This is the reason i simmed in smaller batches for a long time. Seeying 0/10 just sucks even if you get 6/10 the next batch.
Got alot off 0/10 and never saw 10/10.
Clearly, simmind small amount is way better.
Whether joking or serious, you do bring up an interesting point. I've been playing for a while now and I've not seen max sims match shards/purple gear/gold gear above 6 sims...
Hi guys
In my guild we have started looking into whether single sim'ing a bunch of times or max sim'ing a single time will yield better results.
I am almost certain that others have done some research into this subject..?
What were your results?
Officially there is no difference tbh but its all psychological what feels better/worse to you as a player,
I SIM in x2 batches because if and when I did the max available and got nothing I needed/wanted it feels crap (8 Sims no shards feels like a kick to the nuts).
But hit a 2x SIM and get nothing not so bad I still 6 trys left.
The end total over time will be roughly the same but how you feel on the jurny is totally different with each path.
(This is not a blanket rule gw does payout more credits if you hit each of the fights over SIM completion of the 12 fights the other stuff remains the same gw credits shards training droids etc)
This is the reason i simmed in smaller batches for a long time. Seeying 0/10 just sucks even if you get 6/10 the next batch.
Got alot off 0/10 and never saw 10/10.
Clearly, simmind small amount is way better.
Whether joking or serious, you do bring up an interesting point. I've been playing for a while now and I've not seen max sims match shards/purple gear/gold gear above 6 sims...
I have. It's rare, but glorious.
It is, however, NOT an interesting point. With a 50/50 drop rate you would expect to see 0/10 as often as you see 10/10. With a drop rate significantly below that you would expect to see 0/10 more often, and just more 0s in general.
An interesting point would be asking how often people who swear by doing one at a time have seen 1/1 ten times in a row. The answer would be mystifyingly close to how often you would see 10/10. Almost never.
No reason to give people false expectations! People forget that with five attempts per refresh, the "normal" shard drop rate is 1/5 so zero drop days are much more common than with eight attempts (without refreshes). It feeds threads like this where people struggle with statistical probability.
I love how people are extreme saying there is no difference or there is a massive difference and have not seen the game code to see if drops differentiate. But to each their own.
Warrior talked about this in one of his Scissorman streams, since he prefers small batches (I think he does 3 or 4). He actually made a compelling argument. Here's his claim: your long term average is the same either way, 1/3 or 30% or whatever it is. However, (he claims) the drops are not purely determined by RNG. Instead, there is a built-in adjustment to prevent extended streaks of bad drops (or good drops). If you get some bad drops, the program will compensate with better odds until you are back at 1/3. This adjustment happens after every batch of sims. The result is that your results are more consistent with small batch sims. Same mean, lower variance.
Now this is his theory. Personally, I just do max sims. Even if Warrior is right, I'll still do max sims. I can handle some RNG swings and trust the Law of Large Numbers.
I love how people are extreme saying there is no difference or there is a massive difference and have not seen the game code to see if drops differentiate. But to each their own.
We have over 3 years of empirical evidence. Those two groups you describe are:
1. People who know that the amount you sim doesn't matter in the long run because they accept facts and evidence.
2. People who sim small amounts because they believe in leprechauns.
Do you go to the casino and play the penny slots? or go straight to the high limit room for just a single pull?
Depends on what my goal is. If I want to sit for a long time and pull a lever while watching a bunch of losing pulls with an occasional winner before walking out with no money I'll do penny slots. If I have better ways to spend my time I'll go to the high limit room, give them all my money, then get on with my life.
I love how people are extreme saying there is no difference or there is a massive difference and have not seen the game code to see if drops differentiate. But to each their own.
Because we don't have to see the code. We only have to see the results and those results have born out in all circumstances and under all variables--Double drops, energy cost, 3 tries per day, 5 tries per day, 8 tries per day, light nodes, dark nodes, fleet nodes, cantina nodes, new characters, old characters. To quote the Barenaked Ladies: It's all been done.
Only a flat-earther would ignore that much empirical evidence.
Is the drop rate the same across all nodes, for all items- gear, character shards, ship shards, etc? Data please, not speculation. It definitely does not feel that way to me, although I am not keeping records. There are some fleet nodes that seem to never drop low level character gear
Are the rates continually being logged? Because I’d really like to trust CG to not change them without telling us, but then I’d really like to poop gold bricks
I am the jedi dog battle droid C-3PO could not kill.
Warrior.... claims the drops are not purely determined by RNG. Instead, there is a built-in adjustment to prevent extended streaks of bad drops (or good drops). .
There are quotes out there from CG Carrie denying this is the case (at the moment). This one for example
We do not do any A/B or multivariate testing whatsoever.
We have in the past employed a few "different offers" based on regions, but quite frankly that was more problematic than it was worth and run for a limited time.
Does that mean we'll never do it? I can't say that, but we have no mechanism in place to do it and we don't currently. One thing we talk about is some kind of Monte Carlo system, where after X number of "bad rolls" we would guarantee a "good outcome" - but we don't have anything in place to give different or predictive offers to different people or the same person based on some criteria.
Note it's a bit of a confused comment as it talks about mean reversion in the middle of a response to a relatively unrelated question about A/B group drop rate manipulation which she flat out denies has been done in the game to date.
I don't have data to share at the moment but my experience tracking bears this out. Characters that get a bad run at the start of a farm remain (for the most part) delayed vs expected completion date and characters that get a good run remain ahead of the expected completion date. The lack of any mean reversion is likely what leads some of the perception that some farms are easier/harder than others (as if you get ahead of expected completion date you stay ahead and if you fall behind you stay behind without extra investment).
Also, I'm really not sure how simming in small groups would change this anyway as although it's not confirmed it's pretty clear each sim is an independent bernoulli trial and they could implement mean reversion independently of the sim group size quite easily if they were going to implement it.
I prefer to sim in small increments (unless hard node, then I just sim it all). When simming in small increments, i do it for gear farming for the sole fact it LOOKS like I'm getting better luck. Especially if I only need a few pieces, I dont want to accidentally do 8 sims and only need the 1 shard/gear piece.
However, I do not believe it yields any better results, it just helps me feel better about doing 3 refreshes for 5 stun cuffs because I'm impatient...
This is the reason i simmed in smaller batches for a long time. Seeying 0/10 just sucks even if you get 6/10 the next batch.
Got alot off 0/10 and never saw 10/10.
Clearly, simmind small amount is way better.
You're clueless of how probabilities work. You have never seen a 10/10 that's true but how many times have you seen 1/1 happen 10 times in a row? Probably never, but I bet if you did 1 attempt 10 times you will see 0/10 10 times in a row.
I do like doing 7 attempts per sim, just a lucky thing for me. When I do 7 attempts I usually get 2-3 which feels oddly satisfying.
Now let me respond to some of you like @Ultra who love to state things as facts all the time. There is no possible way for you to know what is happening. This will remain the truth as long as the devs refuse to reveal the true drop rates and state that they aren't being manipulated in any way imaginable. Nobody knows if it's hard coded in the system to have a maximum cap on the drops. For example they can hard code it so that if you do up to certain number of attempts, like say 7, you CAN, if you're lucky, get 7. But if you do let's 8 -12 you can only get up to 7. And if you do more than 12 you can go over 7.
The truth is that you're taking it all in good faith and you have trust that the devs would never do anything to screw the players. However a sceptical player would not take anything on faith and they would perform perfectly reasonable and justified tests to see if that they are told of being true, are actually true. Where money is an incentive you must always assume that manipulation is a possibility
The fact is numbers don’t lie and we have people testing with a different SIM values over a long period of time and then recording the results,
and contrary to feelings it’s the same result (33% drop rate)
This is confirmed by the developers and is in line with what you said (2 or 3 drops when doing 7 sims) and they have gone on record and said that they don’t manipulate drop rates at all at any point
You can suspect them all you want but don’t make arguments like “well they never told us” followed by “if they did don’t trust them”
It seems like you want so and so as proof but preemptively denying the proof if given
Devs have stated one thing, people have tested to see if they were lying. They aren’t.
I do like doing 7 attempts per sim, just a lucky thing for me. When I do 7 attempts I usually get 2-3 which feels oddly satisfying.
Now let me respond to some of you like @Ultra who love to state things as facts all the time. There is no possible way for you to know what is happening. This will remain the truth as long as the devs refuse to reveal the true drop rates and state that they aren't being manipulated in any way imaginable. Nobody knows if it's hard coded in the system to have a maximum cap on the drops. For example they can hard code it so that if you do up to certain number of attempts, like say 7, you CAN, if you're lucky, get 7. But if you do let's 8 -12 you can only get up to 7. And if you do more than 12 you can go over 7.
The truth is that you're taking it all in good faith and you have trust that the devs would never do anything to screw the players. However a sceptical player would not take anything on faith and they would perform perfectly reasonable and justified tests to see if that they are told of being true, are actually true. Where money is an incentive you must always assume that manipulation is a possibility
It's not a matter of taking everything in good faith, or being highly skeptical. It's a matter of realizing how ridiculouos these sorts of things would be to code into the game for the purpose of generating more revenue. Yes, a programmer could absolutely code these sorts of things quite easily, but in terms of generating more revenue its just not a good way to do it. It would generate minimal amounts increase in revenue at risk of infurating the paying members of the player base if the truth about it ever came out. Simply not worth it. Releasing new characters that players have to pay to get the first time around, releasing new content that players can't complete till they get certain characters to max level/gear/etc, are far more effective - so that is where they want the development team to spend their time.
If they really really wanted to fiddle with the amount of shards players collected, it would be far easier to just change the drop rate on shards from 33% to 25% instead of working with a more complicated formula that:
limits max drops in a multi-sim
or reduces drop rates when players are close to the final few shards needed for the next star level
or tracks individual player drops over time and insures they average out over short time frame
or any number of other possible more complicated ways of doing it
Or you know, they could also just change the number of sims you can do per day from say, 8 to 5.
I will say this. I took one math class in college. Finite Math. This is like pre - statistics. The biggest takeaway I got from it- you don’t need to nerf or be shady with probabilities. At all.
If you want to investigate this further - and actually do some testing, please do a proper experiment of it. People have done various tests before, but I haven't seen anything where anybody did full-on proper statistics on it. (Using the appropriate statistical tests, do your runs in batches, figure out how many you need to calculate the statistical significance of your resut, etc.) So that at the end of the day you can conclude not only what the drop rate seems to be, but also the confidence in the answer.
(Not difficult, but does require a little background/reading in statistics.)
Probably won't convince those who believe (just as there are still plenty of fools to play the lottery),
but at least it would be adding something new to the mix. (Unless you find somebody has already
done this.)
Replies
Whether joking or serious, you do bring up an interesting point. I've been playing for a while now and I've not seen max sims match shards/purple gear/gold gear above 6 sims...
Officially there is no difference tbh but its all psychological what feels better/worse to you as a player,
I SIM in x2 batches because if and when I did the max available and got nothing I needed/wanted it feels crap (8 Sims no shards feels like a kick to the nuts).
But hit a 2x SIM and get nothing not so bad I still 6 trys left.
The end total over time will be roughly the same but how you feel on the jurny is totally different with each path.
(This is not a blanket rule gw does payout more credits if you hit each of the fights over SIM completion of the 12 fights the other stuff remains the same gw credits shards training droids etc)
Its the exact same result. Still 33%
I have. It's rare, but glorious.
It is, however, NOT an interesting point. With a 50/50 drop rate you would expect to see 0/10 as often as you see 10/10. With a drop rate significantly below that you would expect to see 0/10 more often, and just more 0s in general.
An interesting point would be asking how often people who swear by doing one at a time have seen 1/1 ten times in a row. The answer would be mystifyingly close to how often you would see 10/10. Almost never.
I think you mean
1/5 + 2/5
No reason to give people false expectations! People forget that with five attempts per refresh, the "normal" shard drop rate is 1/5 so zero drop days are much more common than with eight attempts (without refreshes). It feeds threads like this where people struggle with statistical probability.
Now this is his theory. Personally, I just do max sims. Even if Warrior is right, I'll still do max sims. I can handle some RNG swings and trust the Law of Large Numbers.
Edited: missed the dig on the upcoming change
We have over 3 years of empirical evidence. Those two groups you describe are:
1. People who know that the amount you sim doesn't matter in the long run because they accept facts and evidence.
2. People who sim small amounts because they believe in leprechauns.
Depends on what my goal is. If I want to sit for a long time and pull a lever while watching a bunch of losing pulls with an occasional winner before walking out with no money I'll do penny slots. If I have better ways to spend my time I'll go to the high limit room, give them all my money, then get on with my life.
Because we don't have to see the code. We only have to see the results and those results have born out in all circumstances and under all variables--Double drops, energy cost, 3 tries per day, 5 tries per day, 8 tries per day, light nodes, dark nodes, fleet nodes, cantina nodes, new characters, old characters. To quote the Barenaked Ladies: It's all been done.
Only a flat-earther would ignore that much empirical evidence.
Proof that batches still work.
Are the rates continually being logged? Because I’d really like to trust CG to not change them without telling us, but then I’d really like to poop gold bricks
Obviously the drop rate for Zaalbar shards is 88% if you sim 8 at a time on April 10th. If only it were April 10th all year long!
There are quotes out there from CG Carrie denying this is the case (at the moment). This one for example
from this reddit thread https://www.reddit.com/r/SWGalaxyOfHeroes/comments/ask8hw/what_physical_pain_looks_like_im_swgoh/
Note it's a bit of a confused comment as it talks about mean reversion in the middle of a response to a relatively unrelated question about A/B group drop rate manipulation which she flat out denies has been done in the game to date.
I don't have data to share at the moment but my experience tracking bears this out. Characters that get a bad run at the start of a farm remain (for the most part) delayed vs expected completion date and characters that get a good run remain ahead of the expected completion date. The lack of any mean reversion is likely what leads some of the perception that some farms are easier/harder than others (as if you get ahead of expected completion date you stay ahead and if you fall behind you stay behind without extra investment).
Also, I'm really not sure how simming in small groups would change this anyway as although it's not confirmed it's pretty clear each sim is an independent bernoulli trial and they could implement mean reversion independently of the sim group size quite easily if they were going to implement it.
Had this beauty yesterday. #maxsims4life
However, I do not believe it yields any better results, it just helps me feel better about doing 3 refreshes for 5 stun cuffs because I'm impatient...
You're clueless of how probabilities work. You have never seen a 10/10 that's true but how many times have you seen 1/1 happen 10 times in a row? Probably never, but I bet if you did 1 attempt 10 times you will see 0/10 10 times in a row.
Now let me respond to some of you like @Ultra who love to state things as facts all the time. There is no possible way for you to know what is happening. This will remain the truth as long as the devs refuse to reveal the true drop rates and state that they aren't being manipulated in any way imaginable. Nobody knows if it's hard coded in the system to have a maximum cap on the drops. For example they can hard code it so that if you do up to certain number of attempts, like say 7, you CAN, if you're lucky, get 7. But if you do let's 8 -12 you can only get up to 7. And if you do more than 12 you can go over 7.
The truth is that you're taking it all in good faith and you have trust that the devs would never do anything to screw the players. However a sceptical player would not take anything on faith and they would perform perfectly reasonable and justified tests to see if that they are told of being true, are actually true. Where money is an incentive you must always assume that manipulation is a possibility
and contrary to feelings it’s the same result (33% drop rate)
This is confirmed by the developers and is in line with what you said (2 or 3 drops when doing 7 sims) and they have gone on record and said that they don’t manipulate drop rates at all at any point
You can suspect them all you want but don’t make arguments like “well they never told us” followed by “if they did don’t trust them”
It seems like you want so and so as proof but preemptively denying the proof if given
Devs have stated one thing, people have tested to see if they were lying. They aren’t.
It's not a matter of taking everything in good faith, or being highly skeptical. It's a matter of realizing how ridiculouos these sorts of things would be to code into the game for the purpose of generating more revenue. Yes, a programmer could absolutely code these sorts of things quite easily, but in terms of generating more revenue its just not a good way to do it. It would generate minimal amounts increase in revenue at risk of infurating the paying members of the player base if the truth about it ever came out. Simply not worth it. Releasing new characters that players have to pay to get the first time around, releasing new content that players can't complete till they get certain characters to max level/gear/etc, are far more effective - so that is where they want the development team to spend their time.
If they really really wanted to fiddle with the amount of shards players collected, it would be far easier to just change the drop rate on shards from 33% to 25% instead of working with a more complicated formula that:
Or you know, they could also just change the number of sims you can do per day from say, 8 to 5.
(Not difficult, but does require a little background/reading in statistics.)
Probably won't convince those who believe (just as there are still plenty of fools to play the lottery),
but at least it would be adding something new to the mix. (Unless you find somebody has already
done this.)