I’ve been tracking my tokens for quite some time now - when they changed the token odds significantly I started my tracker over from scratch (stored the old data,) so the odds I have are based on token draws from May onward. The sample size is still reasonably low, but I’m definitely noticing a trend.
I’m coming out siginificantly low on golds on every type of token. Now, I’m not one to cry foul at a moments notice, but it’s definitely looking weird. Are other people who track their tokens noticing the same thing? Is this an actual problem? Or have I just had a few bad weeks.
For those who are curious, here’s a table:
Heroic Tokens:
Typical odds: ** 74.7% *** 22.8% ****2.7%
My pulls: ** 78.9% *** 19.3% **** 1.8% (out of 57)
PVE Tokens:
Typical Odds: 71.3% 26.2% 3.2%
My Pulls: 72.0% 24.0% 4.0% (out of 75)
PVP Tokens:
Typical Odds: 72.8% 26.0% 2.7%
My Pulls: 80.0% 15.0% 5.0% (out of 60)
So yeah, I wouldn’t think twice about this, but its across the board.
Please don’t misinterpret this as trolling but I would suggest that the variance you observe isn’t statistically significant. It’s difficult to draw a conclusion from that size of data set.
On the plus side of your 4* pulls are above trend.
You’re 100% on the mark. It’s actually amazing that someone is drawing 4*s at a better-than-expected rate, and then starts a thread implying that the devs are lying about token odds.
It’s only 200ish tokens skald. This data doesn’t look statistically significant to me. You aren’t very far off the expected rates, and your sample size isn’t all that large.
Another odd question.. with the Q&A stating with no uncertainty that they are increasing token odds soon.. why would you open them all now? (DDQ and Ultron aside)
I agree completely, actually. If I was above the board on one type and below on others (even if the above was a smaller sample,) I wouldn’t have said anything. The fact that I was below on everything struck me as odd, and I was wondering if this was pervasive. I suspect it’s not. 192 tokens is a decent sample, but not a great one from a statistical standpoint.
As far as the 4s go, I have been reasonably lucky with them, but the chance on them is so small that I don’t count them one way or the other. I’m pretty much just looking at the 2 percentage and noting it’s high.
I’m trying to decide if I want to reset (again) when they up the % for the 4s. I’ve already reset twice - once when they upped the chance for the 3s during vaulting, a second time when they did away with vaulting.
I think you’re reading too much into my question - I’m just looking for more data. Also, I simply don’t count the 4*s in my calculations - sure I did pretty well by them, but the chance is so small a variance of even one throws the odds way off. I’m just looking at golds vs. non-golds.
Could be, I’m basing it strictly off of the information in the token packs - 9 4*s with a .3 chance each. It would be nice if they gave us more in the way of fractions…
If it is in fact 2.3% then the real percentage is something like .2555 (etc.) So .3 only by the most charitable of interpretations…
that’s how last event started for me. this one not so. about 16 tokens in 2-s and 1-* (10k iso). only ‘usable’ one was doc ock, and the 10k iso is nice, but not worthy of a **** label. 18% is dismal compared to last event (guess I was due for a correction?), other’s, and the posted odds.
Mostly because they told us not to hoard, which tells me they won’t be doing this for a while. It may sound strange, but opening tokens is a lot of the fun I have in this game - holding off for months for a slight bump in a percentage chance is just not worth it to me. That, and I can really use the Iso-8 now.
My Ultron tokens have been coming out crap for me - one gold so far out of probably 10ish tokens. But I didn’t include those because the number is so low.
I also agree regarding the 10k Iso not deserving a **** label. I got one of those last time - I got excited when I saw the blinking gold, and the letdown when it was 10k Iso was palatable. Whatcanyado?
I think the hard thing about tokens (and why I detest their overusage) is because even if you feel like your odds are terrible, the starting odds are terrible to begin with, so you’re probably doing just fine statistically, but still terrible in practice.
For that matter, single token odds are really hard to translate with any statistical significance unless you’ve got a large enough pool. Put simply, extrapolating from a pool of 100 (per token type) is at or very near the bottom end of what could be considered a viable study.
To that end, coming up with a viable “n” value is going to be challenging, due to the different tokens having different probabilities, arguably necessitating separate “n” values for each, making the overall sample pool size even larger. Again arguably much larger.
All of that said, my pull rate feels like -99.00% (yes, negative), but I realize that it has likely been a lot closer to what is listed… which is still just as terrible.
tl;dr: I hear ya, feel ya, and even agree on a base level, but I think we’re both wrong about the facts, but not wrong about how it feels to operate under such relatively stingy odds.
If the odds were off… What would that mean? I’m not saying they are, but if they were?
If you flipped a coin 10,000 times and got heads 40% of the time, it starts to seem like the coin might be weighted. Although, honestly my statistics classes are so far in the past that I can’t remember the formula for determining this.
