Nerfed Claymore Trap: A mess of misinformation and incompetence

[Fluorite~]

tl;dr: Pre-renewal Claymore Trap, Blast Mine, and Fire Pillar are supposed to do extra stacking damage when more than one monster triggers it.  This was removed from official servers because there was a bug where stacking too many monsters would crash clients and they didn't feel like fixing it, knowing that traps would be buffed in renewal shortly after.  Emulators have unnecessarily reproduced this change, even though that bug never existed; a case of a false premise and an incorrect conclusion.


I usually decide whether a pre-renewal is worth playing or not based on one thing:  Claymore Trap damage.

The good old days of proper skill implementation:
For those of you unaware, Claymore Trap (alongside Blast Mine and Fire Pillar) are supposed to be like this, from back in November 2007:

http://code.google.com/p/rathena/source/detail?r=11634

Modified trap related code so it behaves as explained by Playtester:
Traps trigger on everyone in their 'effect range' (skill_db's splash) for
every target that lies within its trigger range at trigger time (range and
interval in the skill_unit_db). [Skotlex]

What does this mean?
*Effect Range:  The 5x5 square around Claymore Trap, that deals damage. (i.e. you can't trigger Claymore Trap from 2 cells away, but you can take damage from it)
*Trigger Range:  The 3x3 square around Claymore Trap that causes the trap to explode.  This range deals damage AND counts the number of monsters hit within this range.
*Additionally, Fire Pillar will do the same thing (2008/03/04 update)

This made Claymore Trap, Blast Mine, and Fire Pillar very interesting to use.  They're trap AoE skills that require precision to use and are forced into specific elements.  You get the most out of them by doing risky mobbing where all monsters must be traveling close together, or else you're wasting time and traps.  Blast Mine was definitely the worst one to use properly, since its much smaller than Claymore Trap or Fire Pillar.

Basically, it looks like this.  I died several times in the making of this video because of how dangerous it is.  Also, mute the music for a more enjoyable experience (damn teenage years...)
https://www.youtube.com/watch?v=8R3WBuXMI-s


But then there was a change to official servers.
http://forums.irowiki.org/showthread.php?t=17175
"Fixed Blast Mine and Claymore Trap to no longer apply duplicate damage when striking multiple targets."

Reason:  With massive monster stacking, Claymore Trap and Blast Mine could cause disconnects/server instability.
See also:  Orc Hero Firepillar Kill .
Fire Pillar has the same multi-hit function as Claymore/Blast Mine.  If you don't feel like watching that, just know it involved spawned mushroom stacking and Fire Pillar to one-shot kill Orc Hero...and disconnect everyone in the vicinity of the explosion.

This official change removed the extra damage from splash damage from these three skills, but it's just a really lazy way of fixing a bug.  The staff also knew that Claymore Trap, as well as other traps, would end up doing only one hit in Renewal and deal approximately 5-8 times more damage than in pre-renewal.  I cannot emphasize that message enough, because of another fact:

Emulators have never experienced disconnects due to Claymore Trap damage stacking like in the video.  eAthena never disconnected due to mob size, and neither did rAthena.  This is somehow the premise for the following change.

Trap damage on pre-renewal emulators was changed to follow what official servers did: damage traps now do 1 hit of damage to every monster struck, with no extra hits for stacked monsters.  You're left to wonder...why?!  The reason for this change was because of a crash on official servers...but this crash doesn't happen on emulators.  Therefore, there's no reason to implement the change in the first place on emulators.

This just blows my mind on how stupid the logic is.  Compare it to these scenarios:
*You've just been diagnosed with a terminal illness and you must undergo extensive chemotherapy to the point that you can't walk.  However, you're expected to have a full recovery in a year and this treatment will just be a thing in the past.
*You've just been diagnosed as being perfectly healthy, but you undergo said extensive chemotherapy to the point that you can't walk.  You're not expected to ever recover from this.

Emulators decided to go with the second scenario, possibly to have their pre-renewal code be as up-to-date with officials as possible.  But with changing pre-renewal damage traps, being up-to-date has ironically made them out-of-date, because they will never get the expected buff in renewal.  And this happens for every new pre-renewal server that opens nowadays!


Recently...
But what really pisses me off now is how it was incorrectly branded on ggRO and now AevaRO:  Doing multi-hit splash damage, like how the skill is supposed to work, is supposedly a bug.
From http://board.aeva-ro.com/index.php?/topic/44-behind-the-scenes/?hl=claymore#entry287

- Fixed claymore trap bug as well, as per official. Claymore can no longer be exploited to kill a large quantity of mobs with a single blow.

What the hell kind of misinformation was read to come to that conclusion?  In order for that to be correct, you'd have to tell 5+ years of Hunters, Rogues, and Wizards that their skill was somehow incorrectly implemented and that Gravity was too stupid to realize it all that time.

These skills are supposed to do extra damage for mobbing; their damages depend on that.  Just compare Land Mine's damage to Blast Mine's damage on a Marc.  Land Mine is Earth element.  Blast Mine is Wind element.  Without the extra hits that Blast Mine should be getting from mobbing, Land Mine does more damage to the Marc than the Blast Mine, despite not getting the 175% elemental damage boost!
Land Mine damage:
http://calcx.wushuang.ws/?cakbLaYabbiagbwbuabukaabNaaaaaaaaae5aaaaaaaaacPaad5aaeueFaafbaafhbDfwbMfqaaaaaaahkkkaaaabkkaaaaaabSfaaaeMaaaaHaaaa
Blast Mine damage:
http://calcx.wushuang.ws/?cakbLaYabbiagbwbuabukaabNaaaaaaaaae5aaaaaaaaacPaad5aaeueFaafbaafhbDfwbMfqaaaaaaahkkkaaaabkkaaaaaabYfaaaeMaaaaHaaaa



What can be done?
Most servers I've seen choose not to do anything with the existing trap damage code.  I highly speculate this to be incompetence or lack of time on the development staff.  Sadly, nothing can be done if your server's development staff is terrible.

Some might argue that the trap is overpowered (despite it being a very niche way of leveling, not good in PVP/GVG/BG/MVP, and a huge strain on Sniper skill trees).  But as we see from the Land Mine/Blast Mine example from above, completely removing the extra damage makes no sense either.  The trick then is to strike a balance; if you think these traps scale too well with mob size, then put a limit on how many hits they can do, such as 10 maximum.  It's enough to cause panic against mobs that aren't completely killed off (due to certain mob skills being activated at low health), and does not encourage infinite mobbing.

Some might argue that it's too affordable or easy to use at lower levels.  Then increase the number of traps needed per cast or have a level requirement.  Hell, you could even have the skill scale with base level.


But in the end, it's a server's decision to implement any changes to these damage traps or not.  I can only hope that they aren't deliberately keeping it nerfed for the wrong reasons.


And if you somehow don't like Claymore Trap and it should stay in its default pre-renewal setting, ask yourself something else: Has anyone ever complained that Blast Mine or Fire Pillar were doing too much damage?


EDIT:  Extra reading for how to possibly revert the change.  https://rathena.org/board/tracker/issue-4826-traps-multi-trigger/

[Regret]

My thoughts exactly.
Personally, I declare every server that has unnecessarily nerfed Claymore Trap garbage, it just doesn't make any sense.

Oh, you're taking away a niche way of leveling?
Why thank you.

[Playtester]

Well, it's quite debatable how it is really supposed to be. Gravity literary called the multi-triggering a "bug".

Emulators are simply based on the "most recent pre-renewal version that existed". There is currently no emulator that focuses on an episode older than 13.2, even though from many gameplay viewpoints, episode 12 would actually be best (or even 11.3 to prevent guilds quitting over Satan Morroc) for pre-renewal.

I once thought about at least adding a config option for the old behavior, but the old eAthena behavior wasn't even accurate enough. It was too easy to trigger multi-hitting without the use of knockback effect on eAthena. I'd need to implement it properly and for all three trap types (even though Firepillar is kind of glitchy if you use a renewal client as it can no longer be knocked back and consequently won't send position update packets).


I once wrote a better implementation of Claymore Trap here:
http://hercules.ws/board/topic/7198-claymore-trap-bugged/
This makes it possible to multi-trigger, but also very hard to accomplish if you don't use knockback effects. It's still easier than officially though, because the unit timer on officials is 20ms and on emulators it's currently set to 100ms.

But that's also easy to change in skill.c:

#define SKILLUNITTIMER_INTERVAL 100

(change to 20)
Just takes a lot of CPU load.

[Hyperactive]

I knew who started this thread the moment I read the title since I've seen you mobbing all sorts of stuff on EternityRO and PureRO

The skills you mentioned are supposed to be doing more damage if multiple enemies in the same area. It's still listed as such on the iRO Classic Wiki as well:
Damage Formula: Skill Level * (75 + [ DEX / 2]) * {1 + (INT / 100)} * Number of stacked monster

It's silly to say that the skill was bugged and that it you've fixed a bug by disallowing stacked monster multipliers. It's never been a bug and that's how the skill was designed. Maybe you've followed officials in their changes but they didn't fix a bug with the skill and neither did you. The bug, which existed only on officials, was that large quantities of monsters caused a server crash. The bug was never "claymore is multiplying on stacked mobs" because that would imply that the stacked mob multiplier was an accident and it wasn't. I agree that it doesn't make sense for emulators to follow an official server's "hot fix" for their issues.

I don't think I've ever seen anyone complain that claymore/fire pillar leveling needed to be nerfed. It's a pretty niche build and is mostly useful for high def or very mobby maps. Even then, your typical AGI/DEX sniper probably farms equally fast if not faster with FAS or normal 1v1 killing. Claymore mobbing is very high risk/high reward.

However, I can see how using cultivate to get a ridiculous multiplier on an MvP can be seen as "exploiting" the skill. For that reason I think your suggestion to cap the multiplier at 10 mobs seems perfectly fair. A cap of 10 is not enough to really affect your normal PvM claymore mobbing, but is enough to prevent someone going to extremes to get a *200 multipler or more.

My first question is, why do you not suggest this on the servers? I never saw the suggestion pop up at ggRO (though I might have missed it since there were like 10 pages of suggestions) nor at AevaRO.

My second question is, why do you want it so bad? Do want it just because you are particularly good at using the claymore mobbing method? Or are you using it as a sign of staff incompetence only and are okay playing without it?

[Playtester]

The bug was never "claymore is multiplying on stacked mobs" because that would imply that the stacked mob multiplier was an accident and it wasn't.

Well, according to Gravity it was unintended and a bug.

My first question is, why do you not suggest this on the servers? I never saw the suggestion pop up at ggRO (though I might have missed it since there were like 10 pages of suggestions) nor at AevaRO.

It was suggested on ggRO. I even told the staff how to fix it. They said because Gravity officially called it a bug, they won't unfix it, even though they were based on a episode before it was fixed.

[Hyperactive]

Well, according to Gravity it was unintended and a bug.

That doesn't make sense. Why would they include this "bug" in the description of the skill? I wish I had old files to look at the description but all I have is this description from IGN back in 2004. 2004!

"DESCRIPTION: This trap explodes over a 5x5 area and gives stronger damage when there are more enemies in the 5x5 area."
Source: IGN Hunter Guide from 2004

I think it's pretty obvious that the skill was designed and coded to increase damage on stacked mobs. That doesn't just happen by accident as the word "bug" implies.

It was suggested on ggRO. I even told the staff how to fix it. They said because Gravity officially called it a bug, they won't unfix it, even though they were based on a episode before it was fixed.

I don't think they really cared about being correct for the episode since monsters and skill mechanics were all over the place. GC hitting all mobs 3 times, every time, but no stacked mob multiplier for claymore. What?

I guess if you're going by what Gravity has done with the game the correct thing to do would be to disable mob multiplier damage and make the traps completely useless. The better judgement call would be to realize that the skills were likely intended to have mob multiplier damage and no "bug fix" is needed.

[snowblind]

One thing I never understood is why athena must emulate all official behaviour.. Who cares what gravity said? Officials are simply inferior, which is why private servers were so successful and even spawned a whole new scene with new playstyles that could never have been explored otherwise.

But emulator standard becomes community standard, and sadly, most players tend to reject anything that's somehow different.

[Rider]

Application of the claymore trap fix is optional and since we added it, I included it in our changelogs prior to launch. I personally don't think it is a bug but since we follow official servers, we decided to keep it.

As a player myself, in the past I never thought that claymore trap was really overpowered and it was also in the skill description on how the skill functioned but over the years when new content was added the skill became far more exploitable. Luring an entire map of mobs and finishing them off with just a trap seems a bit broken.  Since the number of hits is triggered by the number of mobs, even those with over 40k HP died to a single trap. When one class is able to kill an entire map in scaraba hole with ease, how can one not question the way the skill works? And when it comes to class balance, doesn't that make one class better than the rest considering that there was hardly any need for high end equipment and such?

I can agree that having a limit to prevent exploit of the said skill like a max 10 mob stack might reduce the effect. There needs to be a lot of testing to figure out the limit that will prove efficient because you need to consider that simple stacked trap can bypass this limit. Then again it can be made so that the stacked traps only affect the first 10 or so mobs. Either way, the effect of the skill would be changed and people might not agree with how the stacked effect works.

[Fluorite~]

My thoughts exactly.
Personally, I declare every server that has unnecessarily nerfed Claymore Trap garbage, it just doesn't make any sense.

I honestly do the same.  ClaymoreTrap/10 is a good score.  NoClaymoreTrap/10 is a terrible score.  I have even chosen to play not-so-great servers simply because their development staff gave back Claymore Trap's multi-hit function.

Well, it's quite debatable how it is really supposed to be. Gravity literary called the multi-triggering a "bug".

...

I once wrote a better implementation of Claymore Trap here:
http://hercules.ws/board/topic/7198-claymore-trap-bugged/
This makes it possible to multi-trigger, but also very hard to accomplish if you don't use knockback effects. It's still easier than officially though, because the unit timer on officials is 20ms and on emulators it's currently set to 100ms.

But that's also easy to change in skill.c: (change to 20)
Just takes a lot of CPU load.

I'm having trouble finding where Gravity called the multi-triggering a "bug".

As for the better implementation of Claymore Trap, I'm surprised I haven't seen it used anywhere yet.  I have played with the malufett version, but that version was horribly overpowered; stack 20 Claymores and put an MVP in an Ankle Snare, then watch it get hit by the damage of 20 Claymore Traps every second.  Dead Osiris in 8 seconds! (This was the reason why I quit TrivalRO after one day)

I would love to play a low-rate server with your version.  Yeah, it's a shame how it isn't quite the same because of the unit timer, but it's pretty close.  I swear, the only reason why I even played PureRO was because their Claymore Trap did what your version does.

My first question is, why do you not suggest this on the servers? I never saw the suggestion pop up at ggRO (though I might have missed it since there were like 10 pages of suggestions) nor at AevaRO.

My second question is, why do you want it so bad? Do want it just because you are particularly good at using the claymore mobbing method? Or are you using it as a sign of staff incompetence only and are okay playing without it?

I don't suggest this on most servers because I don't trust the staff to understand or be competent enough to change it.  There was a suggestion thread about this on ggRO, and I was part of the discussion...but the suggestion was struck down because Claymore Trap's multi-hit damage was said to be a bug (...really, even after I provided evidence...) and was apparently too overpowered.

I can play without Claymore Trap; I usually just play a Beast Strafer instead (except in ggRO's case, where Beast Strafing was either bugged or not available!).  I find normal Dex/Agi builds to be too boring to play; I'd rather have fun blowing stuff up around the world.  It really is my favorite skill in the game, and the reason why Hunter/Sniper is my favorite class in the game.  It's an exciting mixture of smart gameplay: you need to pick a good map/mob targets, you need to mob in such a way that monsters walk close together, and then you need to finish by precisely aiming your skills with perfect timing.  The satisfying explosion is just...addictive.

Application of the claymore trap fix is optional and since we added it, I included it in our changelogs prior to launch. I personally don't think it is a bug but since we follow official servers, we decided to keep it.

As a player myself, in the past I never thought that claymore trap was really overpowered and it was also in the skill description on how the skill functioned but over the years when new content was added the skill became far more exploitable. Luring an entire map of mobs and finishing them off with just a trap seems a bit broken.  Since the number of hits is triggered by the number of mobs, even those with over 40k HP died to a single trap. When one class is able to kill an entire map in scaraba hole with ease, how can one not question the way the skill works? And when it comes to class balance, doesn't that make one class better than the rest considering that there was hardly any need for high end equipment and such?

I can agree that having a limit to prevent exploit of the said skill like a max 10 mob stack might reduce the effect. There needs to be a lot of testing to figure out the limit that will prove efficient because you need to consider that simple stacked trap can bypass this limit. Then again it can be made so that the stacked traps only affect the first 10 or so mobs. Either way, the effect of the skill would be changed and people might not agree with how the stacked effect works.

The reason why I didn't drop ggRO immediately in favor of AevaRO was because of your "fix" to Claymore Trap's "bug".  I only recently downloaded your client, played it for two days, and then got so angry at how Claymore Trap wasn't causing massive explosions that I made this post on RMS.  No jokes.

I also can't think of much "new content" that can be easily exploitable with Claymore Trap anymore.  Scarabas was exploitable for a little bit, but that was when Scaraba Dungeon was giving stupidly high amounts of EXP and the walk/aggro AI for both One-Horn and Two-Horn Scarabas was different than today (plus the introduction of monster fatigue kills mobbing there.  Scarabas, by default, stop following much sooner than other monsters).  Cursed Abbey's Zombie Slaughters maybe, but Floor 2 has so many Ragged Zombies and surprise Necromancer quagmires that it's very difficult to get a good mob going.  Most New World monsters walk too fast, have ranged skills, or are just too lethal to mob, so they're out of the question.  Thor, Abyss Lake, most of Thanatos Tower, Kiel Dungeon 2, Odin's Temple, Bios3...nope, I don't see it.  The only really exploitable place I can think of is Juperos (which is why I suggested a maximum amount of hits).

Claymore Trap just doesn't stack up well against end-game...anything.  Players still need their high-end characters and gears to do end-game content.  Snipers with Claymore Trap will never replace the use of Acid Demonstration, Asura Strikes, Sharp Shooters, or even Cart Termination/Soul Breaker while MVPing.  Claymore Trap has no place in War of Emperium.  And if you want a class that can do a lot with very little in PvM, just check out High Wizards.

[Playtester]

That doesn't make sense. Why would they include this "bug" in the description of the skill? I wish I had old files to look at the description but all I have is this description from IGN back in 2004. 2004!

I don't think this was included in the official description of the skill ever. It was included in the translation project of eAthena, however. And all guides of course explain this trick too. It was an essential part of the game.

Also, just because Gravity officials states it was a bug, doesn't mean it was one in the original design. Because, y'know, they could just be lying.

I don't think they really cared about being correct for the episode since monsters and skill mechanics were all over the place. GC hitting all mobs 3 times, every time, but no stacked mob multiplier for claymore. What?

Server owners not fixing things usually means they don't have the coding skills or time to fix it rather than not wanting to. I don't have any proof, but as I'm very well aware of the code, I have the impression that admins fix things that are very easy to fix, but anything more complicated to fix usually gets rejected.

One thing I never understood is why athena must emulate all official behaviour.. Who cares what gravity said? Officials are simply inferior, which is why private servers were so successful and even spawned a whole new scene with new playstyles that could never have been explored otherwise.

You don't see the bigger picture. If you want a lot of people working voluntary on the same project, you need to define a common goal or else the project will fall apart. A common goal needs to be objective and emulating exactly how it is officially is the only objective definition possible. A goal "Make it so it's most fun for the players" is simply too subjective and won't work for long.

Luring an entire map of mobs and finishing them off with just a trap seems a bit broken.

On officials that isn't possible due to all overhunted monsters switching to passive mode after a while of chasing. Of course every single private server switches that feature off eventually due to high player demand. *sigh*

Also you might want to check the fix I posted earlier in this thread. It's much much less exploitable.

I'm having trouble finding where Gravity called the multi-triggering a "bug".

Here you go: http://forums.irowiki.org/showthread.php?t=17175
Even though Doddler posted that, he literary translated it from the Korean update news.

As for the better implementation of Claymore Trap, I'm surprised I haven't seen it used anywhere yet.  I have played with the malufett version, but that version was horribly overpowered; stack 20 Claymores and put an MVP in an Ankle Snare, then watch it get hit by the damage of 20 Claymore Traps every second.  Dead Osiris in 8 seconds! (This was the reason why I quit TrivalRO after one day)

The problem I see with many emulator developers is that they implement something without really understanding how something works internally on Aegis and without testing it in-depth to compare and instead just implement something based only on iRO (which is in 80% of the cases at least partially incorrect). This causes stuff like that. But I kind of gave up trying to correct everyone. I overworked myself in the end.

I'd probably be more suited as director. Like everyone has to get the "ok" from me before applying a change. =p But really not going to happen. I could really also just tell people how it works as I've run so many tests, but most never ask me or even bother creating a bug report I could reply to. *sigh*

I would love to play a low-rate server with your version.  Yeah, it's a shame how it isn't quite the same because of the unit timer, but it's pretty close.  I swear, the only reason why I even played PureRO was because their Claymore Trap did what your version does.

Feel free to save the link to the fix I posted above and give it to the server owner of your (next) server.

I can play without Claymore Trap; I usually just play a Beast Strafer instead (except in ggRO's case, where Beast Strafing was either bugged or not available!).

Sometimes I really wish some servers would be based on rAthena again so that I could show some players the fixes I did the last 10 months. Like Beast Strafing being completely different now (it's targeted now and DS does not actually need to hit a beast for the combo to be available).


And Scarabas are not pre-renewal monsters so if they end up being exploitable, then maybe a less "fortunate" stat recalculation was used.

[Fluorite~]

Here you go: http://forums.irowiki.org/showthread.php?t=17175
Even though Doddler posted that, he literary translated it from the Korean update news.

I hate to be splitting hairs on this, but that doesn't necessarily mean that multi-hitting is a bug.  That one line about traps is very ambiguous.

"Fixed Blast Mine and Claymore Trap to no longer apply duplicate damage when striking multiple targets." can mean:

*Changed Blast Mine and Claymore Trap to no longer apply duplicate damage when striking multiple targets.  I'm not sure if it was literally the word "fixed" in Korean or not.  If the word "changed" was a better translation, this line would mean Gravity consciously believed that multi-hit damage should not be allowed...after 5+ years of allowing it.  Even then, the line would be better written as "Removed duplicate damage when striking multiple targets with Blast Mine and Claymore Trap."  "Fixed" implies the other option:

*Fixed Blast Mine and Claymore Trap's disconnection bug by no longer applying duplicate damage when striking multiple targets.  Just what exactly is being "fixed" is not clear.  They say they're "fixing" a skill, but they're not addressing what the "bug" actually is.  Based on all the evidence I've seen, they "fixed" the disconnection bug by removing the multi-hit damage.  It's about as terrible as breaking a bone in your arm, but instead of treating it properly, choosing to amputate the arm because a stronger prosthetic one was already ordered (i.e. renewal trap damages).  Unfortunately for pre-renewal private servers, no such prosthetic arm will ever show up.

In short, what the "bug" that promoted the "fix" is not clear from that one line.  But I find it very likely that they were masking their disconnection bug in their maintenance notes to save some face.

[annaquin]

No, it was actually a bug in leaked code source .

The skill should have been triggered on the first monster to collide. The problem was the resolution of collistion box that allowed any amount of monster to trigger the collision event. Message callback was for speed reason posted in a queue of fixed size.

The crash occur because the death_event was posting a respawn_callback which overflow the queue by doubling the message. So the bug was fixed twice, by disabling the sub call of respawn into death_event callback and by nullify the collision queue. Which triggered another bug on official, has stacked monster no longer collide multiple time , so does players. So the N-1 players on a stacked collision are not receiving the event "onhit" and the caster will only receive a single "onattack" event.

But again, if you are going that far , you should also remove the rebirth part of the process.

Gravity invented that rebirth thingy fearing players would go too fast in finishing the game content and so proposed to replay , giving them time to fix new classes bug and prepare RO2.

Why private server are making server rebirth .. is plain dumb.