Talk:Protection

RS erratic progression

I'm going to be removing this from the Protect page, so I'll leave it here for easy access in case someone really wants to add it later. If you want to add it, I'd suggest putting it in a pre-collapsed table. It'd be most useful if you got data for FRLG and Em too.

The following success rates are accurate to Ruby and Sapphire; from the 53rd turn onward, FireRed and LeafGreen have a different sequence of values, as does Emerald<!-- meaning all three are different from each other, but members of a pair are the same -->.<ref>http://forums.glitchcity.info/index.php?topic=6603.0</ref> These numbers represent success rates out of 65536, which means that the move has a 1/65536 chance of failing on the first use.<ref>[http://www.nicovideo.jp/watch/sm19879961 TASさんがまもるの限界に挑戦(裏)] (Japanese)</ref>

{| style="background: #{{normal color light}}; {{roundy|1em}}; border: 3px solid #{{normal color}}"
|-
! style="{{roundytl|5px}}" | Turn
! style="{{roundytr|5px}}" | Success Rate
! style="{{roundytl|5px}}" | Turn
! style="{{roundytr|5px}}" | Success Rate
! style="{{roundytl|5px}}" | Turn
! style="{{roundytr|5px}}" | Success Rate
! style="{{roundytl|5px}}" | Turn
! style="{{roundytr|5px}}" | Success Rate
! style="{{roundytl|5px}}" | Turn
! style="{{roundytr|5px}}" | Success Rate
|-style="background:#FFF"
| 1
| 65535
| 17
| 168
| 33
| 89
| 49
| 2000
| 65
| 110
|-style="background:#FFF"
| 2
| 32767
| 18
| 266
| 34
| 56
| 50
| 100
| 66
| 90
|-style="background:#FFF"
| 3
| 16383
| 19
| 289
| 35
| 57
| 51
| 65535
| 67
| 64
|-style="background:#FFF"
| 4
| 8191
| 20
| 270
| 36
| 61
| 52
| 65535
| 68
| 95
|-style="background:#FFF"
| 5
| 118
| 21
| 343
| 37
| 157
| 53
| 22
| 69
| 69
|-style="background:#FFF"
| 6
| 165
| 22
| 271
| 38
| 247
| 54
| 30
| 70
| 99
|-style="background:#FFF"
| 7
| 166
| 23
| 264
| 39
| 129
| 55
| 23
| 71
| 187
|-style="background:#FFF"
| 8
| 102
| 24
| 65535
| 40
| 129
| 56
| 40
| 72
| 1
|-style="background:#FFF"
| 9
| 65534
| 25
| 51201
| 41
| 100
| 57
| 2
| 73
| 3084
|-style="background:#FFF"
| 10
| 68
| 26
| 38404
| 42
| 20
| 58
| 50
| 74
| 2820
|-style="background:#FFF"
| 11
| 243
| 27
| 25609
| 43
| 250
| 59
| 68
| 75
| 2827
|-style="background:#FFF"
| 12
| 182
| 28
| 20496
| 44
| 40
| 60
| 60
| 76
| 1285
|-style="background:#FFF"
| 13
| 197
| 29
| 10272
| 45
| 500
| 61
| 19
| 77
| 0
|-style="background:#FFF"
| 14
| 203
| 30
| 5168
| 46
| 60
| 62
| 70
|-style="background:#FFF"
| 15
| 194
| 31
| 78
| 47
| 1000
| 63
| 24
|-style="background:#FFF"
| style="background:#FFF; {{roundybl|5px}}" | 16
| 214
| 32
| 75
| 48
| 80
| 64
| style="background:#FFF; {{roundybr|5px}}" | 80
|}{{left clear}}

Tiddlywinks (talk) 22:47, 25 December 2022 (UTC)

Gen V success mechanics

This was in HTML comments on Protect. It really belongs on a talk page, not hidden. (It's referencing this Smogon post.)

IN NEED OF MORE RESEARCH

The bit shifting in the Smogon link implies that should Protect work 33 times in a row (success rate 1 in 4 billion on the last one), overflow will cause the success rate to start doubling again, until the 65th consecutive use and all future uses are guaranteed to succeed. Is this correct? (Protect only has 16 max PP, but one can alternate between four different protection moves.)

UPokeCenter is normally pretty accurate about these kinds of things, but the site owner has made math mistakes before, albeit very rarely. That site says that after the eighth Protect (1/128), ninth and future Protects are automatically forced down to a 1 in 4 billion success rate, which would actually match Gen II very closely. It doesn't match Smogon though and I'm inclined to trust its bit shifting derivation more. But neither site includes the raw assembly like Gens I, II, and VI do.

This would be worth independent checking, if anyone has the will and the means. Tiddlywinks (talk) 03:16, 26 December 2022 (UTC)

Quick Guard and Wide Guard

Hey, I read the edit note already, but I still think Quick/Wide Guard should be "special protection moves":

• They're variations of Crafty Shield and have a lot more limitations on which moves they block than the rest of the "damage-protection moves".
• Like Crafty Shield and Mat Block, they affect all Pokémon on the user's side (no variation of Detect does that).
• They don't have a rising chance of failing with each consecutive use since Gen VI (much like Crafty Shield and Mat Block).

Yes, they still influence on the chance of other protection moves failing, but I think they should be on the "failure chances" section as a mention, like Encore. Shinka (talk) 18:10, 26 December 2022 (UTC)

The thing about those sections is that the current names/descriptions are kind of the least important aspects. I made those sections solely to group moves with a linked failure chance and other moves. I tried to name them something that I thought was a little more insightful or something, but if people think it makes the groupings confusing, then perhaps the sections should just have more explicit names so that the intention is very plain. Tiddlywinks (talk) 16:39, 27 December 2022 (UTC)

Right now, the section names are very clear IMO. If the Guards will stay with the variations of Detect, the description of that section will have to change, as they do not have a failure chance since Gen VI (according to their own pages). And, if that's the case, I think the description will become a bit confusing or unnecessarily detailed, something like: "These moves mainly protect from damaging moves, some with more limitations. These moves (plus Encore, and except Quick Guard and Wide Guard) have a chance to fail when used consecutively. This applies across moves, including the Guards: if a Pokémon uses Quick Guard consecutively, the chance of Quick Guard or Wide Guard failing will not increase, but it will increase the chances of Protect/Endure/Obstruct/etc. failing". I recognize that this can be better rewritten to clarify the data, but, nonetheless, I just think it'll be simpler to list them as "special protection moves" and have them as a mention in the "failure chance" subsection. Shinka (talk) 17:55, 27 December 2022 (UTC)

Oh shoot, they don't? That half-state doesn't make a whole lot of sense, but it's no wonder I didn't notice since I only looked real close at Protect. I'm sorry, I also didn't notice that that was your last point in your opener; that should be the most important part!

Now I'm wondering about the details of how they affect Protect/etc (if every use of the Guards is actually counted, or maybe it only adds 1...? That makes a certain sense to me), but overall you probably had it right. Sorry again!

(With that mistake, there's a small part of me that's wondering if Encore also has some connection I didn't notice instead of Endure.) Tiddlywinks (talk) 22:08, 27 December 2022 (UTC)

No problem, man. I edited the page already, putting the Guards under "moves that protect the user's side of the field", but feel free to edit or reword it the way you want. I changed the section names too, and, again, feel free to change it if you find a better-suiting option. Sadly I can't say anything more technical about the failure chances between the Guards and the variations of Detect, as I know next to zero about the topic. Shinka (talk) 22:46, 27 December 2022 (UTC)