IMO the system does a query on the db for multisims and determines drop rate all at once. It doesn't do each multi sim independently. that's why i never multisim more than 3. If the true shard drop rate is supposed to be 33% then I at least should get one.
thats why i would never hit 20 multisims at once. So in that case you rolled a 5% drop rate. next time you could have rolled a 100% drop rate. It evens out. but not worth the risk.
superstition? probably. still i feel better doing it this way.
IMO the system does a query on the db for multisims and determines drop rate all at once. It doesn't do each multi sim independently. that's why i never multisim more than 3. If the true shard drop rate is supposed to be 33% then I at least should get one.
thats why i would never hit 20 multisims at once. So in that case you rolled a 5% drop rate. next time you could have rolled a 100% drop rate. It evens out. but not worth the risk.
superstition? probably. still i feel better doing it this way.
Agreed, I never sim over 3 rolls. Superstition or not, in my experience the drop rates at large rolls are dismal.
Well thsi is just my thoughts but some characters for me have some kind of golden number,yesterday i was farming pk,and i found out i get a shard EVERY TIME that i did a normal non ticket battle,maybe this was rng but i got 32 shards and unlocked him
If my maths are right, that's like 0.03% probability. 1/3000ish. Certainly not impossible. (Technically that's prob for 0/20, but 1/20 should be really close).
The probability for that should be (2/3)^19*1/3 = 0.00015.
IMO the system does a query on the db for multisims and determines drop rate all at once. It doesn't do each multi sim independently. that's why i never multisim more than 3. If the true shard drop rate is supposed to be 33% then I at least should get one.
thats why i would never hit 20 multisims at once. So in that case you rolled a 5% drop rate. next time you could have rolled a 100% drop rate. It evens out. but not worth the risk.
superstition? probably. still i feel better doing it this way.
Agreed, I never sim over 3 rolls. Superstition or not, in my experience the drop rates at large rolls are dismal.
it make sense to do it the way i described. less stress on the servers. So it takes the RNG % and times it by the # of multi sims i requested. If I get a roll of 32 or below i get zero shards, 40 i get 1 shard, if i get a roll of 68 i get 2 shards and if i roll a 90 or above i get 3 shards.
that's the only way i believe someone doing 20 sims could get one shard. they rolled a 5x20=1.
i have trouble believing the system did this instead .03x.03.x03 etc.........
i work with developers every day and i am no expert but everything they do is to give you full functionality with the least amount of load on the servers. everything they do is to try and minimize the strain on the servers. independent drop rate checks make no sense in a game this big and with this many users
When I'm farming character shards, I multi-sim 3 battles. In theory, I should get 1 every time. If I got 3 pulls and don't get a shard, I stop and try again later.
For purple gear, I do a multi sim of 5. Seems to get 1 each pull for the most part.
I multi sim all the time. (If I have the sim tickets). Haven't really kepted track of how often I receive a shard. I'm at 2/65. Two refreshes a day. Since the day he became available. My opinion, Rngesus is a big factor.
I was 3 for 4, 2 for 4, 1 for 4 just now. TFP was pretty good to me today. I haven't noticed him being horrible drop rate. Sounds like it's random, based on both our experiences.
The fact that you guys are still trying figure out drop rates is incredible.
It's 1/3 average. It has been proven time and time again.
I got 11 shards on a 20 sim this morning. So clearly the drop rates are fine. Yeah I know that's over 50% but I have had some o/4 and such. But overall its about 1/3 tries.
Use multi-sim, drop rates are better, I am not going to explain the probabilities yet again.
To whoever said they think the drop rates are better for TFP than B2 or Wedge, well no kidding. TFP is a 10 energy sim, where B2 and Wedge are 12. You get a few more sims with the same energy totals.
You naturally regen about 120 cantina a day, you get 45 at midday, and 2 refreshes gives you 240. So that's 405 cantina energy a day. That's 40 sims, 1/3 of that is about 13. On average, you should net about 12-13 shards a day. It's still going to take a solid 20 days to get to 7* unless you are refreshing more. If you get 12 in 24 hours you are doing just fine. You will get 18-20 some days, and you will get days with 6. But if you track it you will find you get 12 a day on average.
Sometimes drop rates seem lower for different toons for the simple reason that some nodes are 8, some 12, etc. B2 was a 12 energy node, and that only got 10 battles and an average of 3 shards each sim. However, now I'm farming SRP who is an 8 energy node, and therefore get 15 battles each time. As a result he averages more, usually 5, sometimes even 8 or 9.
@AlexHermanssonGr@goldenarmadi110 actually that answer is not correct. you need to think about the fact that there are a number of permutations for getting only 1 out of 20 (e.g. it could be on the first pull, second pull, etc).
The true answer should be 0.00015 as you calculated then multiplied by 20 (the binomial coefficient corresponding to picking 1 from 20) = 0.003 (rounded for simplicity) = 0.3%.
The probability for that should be (2/3)^19*1/3 = 0.00015.
Thanks...probability was never my strong suit!
It's not his strong suit either, lol. You draw 20 times with a probability of 1/3 each time, this is a standard binomial distribution. You can easily calculate the probability of 0, 1, 2 etc. hits with Excel using the BINOMDIST function like BINOMDIST(1;20;1/3;wrong). This results in 0.3% for 1 hit. To be fair, you should include 0 hits as well since you'd be even more upset then, so it's 0.33% for 0 or 1 hits out of 20 tries. That's 1 in 300, so indeed quite unlucky.
And as a programmer myself I call **** on all peeps who say multi-sims are worse. It has NOTHING to do with server load as the extra "load" from calculating 20 sims compared to 3 sims is pratically zero while the extra load from servicing multiple 3 sim calls is noticable (compared to a single 20 sim call).
Humans don't handle probabilities well. Ask a human to write down a made up list of the result of 20 coin tosses. Then actually toss a coin 20 times and write down the results. Show the two results to someone who doesn't know which is which and it's usually quite easy to tell which is which. The human made up one will have less long streaks as we don't believe this is "random" but it happens all the time. All the crap theories on here about whether multi-summing gives better results or not are just a manifestation of the same thing. Anyone who has bothered keeping a record of results has seen that the overall probability is the same whatever you do.
The probability for that should be (2/3)^19*1/3 = 0.00015.
Thanks...probability was never my strong suit!
It's not his strong suit either, lol. You draw 20 times with a probability of 1/3 each time, this is a standard binomial distribution. You can easily calculate the probability of 0, 1, 2 etc. hits with Excel using the BINOMDIST function like BINOMDIST(1;20;1/3;wrong). This results in 0.3% for 1 hit. To be fair, you should include 0 hits as well since you'd be even more upset then, so it's 0.33% for 0 or 1 hits out of 20 tries. That's 1 in 300, so indeed quite unlucky.
And as a programmer myself I call **** on all peeps who say multi-sims are worse. It has NOTHING to do with server load as the extra "load" from calculating 20 sims compared to 3 sims is pratically zero while the extra load from servicing multiple 3 sim calls is noticable (compared to a single 20 sim call).
True, forgot about the different permutations of where the 1 "hit" can be. Then we have the binomial distribution as you said, thanks (:
If my maths are right, that's like 0.03% probability. 1/3000ish. Certainly not impossible. (Technically that's prob for 0/20, but 1/20 should be really close).
Actually no, it's not really close. It's an order of magnitude more.
1/3 chance of a shard, means that:
0/20 = (2/3)^20 = 0.0003, or 0.03% as you said.
However
1/20 = 20 * 1/3 * (2/3)^19 = 0.003, or 0.3%. It'll happen 3 times in 1000, rather than 3 times in 10,000. It shouldn't be common in either case (3/1000 means that you can expect something like this to happen once every 333 times you multisim 20 runs, however it should happen 10x more often than 0/20, which should only happen once in every 3,333 runs.
Replies
thats why i would never hit 20 multisims at once. So in that case you rolled a 5% drop rate. next time you could have rolled a 100% drop rate. It evens out. but not worth the risk.
superstition? probably. still i feel better doing it this way.
Agreed, I never sim over 3 rolls. Superstition or not, in my experience the drop rates at large rolls are dismal.
The probability for that should be (2/3)^19*1/3 = 0.00015.
it make sense to do it the way i described. less stress on the servers. So it takes the RNG % and times it by the # of multi sims i requested. If I get a roll of 32 or below i get zero shards, 40 i get 1 shard, if i get a roll of 68 i get 2 shards and if i roll a 90 or above i get 3 shards.
that's the only way i believe someone doing 20 sims could get one shard. they rolled a 5x20=1.
i have trouble believing the system did this instead .03x.03.x03 etc.........
i work with developers every day and i am no expert but everything they do is to give you full functionality with the least amount of load on the servers. everything they do is to try and minimize the strain on the servers. independent drop rate checks make no sense in a game this big and with this many users
For purple gear, I do a multi sim of 5. Seems to get 1 each pull for the most part.
Thanks...probability was never my strong suit!
+1
Just got him to 5 stars... every star and he kicks just a little more butt.... aww yeah
The true answer should be 0.00015 as you calculated then multiplied by 20 (the binomial coefficient corresponding to picking 1 from 20) = 0.003 (rounded for simplicity) = 0.3%.
And as a programmer myself I call **** on all peeps who say multi-sims are worse. It has NOTHING to do with server load as the extra "load" from calculating 20 sims compared to 3 sims is pratically zero while the extra load from servicing multiple 3 sim calls is noticable (compared to a single 20 sim call).
True, forgot about the different permutations of where the 1 "hit" can be. Then we have the binomial distribution as you said, thanks (:
Actually no, it's not really close. It's an order of magnitude more.
1/3 chance of a shard, means that:
0/20 = (2/3)^20 = 0.0003, or 0.03% as you said.
However
1/20 = 20 * 1/3 * (2/3)^19 = 0.003, or 0.3%. It'll happen 3 times in 1000, rather than 3 times in 10,000. It shouldn't be common in either case (3/1000 means that you can expect something like this to happen once every 333 times you multisim 20 runs, however it should happen 10x more often than 0/20, which should only happen once in every 3,333 runs.