This is topic Roar failing for no reason? in forum Research Lab at The Azure Heights Forum.


To visit this topic, use this URL:
https://www.math.miami.edu/~jam/azure/forum/buzz/ultimatebb.cgi?ubb=get_topic;f=1;t=001399

Posted by MasamuneXGP (Member # 3412) on 01-10-2004, 02:22 AM:
 
Heya all. I'm noticing a perculier aspect of Roar/WW. It seems to fail without reason every once in a while. I don't know if it's luck, but it seems to happen in a very specific scenario.

On one side, a lv.42 Skarmory and an irrelevent pokemon. On the other side, a lv.20 Aron and another irrelevent. The two irrelevents move, then Skarm uses Roar on its ally. The ally flees, and my lv.70 Rayquaza is dragged out, making it Skarm and Rayquaza vs Aron and an irrelevent. Then Aron uses Roar on the newly arrived Rayquaza. Aron's roar fails.

The crazy part is, Aron's roar works if any other poke BUT rayquaza is dragged out by the Skarm's roar. Again, this may be mere concidence, but it happened 4 times in a row. There was one occation where my Skarm's roar failed against Rayquaza as well. My primary theory is that it has to do with level. Perhaps something akin to the OHKO moves, where accuracy changes with level difference. But... I need something concrete. Does anyone have any info on this?
 
Posted by Uiru (Member # 437) on 01-10-2004, 11:37 AM:
 
Did your Roar ever work against Rayquaza?
~Uiru
 
Posted by MasamuneXGP (Member # 3412) on 01-10-2004, 04:37 PM:
 
Yes, in any other situation except that one. Although that could be concidence as well.
 
Posted by Continue (Member # 31) on 01-11-2004, 04:15 PM:
 
I had originally discounted this as a bug in the standard part of the game, since I never encountered it during link testing. I always had the same level pokemon in link testing, though.

I tried it with different level pokemon and found that the OHKO accuracy bonus seems about correct. There is no failure on pokemon that are the same level or lower than the pokemon using Roar/WW. For pokemon that are of a higher level, there is a definite failure rate, regardless of the species of pokemon involved.

I don't currently have enough results data to see if it is the same rate as the OHKO bonus. What I have right now is close, but it could just appear that way due to a small sampling size.
 
Posted by White Cat (Member # 42) on 01-12-2004, 03:44 PM:
 
Back in RBY, I noticed that CPU Pokemon would sometimes have moves like Growl and Tail Whip fail, but could never figure out why. I wonder if this was also because of a difference in level (or perhaps Speed)?
 
Posted by 137 (Member # 1465) on 01-14-2004, 02:37 AM:
 
Well, back in the day, didn't the moves listed as 100% mean 127/128? Or was this a bit more frequent than that? Or am I totally wrong there?

edit: but a fail is different that a miss most times, right? Maybe the fail was due to the fact that Growl had been used 6 times already, and the stat can't be lowered further. Computer opponents don't care about that sort of thing. The new games have a specific message for that, but I don't know if the old games do.

[ 01-14-2004, 02:40 AM: Message edited by: 137 ]
 
Posted by White Cat (Member # 42) on 01-14-2004, 01:31 PM:
 
No, it happened far more frequently than 1 in 256 that the CPU would fail with those moves, even if they had been used less than six times (or not at all).
 
Posted by LanderZRPG (Member # 1615) on 01-14-2004, 08:15 PM:
 
Isn't there the bug of:

Roar needs to go last, unless another Roar is being used, then it needs to go before that Roar.

Or some such variation? Hence why Vital Throw vs Skarmory in GSC, if I recall correctly, got used.
 
Posted by Continue (Member # 31) on 01-15-2004, 11:01 AM:
 
That behavior, Roar/WW having to go last in order to work, is not present in RS.
 
Posted by Continue (Member # 31) on 01-17-2004, 12:45 PM:
 
I've performed some organized testing (as in actually keeping track of the exact results) and I no longer believe that the failure rate follows the OHKO bonus formula.

The Roar user for my trials was level 31.

The first set of 100 trials was against 3 level 100 pokemon. The expected rate of failure by the formula would be 26.9%. Roar failed 28 times. This was pretty close to the predicted value and made me think that might be correct.

The second set of 100 trials was against 3 level 50 pokemon. The expected rate of failure by the formula would be 7.4%. Roar failed 21 times. Halfway through the set it was pretty apparent that it wasn't even going to be close to the formula-predicted value.

200 trials really isn't a good enough sampling size, but I'm going to settle on it unless I end up with a huge chunk of spare time soon, or somebody else wants to try this. For now, I'm taking my conclusions from what I've gathered so far.

The failure rate for both sets isn't that far apart, 28 and 21. Considering the level difference, these values seem to be too small a range for a formula-generated failure rate. It looks more like there's a flat percentage failure rate regardless of the level difference. Going on that basis and averaging out the number of failures, the number comes out to 24.5. It's close enough that I'm willing to call it a flat 25% failure rate.

I know that 25% isn't a rate that has been previously found significant in the pokemon games, but this situation is unusual enough for me to think it is appropriate.
 
Posted by Anonymous 2003 (Member # 3658) on 08-12-2004, 04:10 PM:
 
quote:
Originally posted by Continue:
That behavior, Roar/WW having to go last in order to work, is not present in RS.

I think in the Nintendo guide they said it did go last. Are you sure?

~Anonymous 2003
 
Posted by TeLeFonE (Member # 2293) on 08-13-2004, 02:52 PM:
 
quote:
Originally posted by Anonymous 2003:
I think in the Nintendo guide they said it did go last. Are you sure?

~Anonymous 2003

What Continue meant is that it is not required that Roar go last to succeed. Here's the example:

In GSC, two pokemon could both use Roar and the faster pokemon's Roar fails and the slower one's works. Hence the Curse/Roar combo to stop pseudo-hazers. This is also the same with other must go last moves like Counter and Vital Throw.

In RS, however, if the above senario happens, the pokemon that is faster and uses Roar first is successful in Roaring away the other pokemon.
 


Karpe Diem