Sign in to follow this  
Sevenless

Some thoughts on the skillgain system

Recommended Posts

Back with some more small sample sizes. This is as always not meant to be an exhaustive study, as much as it is a guideline on how/where you can do testing yourself if you're so inclined. Otherwise feel free to take this with a grain of salt. I do this for myself, but I publish my thoughts in case anyone else wants to read.

 

Mining:

 

It's a relatively known fact (so I'm told) that mining skillgain only happens when the ore produced is above 1ql, and below 40ql. It seems quite likely that mining produces results closer to the surface, and as such I believe that's actually the skill roll you achieve (it takes into account your skill, your characteristics, your tool and is hard limited to your skill or the ore vein's quality).

 

I did some testing to verify, and indeed can confirm that 39.99ql ore gives skill, 40.03ql does not and constantly. Interestingly, this does not apply in any way to digging, which clearly indicates not all skills are working off this as a base model. Hats off to the devs for the obscurity this system manages.

 

I've also noted that using skiller picks below lvl 50 mining is a waste. I'm getting *very* good gains at 52 mining off rock tiles with a q1 pickaxe. Higher than I've ever had off rock in the past using my q1 pick, and far better than iron or lead. Since rock is the easiest minable substance, it's pretty clear I shouldn't have been using a skiller pick up to this point.

 

Final point to note: Skill tick sizes are purely based off action time (excluding CoC) in mining. Doesn't matter what you mine, gain value is the same size per second for your skill level. Success range for tick values is unaffected by CoC (40.01 still does not give skill with a 50coc pick)

 

Farming:

 

So the first thing I noticed about farming was the tendency of the messages to have an increasing order of... complimentariness(?)

  • The field is tended
  • The field is now tended
  • The field looks better after your tending
  • The field is now groomed
  • The field is now nicely groomed

So I thought to myself, what if these aren't random? Sure enough, the pattern across crops is different. Hard crops I'm far too underskilled for (garlic) always gave "is tended" while easier crops like cotton gave me a mixture. And so I revised the descriptions with success ranges that made sense.

  • The field is tended - 1, failure
  • The field is now tended - 1.01->25, weak success
  • The field looks better after your tending - 25.01->50, mediocre success
  • The field is now groomed - 50.01->75, good success
  • The field is now nicely groomed - 75.01->100, excellent success

And interestingly enough, it shows the same pattern as mining does (albeit rounded because the success is harder to read directly in the 25->50 range).

 

  • The field is tended - never skill gain
  • The field is now tended - always skill gain
  • The field looks better after your tending - usually skill gain
  • The field is now groomed - never skill gain
  • The field is now nicely groomed - never skill gain

Before you ask, no the value of success doesn't seem to influence crop output volume. I get the same volume of cotton as I do corn even though I succeed a lot on cotton and not nearly as much with corn.

 

And to sum all of this up, yes rake quality does seem to have an impact. It also works like mining with regards to that. This is my data for raking corn with 54 farming skill.

KcyZ0ni.png

 

Where would optimal skillgain be? I'm estimating that 1 "is tended" per 3 "looks better" since ~1/3 looks better will be overkill on the skill and fail. But the skillcurve might not be that easy to guess. Certainly this requires tinkering with rake quality in order to optimize your gains.

 

Characteristics:

 

These seem to work quite differently than skills do. Namely, it seems that the success check for them to gain skill appears to ignore tool quality/skill level entirely, and instead only goes off the relative difficulty of the task compared to your current Characteristic level. This is predominantly seen in priest Soul Depth gain. Stew holds as a steady source of Sdep, and outperforms meals, until much higher Sdep value (

According to mamadarkness, by 50 Soul depth meals had edged out stews). I've also noticed further support for this when doing my raking: Farming corn gives much worse Soul Depth to my 27 soul depth priest than farming cotton seems to. And doing high level tasks like raking garlic almost never gives gain of any characteristic.

 

My Conclusions:

 

Farming does have skill success feedback

Rake quality matters and adjusts crop difficulty

Characteristics will likely be benefited most by doing tasks that are well below high end skill values

Skiller tools shouldn't be used for mining (possibly other tasks) until 50 skill roughly.

Edited by Sevenless

Share this post


Link to post
Share on other sites

Back with some more small sample sizes. This is as always not meant to be an exhaustive study, as much as it is a guideline on how/where you can do testing yourself if you're so inclined. Otherwise feel free to take this with a grain of salt. I do this for myself, but I publish my thoughts in case anyone else wants to read.

 

Mining:

 

It's a relatively known fact (so I'm told) that mining skillgain only happens when the ore produced is above 1ql, and below 40ql. It seems quite likely that mining produces results closer to the surface, and as such I believe that's actually the skill roll you achieve (it takes into account your skill, your characteristics, your tool and is hard limited to your skill or the ore vein's quality).

 

I did some testing to verify, and indeed can confirm that 39.99ql ore gives skill, 40.03ql does not and constantly. Interestingly, this does not apply in any way to digging, which clearly indicates not all skills are working off this as a base model. Hats off to the devs for the obscurity this system manages.

 

I've also noted that using skiller picks below lvl 50 mining is a waste. I'm getting *very* good gains at 52 mining off rock tiles with a q1 pickaxe. Higher than I've ever had off rock in the past using my q1 pick, and far better than iron or lead. Since rock is the easiest minable substance, it's pretty clear I shouldn't have been using a skiller pick up to this point.

 

Final point to note: Skill tick sizes are purely based off action time (excluding CoC) in mining. Doesn't matter what you mine, gain value is the same size per second for your skill level. Success range for tick values is unaffected by CoC (40.01 still does give skill with a 50coc pick)

 

Farming:

 

So the first thing I noticed about farming was the tendency of the messages to have an increasing order of... complimentariness(?)

  • The field is tended
  • The field is now tended
  • The field looks better after your tending
  • The field is now groomed
  • The field is now nicely groomed

So I thought to myself, what if these aren't random? Sure enough, the pattern across crops is different. Hard crops I'm far too underskilled for (garlic) always gave "is tended" while easier crops like cotton gave me a mixture. And so I revised the descriptions with success ranges that made sense.

  • The field is tended - 1, failure
  • The field is now tended - 1.01->25, weak success
  • The field looks better after your tending - 25.01->50, mediocre success
  • The field is now groomed - 50.01->75, good success
  • The field is now nicely groomed - 75.01->100, excellent success

And interestingly enough, it shows the same pattern as mining does (albeit rounded because the success is harder to read directly in the 25->50 range).

 

  • The field is tended - never skill gain
  • The field is now tended - always skill gain
  • The field looks better after your tending - usually skill gain
  • The field is now groomed - never skill gain
  • The field is now nicely groomed - never skill gain

Before you ask, no the value of success doesn't seem to influence crop output volume. I get the same volume of cotton as I do corn even though I succeed a lot on cotton and not nearly as much with corn.

 

And to sum all of this up, yes rake quality does seem to have an impact. It also works like mining with regards to that. This is my data for raking corn with 54 farming skill.

KcyZ0ni.png

 

Where would optimal skillgain be? I'm estimating that 1 "is tended" per 3 "looks better" since ~1/3 looks better will be overkill on the skill and fail. But the skillcurve might not be that easy to guess. Certainly this requires tinkering with rake quality in order to optimize your gains.

 

Characteristics:

 

These seem to work quite differently than skills do. Namely, it seems that the success check for them to gain skill appears to ignore tool quality/skill level entirely, and instead only goes off the relative difficulty of the task compared to your current Characteristic level. This is predominantly seen in priest Soul Depth gain. Stew holds as a steady source of Sdep, and outperforms meals, until much higher Sdep value (

According to mamadarkness, by 50 Soul depth meals had edged out stews). I've also noticed further support for this when doing my raking: Farming corn gives much worse Soul Depth to my 27 soul depth priest than farming cotton seems to. And doing high level tasks like raking garlic almost never gives gain of any characteristic.

 

My Conclusions:

 

Farming does have skill success feedback

Rake quality matters and adjusts crop difficulty

Characteristics will likely be benefited most by doing tasks that are well below high end skill values

Skiller tools shouldn't be used for mining (possibly other tasks) until 50 skill roughly.

Characteristics also ignore enchants. 

Share this post


Link to post
Share on other sites

I'm too stupid to comprehend most of the words coming out of your fingers, but um, is this why I got squat in skill gain from my 99 ql rake?


Share this post


Link to post
Share on other sites

I'm too stupid to comprehend most of the words coming out of your fingers, but um, is this why I got squat in skill gain from my 99 ql rake?

Yup

Share this post


Link to post
Share on other sites

So..... Does this mean that mining on a poor or acceptable vein will always result in skill gain?

Share this post


Link to post
Share on other sites

So..... Does this mean that mining on a poor or acceptable vein will always result in skill gain?

 

No, even though you're limited by the vein ql, it can still register a more successful mining action, as if you mined a 60 ql ore, or at least that's my understanding.

Share this post


Link to post
Share on other sites

Gotcha. So high ql vein and low ql pick, or just low ql pick? Also, do we know these hidden difficulties or is it just understood?

Share this post


Link to post
Share on other sites

No, even though you're limited by the vein ql, it can still register a more successful mining action, as if you mined a 60 ql ore, or at least that's my understanding.

 

That is correct.  Skill gain is what you would have gotten.

 

This true with rares, for example, that work the other way around.  A rare will boost the QL of the shard you get, so even if you "fail" the mining action (get a 1 QL shard), but the rare boosts it up to a 2 QL shard, you still get no skill even though it falls within the 1.01-39.99 range.  For skill purposes, you still got a "1" QL shard.

Share this post


Link to post
Share on other sites

I'd say the low ql pick is the most important thing and just focus on the difficulty of the vein type, not the ql.  To be fair, I haven't bothered to do extensive testing on it, just a gut feeling.


Share this post


Link to post
Share on other sites

I'd say the low ql pick is the most important thing and just focus on the difficulty of the vein type, not the ql.  To be fair, I haven't bothered to do extensive testing on it, just a gut feeling.

 

If you do this before 50-55ish skill, you'll end up with a lot of failures.  You're hurting your own mining at this point.

Share this post


Link to post
Share on other sites

The skill check ignores quality cap of the vein and your skill cap (the ore does not, which is why you can gain skill just fine off a q10 vein).


 


Mining is by far the best skill for directly observing the formula at work, but that skill cap muddies the water at low skill since you can't observe the required range to monitor progression.


 


There is no perfect pickaxe. You'll always be suboptimal without doing your own testing and modifying of your pick quality. Q1 pickaxe after 50 is acceptable if you start on rock at that point and move to iron when appropriate.


Edited by Sevenless

Share this post


Link to post
Share on other sites

Final point to note: Skill tick sizes are purely based off action time (excluding CoC) in mining. Doesn't matter what you mine, gain value is the same size per second for your skill level.

So at 72 mining I should get the same gain from rock, zinc or gold, regardless or the vein quality? Strange, this contradicts the wiki, discussions here, and my own testing a few months back.

Share this post


Link to post
Share on other sites

i did read at some point an advice on using a pickaxe of ql = skill/5 an i seems to work for me, I´m at 74ish mining a 50ql caped lead vein with a 15ql pickaxe, and about 60-65% of the mined ores fall in the range of 1.01-39.99


 


I didn´t tried to imp or decrease the ql of the pickaxe, but i have the feeling that if I go lower I will get more 1 ql crit fails, and if I go higher I will get more 40+ or 50 ql caped ores.


Share this post


Link to post
Share on other sites

So at 72 mining I should get the same gain from rock, zinc or gold, regardless or the vein quality? Strange, this contradicts the wiki, discussions here, and my own testing a few months back.

 

He said the skill tick will be the same.  Nothing about the frequency.

 

 

i did read at some point an advice on using a pickaxe of ql = skill/5 an i seems to work for me.

 

That doesn't make much sense to me.  If I'm at 80 skill with a 16 QL pick (80/5=16), that would be total overkill for rock and I'd be failing almost everytime with gold.

 

You have to take into account skill, QL of pick, and the substance you're hitting.  You can't look at one variable and call it a day.

Share this post


Link to post
Share on other sites

So at 72 mining I should get the same gain from rock, zinc or gold, regardless or the vein quality? Strange, this contradicts the wiki, discussions here, and my own testing a few months back.

 

As hailene said, not the same total skill gain, but the same tick size per second spent on an action. The length of the tick is modified by your pickaxe quality so you'd need to get ticks off multiple sources with the same pick. This does not hold true for all skills, but mining and farming it does.

 

Now I'm curious if longer ticks are perfectly proportional or not. Hm Hm, testing time.

 

19.5s - First Action

Rock

[07:16:56] Mining increased by 0.00597 to 52.7069

[07:18:13] Mining increased by 0.00597 to 52.7129

Iron

[07:23:15] Mining increased by 0.00598 to 52.7432

[07:25:10] Mining increased by 0.00597 to 52.7643

30.2s - Second action

Rock

[07:18:42] Mining increased by 0.00919 to 52.7221

[07:20:43] Mining increased by 0.00918 to 52.7372

Iron

[07:23:45] Mining increased by 0.00917 to 52.7524

[07:25:39] Mining increased by 0.00917 to 52.7735

 

There's your proof. Q1 pickaxe at 52.7 mining on rock vs iron, identical gain size, not identical frequency however (had to mine x2 as much iron to get those ticks). Also they are perfectly proportional to my action timer. The slight dip over time coming from my gaining skill which lowers the overall tick size accounted for that is.

Edited by Sevenless

Share this post


Link to post
Share on other sites

Interesting post. I have always found it funny we use low ql tools for the best gains.

Share this post


Link to post
Share on other sites

Interesting post. I have always found it funny we use low ql tools for the best gains.

 

Part of that has to do with eyeballing things. A cursory view of low Q tools shows larger gains (because they have longer timers, therefore more gain per action if not per second). Also, most people who bother to test are high skilled players, and after 50 skill skiller picks become useful in some situations.

 

So yeah, human tendency to see patterns doesn't always lead us to the correct pattern.

  • Like 1

Share this post


Link to post
Share on other sites

I would like to add that skiller picks are basically picks with high CoC, saying skiller pickaxes are useless before 50 mining is false imho since a ql40 pickaxe with pure CoC is also a "skiller" and will net you more skill than an equivalent pickaxe with no CoC.


Not a big issue but it just bugged me to read that :P


 


Also thanks for the info on farming, that really helps me find out which crops I should be farming.


Share this post


Link to post
Share on other sites

A skiller pick as Revelation pointed out is just a pickaxe that you keep at a certain quality depending on your mining. Using a top QL pickaxe before 50 mining will still give you poorer results than a pick at your ideal QL for your skill. You have to play around and find your optimal QL. When I was about 80 mining I found a 15 or so QL pick gave me the best skill and mining silver ore at about 85 skill gave me the best mining.

Stats raise the same no matter the QL of pickaxe, CoC does help I'm pretty sure and WoA has no effect on stat gain. I also find that stats raise the same regardless of if you mine rock or ore, so if you wanted purely stats you could use a 90ql 99woa 99coc on rock/ore and you'd get just as good stats as anything else, aslong as you have the coc. Your mining would raise hell slow though so it's better to just use a skiller pick and mine the appropriate ore/rock and get the best gains in both stats and mining.

This is why you should put CoC even on speed pickaxes, because while you use it for mining good QL materials you'll still get some nice stats, then switch to your skiller when you want to grind mining too.

Share this post


Link to post
Share on other sites

I would like to add that skiller picks are basically picks with high CoC

 

They are most of the time referred to as skiller picks when they are below quality 10, often below quality 5, when being sold by priests. True they are enchanted, but a q90 pick with high coc is not called a skiller pick 95% of the time in my experience.

 

As such, I interpret the term "skiller" to refer to tools intended to be used at low tool quality. And I still hold true that skiller picks in that sense are useless before 50 because q1 pick is ideal for mining rock at 50, not before. CoC is never bad to have on picks, and you're right that 100 CoC on a pick that is the wrong quality for the ore you mine can outperform a 0 CoC pick that is in the correct range.

 

Our use of the terms are different. I believe I have reason to use the term the way I do, but I respect that you interpret it differently.

 

Edit: Likely this concept of "only tools below 10 are skiller" comes from the somewhat blind assumptions most players have about the skill system. It's more complicated than the average player wants to think about so we go with simple mantras like "always use low Q tools for skilling".

Edited by Sevenless

Share this post


Link to post
Share on other sites

He said the skill tick will be the same.  Nothing about the frequency.

 

 

 

That doesn't make much sense to me.  If I'm at 80 skill with a 16 QL pick (80/5=16), that would be total overkill for rock and I'd be failing almost everytime with gold.

 

You have to take into account skill, QL of pick, and the substance you're hitting.  You can't look at one variable and call it a day.

 

I never claimed the only factor of influence in skill gaining is the ql of pickaxe.

 

But I think using a skill/5 in the corresponding ore as per your level, will yield you a decent gain.

 

Of course you can always make your own experiments and play with different tool ql/enchants/vein type & ql until you get the super sweet spot for your particular skill.

Share this post


Link to post
Share on other sites

Of course you can always make your own experiments and play with different tool ql/enchants/vein type & ql until you get the super sweet spot for your particular skill.

 

This doesn't work for a couple of reasons.

 

1.  The "corresponding ore" depends on mining skill and pick QL.  So the right vein you're swinging at depends on those two.

 

If you meant the ore order listed in the Wiki (1-50 rock, 51-60 for iron, ect.) then that still doesn't make sense.  At 10 skill, using a 2 QL pick will cripple your skill gain.

 

 

2.  Improving your pick as you approach the end of an ore is actually the exact opposite of what you want.  According to your system, you ought to have a 10 QL pick when you start mining iron (50 mining skill) and then a 12 QL pick when you hit 60 mining skill (as you finish iron).

 

It's the opposite.  When you break into a new vein, you want your pick to be higher (I'd say around 25 if you're going to dive into iron at exactly 50).  When you leave iron at around 60, the QL will probably drop down to 10-12.

 

So, in short, your system, while simple to implement, is by no means optimal.

Share this post


Link to post
Share on other sites

Lets assume the QL chance is actually a bell curve. For perfect skill gain the center of the bell curve would have to be right at 20QL. This leaves a range of 20QL to both sides before you get fails (1.0QL) or to good ore (>40QL). As a consequence the best skill gain would be possible when you get the same amount of 1QL ore as >40QL ore.

During the last mining session I was still getting about twice as much 40QL ore than 1QL ore and yet I got close to a 50% skill gain ratio. Lowering the pickaxe QL a litte would move the center of curve towards the lower QLs and should give more fails but less 40QL ore.

An estimated optimal skill gain of 50% and an equal distribution between 1QL and 40QL gives us an good estimate for mining capped ore. If you get about 25% 1.0QL Ore, you would also get about 25% capped would-be-40QL ore and about 50% inbetween which give skill.

Share this post


Link to post
Share on other sites

Lets assume the QL chance is actually a bell curve. For perfect skill gain the center of the bell curve would have to be right at 20QL. This leaves a range of 20QL to both sides before you get fails (1.0QL) or to good ore (>40QL). As a consequence the best skill gain would be possible when you get the same amount of 1QL ore as >40QL ore.

During the last mining session I was still getting about twice as much 40QL ore than 1QL ore and yet I got close to a 50% skill gain ratio. Lowering the pickaxe QL a litte would move the center of curve towards the lower QLs and should give more fails but less 40QL ore.

An estimated optimal skill gain of 50% and an equal distribution between 1QL and 40QL gives us an good estimate for mining capped ore. If you get about 25% 1.0QL Ore, you would also get about 25% capped would-be-40QL ore and about 50% inbetween which give skill.

 

Even if it's a standard distribution, there's no reason why that curve needs to have only 50% between 1 and 40 though. The standard deviations could be larger or smaller than that, and who knows if pick/skill influence its breadth.

 

That's what I'm curious about. If I used a Q90 pick on gold, would the curve be the same shape but shifted horizontally only? If so I could probably mine silver/gold with a high quality tool if I so chose. And yet with a q1 tool, I'm mining rock for great gains. But if the band gets compressed, higher would be better. If it gets extended, lower difficulty is better. Ignoring of course the lower actions = better characteristics gain, which would probably skew my choice anyway.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this