Event reward system

[ET-inf3rno]

Technically I'm talking about Palavyn saying getting the shuffle doesn't reduce your number of dailies when you said it reduces enjoyment. I have no idea why you and I are arguing, since I was trying to back your point.

Me neither lol.

 

[Deleted User - 4646370]

Actually, the shuffle reduces the odds to get the daily exclusive from 1/16 to 1/17.
Proof : In each run (=between two shuffles) you always get the shuffle as last price, while you have a 50% probability to get the daily exclusive or any other price (it can be either before or after the shuffle in the order you pick the chests). So you get the shuffle twice as much as any other prize, which means the odds are 2/17 for the shuffle and 1/17 for other prizes.
Then there is the ×2, which adds its odds to all other prizes proportionally to their odds (the shuffle still is twice as likely to be after the ×2 as any other prize, as it's always in the remaining prizes) so in 16 prizes you actually get averagely 1 daily exclusive (and 2× 10 seeds from shuffle, and 1× all other 13 prizes, not including the ×2 as it's not an actual prize). All this considering you're not using the informations from the "reveal two", what would even more complicate the math.

 

[Kekune]

So you get the shuffle twice as much as any other prize

Well that explains my animosity towards it.

 

[Pheryll]

The thing is that the shuffle returns the amount spent on it (or the amount for the x2 and it if done consecutively). So the shuffle does not make the daily prize cost more event currency; it only makes you spend more time clicking.

 

[Enevhar Aldarion]

The thing is that the shuffle returns the amount spent on it (or the amount for the x2 and it if done consecutively). So the shuffle does not make the daily prize cost more event currency; it only makes you spend more time clicking.

Right, the shuffle is basically a free turn that gets you one of the event item needed for a grand prize.

But the shuffle will always be a big pain when it happens before you get the daily prize, if there are a lot of different daily prizes you want to try for. In the May event, there were only three daily prizes I wanted. I had enough event currency on those days to get three or more of each of those three, even with hitting the shuffle several times before hitting the daily and a couple of times not hitting the shuffle until the 14th or 15th chest, meaning less tries for that daily. But the Winter event was a lot worse for me and I got hardly any daily prizes because the shuffle came up much more often before the daily had a chance to. Yes, I know that was just my horrible luck at picking the wrong chests over and over, but that also makes this system very unfair.

After thinking about it some more, I think the first set of chests each day should not include a hidden shuffle, but should instead auto-shuffle when you get the daily prize, and then after that, the hidden shuffle is included in the 16 chests. That way, even if you have to open all 16 chests in the first batch, you are guaranteed at least one of the daily prize. This could make a decent compromise between making the shuffle a manual choice and having it be hidden and random.

 

[Pheryll]

But the shuffle will always be a big pain when it happens before you get the daily prize, if there are a lot of different daily prizes you want to try for.
...
Yes, I know that was just my horrible luck at picking the wrong chests over and over, but that also makes this system very unfair.

The shuffle makes the distribution simulate a singular chest from the chest system. The deviation created by the shuffle is no greater than what is present when selecting a chest where each outcome is given a probability. In fact, the reveal tile helps you do much better than a blind random draw even if it only reveals items that you wish to avoid. If you had to choose a chest with a 1/17th chance of the daily, you would not be doing as well as using the 4x4 grid.

 

[michmarc]

Regarding the shuffle, the simple cases can be reasoned about with math pretty directly, but adding in the x2 and especially the reveal both change the dynamic a lot. I wrote an app to simulate this to see which combination of rules is actually the most beneficial and by how much:

----: Grand: 1000000 Daily: 62500
---R: Grand: 1000000 Daily: 62500
--D-: Grand: 1000000 Daily: 117463
--DR: Grand: 1000000 Daily: 131502
-S--: Grand: 1133275 Daily: 66801
-S-R: Grand: 1142910 Daily: 81208
T---: Grand: 1058618 Daily: 66373
T--R: Grand: 1060055 Daily: 66819
T-D-: Grand: 1058975 Daily: 125130
T-DR: Grand: 1066201 Daily: 140102
TS--: Grand: 1209772 Daily: 71280
TS-R: Grand: 1222986 Daily: 87924 <-- This is what the event actually uses

T = 'Two X'; S = Shuffle; D = Daily Shuffle (winning the daily prize automatically shuffles the board); and R = Reveal.

The reveal strategy is (in order)
* If the 2x is visible and you haven't drawn the daily, draw the 2x.
* If the daily is visible, then pick the daily.
* If the shuffle is visible and you've already found the daily, draw the shuffle.

The numbers are for a million 'picks' from the board. 'Grand' is the number of 'grand prize' tokens you get (normally it's 1:1 but both 2x and reshuffle give you a free pick).

It's trivial to see where some of the numbers come from: If there is no reshuffle at all, you'll gather exactly 62,500 prizes (one every 16 picks). If you only have the reshuffle, because that's basically a free square, you'd average one prize for every 15 picks, which is what the numbers show.

Making the board automatically reshuffle when you draw the daily prize (which is virtually equivalent to having a separate 'reshuffle' button assuming you'd only press it after you hit the daily) almost doubles your rate of return.

Aside from that rule change, note that all of the other rules help you out. The rate of 87924 per million picks is about one daily prize every 11.4 plays. Without a way to reshuffle, you're always drawing all 16 presents before moving on. Reveal helps you tactically grab dailies or reshuffles or helps you avoid junk. And it's obvious why 2x is a good thing.