[Lisbeth] Your own personal crafter!

most people meld offhands with 4 control, 15 craft, 3 cp. which in certain conditions that extra 3 cp comes in very handy. Removing the cp for 1 extra craft isnt viable imo.. Now if you do sb your own materia and want to waste countless IVs on melding, 16 craft would be ideal, but I do not see many people wasting IVs on a 5-7% meld chance.

As for the other topics.. regarding using rapid/pbp .. you'd have to take in the chance of it failing. Alot of recipes do start with pbp.. and if one of them fails, people normally reclaim.. Rapid synth having a base of 50(?) success, even with steady hand 2 increasing it slightly has a higher chance of hitting, but must take into account if rapid synth misses.. you will most likely want to reclaim and start over.. Alot of the 3 stars i've crafted end up ending with ingenuity 2, with 3 careful synth IIs.. innovation and ingenuity are good for fishing for a good/excellent orb but ingeunity 2 would be needed to finish it with 3 careful 2s, which is safer imo than using rapid synth... but pbp would be a good way to start the craft.. two of those after a steady hand, and the odds of missing it are very low..

Reclaim would be a must if you want to focus the craft off of using a rapid synth.. since in most cases if you miss it, you will want to start over.

From what I have noticed.. Rotation#1 @http://i.imgur.com/1MQpZwq.jpg has constantly given me roughly 90-92% HQ on the artisan offhands.


The ideal stats for fully melded artisans/offhand/lucis is 492 craft, 429 control. and something like 390(ish) CP including food. Alot of people debate over best meld for artisans body with either using control+2 or cp+2 for the last meld.. just some rotations out there for 3-4 star crafts literally require that much cp for a high % chance on quality... but I have sold countless bodies on AH with control+2 as the last meld over having the extra cp on it.

 

My post was based not on a rotation, but rather then the skills themself. I explained in great detail to Neverdyne of how my rotations work, to which I always use PBP under SH2(110% chance), never use Rapid Synth, and ING2 is optional if your Careful Synth II's give you at least 70 progress without it. With my rotations I have almost always 100% HQ 4* crafts with NQ mats, and I have a 100% rate on anything 3*. When your not using ING2 on your crafts you use observe to help bait that good/excellent.

I have 1 character that is Craftsmenship focused and the other that is control focused, I have tested many of the "ideal" melds out there, and have my preferences as well, the diffrence between ~10 points in either control or craftsmenship will not change much. And the CP you speak of is either 352 or 355 without food and under food it is 395 or 398. I have no idea of any crafter who would meld CP 2 or Control 2 on the last meld slot since they can both be 3's. You ideal stats you claim are personal preference as well. The only reason to push craftsmenship is to try and gain a extra step by only needing 1 less progress step, or to try and push as close as possible to next tier crafting that is coming up. After a certain point with craftsmenship you start to get diminishing returns, Up at 490+ craftsmenship the diffrence between CS2 under ING2 and not is a mere 6 points.

The common and "ideal" melds of the chest for most top tier crafters is Craft IV, Craft IV, Craft IV, Control IV, CP 3, and if you want to get into a debate over off-hands technical ideal melds, your looking at all Tier IV materia which just isnt "ideal" anyways. I have seen offhands go 1 of two ways: Craft IV, Craft 3, Craft 3, Control 3, Control 3 OR Craft IV, Control 3, Control 3, Control 3, CP3. Which is more cases have that 3 CP has saved my ass more then the craftsmenship ever has.

Now don't get me wrong, I am not saying your wrong or anything of the such. but I am just speaking from experience, I had all 50 crafts before I even had 1 50 DoW/M and I am one of the first people on my server to get to new crafting content and have been since the re-release

I have warned Neverdyne about rapid synth because of the reasons you have stated, its honestly too much of a gamble to rely on a 80% success rate skill for progress, I have only ever used RS when I was doing Master Book 2's, and only then.

But while we are end game crafters, we have to keep in mind that the goal of this plugin is to provide rotations and the ability for anyone to reasonably use it.

I would enjoy seeing and hearing what other crafters do for their rotations as I have yet to find anything that gives me a higher success rate then what I am currently doing.

 

I apologize, my entire post wasnt aimed at you. Just the craftmanship part.. to save from an extra post, I added information about the rotation point of view for neverdyne, in regards to his post. I'm sorry for any type of confusion.

as for the meld for control 2 or cp2.. that is because the artisan body caps at +6 control.. so that last meld, even if you do a tier 3, will only give you +2 control.

and you are very much correct, the stats i listed were preference, based on my own usage, and a majority of the crafters on reddit, and official forums.

 

I assumed it wasn't entirely aimed at me, but I felt to explain in depth to avoid any confusion it might have given you or other users

Last slot would only give +2 assuming you do a tier IV on the 4th slot, which not many do, but that wasn't defined in the discussion so I assumed the 4th slot would be melded with a grade 3 like most are.

Though melds aside, as long you do something like what either of us have posted, the rotations that I am suggesting will work perfectly The only requirement is that you have 451 craftsmenship and 407 Control, anything more is just better. And there is no "real" CP requirement but ~395 is ideal for it, have less and you just need more ToT to proc.

 

Thanks for all the replies guys, they are really helpful since I'm kinda new to high level crafting. It seems like every high level strategy revolves around maximizing Inner Quiet stacks and finishing with a Byregot killzone. Miter's, Kage's, and even Yasuko's picture seem to depict this, but differ on the details on how to maximize the inner quiet stacks. The problem with simple macros is that it's hard to account for the variable extra CP you get from Tricks, and the randomness associated with HT, things that can easily be solved with behavior trees like the strategies Lisbeth uses. When I made Lisbeth, I did it to level up my jobs and was too lazy to do it manually. Now that I have everything maxed and am starting 3-4 star crafts, I'm still to lazy to do it manually, so my goal is to make Lisbeth the best crafter I can. The problem is that I don't want to make Lisbeth too specific, as in "it is great in 4 star 80 dura crafts when you have 490 craftsmanship, but sucks on other situations" kinda deal.

I've been thinking there's two roads Lisbeth can go to. One would be to create specific strategies for specific situations that arise, based on stats, which recipe it is, etc. The benefits is that Lisbeth would be capable of doing the best strategy for that specific case. The downside is that it would require many strategies to be coded in, one for each situation. The other road I could take is to try and make a more general strategy that revolves around the Byregot killzone, but the details on how to maximize inner quiet stacks beforehand would be searched for with some generalizing AI techniques; quite possibly search a tree for an ideal sequence of skills. The downside is that Lisbeth might not perform as good as the best strategy for each situation, but the benefits is that it can do all of them hopefully relatively well.

After a little bit more thought, I think I'll try and combine both roads. Realistically, I don't have the time to code in each strategy for each situation, accounting for each person's tastes. But what I can do, is make Lisbeth more of a "crafting engine" that you feed in a strategy, and provide easy ways to make strategies so that other people can customize them to their liking. It would be like a more powerful version of macros, since it would allow for more complex logic (things like CP prediction, Tricks, reclaim logic, etc). It would be like a template that you can customize. Then I would also offer a general Lisbeth strategy that revolves around a Byregot Killzone, but one that you personally might not like the details of.

I would really like your input on this.

Edit: As a side note, this will probably take some time to be finished, since I have to remake a bunch of stuff, but my goal is to have it up and running before the expansion hits. It would be a tremendous help if any of you would be interested in testing it.

 

Im here to help you test and even come up with rotations and strategies that will work for a broader spectrum of crafters as well as specifically targeted ones as well.

 

Just like counting coins.. you can have many different types of coins.. but the final amount will never change.

That being said.. as long as it works and gets highest HQ chance possible, it shouldnt matter on rotation speculation. There are many crafting rotations out there, and many that work well, but in the end.. they all do the same thing.. 3-4 star HQ crafting is all about inner quiet stacks.. the more stacks, the better. That is why people use hasty touch.. its a free ability and will give stacks of inner quiet.. the downside is the horrible rng that it seems to throw at people often.. thats why in the picture i posted, they say to replace hasty touch with a basic touch if you proc a ToT.. just gives it better chances of hitting.. This is another reason why Iyake's profile for lucis weapons would be a perfect example of ways to make strategies for Lisbeth.

The rotation works very well, and is basically flawless on how he coded it. That is why I asked in my previous post if you planned on making Lisbeth like that. When looking at your strategies, it just appeared that it was basing the rotation being used on pre determined crafting stats, then setting a rotation and not really watching the craft itself.. if that makes any sense..

having the ability for Lisbeth to see if ToT procs, and to count them and base its next ability off of certain conditions will make Lisbeth the most wanted plugin for alot of people.

As for the roads to take lisbeth down.. its a really good botbase for crafting as is.. you give people the option for macros, you made setting up the orders damn easy (the only thing i would suggest with that is to have an extension file type specifically set for lisbeth, so people could make orders for multiple items and save them to allow simple selection of orders that are most commonly used.. for example.. my sb sets. I have orderbot profiles for crafting/gathering/dps SB'ing gear that will create every piece.. and all i do is select the set i want to craft). The strategies that it has now are perfect for every day crafting, and leveling the crafts in themself.. even with the vendor artisan gear that can be bought.. doing 2 star using current strategies would be extremely simple to accomplish. But when going into the realm of 3 and 4 star crafts.. more attention needs to be given if an ability lands or not.. and what to do if it doesnt.. instead of blindly creating a rotation based off current stats.

As Kage and I were discussing, many people have their preferred ways to meld their gear, and rotations that they use.. but in the end the outcome is.. everyone and their mother just wants to HQ the damn thing lol.. giving the strategies the ability to know if previous ability were successful or not, and determining future abilities based on current stacks, number of ToTs that proced, number of hasty touches that land, ect.. will be a big landmark for Lisbeth.

There is alot of potential in it, and I hope I can see it continue to amaze.. I will shut up now and go to bed.

 

Sounds great to me One thing with 3/4 start crafts is as long as the bot can workout how much craftsmanship is needed to complete the craft and leave 1 step just for the end the rest is just fitting in as many touches as possible. There are a few ways of doing this which I have tried as im a 4 star crafter. But in general it relies on ToT and using Masters Mend, Manipulation to restore Durability. Id just like it to be able to calculate how to get as many in as possible and have enough CP for the "big finish". If we could say use Reclaim if not hitting a certain amount of stacks that would be great.

What would be real nice would be if it could "bait" the system. So if you have enough craftsmanship to complete the craft and only need 1 step to finish. It then builds stacks but you are in a situation where your only going to get a good HQ% if you hit a Good condition or Excellent. I guess im saying that this is a craft that your going to reclaim unless you can hit good or excellent which could therefore take you up to a acceptable HQ%.

After writing that an option specify a % HQ or reclaim would be nice.

I dont feel like I have explained it that well but hopefully somebody will understand and translate!

 

Between Miter, Yasuko and myself, Neverdyne has more then enough resources for crafting, though its up to him on how he takes it and codes and makes it work best for Lisbeth. Though regarding your HQ % idea, we can not read the HQ% of a craft as far as I am aware, we can see progress but not quality, though I could be wrong. In my message to Neverdyne I had suggested something along the lines of casting Reclaim if the stacks of IQ arnt high enough, though is assumes that the rotation would not take risks, so I also suggested some things that would take risks and never use reclaim.

 

Oh thats a shame about the HQ%. Yeah it all sounds great, its exciting to see such things and people wanting to have input. Will make what already is an excellent plugin even better!

 

@Yasuko Lisbeth already does most of that. All three strategies already take into account tricks and such, it does mend - hasty cycles until there isn't enough CP for another and then does the byregot finisher. They also take into account excellents and what to do if a good appears in the byregot finisher. They also predict what each synthesis skill will give you in progress based on your stats. Lisbeth's strategies have never been a "look at your stats, then set a rotation", but rather are based on behavior trees. But as I said I'm not that experienced at crafting so I didn't know if the current decision making was the correct one.

You can actually see quality. Right now with the last update, the strategies will finish the craft straight up if the quality is full. What I haven't yet implemented is quality prediction based on a formula, don't know how useful that would be.

 

Oh I didnt know it took into account excellents and goods, thats awesome. Thanks.

 

I coded in food and some updates to the Starry Night Miter profile, but I'm not sure how best to share or collaborate with them.

 

Hmm, I'm not entirely sure. I think I might try using GitHub again and I could make you a contributor so you can make updates, but the last time I tried GitHub it was giving people problems with the Yaml files. Otherwise, you could send it to me and I'll add it.

 

I'm wondering if there's some way to make the choice between profiles based on the durability of the item. I was trying to make some 80 durability crafts which required components with 40 durability, and it was stuck using the same strategy for both which ended up not being the best use of CP unless I made orders separately for each one part. If I tried to add them to the same order together it complained about not having enough materials because it didn't know it was going to make the components I needed with the other strategy.

 

All components are made using Hasty And Kinky, which might not be the best for high level components. If you want to use a different strategy for them, the best thing is to make the order for the components first, and once finished change the order to the final item. You're right that it won't take into account components made by orders themselves when calculating what it needs, this is on purpose but as I said to another user here, I'll try give you the option on settings to change the behavior so it does take into account what previous orders produce when calculating for the next. My reasoning was that, say you need 5 Iron Ingots and 1 Iron Sword for a quest, and you put that on an order, that's what you'll get. Else you'll end up with less ingots because they were used for the sword.

 

How do I express my love for you? I will have your babies!

Ok so after some testing I have noticed quite a few inefficiencies being a main crafter.

1. Inner Quiet needs to be the 2nd thing you do in any craft that you will use it with. It's highly inefficient to use that mid craft as you are likely to waste something be it SH2 or some other buff.

2. Comfort Zone -- At least for the starry miter profile, its using it once too much imo. CZ should typically only be used twice and if you are really lucky with ToT procs 3 times. However the miter is using it even when it will keep you from using MM2 a second time.

3. Rapid synth, while nice, is unreliable to use 4 times. For my stats I use pbpx2, rapid synthx1, basic synthx1. Yes this will increase your xp use by 45 but you save it without using comfort zone and is much more reliable.

4. With 20 durability left and an excellent, the profile chooses to use MM2 instead of a touch.

I'm no pro at coding but this is just my input. #3 is pure prefrence but the others are pretty standard imo.

 

View attachment updates.zip

This is the patch, the food is hardcoded in, didn't mess with the settings. And there's still an outstanding issue with Starry Night Miter where sometimes it doesn't leave enough CP for Byregots.

 

Hi, last night was the first time I attempted to use Lisbeth but everytime I try to load it up, I am plagued with RB crashing. Here is the part of the log that occurs right when I start it.

Code:

[10:47:38.462 N] [Lisbeth] Loaded 3136 recipes.
[10:47:38.475 N] [Lisbeth] Loaded 1 macros.
[10:47:38.476 N] [Lisbeth] Craftsmanship: 296 | Control: 327
[10:47:38.491 N] [Lisbeth] ROOT
[10:47:38.492 N] [Lisbeth] Craft | Natron (5522) | Amount: 4 | Retrieved: 0 | Crafts: 4
[10:47:38.492 N] [Lisbeth] As you wish, master!
[10:47:38.689 N] Connection to 192.99.148.87:31214 established.
[10:47:38.691 N] [Lisbeth] Agil, get your ass over here and help me.
[10:47:38.696 D] System.NullReferenceException: Object reference not set to an instance of an object.
   at Lisbeth.LisbethModule.GetAgil() in f:\Munkee\Desktop\RB\BotBases\Lisbeth\General\LisbethModule.cs:line 61
   at Lisbeth.LisbethModule.GetBehavior() in f:\Munkee\Desktop\RB\BotBases\Lisbeth\General\LisbethModule.cs:line 17
   at Lisbeth.LisbethBot.get_Root() in f:\Munkee\Desktop\RB\BotBases\Lisbeth\General\LisbethBot.cs:line 42
   at ff14bot.TreeRoot.()
   at System.Threading.ThreadHelper.ThreadStart_Context(Object state)
   at System.Threading.ExecutionContext.RunInternal(ExecutionContext executionContext, ContextCallback callback, Object state, Boolean preserveSyncCtx)
   at System.Threading.ExecutionContext.Run(ExecutionContext executionContext, ContextCallback callback, Object state, Boolean preserveSyncCtx)
   at System.Threading.ExecutionContext.Run(ExecutionContext executionContext, ContextCallback callback, Object state)
   at System.Threading.ThreadHelper.ThreadStart()

 

Lol, thanks!

@Munkee Yes, that's my bad, but there's a workaround. The crash happens when "UseAgil" is set to True on the settings, but you don't have Agil installed. To avoid it, change it to false like so: "UseAgil: False" I'll add a check next version so it never happens either way.

@Miter Thanks, I'll add food support through the settings soon. That's the good thing about using yaml files, it's easy to add new stuff. Right now I'm reworking the way Lisbeth does crafting completely, so we'll see how it goes.