Rewards for raid simming (data table)

vksg

In addition to announcement https://forums.galaxy-of-heroes.starwars.ea.com/discussion/198056/rewards-for-raid-simming by @CG_SBCrumb, I created a data table.

Why? Because guild members still have to join* the raid to get rewards, decisions by officers (simming vs. not) should take into account the average participation within that specific guild instead of comparing total (50x) figures.

Simming yields comparatively worse results the smaller the number of participants**. May the table help you to decide which way to go***. last updated on 2019-04-10

*) see join list screenshot and CONFIRMED: join period for sim is MANDATORY
**) I would really appreciate your help with those salvage numbers. I recorded HPIT rewards for some time so I am pretty sure, I got H1-01 and H1-02 right. Might have mixed up the others and have no clue for H1-06.
***) see detailed explanation of this table in my post below

• Edit v0.2 - Highlighted missing information, clarified column for counting joined players
• Edit v0.3 - Changed H1-03 to 25+12 (was 25+10); Removed box column from cumulative data (value was not cumulative); added guessed values for smallest box
• Edit v0.4 - Added chances for fully crafted gear pieces
• Edit v0.5 - Updated chances according to new data

vksg

v0.3 of the table (last version without chances for fully crafted gear) for reference. Easier to read on small devices (fewer columns).

Lor_San_Teka

I will try to remember when our sim is done tomorrow morning. Were you able to do a raid without a join period and get rewards to all members?

Jarvind

So if I'm reading this correctly, simming yields less of everything except shards, the one thing that people with the ability to sim definitely don't need?

TaviRus

Jarvind wrote: »

So if I'm reading this correctly, simming yields less of everything except shards, the one thing that people with the ability to sim definitely don't need?

Simming provides more guild tokens for the guild once you have at least 30 people signup for Hpit sim.
I really don't see why we couldn't get first place or top 3 rewards for this. (or even 10th) Could tie it to how many people have the #1 Pit damage achievement and title to give the guild something to work towards.

Vos_Landeck

Jarvind wrote: »

So if I'm reading this correctly, simming yields less of everything except shards, the one thing that people with the ability to sim definitely don't need?

No, simming also gives more guild currency....830 per person vs. an average of 761 per person (assuming 50 players join).

vksg

Jarvind wrote: »

So if I'm reading this correctly, simming yields less of everything except shards, the one thing that people with the ability to sim definitely don't need?

It depends. That is why I made the table for our guild to decide.
The blue part on the right side (cumulative data) calculates the advantage of simming for every possible number of participants (first row 1 participant, last row 50 participants).
In detail that means:

• Credits: No matter how many players join, simming is always bad. However the net loss for the guild varies. It is worst (2.890.000) around 25 players participating compared to a 1.390.000 loss if all 50 members are participating.
• Guild Tokens: Break-even is between 29 and 30 players joining. <=29 means a net loss in tokens, >=30 means a net gain. Highest possible gain: 3.450 tokens with 50 joined. Biggest possible loss: -1.300 tokens with 10-15 joined.
• Shards: Obviously always a net gain, no matter how many players join. Hugh gains up to 314 shards (with 50 joined).
• Gear: I am missing the figures for the smallest box (H1-06), but as it seems, there is always a net loss in gear/salvage no matter how many players join.

So you have to answer probably two questions:

• Do we badly need more Han Solo shards? Most likely no (most, >=35, or all players already have Han Solo 7*). If yes for some reason, then go for simming.
• Do I expect more that 30 players to join?
• • No: Do not sim the raid
  • Yes: Are those additional guild tokens worth the loss in credits and gear/salvage? That's up to you to evaluate the trade-off: Example: With 37 joined players, 1.100 more guild tokens 'cost' you 2.540.000 credits and 170 salvage.

TL;DR

• Credits: Simming is bad, always.
• Guild tokens: Simming pays off if 30 or more players join the raid.
• Shards: Simming is good, always.
• Gear: Simming is (most likely) always bad.

vksg

VonZant

vksg wrote: »

Jarvind wrote: »

So if I'm reading this correctly, simming yields less of everything except shards, the one thing that people with the ability to sim definitely don't need?

It depends. That is why I made the table for our guild to decide.
The blue part on the right side (cumulative data) calculates the advantage of simming for every possible number of participants (first row 1 participant, last row 50 participants).
In detail that means:

• Credits: No matter how many players join, simming is always bad. However the net loss for the guild varies. It is worst (2.890.000) around 25 players participating compared to a 1.390.000 loss if all 50 members are participating.
• Guild Tokens: Break-even is between 29 and 30 players joining. <=29 means a net loss in tokens, >=30 means a net gain. Highest possible gain: 3.450 tokens with 50 joined. Biggest possible loss: -1.300 tokens with 10-15 joined.
• Shards: Obviously always a net gain, no matter how many players join. Hugh gains up to 314 shards (with 50 joined).
• Gear: I am missing the figures for the smallest box (H1-06), but as it seems, there is always a net loss in gear/salvage no matter how many players join.

So you have to answer probably two questions:

• Do we badly need more Han Solo shards? Most likely no (most, >=35, or all players already have Han Solo 7*). If yes for some reason, then go for simming.
• Do I expect more that 30 players to join?
• • No: Do not sim the raid
  • Yes: Are those additional guild tokens worth the loss in credits and gear/salvage? That's up to you to evaluate the trade-off: Example: With 37 joined players, 1.100 more guild tokens 'cost' you 2.540.000 credits and 170 salvage.

TL;DR

• Credits: Simming is bad, always.
• Guild tokens: Simming pays off if 30 or more players join the raid.
• Shards: Simming is good, always.
• Gear: Simming is (most likely) always bad.

You have done great work. Thank you.

Dk_rek

They are being cheap imo but just my IMO

every once in a while you have to throw a bone to a starving dog

Lor_San_Teka

vksg wrote: »

Nobody out there who can fill out those missing salvage numbers for me? Searched the forums and reddit but did not find anything useful. There should be reward screenshots out there but my searches don't bring them up.

Our first sim

HarryTDF

vksg

Wow, thanks @ both of you. That explains why I thought my own data were inconsistent. So it can be either 10 or 14 from the same pack for the smaller stack of salvage. I updated the table with the average.

BillyTheKid1981

Our first sim rewards. 47/50 joined

LordAnubis

Blackbeardpepe

Fathertaylor

Thanks for this. It’s some really good information.

Snake2

Not sure how you might work this into your numbers, but based on 60 raids of full guild reporting I believe the chance of a full piece of rancor specific gear from h1-02 is ~12%. H1-03 is 9%. H1-04 is 6% h1-05 is 3%.

I recorded full drops only and only if they were rancor specific or not. I might be able to use my info to come up with a drop rate for non rancor guild. If that would interest you let me know.

vksg

Engels

im one of the guts who can auto a rancor this new system is a kick in a n**s. I love when they say they are making stuff better but in reality its worse

Snake2

I meant to type non rancor gear instead of guild. (Like carbantis, cuffs, etc etc)

The 12% is kind of an estimate because it's hard to differentiate the 1-3 drops since it's two different boxes in one rewards screen.

The drop rate for a full piece of any kind for the top 3 was roughly 30-35%. 4-6 was around 15 ish percent. When I reduced it to just callers, furnaces, discs and scanners was how I ended up at the 12%

LORDKAI

48/50 joined.

vksg

vksg

An additional remark on fully crafted gear, as there seem to be more implications with this update.
For the sake of simplicity, for the following examples I will stick to runs with 50 players joined.
The averages for a guild to get some fully crafted gear are inconspicuous and comparable.

• A guild gets an average of 3,45 fully crafted pieces on a manual run.
  Check for the simmed figure is rather simple: 9% H1-03 * 1 piece * 50 players => 4,5 pieces expected value.
• A guild gets an average of 4,5 fully crafted pieces on a simmed run.
  Check for the manual figure requires chances of different boxes so head over to my table to see details: 6,9% H1-several * 1 piece * 50 players => 3,45.

Again, nothing conspicuous here. Now this:

CG_SBCrumb wrote: »

• Raid Simming: Veteran Guilds can now sim the Heroic Raid, The Pit! Simming a Raid automatically delivers identical rewards to everyone in the guild. Guilds must have completed the Heroic Rancor 50 times, and there must be at least 35 players with a 7* Han Solo in the guild at the time of battle start in order to sim.

This allowed some small space for interpretation reading it in two ways:

1. Everyone gets his own draw from the same box (equal chances, different reward).
2. Everyone gets the same draw from the same box (equal chances, equal reward).

I really hoped they went for option 1 yet I was afraid they chose option 2. As it turns out, they went for the 'one-draw-for-all' reward (option 2).
What does it mean? They introduced some heavy leverage by reducing the sample size to 1 and adding a multiplier of up to 50 for the result set. There is no longer a binomial distribution for player rewards of a single simmed run (simplified: no longer right-skewed bell curve/Gauss/whatever; I will add some charts afterwards). Okay, so what does it mean again? Let's compare:

• Manual raid run: Rewards to players are normally distributed. Results around the expected value (3,45) are the most likely. The more the deviation from that value, the less likely the result is.
• Simmed raid run: It's black or white. Get 50 fully crafted pieces if you are lucky. Get nothing most of the time.

Doesn't it balance out? Sure, if you look at multiple raids rather than rewards within one raid. However with such a bias in draws we are now talking of years/decades/centuries... to smooth out results.

Not convinced?
With manual runs, the chance for a guild to get 50 pieces in one run pretty much equals zero. Now with simmed runs, this chance is about 9%. Isn't that amazing?
Obviously, this works the other way around as well. With manual runs, the chance for a guild to get no full pieces at all is quite low (~2%). Now with simmed runs, your chance for the whole guild to hit the worst case is some sweet 91%.

Why is that a problem?
It throws off the balance between different guilds (in terms of competitiveness and what you get for equal efforts).
Your simmed chance to hit a streak of three best case draws in a row is 9%^3 = 0,07%. Seems low, isn't it? Well no, actually it's way to high. Out of 1000 guilds, 72 guilds will get this jackpot. Might be some shiny new 150 full Mk5 Droid Callers. Now there are also 72 out of 1000 guilds, which have the same chance (91%^75 = 0,08%) to hit the worst case 75 times in a row!

Which 'group of 72' will your guild belong to?

The more I think of it, anything but a revision of this update would really surprise me.

vksg

Some charts to visualize the difference last updated on 2019-04-10

EdSolo

Kind of a subjective analysis, since it isn't like the gear is pooled throughout the guild. It is a matter of the guild overall getting x pieces of gear manually or y pieces of gear simmed. Under the simmed situation everyone gets the same thing, it is just a matter if you are lucky or not. I have had many a top ten finish with crappy gear and lower finishes that have given better gear. The extra shards are nice for the shard shop (which I swear has not had stun gun pieces show up in over a week it feels like). The credits are meaningless to me at this point. I think I'm sitting on 250 million or so with all my toons at lvl 85. The ease of just having to join the raid and not worry about it being on in the 15 minutes my guild allows for soloing it is nice since if you miss out on putting up a non zero score, you end up generally below rank 25.

vksg

EA_Rtas

vksg wrote: »

@EA_Rtas My long message above the charts was eaten by the forum spam filter. Would be great if you could resurrect it. Without it the charts don't make any sense. Thanks

Should be restored now, just as a heads up we generally will get to these as fast as we can when it happens try not to discuss it if you don't mind as it does come under discussing moderation.

vksg

Keydash1

In all the discussion of "good" and "bad", you neglected to factor in the time cost. This game is a huge time sink. Raids are worse, a time sink attached to a fixed schedule. Our guild will sim the raid as long as the rewards are reasonably comparable on average, just to remove a chore from the game. And the rewards are reasonably comparable.

vksg

Keydash1 wrote: »

neglected to factor in the time cost. This game is a huge time sink. Raids are worse, a time sink attached to a fixed schedule.

Agreed on the time sink. However, time cost is something I refer to as 'individual conditions within a guild' to be taken into account. Nothing I would have to add to the analysis of the reward structure. If a guild has a system in place where only one player does the Rancor and the rest posts 0, we are talking about 1 player investing 6-8 minutes more for a manual run instead of simming. If you have a system of 50 players posting 10M damage they only get the exact same results as the 1-player-run but the time investment and fixed schedule might feel a lot worse compared to simming. It's up to your preconditions.

Keydash1 wrote: »

Our guild will sim the raid as long as the rewards are reasonably comparable on average, just to remove a chore from the game.

That's absolutely fine.

Keydash1 wrote: »

And the rewards are reasonably comparable.

No, unless you add a 'for our guild' to it. I compared the different runs and in most cases they do not give approximately equivalent results. Reasonableness is highly opinionated so I would never add something like that to my data and hence patronizing you/others.

In my above comments I only went so far to propose a unilateral (opinionated) conclusion for two cases:

vksg wrote: »

So you have to answer probably two questions:

• Do we badly need more Han Solo shards? [...]. If yes for some reason, then go for simming.
• Do I expect more than 30 players to join?
• • No: Do not sim the raid

I stick by that.