Strategy to make a 91+ LP genie

Oups_Dead - Harshlands

not sure if it helps but once i had 71/90 genie ppl told me if i wanted an 81+ genie to stop lvlimg it and sell it right away "cuz i had no chance" i was also almost out of spirits but i didn t listen nd guess what ? 81/100 genie i linked it in guide and was like "take that in ur faces guys !!!!!"

Tide_Surfer - Archosaur

Just make a rule up and hope Lady Luck pays you a visit in my own opinion, I've never made a genie over 77/100 LP (My current one which I am keeping, it was my lvl 1 genie) p.s. big fan of your youtube Adroit

HrunsPanda - Archosaur

Very interesting, problem.

With the medium scores being very rare, i think it is mainly a question of where do you want to make the swich from having a score that alows for 1 low score to come in. like for example you have an 80/80, then you can have a low score and still get to your 91 while a 77/80 cannot have a low score and still make 91. Since a medium score is very rare, a 77/80 therefore is barely better than a 73/80.

So you could start out accepting this 27/30 knowing that every time you will need to score a 9 or 10. Or you could in he beginning accept only 30/30 allowing you to get 1 low score and all others 10 except for 1 or 2 9s. Since the first levels cost so little exp, i think it would be best to ditch anything that doesnt get 10 points, but up to what level ? This is something that we should be able to solve mathematically in excel.

This however will create a godawfull amount of faillures in your inventory. So it may in the end come down to that, inventory space. Just get an inventory full of genies. Make em all level 10. Then take the 10/10s to level 20. Take the 20/20s to level 30 etc. Until you have no genie that theoretically can get to 91. Then proceed to take the ones that are higest level and least points away from perfection. Keep working your way backward until eventually maybe you will actually get to leveling the 9/10s. Again of course you will need to make choises here. After for example you have tried your 49/50s, whats next, the 48/50s or the 39/40s ?

So ye.... good question If i ever get the answer ill let you know b:pleased Sorry for wasting your time with this useless thinking out loud for now

Asterelle - Sanctuary_1381265973

Adroit - Lost City wrote: »

I really like the way you did that, a very simple/elegant solution to something I was making overly complicated.

I'm curious though, I'm trying to think of a way to use this GRI information in determining whether to give up or continue on a genie.. (or really just any method that makes sense), I'd assume you don't have the perfect answer or you'd likely have already shared (I've noticed you seem to be a little picky with your posts.. usually just dealing with facts and making sure whatever you say is correct.. a quality I really admire), but do you have any thoughts/ideas on determining a criteria for leveling genies? I can't tell if this is a really difficult problem or running off 3 hours of sleep is affecting me.. but I just can't come up with any ideas that I'm happy with atm

Well an actual strategy is kind of tricky. With infinite spirit you can continue upgrading every genie until it has dropped 10 points or more from the maximum but that could be wasteful.

On average you need a little more than 9/10 LPs every 10 levels so I would just stop if you ever fall below that average. There's not much point in continuing a 33/40 genie when you could just restart.

The best genie I ever made was just 85/100. Then I bought a 93/100 for like 70m.

CRYSTY - Sanctuary

Chigenkaiona - Dreamweaver wrote: »

Take your wallet out, max out all your credit cards, max out your debit account, charge as much gold as you can. Say hello to your 91/100 genie. Then donate the spare gold to the rest of the community and you might get a 100/100 while your at it.

This ^.
I've read on forum or pwi blog somewhere ,long time ago when the genies were launched ,that if a genie would reach 100/100 points would be called "The Queen" if i remember.Did someone from any server got the queen ?? I doubt that b:chuckle

Anyways any genie same like any cash shop boxes or lucky items has "presets", the item/points are already chosen in server. (see those percentages on the wiki - Genie rarity by Asterelle)
The only real randomization in real time would be in an offline game not in an online one.Imagine if something from an online game would be randomized on the client' s computer, someone,somewhere, a curious hacker might try to modify the randomization process b:chuckle

Morale: I don't believe in random chances in online games , only in " custom server presets" for the money's sake b:chuckle

Speaking about genies: i hate when they nerfed tree of protection ability.

Mizuryu - Harshlands

It's easy.

All you do is draw a pentragram, chant a spell,
And ask the devil for one.



No kidding, but in all honest it's called lucky points.

HrunsPanda - Archosaur

CRYSTY - Sanctuary wrote: »

This ^.
I've read on forum or pwi blog somewhere ,long time ago when the genies were launched ,that if a genie would reach 100/100 points would be called "The Queen" if i remember.Did someone from any server got the queen ?? I doubt that b:chuckle

Anyways any genie same like any cash shop boxes or lucky items has "presets", the item/points are already chosen in server. (see those percentages on the wiki - Genie rarity by Asterelle)
The only real randomization in real time would be in an offline game not in an online one.Imagine if something from an online game would be randomized on the client' s computer, someone,somewhere, a curious hacker might try to modify the randomization process b:chuckle

Morale: I don't believe in random chances in online games , only in " custom server presets" for the money's sake b:chuckle

Speaking about genies: i hate when they nerfed tree of protection ability.

I dont agree with your morale.

Many people have statistically analysed online casinos and pokerwebsites and found them to be random to a satisfactory degree. (true randomness doesnt exest, but it doesnt exist in real life either or well, thats where the whole philosophical discussions come in about things like freedom of chose and destined futures) Why would an MMORPG server not have a proper random number generator ?

Magnanimous_ - Heavens Tear

You can buy your own USB-TRNG (True Random Number Generator) as cheap as 100 bucks.. ijs

serbetel

Getting lucky has nothing to do with a plan.
A 1/10 has same change to hit 91 as a 10/10.

Nearly all my "good" genie s started with "bad" stats
but that just shows randomisation works well
its not a rule or somehting.

Asterelle - Sanctuary_1381265973

serbetel wrote: »

A 1/10 has same change to hit 91 as a 10/10.

You're off by a factor of 1,391.

serbetel

Asterelle - Sanctuary wrote: »

You're off by a factor of 1,391.

That depends on the statistical dogma
but considering you might use only 10/10 genies..
..how many chances would you throw away and how would it influence your statistic ?

EDIT:
In Online-Games there s only 2 ways to hand out "random" items,
fixed or dynamic.

Fixed would mean pwi creates 50 mio genie s and hand them out one by one,
that would mean they not only risk to run out of prebuilds,
it would also become exploitable cause once the "Queen" is handed out
nobody would need to try anymore.

Dynamic creation means that ofc such a initializer as such a table can be used
but the result is open .. that means there really can exist randomness.

If you really try to predict random numbers
then its more likely that a random sequence hit s 91
for example 9, 7, 10 ..
then a Sequence like 10, 10, 10, 10 ..
simply because more variations exist.

But again thats just statistical dogma,
luck is luck.

Oups_Dead - Harshlands

CRYSTY - Sanctuary wrote: »

This ^.
I've read on forum or pwi blog somewhere ,long time ago when the genies were launched ,that if a genie would reach 100/100 points would be called "The Queen" if i remember.Did someone from any server got the queen ?? I doubt that b:chuckle

Anyways any genie same like any cash shop boxes or lucky items has "presets", the item/points are already chosen in server. (see those percentages on the wiki - Genie rarity by Asterelle)
The only real randomization in real time would be in an offline game not in an online one.Imagine if something from an online game would be randomized on the client' s computer, someone,somewhere, a curious hacker might try to modify the randomization process b:chuckle

Morale: I don't believe in random chances in online games , only in " custom server presets" for the money's sake b:chuckle

Speaking about genies: i hate when they nerfed tree of protection ability.

actually on HL someone had the queen genie ... however this guy got banned later idk why :P

TPstar - Sanctuary

Make a few alts, and level them to level 40, and use them to do BH29 and feed all exp/spirit into genie. You can make many genies and be more picky on lower levels, plus you won't feel so bad with the reduce ratios on feeding lower level genies. Those genie doesn't meet the requirement gets turn into exp cube on feeding those genies who made the cut.


You can even level a few more toons into the 70s (BH59) if you feel like to have another set of toons to level those genies who pass the first set of elimination.

XSonOfCircex - Sanctuary

serbetel wrote: »

Getting lucky has nothing to do with a plan.
A 1/10 has same change to hit 91 as a 10/10.

That's just stupid. b:surrender

To get 91+ with a 1/10 genie means you need to get 90/90 of the remaining lucky points, while a 10/10 requires only 81/90.

If you had an infinite amount of spirit it might be worth it, but even at 300m or however much you have there is a finite amount left. Why waste it on genies that are less likely to be good?

HrunsPanda - Archosaur

TPstar - Sanctuary wrote: »

Make a few alts, and level them to level 40, and use them to do BH29 and feed all exp/spirit into genie. You can make many genies and be more picky on lower levels, plus you won't feel so bad with the reduce ratios on feeding lower level genies. Those genie doesn't meet the requirement gets turn into exp cube on feeding those genies who made the cut.


You can even level a few more toons into the 70s (BH59) if you feel like to have another set of toons to level those genies who pass the first set of elimination.

I havent researched it, but i am fairly sure that the exp you gain at lvl 100 compared to what you gain on level 40 makes up for the reduced ratio many times over since its an exponential exp curve and at lvl 100 i can kill mobs a dozen time more efficient than i could at lvl 40.

What i do however is use my lvl 77 zhenning seeker. When it is zhenning in the pits, it gains all this exp, but it has to stay at level 77. So genies are a nice by-product.

serbetel

XSonOfCircex - Sanctuary wrote: »

That's just stupid. b:surrender

To get 91+ with a 1/10 genie means you need to get 90/90 of the remaining lucky points, while a 10/10 requires only 81/90.

If you had an infinite amount of spirit it might be worth it, but even at 300m or however much you have there is a finite amount left. Why waste it on genies that are less likely to be good?

Look lets try make a 11/20
you go for [10, 10]
i take the remaining valid solutions:

[1, 10]
[10, 1]
[2, 10]
[10, 2]
..
[5, 6]
..
[9, 10]
[10, 9]

Now you can choose if you wanna try 100 times for something wich you can (statistically) get in 1 or 2 trys.

By throwing away all those potential candidates you not only waste ressources
you also waste chances (theoretically - since luck is luck).


Edit:
There are 2 wrong assumptions floating around:
1. since the LP are random you have the same chance at 80 LP like at 71 LP to hit 91 LP.
2. you are comparing a 80/80 with a 71/80 ignoring the costs the 80/80 produced.

Golden Rule of thumb - as long the Genie suits the minimum requirements of a set goal its worth it.

TPstar - Sanctuary

HrunsPanda - Archosaur wrote: »

I havent researched it, but i am fairly sure that the exp you gain at lvl 100 compared to what you gain on level 40 makes up for the reduced ratio many times over since its an exponential exp curve and at lvl 100 i can kill mobs a dozen time more efficient than i could at lvl 40.

What i do however is use my lvl 77 zhenning seeker. When it is zhenning in the pits, it gains all this exp, but it has to stay at level 77. So genies are a nice by-product.

Remember, the OP is 105. No longer gaining exp.

Adroit - Lost City

I may not be gaining exp, but I can still farm ~12-15m spirit a day if need be (although I hope I don't have to).

I have started simulating a couple strategies, will spend some more time tomorrow but thought I'd post em here in case anyone is curious. I haven't added exp cubes to my script yet, so that is one possible way to reduce the overall SP cost.

For my first strategy where I just continue on as long as I have at least 90% of the possible lucky points.. I simulated making 9373 91+ genies using this method and found that:

Average LP value: 94
Average Mirage Cost: 32220
Average SP Cost: 1079m (with a standard deviation of 800m SP)
Average # of 81-90 LP genies: 1.55
Proportion of these genies that were made with 630m or less spirit is: ~.434

Just for fun I also simulated using a method where as long as it was still possible to create a 91+ genie, I continued trying (which means would keep going on a 1/10 genie lol.. wouldn't stop until it was missing more than 9 lucky points. This method probably uses the minimum number of mirages, but I doubt is the most efficient method. I made 10012 91+ genies using this method to come up with the following numbers:

Average LP value: 93
Average Mirage Cost: 23460
Average SP Cost: 1720m (with a standard deviation of 1093m SP)
Average # of 81-90 LP genies: 1.90
Proportion of these genies that were made with 630m or less spirit is: ~.300


I'll have to come up with some new ideas to test tomorrow, if anyone has any they think would be efficient or just fun to try out.. lemee know I might give it a shot

serbetel

It is an illusion to assume a 80/80 would have a higher chance then a 71/80.
Simply because the chance to 9x 10 LP in a row is so much smaller then any other permutation.

All those ppl who aim for 80/80 seem to assume luck grows the more one is lucky,
thats just wrong.



The only thing wich is guaranted is the amount of existing permutations and their partials (chances).


Edit:
Adroit: Calculate all existing valid permutations (wich includes 80/80)
and set them in relation to that one solution 80/80
then you got a factor to calculate with.

It s like comparing the chance to win with 1 lottery ticket over a chance to win with 1000+ tickets.

Adroit - Lost City

serbetel wrote: »

It is an illusion to assume a 80/80 would have a higher chance then a 71/80.
Simply because the chance to 9x 10 LP in a row is so much smaller then any other permutation.

All those ppl who aim for 80/80 seem to assume luck grows the more one is lucky,
thats just wrong.



The only thing wich is guaranted is the amount of existing permutations and their partials (chances).

There are fewer possible ways to go from 71/80 to 91+/100 (only possible way is to get 10 twice) than it is to go from 80/80 to 91+/100 (could get 10 twice, or 10 then a 9, or 10 then an 8, down to 10 then 1.. or 9 then 10, 9 then 9, 9 then 8.. all the way down to 1 then 10). It is clearly much more probable to end up with a 91+ genie if you are starting from 80/80 than it is from 71/80..

Fissile - Archosaur

Adroit - Lost City wrote: »

I may not be gaining exp, but I can still farm ~12-15m spirit a day if need be (although I hope I don't have to).

I have started simulating a couple strategies, will spend some more time tomorrow but thought I'd post em here in case anyone is curious. I haven't added exp cubes to my script yet, so that is one possible way to reduce the overall SP cost.

For my first strategy where I just continue on as long as I have at least 90% of the possible lucky points.. I simulated making 9373 91+ genies using this method and found that:

Average LP value: 94
Average Mirage Cost: 32220
Average SP Cost: 1079m (with a standard deviation of 800m SP)
Average # of 81-90 LP genies: 1.55
Proportion of these genies that were made with 630m or less spirit is: ~.434

Just for fun I also simulated using a method where as long as it was still possible to create a 91+ genie, I continued trying (which means would keep going on a 1/10 genie lol.. wouldn't stop until it was missing more than 9 lucky points. This method probably uses the minimum number of mirages, but I doubt is the most efficient method. I made 10012 91+ genies using this method to come up with the following numbers:

Average LP value: 93
Average Mirage Cost: 23460
Average SP Cost: 1720m (with a standard deviation of 1093m SP)
Average # of 81-90 LP genies: 1.90
Proportion of these genies that were made with 630m or less spirit is: ~.300


I'll have to come up with some new ideas to test tomorrow, if anyone has any they think would be efficient or just fun to try out.. lemee know I might give it a shot

A couple points to consider. It looks like you let your script run, producing and killing genies, until it generates ~10,000 in the 91+ range. If that is the case, you may want to take a look at the random number generator you are using. Depending on your language of choice, the pseudorandom number generator may have a period as low as 16777216 (*glares at Visual Basic*). Making hundreds of millions of calls to a random number function with such a small period will give misleading results. I use an implementation of Mersenne twister that I wrote in VB (Excel) in 1996.

Also, I thought your number of 81-90 genies seemed really low, but I realize why. You might also want to track the number of 71+/90 genies you make that your algorithm throws out. Ordinarily you would save these and level them or sell them.

serbetel

Adroit - Lost City wrote: »

There are fewer possible ways to go from 71/80 to 91+/100 (only possible way is to get 10 twice) than it is to go from 80/80 to 91+/100 (could get 10 twice, or 10 then a 9, or 10 then an 8, down to 10 then 1.. or 9 then 10, 9 then 9, 9 then 8.. all the way down to 1 then 10). It is clearly much more probable to end up with a 91+ genie if you are starting from 80/80 than it is from 71/80..

Agree, but you cannot exclude the way to get to 80/80.

Especially since all those lower permutations like
7.5.10 or 10.2.8 are so much cheaper to get.

Consider them like lottery tickets.
The 80/80 People want to guess a number out of 1000+ just with 1 try.
The permutation people accept all valid combination of numbers wich is MUCH higher.


Maybe its easier to think of Lucky points as Symbols
(a=1, b=2, c=3 .. ) then imagine all the valid combinations.

soundslegit

serbetel wrote: »

It is an illusion to assume a 80/80 would have a higher chance then a 71/80.
Simply because the chance to 9x 10 LP in a row is so much smaller then any other permutation.

I think your problem is your mistakenly assuming that there is an equal chance of getting 1,2,3,4,5,6,7,8,9,10 lucky points every 10 levels. There isn't.

serbetel

soundslegit wrote: »

I think your problem is your mistakenly assuming that there is an equal chance of getting 1,2,3,4,5,6,7,8,9,10 lucky points every 10 levels. There isn't.

Thats my point.
But it starts much earlier not at 80/80.

You simply cannot calculate luck but every valid combination.

Let X be 10, what s wrong with:
1.x.x.x x.x.x.x
x.1.x.x x.x.x.x
x.x.1.x x.x.x.x
..
Simply nothing - all valid combinations.


It s not only math s its also personal experience,
no matter what chance to get a number - by reducing your chance to only 1 valid combination
you will 99 % sure drop the chance to really reach 91+.

Why would make pwi it that "easy" to onloy focus on 10/10/10/10..


Edit:
The chance come very equal but if you only choose 10.10.10.10 genies for example
then ofc the chance to get on is much lower
then to use all those variations of 9.10.10.10 / 8.10.10.7 / etc.

It s an illusion to think just cause 5x 50/50 come up with a 6th low number
all other valid combination would come up with a low number too.
Simply because you dropt all the other chances.

Adroit - Lost City

Fissile - Archosaur wrote: »

A couple points to consider. It looks like you let your script run, producing and killing genies, until it generates ~10,000 in the 91+ range. If that is the case, you may want to take a look at the random number generator you are using. Depending on your language of choice, the pseudorandom number generator may have a period as low as 16777216 (*glares at Visual Basic*). Making hundreds of millions of calls to a random number function with such a small period will give misleading results. I use an implementation of Mersenne twister that I wrote in VB (Excel) in 1996.

Also, I thought your number of 81-90 genies seemed really low, but I realize why. You might also want to track the number of 71+/90 genies you make that your algorithm throws out. Ordinarily you would save these and level them or sell them.

Hmmm, well this is taken from the docs of the random number generator I'm using (and I haven't added reseeding)

"If reseeding is never done by the script, the seed starts off as the low-order 32-bits of the 64-bit value that is the number of 100-nanosecond intervals since January 1, 1601. This value travels from 0 to 4294967295 every ~7.2 minutes."

Tbh I'm not really familiar enough with random number generators to know if this is sufficient or not, I should probably do a little more reading to find out. That does sound like a good idea to keep track of 71+/90 genies, I'll probably add that tomorrow or something. Thanks for the feedback

soundslegit

serbetel wrote: »

Thats my point.
.

So your point is that your wrong?

serbetel

soundslegit wrote: »

So your point is that your wrong?

You missunderstood something try again.

I compared "luck" with "valid permutations" not the chance to get to 91 with 80/80.

Even if you got a bad start like 1/10 in the end only the result matters,
thats what most pppl seem to confuse.




Those LP per decade are nothing like symbols,
there s nothing to calculate with.

Fissile - Archosaur

The (misguided) point he is making is that for every one 80/80 try with, you will have hundreds (or thousands) of 71+/80 genies you can try with. Of course, if you try with that many 71/80 genies you have a much greater chance of successfully reaching 91/100. This is backed up by Adroit's own simulation using the naive method (keep going til it is impossible to succeed).

But in doing that, you've also raised the spirit cost by 70% compared to the other strategy. Which is worth more, 10,000 mirages or 700 million spirit (2 months of daily PV, 2 tokens per day, with 4-6 runs per token + about 5-10 hp charms [1-2 charm ticks per run], since a wiz is going to have to take a beating to do runs that fast + any hp and mp food)?

Oups_Dead - Harshlands

can t we just say it depends on ur luck nd close the thread? :P

serbetel

Fissile - Archosaur wrote: »

But in doing that, you've also raised the spirit cost by 70% compared to the other strategy.

That would only be true if every level would require same amount spirit, but exp needed per level increase.

From my personal experience (using the best method available)
exp is the minor factor while time to produce/decompose is the major factor.



Maybe its easier to understand with another example,
if you try to build a garage for your car then the goal is set - 1 spot.
There s absolutely no reason to build 10 spots just for making sure the car gets 1.
( in fact every major company would kick workers that waste ressource that excesiive )

And again - LP are just symbols,
if you only try for the x.x.x.x x.x.x.x
while any combination of x.a.x.x x.x.x.x or x.b.x.x x.x.x.x is discardet
you limit yourself unnecessarly that much that its nearly impossible to hit the goal.