PVP Drops

GrimSkald

So, I'm a data horder. I like collecting data. My spreadsheet started so I could track the PVP rewards, and grew from there. I tracked PVP rewards regularly in 2014 until when it changed in 2015, then started again until mid 2015 when I was satisfied with my data.

Anyway, when the new Champion system was introduced, I decided to start again with the fresh data. People were complaining that 2*s seemed to drop less, I doubted it myself, but decided to verify and get a new set of percentages. Now here's where I noticed something weird.

I seemed to be getting a lot of 2*s. Particularly compared to 1*s (almost the same number, in fact.) What's even weirder is that these numbers seemed to stay even - sometimes I'd have more 2* drops than 1* drops, sometimes less, but never a number that was particularly significant. The old data set had 1* drops at 34.1%, 2* drops at 16.2% (so just under half the characters dropped were 2*s.) Here's the new data set:

Reward

---

# drops

---

%
70 ISO-8

---

262

---

32.3%
140 Iso-8

---

136

---

16.8%
250 Iso-8

---

20

---

2.5%
1* Cover

---

199

---

24.5%
2* Cover

---

194

---

23.9%

Weird, eh? I'm extremely suspicious of this data, but I've collected it myself and I'm now over 800 drops. Is the RNG so streaky that it ends up dropping 2* covers more often than it should? Not that I'm complaining, but it seems extremely counter-intuitive and was definitely not the case with my earlier tracking.

Does anyone else have any data that corroborates or refutes this? I'm honestly interested.

GrumpySmurf1002

I lump the 70/140/250 together, but the 1/2* covers I have listed separately, and I have my number of battles won, so can give the percentage.

I don't date it though, so will have to compare to which PvP it started with. I'll see what I can do. I'll check tonight and update.

DaveR4470

The old data set had 1* drops at 34.1%, 2* drops at 16.2% (so just under half the characters dropped were 2*s.)

That's a third of the characters dropping as 2*s, not half.

Your CURRENT data shows a 50/50 split between 2* and 1* drops. More specifically, it neatly maps (within a reasonable margin of error) to the following possible distribution:

70 ISO: 30%
140 or 250 ISO: 20% (with an 80% weighting towards 140)
2*: 25%
1*: 25%

Which you may have now uncovered!

Pylgrim

I think what he's trying to say is that it is surprising that 2*s drop 25% of times or so when the comparative reward of 250 iso only drops 3% of times or so. I think that the difference is (IIRC) 2* covers don't give you the chance to share them with FB friends.

Ruinate

I don't pay attention to pvp rewards because I am going for speed, but I am shocked at how low the drop% is for 250 iso. I remember during double iso week, 500 was dropping like crazy for me. Was a glorious time.

GrumpySmurf1002

GrumpySmurf1002 wrote:

I lump the 70/140/250 together, but the 1/2* covers I have listed separately, and I have my number of battles won, so can give the percentage.

I don't date it though, so will have to compare to which PvP it started with. I'll see what I can do. I'll check tonight and update.

So here's the data I have for Seasons 23-25 (current)

1005 drops
263 1* covers (26.1%)
249 2* covers (24.7%)

So combined 50.8%

That's just the PvP drops, I filtered out the limited LR data I have.

Orion

It seems to me that the 2* cover drop is very streaky. I could have 5 or 6 of them drop in a row and then not see another one for the rest of the PvP. It's still a little surprising to see it on par with the 1* drop.

Warbringa

All I know is that I have played and won 12 PvP (including the seeds) matches on the new format and I have received 70 iso each time, no other reward. I am not joking and never have seen a streak like this and yes I have kept track of each drop. Did they change the reward drops for the test PvP and I wasn't aware? I usually see 140 Iso almost as much as 70 iso on normal PvPs. Didn't know if difficulty of teams you faced altered PvP drops in the old format or something. I am facing teams closer to my own roster now compared to ridiculous number of maxed 4* teams I was seeing before.

Wish I had the OP luck on drops, I hardly ever see 2* drops from PvP matches. Maybe one every couple of events if I am lucky, of course I don't normally play a ton of matches on each event - maybe seven or eight matches including seeds on the old format.

Edit: Played three more matches this morning and did finally get 2 140 iso rewards and the third was a 70 iso of course.

GrimSkald

GrumpySmurf1002 wrote:

GrumpySmurf1002 wrote:

I lump the 70/140/250 together, but the 1/2* covers I have listed separately, and I have my number of battles won, so can give the percentage.

I don't date it though, so will have to compare to which PvP it started with. I'll see what I can do. I'll check tonight and update.

So here's the data I have for Seasons 23-25 (current)

1005 drops
263 1* covers (26.1%)
249 2* covers (24.7%)

So combined 50.8%

That's just the PvP drops, I filtered out the limited LR data I have.

Huh. That's very close to what I have - so it seems plausible - either the 2* and 1* drops are the same or they're extremely close. I'll keep going - I want over a thousand, at least - but I feel more confident in my data with some corroboration. Incidentally mine include Lightning Rounds (and probably are made up of LRs in a sizable percentage - if I get the seed teams I hit them as hard as I can,) but I really doubt they are different.

GrimSkald

Orion wrote:

It seems to me that the 2* cover drop is very streaky. I could have 5 or 6 of them drop in a row and then not see another one for the rest of the PvP. It's still a little surprising to see it on par with the 1* drop.

Oh, I've definitely observed that as well - I'll deliberately continue playing PVP when I get 2* drops when I would otherwise break. The RNG is very, very streaky. That being said, over the long term the streakiness should average out. After all, as another poster observed, you'll get the 70, 140, and 1*s in streaks as well. 12 data points is almost nothing, I'm afraid.

Incidentally, the 250 Iso is extremely rare (I think of it as an outlier.) This is true of the old data set - it was a bit higher (4.7%) but it's so uncommon that I have no confidence in the accuracy of that number. When it drops that seldomly one or two drops less or more will skew the number wildly....