Soft Stat Caps in ROI?

[TicklerTurin]

I've tried searching for this information, but has anyone identified if there are soft caps with diminishing returns in ROI? Ex: Other than the offense cap, is there a point where might essentially stops modifying offense rating for champs and it would be better to go for vitality or something else? If someone in beta has this information or knows where it is I would really appreciate it.

[Graalx2]

We removed the cap on Offence. There is no reachable (probably) point in which the contribution of your primary stat to Offence is lowered.

There ARE diminishing returns built into all defensive ratings.

[ANewMachine]

We removed the cap on Offence. There is no reachable (probably) point in which the contribution of your primary stat to Offence is lowered.

There ARE diminishing returns built into all defensive ratings.

Any chance of us getting the formula for that? I mean, we'll derive it eventually, but some of the numbers we've gotten are... odd. Like for avoidances, going 3K->6.9K gives *less* avoidance than going 6.9K -> 10.9K. Which is... unexpected. Seeing the curve would clarify how this works, and answer a lot of questions about things like Finesse (is it better to stack one avoidance very high so that the percentage taken is lower because it removes rating that's already seriously diminished, or does that not matter?).

[Ithrien]

Are you guys even aware of what our stats can be stacked to current-game (/pre-RoI)?

1,004 Will and 792 Fate on a level 65 toon (current-game gear, with no will/fate legacies on either LI)?

If gear is AT ALL an upgrade, especially one that reflects the TEN levels we gain... I expect to see some high numbers. I hope you know what you're doing.

[MoonwalkIntoMordor]

Like for avoidances, going 3K->6.9K gives *less* avoidance than going 6.9K -> 10.9K.

That is good news, if the figures are correct and are working as intended. I believe that the player should be rewarded for focusing on a certain stat, not punished. Since you already are handicapping yourself in other ways by neglecting other stats in order to focus on that specific one.

[Happyfish]

We removed the cap on Offence. There is no reachable (probably) point in which the contribution of your primary stat to Offence is lowered.

There ARE diminishing returns built into all defensive ratings.

All of my gear here is lvl 72+. I am wearing 4/6 of the PvP set and 2/6 (boots and gloves) of the PvE set . I'm still using lvl 65 extraordinary relics (player designed runes, gems, and settings). My jewelry is a mix of quest rewards, crafted items, rep rewards, and instance drops. I'm also using the hunter PVP off-hand as well as the r15 cloak.

I don't have any buffs and I'm not in a stance.

-------------------
Here a hunter is posting 71% increase to ranged offense, how do you expect that this balances with threat generation from tank classes? I'm just curious if we are expecting DPS to outpace threat in this xpac.

Edit: I reread this and I dont mean it to be snarky (not sure if it comes across that way). I'm just curious how dps and threat are scaling.

[Ithrien]

Here a hunter is posting 71% increase to ranged offense, how do you expect that this balances with threat generation from tank classes? I'm just curious if we are expecting DPS to outpace threat in this xpac.

And here we have:

And yeah
Champ with 1678 Might. According to the guy, unbuffed.

102.1% increase to melee offense. Curious about RKs. I imagine LMs will be able to stack even higher (tactical offense) because of the nature of their red traitline set bonuses.

[Artanisul]

That has been my biggest concern, even before the offense caps were reported removed. They really need to pull the curtain back and show us the threat mechanics. When offenses go up...but added threat generation on skills does NOT get an increase, then we go back to the mistake made with Moria Launch. We know the threat skills scale, but so do the DPS ones, so it is just a net increase to damage.

What is the plan to keep things under control? Does melee offense modify the threat componant, will it after the expansion? Does tactical offense modify it? If its tactical then why are we not getting tactical offense from might anymore, why would you change that??

We really do need SOME answers....pwetty pweeze?

[Golledhel]

From what I've been reading, threat generation got a HUGE boost as well.

[Stonecrop]

Any chance of us getting the formula for that?

Unless they've changed the formulae for Isengard, this thread is what I've been using and it's been accurate in my testing.

[Happyfish]

From what I've been reading, threat generation got a HUGE boost as well.

What have you been reading?

[ANewMachine]

That is good news, if the figures are correct and are working as intended. I believe that the player should be rewarded for focusing on a certain stat, not punished. Since you already are handicapping yourself in other ways by neglecting other stats in order to focus on that specific one.

Except that doesn't reflect the diminishing returns, which should give you less rating per point as you get higher rating, and which are the intended result. (BTW, if you want to reward stat-stacking, then you have a linear increase - 1 point of rating would give the same percentage increase going from 0->1 as it does going from 10,000->10,001 - otherwise you're not rewarding stacking, you're driving players to it, which actually removes the available choices).

I think it's bad news if the devs say "diminishing returns" and the stats show "increasing returns."

[horus418]

From what I've been reading, threat generation got a HUGE boost as well.

What have you been reading?

Yeah really, because from what I've been told, it's the exact opposite.

[Reillan]

Unless they've changed the formulae for Isengard, this thread is what I've been using and it's been accurate in my testing.

If that's accurate, we'll only need to hit 9917 crit to be at 25%.

[ANewMachine]

Unless they've changed the formulae for Isengard, this thread is what I've been using and it's been accurate in my testing.

From Mysterion's post here, we have a level 75 character with 10,953 Block, 3,028 Parry, and 6,960 Evade. Let's compare expected values to what we're actually observing.
10953 / ((1190/3)*75+10953) = 26.9%. Observed value: 23.7%. Variance: 11.8%.
6960/((1190/3)*75+6960) = 18.96%. Observed value: 17.8%. Variance: 6.11%.
3028/((1190/3)*75+3028) = 9.24%. Observed value: 14%. Variance: 33%.

Worth noting that the Warden has only rating-based bonuses, not percentage-based bonuses (or penalties - looking at you, heavy shields) that would skew the results, making this a perfect class to test the ratings with.

The formula appears to have changed.

[Golledhel]

Yeah really, because from what I've been told, it's the exact opposite.

As a Guardian I never run OP. It's not my job to deal damage tyvm. I didn't stack huge Vitality. It could have been higher. The gear that He is wearing is NOT EnG tavern issue either. The gear was picked to give HUGE mitigations, which it does.

Why do I want HUGE damage and high crit chance? I don't need to hit as hard anymore since Guards got a MASSIVE aggro generation increase. I'm there to hold aggro and take the damage, not kill things.

I COULD have increased my Vitality to 1500+ but instead opted for 1200+ and much higher Mitigations.

He mitigates 47.9% of ALL physical damage. 69.3% of anything common and 51.4% of tactical damage (shadow, acid etc etc)

My incoming healing rating is 24.9% (enough for me thanks) and my block rating is pretty much 25% when using Guards ward. Since most of my aggro skills come from behind a block response i'd much rather block than parry or evade, although If i really need avoidances I'll just obviously hit up my pledge.

Don't judge a book by it's cover until you know everything.

Oh wait... you're a warden

[horus418]

If that's accurate, we'll only need to hit 9917 crit to be at 25%.

I'd say it's not accurate being that a warden posted 10,000+ block and was not yet at 25%.

[horus418]

Oh wait... you're a warden

Oh look... an insecure hater (from this post anyway),.

I'm glad it looks like they hooked the guards up though, being the fair non-biased guy that I am.

[Tolfast]

-------------------
Here a hunter is posting 71% increase to ranged offense, how do you expect that this balances with threat generation from tank classes? I'm just curious if we are expecting DPS to outpace threat in this xpac.

Edit: I reread this and I dont mean it to be snarky (not sure if it comes across that way). I'm just curious how dps and threat are scaling.

This is the thing that worries me the most as a tank, when moria was released tanks were effectively useless, can they really make the same mistake twice? i hope not...

[MoonwalkIntoMordor]

Except that doesn't reflect the diminishing returns, which should give you less rating per point as you get higher rating, and which are the intended result.

I think it's bad news if the devs say "diminishing returns" and the stats show "increasing returns."

Of course I didn't mean that it would be good if it is working in a way the developers don't intend it to. That is why I said "if they are working as intended". If the developers intend to have diminishing returns yet the formula somehow does the exact opposite, that is indeed bad news. Shows there is a serious mistake in the coding.

(BTW, if you want to reward stat-stacking, then you have a linear increase - 1 point of rating would give the same percentage increase going from 0->1 as it does going from 10,000->10,001 - otherwise you're not rewarding stacking, you're driving players to it, which actually removes the available choices).

Linear increase doesn't reward or punish stacking. It's pretty much the neutral way to go.

I can't be that sure about your prediction though. The more you stack the more difficult it gets to go further and the more sacrifices you need to make from other stats. For example, you wouldn't exactly be driving every single champion to stacking only might if it leaves them with 5k morale at level 75. But it would still be a viable option for some situations. That is why I believe something like cubic or exponential growth may be a viable idea.
_________________________

I also took a look at the numbers you posted from the warden related thread. They are not what I expected when I saw your first post. There is something off with 3k rating giving 14%. I can understand the variance in the other two numbers because the formula we are using has been structured to fit a curve only up to 15%. But 3k is about 1.2k less than what we currently need to get 14%.

I also looked at some other numbers later in the thread that showed some data of critical defence, which currently shares the formula with avoidance on live servers. In that bit, 3140 crit defence translates to 9.5%, which seems more accurate.

It just seems like an error to me, not in the formula but of the person who posted the screenshot. I don't really think they have changed the formula after seeing all the other figures.

[Leinad312]

I'm not sure about that parry rating. Her post lists the parry as 3028, but the screenshot shows 4828.

[ANewMachine]

I'm not sure about that parry rating. Her post lists the parry as 3028, but the screenshot shows 4828.

Well, that fixes the perceived diminishing returns problem, but the formula still changed if we accept the evade and block numbers as accurate.

[TicklerTurin]

We removed the cap on Offence. There is no reachable (probably) point in which the contribution of your primary stat to Offence is lowered.

There ARE diminishing returns built into all defensive ratings.

This confuses me slightly. Maybe every might still gives you the exact amount of offense rating, but doesn't the rating to percentage equation still have a diminishing return? Is that equation still the same as the live build?

[ANewMachine]

This confuses me slightly. Maybe every might still gives you the exact amount of offense rating, but doesn't the rating to percentage equation still have a diminishing return? Is that equation still the same as the live build?

Every Might gives you 10 Offense (for Might-based classes), and you're correct, it's the ratings-to-percentage where you get diminishing returns. Some of the formulas may have changed (I've got someone in beta getting me a more robust data set to test this more thoroughly). In live, it's a diminishing returns model already - you get less %age per point of rating as you increase in rating.

[Mysterion]

It just seems like an error to me, not in the formula but of the person who posted the screenshot. I don't really think they have changed the formula after seeing all the other figures.

My original post didn't have the tooltips, and I may have written the numbers down incorrectly. Many beverages were had that evening. It was also from b1 or b0, so it was awhile ago. This post has the tooltips included in the screenshot, and is from the current build of beta:

http://forums.lotro.com/showthread.p...99#post5642899

Hopefully that helps a little more.