Does Slippery Ice work?

Pet Battles
Says will make opponent miss 20% of the time. I know it's a chance, but it's large enough chance that I should see it.

Yet when I was testing it out last night (pve). Ran my penguin for a while vs mostly otters and crickets, there was only 4 misses from opponents in 8 matches (3 opponents each match) being constantly in ice. Seems like I should have seen a lot more misses than just a few... it's a large enough test sample with a really large number of attacks. Yet those crickets and otters missed less than the base miss.

Anyone tried this move lately?
RNG is RNG.

That's still a 75% accuracy rate even w/ the debuff.
You would need a much larger sample size to get a real read on it.
You would need a much larger sample size to get a real read on it.


How large do you mean? 8 battles is a pretty big sample...
05/15/2013 10:42 AMPosted by Kenpo
You would need a much larger sample size to get a real read on it.


How large do you mean? 8 battles is a pretty big sample...


Hundreds my good sir. And that is why I take people who think they know stats with a grain of salt.

Edit: A sample size of 8 can have a variation in the neighborhood of +/- 54%. Just going up to 100 would get you into the area of +/-15%. As you can see one of those numbers is far more reliable than the other and way less likely to be skewed by streaks of variation. Thanks for your time.
yes it works well, i use it quite often. however, by itself isnt all that effective. but u can stack it with sandstorm and it progressively gets better. you can even swap in a skunk for a very large chance to miss. problem with that is you yourself have to do a lot of swapping. it would be a great team if they came out with another pet besides fridged frostling that had a miss ability AND a end of turn ability like ice tom/geyser etc. then swapping would be less damaging.
05/15/2013 10:49 AMPosted by Hearus


How large do you mean? 8 battles is a pretty big sample...


Hundreds my good sir. And that is why I take people who think they know stats with a grain of salt.

Edit: A sample size of 8 can have a variation in the neighborhood of +/- 54%. Just going up to 100 would get you into the area of +/-15%. As you can see one of those numbers is far more reliable than the other and way less likely to be skewed by streaks of variation. Thanks for your time.


His sample size isn't 8, it's the number of attacks made under those conditions during the 8 battles.
It only works when used AGAINST you. When you try to use it against someone else it never works. At least that's been my experience. :)
05/15/2013 10:42 AMPosted by Kenpo
You would need a much larger sample size to get a real read on it.


How large do you mean? 8 battles is a pretty big sample...


8 is a very, very small sample. Try hundreds or even thousands lol >_>
I used it extensively with a winter-themed team and it does appear to increase the miss rate.

However, its more important duty is to make things chilled that are otherwise immune to blizzard (elementals) so you can lay the bonus damage smack-down with the other penguin attacks.
His sample size isn't 8, it's the number of attacks made under those conditions during the 8 battles.


I am still quite sure the time the buff was up was less than 100, probably by a sizable amount. Add in any other variances such as pet accuracy, possible buffs debuffs besides the power in question, pet levels.

The ops far less than scientific attempt to gather data much less ability to analyze it was the point of my post which I very much stand by.
In order to meet the assumptions for a one sample proportion test, he would need to have 25 attacks made against him with the debuff up at LEAST, and this would have a very low confidence.

n(p)>5 and p would be the 20%

n(.2)>5
n>25 attacks

Assuming that in 8 fights the opponent uses more then 3.125 attacks per fight, he does have enough. However, he would have had to kept track of how many hits attacks the opponent used in the if he wanted to figure out what the odds are of there only being 4 misses.
05/15/2013 07:18 PMPosted by Vurheelin
Assuming that in 8 fights the opponent uses more then 3.125 attacks per fight, he does have enough.


You are assuming a lot more controls than just 3.125 attacks pet fight. Again there are many factors that can skew the data. Pet levels, the powers in play and their accuracy, the possible presence of outside bufffs, the potential for the 5% dodge to come into play. Oh and avoidance moves can effect it to. Basically any dodge/avoidances effectively make the sample smaller.

If it was a true controlled test using a 100%(minus the 5% miss rate) accuracy move over and over on an equal level pet with no other buffs or debuffs in place you would still have a 1 in 5 shot of having extreme rng effect your numbers, or close to it, with a sample size that small. And this all assuming the person collecting data isn't using the old sure fire method of I remember that was how many I missed method that is generally the culprit for post such as this.

I feel pretty confident in saying that the controls were far from that managed and therefore would make the sample even smaller than it looks. Plus I am confident in saying that someone who post in a manner similar to the op have a tendency to bias there own 'stats' and improperly gather data.
You are assuming a lot more controls than just 3.125 attacks pet fight. Again there are many factors that can skew the data. Pet levels, the powers in play and their accuracy, the possible presence of outside bufffs, the potential for the 5% dodge to come into play. Oh and avoidance moves can effect it to. Basically any dodge/avoidances effectively make the sample smaller.


As you said less than scientific.

However it was 8 matches, 3 pets each. All fights were in PVE vs some combination of crickets and otters... so no other accuracy buffs/debuffs present. Each fight had a good number of turns, i didn't keep track of the numbers, just misses which were so rare it was easy to track.

Now I just got to run maybe like 50 fights and test it out in a long run. Probably will wait till 5.3 where base miss/dodge will go away so it'll be easier to test.
Oh look - another Kenpo thread where something he doesn't understand is 'broken'.

I almost miss Flaya's 'this move is broken' threads without any care to think about it or research it because he 'can't be bothered'.


You are assuming a lot more controls than just 3.125 attacks pet fight. Again there are many factors that can skew the data.


Whether or he biases the data by how he collects it has nothing to do with him needing an average of 3.125 attacks made against him with slippery ice up to get a big enough sample size.

Join the Conversation

Return to Forum