it has no means of deremining incrimental differences at each point (which is basically the point of a line graph in the first place). At a minimum a unit of measure is needed to make the graph even readable.

[Yora]

not an actual graph meant to accurately model optimization.
Because it is neither an actual graph, nor does it accurately model anything.

It's just a graphical representation, using the visual style of an actual statistical graph to show a hypothesis. What this image says is not "optimization works like this", but rather "if it would be possible to quantify class balance and optimization and put those numbers into a graph, the graph would probably look like this".
But very true: It's not a graph. It's a doodle of a graph.

[true_shinken]

That graph is... I'm hoping it isn't how you imagine CO.
My thoughts exactly.
It does have some value if you consider Mystic Ranger with Sword of Arcane Order, but...
Oh, wait, forget about it.
I now remember, I saw this graph before, it's about 4e. And 4e strikers are really like that.

[Sir Swindle89]

I suppose you really could only do this with somting like 4E where you have classes that are supposed to do an role in a party, not just do everything/whatevertheywant.

Of course the problem with that is you forget about every thing else the class does other than, in this case, single target damage. Like the fact that a lock is about as controllery as he is strikery, thus making him the worst striker but still rather usefull in a group with few controllery types.

Also, to the apparent mathmeticians, I posted the chart up there as an example i don't think anyone would take a simple image seriously. even a well developed graph would need serious text justification to get looked at as anything other than a doodle.