Is Potent Sleep as good as advertised?

Do you ever wonder if an ability really succeeds as often as it’s supposed to?

I tracked Potent Sleep for five matches and 120 casts.  The results were 91 hits and 29 misses.  That’s a 75.8% success rate, so it’s right around the advertised rate (75%).

If you look at the individual stats, though, you can see why confirmation cognitive bias causes us to think something isn’t doing what it’s supposed to be doing.

Match 1 - 11 of 13 (85%)

Match 2 - 21 of 29 (72%)

Match 3 - 25 of 30 (83%)

Match 4 - 13 of 15 (87%)

Match 5 - 21 of 33 (64%)

I lost Match 4 (my best sleep rate) after Potent Sleep missed on a purifier and I couldn’t recover.

I won Match 5 (my worst sleep rate, which started out 1 for 4) which was the 4th Gatekeeper.  Of the seven sleeps that I would consider critical to the match, Potent sleep was 7 of 7.

Match 3 had the worst streak with three consecutive misses (even though there were only five misses in the whole match).

If I hadn’t been tracking, I would have told you sleep let me down in Match 4 and really came through for me in Match 5 (which is true in a way), but statistically, it was just the opposite.  

What is and what seems to be aren’t always the same thing. (So why does Rexo always kill himself with Sudden Death then, right?)

One of my matches I used Goldtail the entire time and didn’t miss a single time. Yet on another match I missed 5 times in a row with Goldtail causing a full slept lock to break. Was able to recover with Wraithcaptain. 

If you look at the individual stats, though, you can see why confirmation bias causes us to think something isn’t doing what it’s supposed to be doing.

That’s not really how confirmation bias works. We look to affirm things that we believe are true rather than disprove, so for example using a small sample size to prove it’s 75% when using a much larger sample size would give a more accurate result. It comes down to basically selecting facts we want to use and disregarding ones that stray away from our initial presumption.

That actually sounded sassier than intended, but you get the point.

No, I appreciate the clarification.  If I compare these two…

Confirmation bias - interpreting evidence as confirmation of one’s existing beliefs

Cognitive bias - inferences drawn in an illogical fashion based on perception of input

…I should have used the broader term of cognitive bias.

The post was actually in response to two statements I heard during UC that “Potent Sleep doesn’t seem to be working lately.”  I was trying to show that my perception of input (multiple missed sleeps) could lead to an illogical conclusion (Potent Sleep isn’t working like it should).

I guess confirmation bias would have responded, “I knew it was broken!  I just missed three sleeps in a row.”

Been testing rexo a bit and have had him hit himself 5/7 times so far. Need a bigger sample size.

It’s actually based on a grain of truth in that the sleep inaccuracy scales with the buffs.
I don’t have exact numbers but roundabout the 600 point potent sleep is pretty much the same as normal sleep, cause they both miss about 5 times before they hit.
Also yeah that sounds right. Although cognitive bias is a VERY broad term, there’s like 50 of the damn things, but hey it gets the job done XD

I don’t know if this means anything, but I found in my experience with the past 2 uc’s that potent sleep has about a %50 chance against any monster with knockback. Was my biggest downfall in most matches