Dear Devs: A request re: threat mechanics discussion

[Junts]

I'm not the sort to make a thread that specifically requests a developer response, but I don't see another easy way to address this question effectively.

On DDO, we're usually left to figure out the specific interactions of mechanics ourselves. Generally, I find this pretty engaging. It isn't too difficult, it follows some fairly consistent rules or patterns, and we're usually pretty accurate with determining stuff after some testing and time spent.

From time to time we've deviated from normal interactions in some way. Most of these have been easily tested or have been developer verified (Example: Dodge bonuses interacting like exceptional bonuses do, stacking with each other but not identical effects). Some interactions are completely abnormal but very easy to test and document and consistent (for example: healing amplification's multiplicative interaction with itself).

As a player I generally enjoy this stuff and its one of the things I enjoy on DDO to figure out how this stuff works.

In the last patch, the threat bonuses on Defender stances were finally fixed. I understand this bug was pretty difficult to deal with (and probably moreso because the concept had to be retuned). Part of the delay in it's reporting, however, also came from us: It was extremely hard to test and verify that the stances weren't working properly, and it took me a couple months after they were released to verify that.

We've been given some information on threat mechanics (eg that these enhancements are now typed, stack with each other when typed differently and with treasure effects). However, we don't know how that stacking works, and, for a variety of reasons, it is extraordinarily difficult, in fact, practically impossible, for us to test this effectively and break it down. Here is a non-exhaustive list of reasons.

1: There's enough random variance in damage dealing to require many tests and very high hp mobs for people to test their threat interactions with each other.
2: RE: above, we're also hamstrung but not knowing some basic mechanics of DDO threat (does it decay over time, for example)
3: We don't know what kind of stacking to expect from treasure effects, which is what we need to test most: Is it additive or multiplicative? Does it use only the highest source (typical), or does each non-identical benefit stack (if I have treasure effects for +15% threat and add a +20% threat effect, do I gain 5% threat (nonstacking), 20% threat(additive stacking), or 23% threat? (multiplicative stacking))
4: Items with threat bonuses are pretty rare and in many cases are exceptionally hard to acquire (epic sets, tower of despair sets, dragontouched sovereign rune 3 piece set effects, etc). This makes finding 2 characters who have the same treasure effects to test with each other difficult, much less ensuring those characters deal close to the same damage before the threat effects!

For this reason, I feel that the underlying mechanics and interaction of threat are outside what we, as players, will be able to realistically deduce like we have with other mechanics. This limits our ability to give feedback on their effectivenes and to report bugs in their behavior or interactions with each other, as we don't know how they're supposed to work.

For this reason, I think it would really benefit the community to get some specific answers regarding threat mechanics and interactions, like we recieved regarding dodge bonuses. With this information we can give constructive feedback and report bugs regarding our character's behaviors and performance. This information would also allow us to give better feedback on monster AI, as there are some monsters who behave in very wonky fashions regarding aggro, especially teleporters (a mob who teleports to his aggro target but cannot see it will drop aggro and pick another target seemingly at random, for example if he teleports and lands facing away from the person he teleported to attack, he'll run after the first person he sees: if he's facing the original person, they'll retain all the aggro they had: this seems buggy, but it's next to impossible to test with our knowledge of threat and aggro).

I'll end this post with a list of questions I think would enable us to do this.

1: How do treasure threat effects interact? Is it highest-source only (wearing 20% and 10% threat bonuses means 10% does nothing) or all unique effects (wearing 20% and 10% causes both to work, wearing two 20% bonuses gives you only one benefit)?

2: How do threat effects interact with each other? Are they multiplicative with each other or additive to each other? (additive seems like it would cause issues with threat -reductions-, since you might be able to stack them to 0% threat, and leads most people to assume they are multiplicative). Are they a combination of the two even? (spell damage being [total metamagic bonuses] * [total enhancement/buff/equipment bonuses] * [base damage], there is precedent for that kind of interaction on DDO).

3: Is threat permanent, based on damage dealt in the last 20-30 seconds, or does it decay in some fashion? Does aggro transfer when one person passes another at all, or when a certain threshold is passed (for example, WoW threat requires someone to reach 130% of the threat of the person who's currently got aggro before aggro will switch). the way mobs can bounce between aggro targets on DDO leads me to believe that there is no buffer zone, but you can't be sure.


Thanks in advance

[Dragaer]

A bump for you Junts. Inquiring minds want to know.

[kamimitsu]

I've been thinking about the issue as well, Junts. I wondered why we don't just have a have a "melee threat" (and a "spell threat") section of the character sheet? It doesn't seem that different from knowing your AC, saves, to-hit, etc. Explanations would be less necessary for specific effects and there would be more transparency in listed and actual effects/items (for both positive and negative threat). I'd love to see "Melee Threat: 56%" on Shizmonkey, or "Melee Threat: 173%" on Shizrhino.

We'd still need some explanation of mob AI in relation to threat: decay over time, mob intelligence or lack thereof (like why do some mobs seem to run straight at healers, and others, like zombies, don't seem to care who they attack at all regardless of threat/damage?), etc.

[Angelus_dead]

2: RE: above, we're also hamstrung but not knowing some basic mechanics of DDO threat (does it decay over time, for example)

And we also don't know how consistent aggro mechanics are supposed to be from creature to creature. Maybe trolls have one aggro rule if their CR is even and another if it's odd! It could be anything going on under there, and players have no ability to report bugs because they have only guesses as to what the desired behavior should be. For example, Intimidate tanking kinda broke recently; was that intentional?

Two interesting aggro complexities I've noticed:
1. When I'm near an orc that's aggroed to someone else, the monster will start hitting me if I press shift. If I'm not pressing shift it wanders around.
2. I was shooting an ogre with arrows from a ledge and was the only one doing damage. A hireling cleric walked by, and the ogre started clubbing him.

[My2Cents]

Just a couple of comments from the peanut gallery : I'm not as experienced with DDO threat and agro and the intricacies of combat as many, but I wanted to sound a word of caution:

One of the reasons I left a previous (very popular) MMO is that I found there became almost an entire religion built around the mechanics of agro and threat (and around the AH too but thats another story.) I definitely agree that knowing how things work is important, but there were two things that bothered me about what happened there.

First, it seemed like if you were not a caster with CC you were expected to tank, period. So the game seemed to boil down to Crowd Control, Tanking and healing. And, each build could only perform one role so until you built several alts you were trapped in a role, period. One of the things I've liked about DDO is there seems to have been a bit more loose philosophy allowing many different builds to try many different roles. Heck, I became so tired of Tank or CC or Heal or don't play. I think the concept of tanking, although apparently necessary, seems to take away a lot of the fun of these MMOs. I don't remember "tanking" from PnP like it seems to be embodied online.

The second outcome was that huge and vast tools were developed around the concept of agro and threat, and a large portion of the game just became a numbers contest of who could dish the msot threat. If you couldn't dish out -the absolute most- DPS or threat, you would have a great deal of trouble finding a group for any higher level runs. While some of this is normal and expected, it can be taken too far, and I'd prefer DDO be careful when addressing threat issues.

Just me, peanut gallery, for whatever these comments are worth, or not...

[sirgog]

And we also don't know how consistent aggro mechanics are supposed to be from creature to creature. Maybe trolls have one aggro rule if their CR is even and another if it's odd! It could be anything going on under there, and players have no ability to report bugs because they have only guesses as to what the desired behavior should be. For example, Intimidate tanking kinda broke recently; was that intentional?

Two interesting aggro complexities I've noticed:
1. When I'm near an orc that's aggroed to someone else, the monster will start hitting me if I press shift. If I'm not pressing shift it wanders around.
2. I was shooting an ogre with arrows from a ledge and was the only one doing damage. A hireling cleric walked by, and the ogre started clubbing him.

Some mobs have random aggro - they pick someone they can see and aggro on it for a set period of time, then choose aggro again randomly. Arraetrikos is the best-known example, some undead do this too.

As for pressing Shift - if you are standing between a mob and its intended target and actively blocking, you will generate threat. (By between, I mean on any plausible path, not on the most direct one). This was brought in to address various exploits such as the 'shield block STK guardian into a corner, caster gets all the aggro and solos him in complete safety', as well as other non-exploitative but 'easy button' play.

On the original questions: A Dev confirmed some time ago in a DDOcast that threat effects stack multiplicatively. IIRC they were explicitly referring to the combination of 2-piece Tharne's Wrath, Treason and Rogue Subtle Backstabber, indicating that at least one threat modifying set bonus, at least one non-set threat modifying bonus and at least one class-based threat modifier stack. However, it's never been confirmed (to the best of my knowledge) whether multiple non-set threat bonuses stack with each other, or whether multiple set bonuses stack.

[Angelus_dead]

Some mobs have random aggro - they pick someone they can see and aggro on it for a set period of time, then choose aggro again randomly. Arraetrikos is the best-known example, some undead do this too.

What makes you say that?

As for pressing Shift - if you are standing between a mob and its intended target and actively blocking, you will generate threat.

That conflicts with the observation I reported, which illustrates why it could be nice to have some real answers for these things.

[sirgog]

1) What makes you say that?


2) That conflicts with the observation I reported, which illustrates why it could be nice to have some real answers for these things.

1) Observe Arraetrikos. His random aggro is a well-known fact and is the reason that squishies in the back lines (in both 4 and 5) get fireballed; sometimes the fireball hits the rest party first, sometimes it gets to the back line squishy. When running ungeared at-level elite runs, I take advantage of this aggro pattern to use a different part 4 strategy that ensures non-evasion toons are seldom fireballed. It works, hence I am certain that his aggro is random. Horoth and Turigulon, despite also being Pit Fiends, have standard threat-based aggro, again easy to test.

2) Note that I said 'on any plausible path'. If the mob is at position A1 and its target is at A5, A1-A2-B2-C2-C3-C4-B4-A4-A5 is a plausible path, so it might take offense to you blocking at C3. Games calculate multiple paths for mobs to move around obstacles such as players.
Shield blocking might generate aggro based on proximity to the mob, proximity to the mob's target, proximity to the closest path between them, or in the way I outlined above. All would be roughly similar in effect.

[Angelus_dead]

1) Observe Arraetrikos. His random aggro is a well-known fact

But why would you talk about Arraetrikos in reply to my comment?

2) Note that I said 'on any plausible path'.

I don't see why you're talking about that.

Games calculate multiple paths for mobs to move around obstacles such as players.

Really, why?

Shield blocking might generate aggro based on proximity to the mob, proximity to the mob's target, proximity to the closest path between them, or in the way I outlined above. All would be roughly similar in effect.

There is a pretty big difference between generating threat and getting aggro.

[SardaofChaos]

Really, why?

In case the players move, obviously.

[Xeraphim]

A bump for you Junts. Inquiring minds want to know.

Indeed.

I'd appreciate the players just bumping the thread instead of spamming it with speculation and wild guesswork and conjecture.

[Itachi83]

Actually, I distinctly remember evidence that some mobs agroed on healer types before others.

Of course I have nothing but anecdotal evidence for this, so it really would be amazing if we could get a dev to comment on the specific mechanics of certain threat situations.

Indeed.

I'd appreciate the players just bumping the thread instead of spamming it with speculation and wild guesswork and conjecture.

That is slightly rude.

[Xeraphim]

That is slightly rude.

I don't understand. I'm being analytical. There is nothing rude about that declaration when the topic of this thread is considered, to my knowledge and perception.

As for Healers, they do seem to be targetted as the highest value target of the group on occasion, and clearly at random in other cases. Individual monster type aggro may be at work here, but that's speculation and blind conjecture as well.

This quandry can only be truly absolved by a Developer with carefully selected information dumping.

[Itachi83]

I don't understand. I'm being analytical. There is nothing rude about that declaration when the topic of this thread is considered, to my knowledge and perception.

As for Healers, they do seem to be targetted as the highest value target of the group on occasion, and clearly at random in other cases. Individual monster type aggro may be at work here, but that's speculation and blind conjecture as well.

This quandry can only be truly absolved by a Developer with carefully selected information dumping.

Eh, I am not going to argue with you.

I can't see it from your perspective, so it is entirely possible that I misinterpreted your declaration, and the intentions that were behind it.


I agree that it would be positive if we had a dev come in and help with this issue.

[Thelmallen]

I nominate Junts for the forum title "High Inquisitor". Great post as always.

[grodon9999]

Excellent post, it would be good to get an answer on this.

[Samadhi]

Unlike you, I never liked the lack of transparency in this game. Easy to test != knowing what is intended and what is accidental.

However, any push for transparency in anything is highly appreciated, including the obvious one listed here (how to decide what loot stacks and what doesn't is a HUGE gap to not have understanding of from our perspective).

[Goldeneye]

I agree Junts.
Aggro management and threat is shrouded in mystery. I'm a complete nerd, and would love some numbers to crunch!

[Dark-Gulrak]

Bump for knowledge and curiosity

[SisAmethyst]

Just a couple of comments from the peanut gallery : I'm not as experienced with DDO threat and agro and the intricacies of combat as many, but I wanted to sound a word of caution:

One of the reasons I left a previous (very popular) MMO ... there were two things that bothered me about what happened there.

First, ... , each build could only perform one role so until you built several alts you were trapped in a role, period. One of the things I've liked about DDO is ... allowing many different builds to try many different roles...

The second ... tools were developed ... the game just became a numbers contest of who could dish the msot threat... While some of this is normal and expected, it can be taken too far, and I'd prefer DDO be careful when addressing threat issues.

...

I see your concerns but Junts is not really asking for absolution but a question to understand the game better and I guess the majority don't want to change the game to become an clone of another MMO. Anyway I haven't seen any plug-in tool (except of the skins) yet for DDO as you may see it for other MMOs (which I am happy with). On the other side the basic rules of DDO are written in the core books, so you can already take an excel sheet to calculate a theoretical DPS value for a specific scenario. Therefore it is enough to get an understanding for your character and how to improve without the need for an in-game DPS-meter which would maybe then lead to stereotype character builds.

But even then I think DDO has the advantage of being much more universal already on the basis, which is why you can see such funny and crazy efforts like an All-Babarian-Shroud. The game itself gives you more freedom, flexibility and options.

However there is a big gap for understanding aggression and thread mechanic. I would exaggerate if I say: "While in other MMOs tanking is the only way to beat the boss, in DDO it is most of the time not an option at all", but I guess one get the point. You can't even give qualitative feedback on this subject as we don't know enough about it. This unfortunately includes as well bug reporting and player feedback to the developers.

Further you can test in the Tavern brawl how much damage a weapon may deal, but we can't simulate the aggression or the effect of thread generation this way. Just imagine you still wouldn't know all the effects and recipes of the Greensteel crafting. It took months to figure it out and endless trials and wasted shards. Even then we needed some hints, like for creating the dual shards. If we still wouldn't know all of this, probably no one would run Shroud more then one time for the favor points, because this crafting would be too unfunny and costly to even start with.
Right now we are in the middle of the dessert and no sign or way could be made out, guessing where north and south is. So what is asked here are some hints like we got back then for the Greensteel recipes to get an idea in which direction we can go. Like painting a road with several crossway's in this dessert. Then we still have to go one or the other direction, but we at least know its a way and not endless walking in circles.

PS: /+1 for this article Junts! (if I could)