Whats better?

[FailureMidgard]

I'm giving up only because Thorny feels like he has some kind of pride or e-peen to defend when all I said was "/DRG isn't as bad as people make it out to be". Not once did I say it's better than /WAR, and in certain situations it's comparable (situation being that the MNK doesn't have the most top end gear available, which is most of the people).

Oh and while you don't take significantly more damage when using counterstance with berserk up, a /DRG doesn't have to make that choice does he? Plus, nobody MNK/WAR at all these days. The point is, /DRG is trading what minimal performance difference for some more survivability.

That's that.
Oh and I bet Thorny haven't even tried MNK/DRG before.

PS: It's great for conflict.

I'll say I've never tried either, as I don't have monk leveled, but do you really think 1 small hate shed is really worth talking about that much? I agree "it's not that bad", but it is still worse in virtually every way possible. I mean, a NQ Amemet Mantle is really "Not that bad" also...in fact, it is an amazing buy for how cheap it is, yet it's frowned upon because an Amemet+1 is still pretty cheap (all things considered) and is, in every way, better.

[el_tio]

The fact is, the more DA you add, the less beneficial it is to you to choose DA over another useful stat.

This is precisely why DA is a constant return. It's not that DA gets less beneficial the more you have, it's that other stats get more beneficial the more DA you have while DA stays the same, which is why other stats become more appealing alternatives the more DA you have. The more you swing the more benefit from hitting more accurately, etc.

Take acc vs. atk. If you model it strictly as a % increase, adding 5 attack to 400 atk when you're at 25% accuracy provides the same benefit percentage-wise as when you have 50% accuracy. Yet obviously you're getting more advantage from attack when you have 50% acc than when you have 25% acc.

If you model the same comparison using marginal returns (how diminishing returns etc. is truly defined) you'll see the increasing return from attack as acc increases. Accuracy is another stat that's conventionally called diminishing returns in FFXI-speak but it's another constant return.

Either model works overall, because if you look at it percentage-wise the point where the percentage increase of a stat "diminishes" below the worth of another stat, that point should coincide where the marginal return of other stats has increased beyond the stat you've already raised.

I prefer the DOT model over the DOT% model simply because it more closely reflects what is actually being measured--DOT. Of course % increase is often simpler for gear comparisons though.

[Jeho]

Plus, nobody MNK/WAR at all these days.

I lol'd

I should clarify a bit. Skilled MNKs with full merits will go to merit parties from time to time as /WAR. And yes, these MNKs usually have top-tiered gear as well.

So what I believe you meant to say is most MNKs will not go /WAR because they are either gimp in merits or gear (pretty much both) or they kinda suck at their job (which is el oh el seeing as how it's lolMNK in the first place).

and;

Before you toss out a "nou" or something like along the lines of that, you said "at all". Now since this is based on an opinion and not fact I wanted to make that clear. Since you don't know every MNK that plays this game on your server alone, you can not empirically say that "nobody goes MNK/WAR at all these days". It's just not true.

[FailureMidgard]

This is precisely why DA is a constant return. It's not that DA gets less beneficial the more you have, it's that other stats get more beneficial the more DA you have while DA stays the same, which is why other stats become more appealing alternatives the more DA you have. The more you swing the more benefit from hitting more accurately, etc.

Take acc vs. atk. If you model it strictly as a % increase, adding 5 attack to 400 atk when you're at 25% accuracy provides the same benefit percentage-wise as when you have 50% accuracy. Yet obviously you're getting more advantage from attack when you have 50% acc than when you have 25% acc.

If you model the same comparison using marginal returns (how diminishing returns etc. is truly defined) you'll see the increasing return from attack as acc increases. Accuracy is another stat that's conventionally called diminishing returns in FFXI-speak but it's another constant return.

Either model works overall, because if you look at it percentage-wise the point where the percentage increase of a stat "diminishes" below the worth of another stat, that point should coincide where the marginal return of other stats has increased beyond the stat you've already raised.

I prefer the DOT model over the DOT% model simply because it more closely reflects what is actually being measured--DOT. Of course % increase is often simpler for gear comparisons though.

And again, if you read my post, all you're debating is the technicality. The only time the term "Diminishing Returns" should not apply, is if someone is saying "I don't want to spend X gil for a 5% DA rate because DA has DR and I already have 25%". Because any time you are comparing gear, you are looking at it relative to something else. If you suddenly add Fighter's Roll to a merit party, Pole Grip -does- get worse relative to the other options available.

I really don't care what model you use, but trying to tell someone that accuracy, attack, or DA does not have diminishing returns is incorrect at worse, and misleading at best. For all intents and purposes, all of those stats do suffer from DR BECAUSE they are linear on their own. Finally, your acc vs attack makes no sense, because again you are only looking at 4 free attack, not at X attack vs Y accuracy vs Z haste, which is what the situation ALWAYS IS WHEN YOU ARE COMPARING GEAR.

[Sanjiyan]

Plus, nobody MNK/WAR at all these days.

That is meant as an exaggeration to the current state of the game. It's akin to saying "nobody chi blasts these days" but there are still MNKs that do. MNK/WAR at meripo with full time counterstance is only viable on birds, and/or with outside party healing ala Rukenshin-styled meripo. In terms of meripo situation across the game, that's pretty much as common as chi blasting MNKs. And guess what? Nobody chi blasts these days.

[Kroot]

Please don't mention warcry in a MNK discussion. No monk is using it since better JA Boost doesn't overwrite it, which any monk will be using about every 30 secs for a WS.

warcry only lasts 30 sec, is beneficial while tping, has no negative effect, and also helps boost other players stats so I don't see a reason not to use it besides being lazy

And guess what? Nobody chi blasts these days.

people no longer use chi blast as a main source of dmg on HNMs
people still use it for random pulling/claiming and mainly use it to inhibit tp gain on mobs with nasty tp moves.
25% less tp gain is pretty awesome

[el_tio]

warcry only lasts 30 sec, is beneficial while tping, has no negative effect, and also helps boost other players stats so I don't see a reason not to use it besides being lazy

If you don't call more enmity than a voke a negative effect, then sure.

[Thorny]

I'd like to jump in between Thorny and Sanjiyan.

First off, overall, I agree with Thorny, but sorry Thorny, saying "DA doesn't have Diminishing Returns" is a kinda silly thing to say. It really depends on your definition of DR.

If you definite DR as "the return you get as opposed to other stats diminishes", then indeed, it has DR.

If you define DR as "the amount of damage you get from 1 point of the stat", then there is no DR.

It's really irrelevant which way you look at it though. You -ARE- making sacrifices in other stats to gain DA. The amount of damage you gain per DA is irrelevant unless you look at the amount of damage you gain per other stat you gave up. Rather than looking at the damage per DA%, look at the "accuracy needed to equal the damage gained from 1% DA", and you will see it as diminishing returns.

The fact is, the more DA you add, the less beneficial it is to you to choose DA over another useful stat.

If you're calling the same increase per point of DA diminishing returns, then what has constant returns exactly? Adding the same amount per point will always do relatively less for you as you get more of it, that's obvious. For something to have constant returns, does it need to scale as you get more of it exactly proportional to how much more of it you got to keep it a standard % increase per point? That makes no sense. If you draw a graph of your damage increase on the Y-axis and your amount of DA on the X axis, it's a straight line. Compare it to haste, which has increasing returns, which is a slow upwards curve.. decreasing returns would be a downward curve >_> I've explained it through linguistics, semi-chart, and now graph examples.. if people still don't get it I'm not really sure what else I can do.

[Kroot]

If you don't call more enmity than a voke a negative effect, then sure.

it is 300 VE per person so most likely half a provoke and it can be used before you engage a mob so it will not count.

[FailureMidgard]

If you're calling the same increase per point of DA diminishing returns, then what has constant returns exactly? Adding the same amount per point will always do relatively less for you as you get more of it, that's obvious. For something to have constant returns, does it need to scale as you get more of it exactly proportional to how much more of it you got to keep it a standard % increase per point? That makes no sense. If you draw a graph of your damage increase on the Y-axis and your amount of DA on the X axis, it's a straight line. Compare it to haste, which has increasing returns, which is a slow upwards curve.. decreasing returns would be a downward curve >_> I've explained it through linguistics, semi-chart, and now graph examples.. if people still don't get it I'm not really sure what else I can do.

You're still set on graphing damage vs DA, for whatever reason, when that is useless. Looking at a graph of Damage vs DA shows you absolutely nothing on it's own, about what DA is doing for you. Let's stop using the term "Diminishing Returns" and instead say that the gain from the double attack becomes less worthwhile.

When you look at a piece of gear, you choose between X DA and Y Accuracy. Let's say it's 2 accuracy (1% hit) vs 2% DA (grip).

If your base accuracy is say 50%, with 0% base DA, the 2% DA will add exactly 1 hit per 100 swings. The 1% accuracy will add also 1 hit per 100 pre-DA swings.
If you go up to 50% DA rate base, 2% DA will still add exactly 1 hit per 100 swings, and yet 1% accuracy will now add 1.5 hits per 100 pre-DA swings.

Assuming a constant Accuracy, the 2% DA item becomes worse when you go from 0%>50% base DA rate. Or more accurately, WHAT YOU HAVE TO GIVE UP BECOMES BETTER. Who cares about terminology, that is literally all you are arguing. It is 100% true that if you add more DA, then accuracy becomes a stronger stat as opposed to DA. On the other hand, as you add more accuracy, DA becomes a stronger stat as opposed to accuracy.

If you really want to graph it, you should be looking at a 3dimensional graph, NOT 2dimensional, because the damage formula is in NO way 2 dimensional. This is a multi-variable equation, treating it any other way is misleading at best.

[Thorny]

You're still set on graphing damage vs DA, for whatever reason, when that is useless. Looking at a graph of Damage vs DA shows you absolutely nothing on it's own, about what DA is doing for you. Let's stop using the term "Diminishing Returns" and instead say that the gain from the double attack becomes less worthwhile.

When you look at a piece of gear, you choose between X DA and Y Accuracy. Let's say it's 2 accuracy (1% hit) vs 2% DA (grip).

If your base accuracy is say 50%, with 0% base DA, the 2% DA will add exactly 1 hit per 100 swings. The 1% accuracy will add also 1 hit per 100 pre-DA swings.
If you go up to 50% DA rate base, 2% DA will still add exactly 1 hit per 100 swings, and yet 1% accuracy will now add 1.5 hits per 100 pre-DA swings.

Assuming a constant Accuracy, the 2% DA item becomes worse when you go from 0%>50% base DA rate. Or more accurately, WHAT YOU HAVE TO GIVE UP BECOMES BETTER. Who cares about terminology, that is literally all you are arguing. It is 100% true that if you add more DA, then accuracy becomes a stronger stat as opposed to DA. On the other hand, as you add more accuracy, DA becomes a stronger stat as opposed to accuracy.

If you really want to graph it, you should be looking at a 3dimensional graph, NOT 2dimensional, because the damage formula is in NO way 2 dimensional. This is a multi-variable equation, treating it any other way is misleading at best.

I'm not comparing to accuracy or other variables that compound with DA though :/ I'm only clarifying that DA doesn't have diminishing returns because it gives the same benefit each point, not trying to say anything else. I'm well aware that everything else is comparatively better the more DA you have, I'm not stupid. You're not conveying anything new to me, I'm just trying to illustrate that DA has constant returns.

Suppose that one kilogram of seed applied to a plot of land of a fixed size produces one ton of crop. You might expect that an additional kilogram of seed would produce an additional ton of output. However, if there are diminishing marginal returns, that additional kilogram will produce less than one additional ton of crop (on the same land, during the same growing season, and with nothing else but the amount of seeds planted changing). For example, the second kilogram of seed may only produce a half ton of extra output. Diminishing marginal returns also implies that a third kilogram of seed will produce an additional crop that is even less than a half ton of additional output. Assume that it is one quarter of a ton.

Read it, apply basic logic, ???, profit

[FailureMidgard]

I'm not comparing to accuracy or other variables that compound with DA though :/ I'm only clarifying that DA doesn't have diminishing returns because it gives the same benefit each point, not trying to say anything else. I'm well aware that everything else is comparatively better the more DA you have, I'm not stupid. You're not conveying anything new to me, I'm just trying to illustrate that DA has constant returns.


Read it, apply basic logic, ???, profit

Yet, he was saying it has diminishing returns (presumably meaning what I'm saying), then you tried to call him wrong based on a technical definition of diminishing returns. It is being compared to accuracy, from a different subjob (which also suffers from the same "Diminishing Returns" or whatever you want to call it).

I fully understand the idea of diminishing returns, and the only thing that has any "true" diminishing returns in this game (afaik) is int:spelldmg, and skill:accuracy. That being said, for all intents and purposes, every stat besides haste has some sort of diminishing returns. Also, in your quoted example there, you could take it further. What if when buying that seed, if you buy 2 kilograms, you get a discount of X%, which now gives you more crop per dollar still than only buying 1, regardless of the diminishing returns. There is diminishing return to the crop per kilogram of seed, but there is no diminishing return (until a higher threshold is reached anyway) in the crop per dollar spent.
Furthermore, assume that some biologist finds some way to mess with the seed and make it so it no longer has diminishing returns. But this costs 20x more per kilogram of seed to manufacture. Saying "it has no diminishing returns", while correct, is both misleading and irrelevant, which seems to be analogous to what you are arguing, right now.

[Sanjiyan]

Assuming 100% acc for 100 attacks.
0 DA = 100 attacks.
10 DA = 110 attacks, or 10% improvement over 0 DA.
20 DA = 120 attacks, or 20% improvement over 0 DA, but 9% improvement over 10 DA.
30 DA = 130 attacks, 8.3% over 20 DA.

Every point you add gives the same number of increased attack rounds when compared to baseline of no DA, but less improvement over a previous DA build. The more DA you previously had, the less the next point of DA is going to do for you in terms of overall damage. Thus, diminishing returns.

/DRG w/ brutal is 5 DA.
/WAR w/ brutal is 15 DA, or 1.15/1.05 = 9.5% improvement.

I think it's already been posted before.

[Suiram]

I fully understand the idea of diminishing returns, and the only thing that has any "true" diminishing returns in this game (afaik) is int:spelldmg, and skill:accuracy.

While I agree with your general point that people might be too pedantic about the use of "diminishing returns" with respect to FFXI calculations, I just want to point out that the two examples you gave aren't quite true. INT:spell damage is the same "linear increase but lower relative % improvements in damage" as DA and other stuff. Likewise, skill:accuracy shouldn't really be called diminishing either, since even though it becomes 10 skill:9acc at 200, it's still a piecewise-linear function.

A good example of true diminishing returns would be MDB (I think Valyana first observed this in a different thread that had this derail), since you can look at the % damage reduction as MDB/(100+MDB). But again, I personally have no quarrel with people who call both instances diminishing returns, and have used it myself on many occasions, although I do still think calling the above functions "bad linear" or something is a nice compromise.

[FailureMidgard]

While I agree with your general point that people might be too pedantic about the use of "diminishing returns" with respect to FFXI calculations, I just want to point out that the two examples you gave aren't quite true. INT:spell damage is the same "linear increase but lower relative % improvements in damage" as DA and other stuff. Likewise, skill:accuracy shouldn't really be called diminishing either, since even though it becomes 10 skill:9acc at 200, it's still a piecewise-linear function.

A good example of true diminishing returns would be MDB (I think Valyana first observed this in a different thread that had this derail), since you can look at the % damage reduction as MDB/(100+MDB). But again, I personally have no quarrel with people who call both instances diminishing returns, and have used it myself on many occasions, although I do still think calling the above functions "bad linear" or something is a nice compromise.

I was always under the assumption the skill thing was actually diminishing returns and not piecewise, but was approximated at .9, my bad then there (though, it seems pretty hard to prove it is one or the other in this game...). As for int, again, I was incorrect in how I thought it worked I guess, and thought that you hit "int tiers" or something of the sort the more int you gained.

[Suiram]

I was always under the assumption the skill thing was actually diminishing returns and not piecewise, but was approximated at .9, my bad then there (though, it seems pretty hard to prove it is one or the other in this game...). As for int, again, I was incorrect in how I thought it worked I guess, and thought that you hit "int tiers" or something of the sort the more int you gained.

Well, two quick things. The 10:9 skill:acc post-100 thing was proven using /check to get exact accuracy changes, not through parsing or something, so it is pretty well-founded. Also, you are technically correct, there are tiers at higher dINT for nukes, but it's one of those things that is never reached in practice (literally only seen on T1 and early T2 spells on lower level mobs, banish spells and bio spells), and it too is a piecewise linear function (but it eventually caps, so at some point you literally get no returns, but until then it's linear).

I wish I had taken more economics or statistics class, because I'm having a hard time finding an exact definition of "diminishing returns". For instance, with these linear functions the relative increases go to 0 as the input grows, and I would think that was what is meant by "marginal returns," but I guess it means something else.

[Thorny]

Assuming 100% acc for 100 attacks.
0 DA = 100 attacks.
10 DA = 110 attacks, or 10% improvement over 0 DA.
20 DA = 120 attacks, or 20% improvement over 0 DA, but 9% improvement over 10 DA.
30 DA = 130 attacks, 8.3% over 20 DA.

Every point you add gives the same number of increased attack rounds when compared to baseline of no DA, but less improvement over a previous DA build. The more DA you previously had, the less the next point of DA is going to do for you in terms of overall damage. Thus, diminishing returns.

/DRG w/ brutal is 5 DA.
/WAR w/ brutal is 15 DA, or 1.15/1.05 = 9.5% improvement.

I think it's already been posted before.

Did you even read the wiki example of diminishing returns? Each point does comparatively less for you because you're doing more, but DA is linear and diminishing returns implies each point gives less than the previous, similar to how each point of haste gives more benefit than the last due to increasing returns.

[Sanjiyan]

They do give less. You don't play this game in a vacuum. You are either comparing the damage to yourself, or to other people. You don't compare it to nothing. Just by that virtue, DA does give less performance for every point you add.

But you don't seem to understand that, so forget it.

[Thorny]

They do give less. You don't play this game in a vacuum. You are either comparing the damage to yourself, or to other people. You don't compare it to nothing. Just by that virtue, DA does give less performance for every point you add.

But you don't seem to understand that, so forget it.

I've understood that since well before this thread existed, but you're using the wrong term for it and refusing to read anything I post. Increasing returns = upwards curve, decreasing returns = downwards curve, constant returns = straight line ... simple enough, yet you can't understand or won't admit that you're wrong. I've said at least a half a dozen times that it does COMPARATIVELY less for you the more you have, but each point adds the same value, meaning it has constant returns.

[Suiram]

Haste and "magic damage taken" are technically linear too. If a mob does 1000 magic damage to you, and (pretending here that it's not actually in 256ths) you add magic damage taken -10% and take 900 damage, a 10% decrease. You add another 10% and take 800 damage, another 10% decrease off your original 1000. The difference is all in the comparative benefits (except with MDB, where the function is itself non-linear).