Galactic Chase. Which energy node are you guys using?

We have Galactic Chase on the run, and we gotta choose which Node we should spend Ship Energy on.
The probability for getting blueprint of Emperor's Command Shuttle is (Energy cost for the node Node)*4%.
On average, the choice of specific node doesn't matter much, because the average amount of bluprint is 0.04*(Total Energy spent), for this Galactic Chase is Bernoulli process, or Binomial distribution.
But for low energy Nodes, we get higher variance for blueprints, so we have a higher chance of getting above-average amount of blueprints as well as below-average amount.
On my side, I am not really the fan of praying to RNGesus, so I prefer higher stability, meaning I prefer 20 Energy Nodes.
What are your choices? Do you prefer high risk&return or low risk&return?

Replies

  • I only need Ani's ship out of all the hard nodes, so refreshing that node twice per day, and spending the rest (normal ship energy + refresh + 3x 100 crystal refreshes) on the normal 10 energy node with the stun gun. Getting useful rewards > superstition.

    I did an 11:55 3x 100 crystal refresh and the 600 crystal pack after midnight hit, so I'm at 72/80 for the unlock now.
  • crzydroid
    7254 posts Moderator
    edited February 2019
    The low energy nodes are right-skewed, so while you can get higher drops thanks to more trials, you will more frequently get low drops. Highest energy have the highest drop rate, so while the mean is balanced by those low drops, you have a better chance of getting high drops for a given number of trials.

    It depends on how many node refreshes you'd be willing to buy. Maybe sit down and do the math for your particular situation.
  • I've done 20 energy hard nodes exclusively. I've seen multiple 8/8. At 47/80 now after 3 energy refreshes, 100 stored energy, energy pack and accumulated energy since reset.

    47 shards, 1165 energy. 4% on the nose.
    Ally Code 766-465-766 swgoh.gg/u/trystansr/collection/
  • EventineElessedil
    6171 posts Member
    edited February 2019
    I have done the math, you get the same shards per energy no matter which node you farm. According to the stated rates, of course.
    So pick a node, any node, with the stuff you want most.
    Nest node for me.
  • Yep, focus on hard nodes for things you dont have first, then normal nodes with gear you need. For me that's Sith Assassin Ship, ETA-2, and Lando's falcon, then normal 5-e for stun guns
  • I'm farming Bistan's U-Wing and Sith Fighter anyway, so those.
    u58t4vkrvnrz.png



  • Same nodes I was already farming, 16 energy Yando and ship and 20 energy Nest. Haven’t been doing refreshes, hoping to do that when it coincides with regular hard energy.
  • The Nest node. And I've been doing pretty well so far.
  • I have done the math, you get the same shards per energy no matter which node you farm. According to the stated rates, of course.
    So pick a node, any node, with the stuff you want most.
    Nest node for me.
    Well, over a long enough stretch of time with a significant enough sample size you'll wind up getting the same. But due to the nature of the event and the cost of energy refreshes, there's a difference. It comes down to risk tolerance--if you're willing to trust your luck, go for low cost nodes and hope you get as many successes as possible. If you're risk averse, go for the most expensive nodes so you'll get as few failures as possible.

    Personally, I'm risk averse--but also I have a farming plan and I don't want to deviate from that. So I'm finishing Bastila, and with the refreshes (which I don't normally do--don't like refreshing for energy that doesn't count towards the daily 600) I'm doing Jolee and Zaalbar.
    Ceterum censeo Patientia esse meliat.
  • ShaggyB
    2390 posts Member
    Mostly 20 energy ones... and i buy 3 refreshes... should be at 5 star by event end
  • Jolee and Enfys when crystal income allows.
  • Nauros
    5429 posts Member
    Not changing my farming habits, so Sith Fighter and Enfys hard nodes.
  • VonZant
    3843 posts Member
    Sith Fighter, Poe, Scimitar. Was doing Nest but only need 10 shards now and Bounties appear in 3 days.
  • Tsuo_Vook
    62 posts Member
    edited February 2019
    Professor of Statistics here. If you want to actually do this properly, you want to be using a Negative Binomial distribution. It counts the number of trials[1] it will take to get to a fixed number of successes (assume 80 successes for an unlock).

    Based on some quick calculations, here are the 90% probability intervals for how much you're likely to have to spend to unlock ECS on different nodes:
    Node Energy              5% will unlock in            95% will unlock in
          8                     1640 energy                 2216 energy
         10                     1660 energy                 2200 energy
         16                     1728 energy                 2128 energy
         20                     1780 energy                 2080 energy
    

    As others have previously stated, there's not a big advantage to picking any energy cost over the others, it's just about risk tolerance. They'll all unlock at about the same point on average, with a little more variability (positive and negative) as node energy cost decreases.

    For me, I'm just farming whatever I have a need for on fleet nodes, primarily flipping between Nest on a 20-energy node and some gear I want at 10-energy nodes.


    [1] Depending on parameterization. Alternately, it can count the number of failures until you reach a fixed number of successes. This is a pretty easy post-hoc adjustment to make, because you know that fixed number of successes.
  • VonZant
    3843 posts Member
    Tsuo_Vook wrote: »
    Professor of Statistics here. If you want to actually do this properly, you want to be using a Negative Binomial distribution. It counts the number of trials[1] it will take to get to a fixed number of successes (assume 80 successes for an unlock).

    Based on some quick calculations, here are the 90% probability intervals for how much you're likely to have to spend to unlock ECS on different nodes:
    Node Energy              5% will unlock in            95% will unlock in
          8                     1640 energy                 2216 energy
         10                     1660 energy                 2200 energy
         16                     1728 energy                 2128 energy
         20                     1780 energy                 2080 energy
    

    As others have previously stated, there's not a big advantage to picking any energy cost over the others, it's just about risk tolerance. They'll all unlock at about the same point on average, with a little more variability (positive and negative) as node energy cost decreases.

    For me, I'm just farming whatever I have a need for on fleet nodes, primarily flipping between Nest on a 20-energy node and some gear I want at 10-energy nodes.


    [1] Depending on parameterization. Alternately, it can count the number of failures until you reach a fixed number of successes. This is a pretty easy post-hoc adjustment to make, because you know that fixed number of successes.

    I dont know what any of that means.

    But I have determined that even if it is wrong or a lie, its a very elaborate one, so you deserve a "like" no matter what.
  • NicWester wrote: »
    I have done the math, you get the same shards per energy no matter which node you farm. According to the stated rates, of course.
    So pick a node, any node, with the stuff you want most.
    Nest node for me.
    Well, over a long enough stretch of time with a significant enough sample size you'll wind up getting the same. But due to the nature of the event and the cost of energy refreshes, there's a difference. It comes down to risk tolerance--if you're willing to trust your luck, go for low cost nodes and hope you get as many successes as possible. If you're risk averse, go for the most expensive nodes so you'll get as few failures as possible.

    Personally, I'm risk averse--but also I have a farming plan and I don't want to deviate from that. So I'm finishing Bastila, and with the refreshes (which I don't normally do--don't like refreshing for energy that doesn't count towards the daily 600) I'm doing Jolee and Zaalbar.

    This is backwards.

    If you are risk averse, go for the low energy nodes.

    All nodes are identical over a large enough sample size. To be risk averse, you want to make your sample size larger. To do that you use the cheapest nodes possible.


    Let’s say you have $500 and a roulette table. To be risk averse you put $1 on red 500 times. You don’t put $5 100 times.
  • VonZant wrote: »
    Tsuo_Vook wrote: »
    Professor of Statistics here. If you want to actually do this properly, you want to be using a Negative Binomial distribution. It counts the number of trials[1] it will take to get to a fixed number of successes (assume 80 successes for an unlock).

    Based on some quick calculations, here are the 90% probability intervals for how much you're likely to have to spend to unlock ECS on different nodes:
    Node Energy              5% will unlock by            95% will unlock by
          8                     1640 energy                 2216 energy
         10                     1660 energy                 2200 energy
         16                     1728 energy                 2128 energy
         20                     1780 energy                 2080 energy
    

    As others have previously stated, there's not a big advantage to picking any energy cost over the others, it's just about risk tolerance. They'll all unlock at about the same point on average, with a little more variability (positive and negative) as node energy cost decreases.

    For me, I'm just farming whatever I have a need for on fleet nodes, primarily flipping between Nest on a 20-energy node and some gear I want at 10-energy nodes.


    [1] Depending on parameterization. Alternately, it can count the number of failures until you reach a fixed number of successes. This is a pretty easy post-hoc adjustment to make, because you know that fixed number of successes.

    I dont know what any of that means.

    But I have determined that even if it is wrong or a lie, its a very elaborate one, so you deserve a "like" no matter what.

    I was wondering if this might need a little more explanation.

    So. Let's say you're farming for ECS exclusively on 8-cost fleet table nodes. If you're quite lucky, you'll be able to unlock ECS in 1640 or less total energy (205 or fewer battles). This will happen for about 5% of players using this farming strategy. It's similar to rolling a natural 20 on a d20 in tabletop roleplaying.

    On the other hand, if you're particularly unlucky, it might take you 2216 or more energy (277 or more battles) to obtain all 80 blueprints needed to unlock the ECS on 8-cost nodes. About 5% of players will be this unlucky, so it's similar to rolling a natural 1 on a d20 in tabletop roleplaying. (95% of players will have unlocked the ECS by the time they spend 2216 energy, if they're farming exclusively on 8-cost nodes.)

    Looking at the table, you might notice that the lucky 5% side increases with node energy while the unlucky side ("95% will unlock by") decreases with increasing node energy. This is what other posters have referred to as the risk-aversion of using higher cost nodes.

    The mistake @slickdealer makes in identifying the risk-averse strategy is an easy mistake to make. Slickdealer is basing their statement on the central limit theorem, which says large samples of data should behave more predictably as sample size grows. It's a valid argument, and it's going to hold in most areas of the game---but not on the particular problem offered by Galactic Chase. What's driving the behavior here is the fact that variability in binomial distributions shrinks as the probability of success on an individual trial moves toward 0 or 1. If we had 24-energy nodes to farm, they would be guaranteed to drop ECS blueprints and you could know precisely how much energy it would take to unlock ECS off of those nodes. Here, the 20-energy nodes give the highest drop rate; and because they have such a high drop rate, farming on them turns out to be the most risk-averse strategy. That is, the strategy where unlocking the ECS depends least on luck.

    If you farm on 8-energy nodes, you might get lucky and unlock ECS a little bit quicker. Then again, you might get unlucky and need to spend extra energy. If you farm on 20-cost nodes, your farming results will be the least "swingy"---that is, RNG will be less of a factor in how long it takes to get, and your farm will wind up being more reliable (although not necessarily any cheaper or faster, since you might have gotten lucky on the low-cost nodes).
  • This is backwards.

    If you are risk averse, go for the low energy nodes.

    All nodes are identical over a large enough sample size. To be risk averse, you want to make your sample size larger. To do that you use the cheapest nodes possible.

    Let’s say you have $500 and a roulette table. To be risk averse you put $1 on red 500 times. You don’t put $5 100 times.
    This is what you'd do if you had A LOT of runs. But we don't have that. I do two refreshes, which is a luxury I'm afforded because I'd stockpiled 1600 crystals before the event started (Gearing up some folks who currently don't need raid gear, so I'm saving it up so that when they inevitably do I can just buy it outright instead of having to wait for Rancor boxes to give me what I need) and with the couple hundred I get a day, I can run at a small loss for a while without impacting too much. So that means I get 405 energy to spend a day, or 20 sims per day, over the course of five days, for a total of 101 sims. That's not enough to get a reliable sample size that would make the cheap nodes worthwhile.

    But, also, if you look at Professor Math Man's table above, you'll find (and I'm going to quote someone else here because this next bit is going to answer their question, too)…
    VonZant wrote: »
    I dont know what any of that means.

    But I have determined that even if it is wrong or a lie, its a very elaborate one, so you deserve a "like" no matter what.
    It means that if everyone ran only 8-energy nodes, then 95% of people who did it would unlock the ship after spending roughly 2216 energy. 5% of players would get lucky with their drops and only have to spend 1640 energy. But if everyone ran it on 20-energy nodes, then 95% of people who did it would unlock the ship after spending roughly 2080, while the lucky 5% would only spend 1780.

    In other words, the 8-energy nodes will be worth it for super lucky people, because they get more individual rolls on the loot table to get the blueprints. But, overall, most of us aren't going to be lucky, so we wind up spending more energy than if we were doing the 20-energy nodes. On the other hand, the super lucky people on the 20-energy nodes aren't getting that much of a discount because they have fewer, but more reliable, rolls per day.

    Long story short: With a limited number of rolls you have two choices--reliability or variability. Hindsight will always be 20/20, so if you do the 20-energy nodes and you only get a 50% drop rate (Not to be confused or conflicted with the 80% they claimed--that's the expected return. IE: If roll 1d10 every time you sim, you'll get a blueprint on a 1-8 and no blueprint on a 9 or 0. Sometimes you're going to roll 10 dice and get five 9s or 0s...) then in hindsight you could say that the 8-energy nodes would have been better for you. But you never know in advance what you're going to get. So just pick the one you feel more comfortable with--a lot of reliable shards, or a good amount of shards with the possibility of even more than the reliable ones.
    Ceterum censeo Patientia esse meliat.
  • NicWester wrote: »
    I have done the math, you get the same shards per energy no matter which node you farm. According to the stated rates, of course.
    So pick a node, any node, with the stuff you want most.
    Nest node for me.
    Well, over a long enough stretch of time with a significant enough sample size you'll wind up getting the same. But due to the nature of the event and the cost of energy refreshes, there's a difference. It comes down to risk tolerance--if you're willing to trust your luck, go for low cost nodes and hope you get as many successes as possible. If you're risk averse, go for the most expensive nodes so you'll get as few failures as possible.

    Personally, I'm risk averse--but also I have a farming plan and I don't want to deviate from that. So I'm finishing Bastila, and with the refreshes (which I don't normally do--don't like refreshing for energy that doesn't count towards the daily 600) I'm doing Jolee and Zaalbar.

    This is backwards.

    If you are risk averse, go for the low energy nodes.

    All nodes are identical over a large enough sample size. To be risk averse, you want to make your sample size larger. To do that you use the cheapest nodes possible.


    Let’s say you have $500 and a roulette table. To be risk averse you put $1 on red 500 times. You don’t put $5 100 times.

    Don’t quit school yet

    (You are assuming 1$ and 5$ pulls have the same chance outcomes, which is wrong...)
  • Actually, also worth mentioning...
    Let’s say you have $500 and a roulette table. To be risk averse you put $1 on red 500 times. You don’t put $5 100 times.

    No, you do put $5 on red 100 times (out of those two options). Or better yet, you bet all $500 just once. Games like roulette favor the house; the fewer times you bet, the likelier it is you actually make a profit.

    But that's not especially related to risk aversion. If you're risk-averse, you just don't play roulette. Or anything other than poker, really.

    I'll see myself out.

  • If you feel luck is on your side, 10 energy nodes are best bet.
    Sure 20 energy has 80%, 10 energy has 40%
    So, in theory you get same amount of shards.
    BUT you will never get more than 1 shard at 20 energy, while there is a decent chance to get 2 shards from 2 10 energy nodes.
    Again, if you feel luck is a bit on your side.
  • Tsuo_Vook wrote: »
    Actually, also worth mentioning...
    Let’s say you have $500 and a roulette table. To be risk averse you put $1 on red 500 times. You don’t put $5 100 times.

    No, you do put $5 on red 100 times (out of those two options). Or better yet, you bet all $500 just once. Games like roulette favor the house; the fewer times you bet, the likelier it is you actually make a profit.

    But that's not especially related to risk aversion.

    Risk Aversion was the entire point, which is why your post is completely incorrect.
  • Emmant wrote: »
    NicWester wrote: »
    I have done the math, you get the same shards per energy no matter which node you farm. According to the stated rates, of course.
    So pick a node, any node, with the stuff you want most.
    Nest node for me.
    Well, over a long enough stretch of time with a significant enough sample size you'll wind up getting the same. But due to the nature of the event and the cost of energy refreshes, there's a difference. It comes down to risk tolerance--if you're willing to trust your luck, go for low cost nodes and hope you get as many successes as possible. If you're risk averse, go for the most expensive nodes so you'll get as few failures as possible.

    Personally, I'm risk averse--but also I have a farming plan and I don't want to deviate from that. So I'm finishing Bastila, and with the refreshes (which I don't normally do--don't like refreshing for energy that doesn't count towards the daily 600) I'm doing Jolee and Zaalbar.

    This is backwards.

    If you are risk averse, go for the low energy nodes.

    All nodes are identical over a large enough sample size. To be risk averse, you want to make your sample size larger. To do that you use the cheapest nodes possible.


    Let’s say you have $500 and a roulette table. To be risk averse you put $1 on red 500 times. You don’t put $5 100 times.

    Don’t quit school yet

    (You are assuming 1$ and 5$ pulls have the same chance outcomes, which is wrong...)

    In this case they have identical outcomes over a sufficient sample size. Read the post from CG.
  • NicWester
    8928 posts Member
    edited February 2019
    Tsuo_Vook wrote: »
    Actually, also worth mentioning...
    Let’s say you have $500 and a roulette table. To be risk averse you put $1 on red 500 times. You don’t put $5 100 times.

    No, you do put $5 on red 100 times (out of those two options). Or better yet, you bet all $500 just once. Games like roulette favor the house; the fewer times you bet, the likelier it is you actually make a profit.

    But that's not especially related to risk aversion. If you're risk-averse, you just don't play roulette. Or anything other than poker, really.

    I'll see myself out.
    Roulette is interesting because with a deep enough bankroll you CAN win every time over a long enough period. You bet on a 2:1 outcome (Red/Black, Odd/Even, 1-18, 10-36, etc) and every time you lose, you double your last bet. Eventually you win and recoup all the previous bets you lost, plus the initial amount you bet. The odds aren't exactly 50/50 on any of those bets (0 and 00 are green, neither odd nor even, and not 1-36) but it's close enough that if you had an unlimited sack of money and just kept betting on the same outcome over and over, eventually it'd work.

    IN THEORY.

    In practice, every time I try this the wheel goes on an improbable run of 8+ outcomes where I've been betting red and it lands on black... The last time I did it, at my company holiday party where we were given fake money to bet with (that bought raffle tickets at the end of the night, so there was an incentive to win) I bet on black and it landed on red 6 times in a row, 00 once, and red another 7 times in a row. That's 15 straight times it was non-black.

    That's some Rosencrantz and Guildenstern Are Dead nonsense right there.
    Ceterum censeo Patientia esse meliat.
  • Tsuo_Vook wrote: »
    Actually, also worth mentioning...
    Let’s say you have $500 and a roulette table. To be risk averse you put $1 on red 500 times. You don’t put $5 100 times.

    No, you do put $5 on red 100 times (out of those two options). Or better yet, you bet all $500 just once. Games like roulette favor the house; the fewer times you bet, the likelier it is you actually make a profit.

    But that's not especially related to risk aversion.

    Risk Aversion was the entire point, which is why your post is completely incorrect.

    We must have very different notions of "risk aversion". In the roulette situation, my notion of risk aversion is "you get to keep the most money". Yours appears to be "you reliably attain the inevitable result of losing lots of money, while minimizing the pesky possibility that you might get lucky and keep some".
Sign In or Register to comment.