Bugs that directly influence the outcome of a game

freezai

Live for the Applause
is a Tiering Contributor
Hello everyone, I want to talk about the policy regarding server errors/bugs that directly impact the outcome of the game. The context of this is a recent DPP LC game in LCPL where an impossible damage roll completely changed the game. Both sides submitted teams prior to the match and it was determined that the roll that occured in the game was mathematically impossible. An administrative decision was taken and server errors were deemed to be a part of the game and no action was taken.

Quoting from the tournament guideline and regulations:
Simulator Glitches: Any battle mechanics not properly implemented in the simulator used for a tournament battle are fair game, unless explicitly banned by that tournament's rules. Illegal movesets that are not restricted by the simulator fall into this category as well. However, this does not include faulty tier scripts; for example, if there was a bug in the BW OU tier script that allowed you to use Drizzle and Swift Swim on the same team, taking advantage of this would result in disqualification in that battle. If a game breaking glitch is discovered, we will make a blanket tournament ban on that glitch via an announcement in the Tournaments forum until it is fixed.
There's not a lot of precedent for handling server errors, but it appears to say that glitches are a part and parcel of playing on PS, which I don't think should be the case. It's not fair for players to be punished for something completely out of their control, especially in important scenarios. I think it is important to address this issue now as although this was just LCPL, imagine the outrage this could cause if a bug were to occur in something like SPL, OST, or another highly prestigious tour.

In my opinion, matches decided by a bug should be taken on a case by case basis, and a rematch should be awarded if the server error is deemed to have a direct influence on the outcome of the game. This is a fair solution in a situation without any obvious answer. It also wouldn't affect situations where, say, someone only won 5-0 or 4-0 instead of 6-0 in a situation where the winning team received a glitch in their favour, because then the host can make a judgement decision regarding it.

Replay: http://replay.pokemonshowdown.com/gen4lc-572160700 (Go to turn 31)

The roll in question: Both sets were known, and the unrevealed pokemon at the end of the game was Mantyke.

196 Atk Life Orb Carvanha Aqua Jet vs. 236 HP / 156+ Def Gligar in Sun: 6-8 (23 - 30.7%) -- 22.8% chance to 4HKO (6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 8)

The opposing Gligar lost 8% of its health! (What actually happened)
 

atomicllamas

but then what's left of me?
is a Site Content Manager Alumnusis a Senior Staff Member Alumnusis a Community Contributor Alumnusis a Top Tiering Contributor Alumnusis a Contributor Alumnus
Are we 100% sure that the EVs/IVs were correct, 0 atk IVs/EVs with a negative nature (the default if carvanha had originally been a special attacker) deals 8% (2 damage) 15/16 times.
 
Are we 100% sure that the EVs/IVs were correct, 0 atk IVs/EVs with a negative nature (the default if carvanha had originally been a special attacker) deals 8% (2 damage) 15/16 times.
That does make sense, so on that level I understand the given win, but I think that the fundamental policy here is still a problem even aside from that; even if this example is not one, it should still be stopped before we actually do have one. Unless the glitch is also in the original games, it shouldn't be allowed to effect tour matches. On Smogon, we attempt to cater to a competitive metagame with the only limits being those of the game itself, and glitches that can't be prepared for are the opposite of that. You shouldn't have to hold your win condition in reserve just in case it turns out Ice Beam isn't going to hit their Garchomp super effectively, and you shouldn't have to lose if it turns out that did happen.
 

Xayah

San Bwanna
is a Community Contributoris a Forum Moderator Alumnusis a Contributor Alumnus
Are we 100% sure that the EVs/IVs were correct, 0 atk IVs/EVs with a negative nature (the default if carvanha had originally been a special attacker) deals 8% (2 damage) 15/16 times.
0- Atk Life Orb Carvanha Aqua Jet vs. 156 HP / 76+ Def Gligar in Sun: 6-8 (24 - 32%) -- 22.8% chance to 4HKO
(6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 8)

Even if TheFenderStory had indeed messed up his EVs and IVs as you suggested, Drew would also have had to mess up HIS EVs to the point of having an extra Defense point:

0- Atk Life Orb Carvanha Aqua Jet vs. 156 HP / 156+ Def Gligar in Sun: 2-6 (8 - 24%) -- possible 5HKO
(2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 6)

We have also confirmed with both players that the spreads they used were as in the calc Serene posted. In addition, the bug is recreatable:
http://replay.pokemonshowdown.com/gen4lc-573390480
Serene's Grace used a 156/76+ spread while I was 196+ with a Life Orb, the exact same situation as in the game.

EDIT: I made a mistake in the recreation, as Fender's Carvanha was Jolly, not Adamant. Here's another recreation but with Jolly Carvanha: http://replay.pokemonshowdown.com/gen4lc-573395376
 
Last edited:

Bughouse

Like ships in the night, you're passing me by
is a Site Content Manageris a Forum Moderator Alumnusis a CAP Contributor Alumnusis a Tiering Contributor Alumnusis a Contributor Alumnus
I mathematically confirmed by hand that 6 hp is the minimum damage that should have occurred based on X-Act's onsite article about damage calculation in DPP.

I'm struggling to see where PS's internal calc could have gone wrong. Maybe it omits a +2 step before applying the modifiers? If you enter the modifier stage with 1 instead of 3, you get 2 damage, not 6.
 

Oglemi

Borf
is a Top Contributoris a Tournament Director Alumnusis a Site Content Manager Alumnusis a Community Contributor Alumnusis a Researcher Alumnusis a Tiering Contributor Alumnusis a Top Smogon Media Contributor Alumnusis an Administrator Alumnusis a Top Dedicated Tournament Host Alumnus
As far as I'm aware the TDs were/have been using this as their policy concerning simulator glitches / bugs / illegalities, at least during my time as TD: http://www.smogon.com/forums/thread...mplementation-for-this-season-of-spl.3494374/

I don't think the guidelines quoted in the OP have been updated in quite some time, tho I could be wrong since I know Zeb did a lot with those threads when he took over, but basically my link is an addendum and expansion to the one in the OP.

Anyway, in regard to the question posed in the OP, if it's something that is known to be incorrect as the game is played or immediately following the match, as in a damage calc occurring that should not have been able to, it would be up to the host to decide the extent to which it affected the outcome of the game and decide how to proceed from there. If it was the game decider, the simulator should be fixed and the game recreated or the match replayed with new teams, depending on when and how the glitch occurred. If the glitch is caught after the week is up or the next round is up, it's basically tough titties unless the host and both players agree to a rematch.

The main thing tho is that these glitches, if they are indeed glitches, need to be reported asap.

Abusing known illegalities is a big no-no, that should be edited in the guidelines.
 
I would like to point out this may or may not have been a direct impact on me winning. Of course it helped, but I had already made up my mind to sack Gligar and revenge kill with Bellsprout before seeing this roll/bug. My play was essentially the same, except instead of sacking Gligar I probably would have brought in Mantyke to try and deal damage to Bronzor or just stayed in and sacked Bellsprout, resulting in the same result if I also crit the Chinchou. I don't really see a clear reason why that bug cost fender the game, as even if he had Carvanha alive, he still would've had a 6.3% chance to win, again presuming I got the crit on Chinchou:

200+ SpA Life Orb Mantyke Hidden Power Electric vs. 0 HP / 36 SpD Carvanha: 22-28 (115.7 - 147.3%) -- guaranteed OHKO
196+ Atk Life Orb Carvanha Crunch vs. 76 HP / 36 Def Mantyke: 18-22 (81.8 - 100%) -- 6.3% chance to OHKO

Anyways that's all I have to and want to say on this subject, and please don't contact me regarding this unless:
a) You are directly disproving this point
b) I have to replay fender for one reason or another
c) You are Coconut or a TD

Cheers!

E: if there is evidence for why fender would've won, please update the OP with it to show the bug's impact.
 
Last edited:

Lemonade

WOOPAGGING
is a Site Content Manager Alumnusis a Team Rater Alumnusis a Social Media Contributor Alumnusis an Artist Alumnusis a Forum Moderator Alumnusis a Top Contributor Alumnusis a Smogon Media Contributor Alumnus
I think the op has the right idea, also someone sinister could discover a sim glitch, not report it, and abuse it. While it might not be possible to determine whether or not they had prior knowledge, basing your game plan off a glitch (dumb idea but /shrug) means it probably has an effect on the outcome, which would be covered by this case-by-case.

On a slightly related note, I don't think giving choice to the player / team about taking replay or win is a very good idea (this exact topic was discussed in one of the biggest esports a few years ago, and tournaments changed their procedure). The "obvious" choice is to take the win, but then other people flame for being bad sports or whatever. But then being a "good sport" could lead to a loss, cause tensions within your team, etc. etc. And based on Ginku's article, people don't take things well. I guess if the host makes the decision they will be flamed, but at least the decision is transparent and hopefully objective (and thus not personal). :blobshrugbutwithagoodface:
 

Bughouse

Like ships in the night, you're passing me by
is a Site Content Manageris a Forum Moderator Alumnusis a CAP Contributor Alumnusis a Tiering Contributor Alumnusis a Contributor Alumnus
FYI y'all hoes better all give this a like for me actually putting in the work y'all seem too lazy to do on actually fixing the problem :)

So to the best of my knowledge this is how damage is supposed to be done in DPP, with flooring done every time multiplication or division occurs: http://www.smogon.com/dp/articles/damage_formula

(I trust X-Act's work pretty much implicitly, but I guess I'll tag a few researchers and a few PS tech people to see if they can sort this out for sure... this is my preliminary work. Marty ih8ih8sn0w OrdA The Immortal V4Victini sirDonovan )
Damage Formula = (((((((Level × 2 ÷ 5) + 2) × BasePower × [Sp]Atk ÷ 50) ÷ [Sp]Def) × Mod1) + 2) × CH × Mod2 × R ÷ 100) × STAB × Type1 × Type2 × Mod3)

Mod1 in this case is 0.5 since Sunny Day was active when a water move was used.
Mod2 in this case is 1.3 because Carvanha was holding a Life Orb
Mod3 in this case is 1 because none of the Mod3 multipliers apply
  1. We start with the Level of Carvanha, 5, and multiply it by 2, getting 10.
  2. Next, we divide 10 by 5, getting 2.
  3. Next, we add 2 to 2, getting 4.
  4. Next, we multiply 4 by 40 (the move power of Aqua Jet), getting 160.
  5. Next, we multiply 160 by 18 (the Attack stat of Carvanha), getting 2880.
  6. Next, we divide 2880 by 50, getting 57.6, which is rounded down to 57.
  7. Next, we divide 57 by 19 (the Defense stat of Gligar), getting 3.
  8. Next, we multiply 3 by 0.5 (Mod1 is 0.5 here), getting 1.5, which is rounded down to 1.
  9. Next, we add 2 to 1, getting 3.
  10. Next, we multiply 3 by 1 (assuming that we didn't get a critical hit here), getting 3.
  11. Next, we multiply 3 by 1.3 (Mod2 is 1.3 here), getting 3.9, which is rounded down to 3. (lol CB > LO in this case in LC which is rare).
  12. Next, we multiply 3 by 85 (the minimum number that the random number can be), getting 255.
  13. Next, we divide 255 by 100, getting 2.55, which is rounded down to 2.
  14. Next, we multiply 2 by 1.5 (since Aqua Jet is Water-type, like one of Carvanha's types is), getting 3.
  15. Next, we multiply 3 by 2 (since Aqua Jet is Water-type, which is super effective against Gligar's first type, Ground), getting 6.
  16. Next, we multiply 6 by 1 (since Aqua Jet is Water-type, which is neutral against Gligar's second type, Flying), getting 6.
  17. Finally, we multiply 6 by 1 (Mod3 is 1 here), getting 6.



This is how it looks like it was applied instead (apologies if I'm reading the PS github wrong. I'm certainly not a javascript expert

baseDamage = Math.floor(Math.floor(Math.floor(2 * level / 5 + 2) * basePower * attack / defense) / 50) + 2;

so plugging in data, we do the following:
  1. We start with the Level of Carvanha, 5, and multiply it by 2 and divide it by 5, getting 2, to which we add 2, getting 4.
  2. Next, we floor 4, getting 4.
  3. Next, we multiply 4 by 40 (the move power of Aqua Jet), getting 160, and by 18/19, the ratio of Attack/Defense of Carvanha and Gligar, getting 151.58
  4. Next we floor 151.58, getting 151.
  5. Next, we divide 151 by 50, getting 3.02.
  6. Next we floor 3.02, getting 3
  7. Next we add 2, getting 5.
So baseDamage = 5

Now it looks for tons of modifiers, and this is the order that they seem to occur to me and if they floor or not, etc.

It looks for a weather modifier, which would halve 5 to 2.5, then floor it to 2.
It then looks for the randomizer, so a min roll multiplies by 0.85, getting 1.7, then floor it to 1 (ie just the max roll results in 2, all other rolls result in 1)
It then looks for a STAB modifier, which would multiply 1 by 1.5, to get 1.5, then floor it to 1. (ie just the max roll results in a 3, all other rolls result in 1)
It then looks for an effectiveness modifier, which would multiply 1 by 2, to get 2. (ie just the max roll results in a 6, all other rolls result in 2)
It then looks for final modifiers, in this case is just LO, so it multiplies 2 by 1.3 to get 2.6, then floor it to 2. (or for max roll 6 goes to 7.8, floored to 7)

I'm pretty sure I've done something slightly wrong here as I believe even here the rolls should be 2,2,2,2,etc,6 not ,7 for the max roll but the point remains that PS is NOT calculating damage as X-Act said damage should be calculated in DPP.


Most importantly the Sunny Day/Rain Dance effects on damage calculation should be occurring BEFORE the +2 that affects the base damage. It looks like LO's multiplier is getting applied at the wrong point too.

Some people who know this stuff better than I do should take a look at the nitty gritty and sort out which methodology is actually 100% correct. All I know right now is they don't match.
 

Xayah

San Bwanna
is a Community Contributoris a Forum Moderator Alumnusis a Contributor Alumnus
As far as I'm aware the TDs were/have been using this as their policy concerning simulator glitches / bugs / illegalities, at least during my time as TD: http://www.smogon.com/forums/thread...mplementation-for-this-season-of-spl.3494374/

I don't think the guidelines quoted in the OP have been updated in quite some time, tho I could be wrong since I know Zeb did a lot with those threads when he took over, but basically my link is an addendum and expansion to the one in the OP.

Anyway, in regard to the question posed in the OP, if it's something that is known to be incorrect as the game is played or immediately following the match, as in a damage calc occurring that should not have been able to, it would be up to the host to decide the extent to which it affected the outcome of the game and decide how to proceed from there. If it was the game decider, the simulator should be fixed and the game recreated or the match replayed with new teams, depending on when and how the glitch occurred. If the glitch is caught after the week is up or the next round is up, it's basically tough titties unless the host and both players agree to a rematch.

The main thing tho is that these glitches, if they are indeed glitches, need to be reported asap.

Abusing known illegalities is a big no-no, that should be edited in the guidelines.
I would like to point out this may or may not have been a direct impact on me winning. Of course it helped, but I had already made up my mind to sack Gligar and revenge kill with Bellsprout before seeing this roll/bug. My play was essentially the same, except instead of sacking Gligar I probably would have brought in Mantyke to try and deal damage to Bronzor or just stayed in and sacked Bellsprout, resulting in the same result if I also crit the Chinchou. I don't really see a clear reason why that bug cost fender the game, as even if he had Carvanha alive, he still would've had a 6.3% chance to win, again presuming I got the crit on Chinchou:

200+ SpA Life Orb Mantyke Hidden Power Electric vs. 0 HP / 36 SpD Carvanha: 22-28 (115.7 - 147.3%) -- guaranteed OHKO
196+ Atk Life Orb Carvanha Crunch vs. 76 HP / 36 Def Mantyke: 18-22 (81.8 - 100%) -- 6.3% chance to OHKO

Anyways that's all I have to and want to say on this subject, and please don't contact me regarding this unless:
a) You are directly disproving this point
b) I have to replay fender for one reason or another
c) You are Coconut or a TD

Cheers!

E: if there is evidence for why fender would've won, please update the OP with it to show the bug's impact.
Of course I appreciate the input on this regarding the TheFenderStory vs Drew game specifically, but I would like that to not be the main focus here. That game is over, the week is over, and despite the bug being discovered immediately after the game, a decision was made that will now no longer be changed. Even though I am a proud member of the Belchsprouts (TheFenderStory's team), I also think it would be ill advised to retroactively force a replay no matter what.

However, I would like to bring up one thing:
it would be up to the host to decide the extent to which it affected the outcome of the game and decide how to proceed from there. If it was the game decider, the simulator should be fixed and the game recreated or the match replayed with new teams, depending on when and how the glitch occurred.
I disagree with this. Even if we assume the host is a complete master of every tier being played in the tournament (I am not aware of how proficient Coconut is with DPP LC, so I'm not making any assumptions here), it can still be extremely ambiguous as to how much a bug decided a game. I hate to be stuck on this game too much, but it's the only example I know of so I have no choice. In the TheFenderStory vs Drew game, this bug not only causes Gligar to survive, thus allowing Drew to later sack it and get a free switch into Ponyta without having to risk one of his more important Pokemon, it also causes Carvanha to go down, thus leaving TheFenderStory without a reliable revenge killer for the Ponyta (the crit on Chinchou was sad but irrelevant here). I am not proficient at DPP LC however, so feel free to correct me.
The point is, at that point in time the game state looks completely different. Gligar alive + Carvanha dead leaves us with an entirely different game than Carvanha alive + Gligar dead. As such, whoever wins the game will once again have to come down to player ability, and maybe TheFenderStory can at that point maneuver himself into a situation where he doesn't lose even to a crit, or maybe Drew gets himself into a position where he longer needs a crit. Unless the game is remade, we will never know either. And as such, the impact of the bug on this game (and potentially the next one a bug takes place in as well) is extremely ambiguous.
So here's my proposal; we need to have a concise and all encompassing policy regarding bugs. I'd definitely prefer it if there was no ambiguity regarding the impact to a game, so I'd like to see something along the lines of:

'Any tournament game in which it is confirmed that a simulator glitch has taken place will be replayed after the bug is found and fixed, unless both players agree to pronounce the game valid.'

This way, bugs will no longer have any impact on our tournament scene whatsoever, as any game in which one takes place will be forcably replayed unless both players agree that the game would have gone the same way with or without the bug. While this could cause a more skillful player to be forced to replay a game if a bug takes place and his opponent doesn't want to pronounce the game valid, it will never cause a more skillful player to lose a game to a simulator glitch (again, this is not necessarily what happened in TheFenderStory vs Drew game).
 

Xayah

San Bwanna
is a Community Contributoris a Forum Moderator Alumnusis a Contributor Alumnus
The Gen 4 damage formula simply hasn't been implemented. Every gen after 2 just uses Gen 5's damage formula.
Outside of me not understanding at all why we've been using the wrong damage formula for years, there's also still something wrong:

196 Atk Life Orb Carvanha Aqua Jet vs. 156 HP / 76+ Def Gligar in Sun: 3-8 (12 - 32%) -- 0.1% chance to 4HKO
(3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 8)

That's the BW calc. The damage dealt was 2. Outside of both players confirming their spreads, I can guarantee that the spreads used in my recreation were correct. As such, there's still a bug there.
 

Oglemi

Borf
is a Top Contributoris a Tournament Director Alumnusis a Site Content Manager Alumnusis a Community Contributor Alumnusis a Researcher Alumnusis a Tiering Contributor Alumnusis a Top Smogon Media Contributor Alumnusis an Administrator Alumnusis a Top Dedicated Tournament Host Alumnus
'Any tournament game in which it is confirmed that a simulator glitch has taken place will be replayed after the bug is found and fixed, unless both players agree to pronounce the game valid.'
The problem with a blanket rule like this is it exposes it to being abused in cases where the glitch doesn't have an impact at all and allowing someone who is otherwise losing to call for a rematch just because a glitch occurred. To put it to its extreme in a hypothetical scenario: me and my opponent both have one pokemon left, me a gliscor and my opp a specs latios with ice beam. No matter what I do I lose the match, but let's say my opp uses ice beam and it doesn't ohko my gliscor like it should. Regardless I lose the next turn when they use ice beam again; however, because of your blanket rule I can call for a rematch since a glitch occurred, and I can support my case by saying maybe other calcs were affected in the match, even if it wasn't apparent that there were. And because it's a team tournament there's no way i'm going to agree to the match being valid.

This is why including some sort of oversight to determine the extent it affected the match is needed, people will abuse a rule like that. Maybe not to the extent presented in my hypothetical but ya.
 

Xayah

San Bwanna
is a Community Contributoris a Forum Moderator Alumnusis a Contributor Alumnus
The problem with a blanket rule like this is it exposes it to being abused in cases where the glitch doesn't have an impact at all and allowing someone who is otherwise losing to call for a rematch just because a glitch occurred. To put it to its extreme in a hypothetical scenario: me and my opponent both have one pokemon left, me a gliscor and my opp a specs latios with ice beam. No matter what I do I lose the match, but let's say my opp uses ice beam and it doesn't ohko my gliscor like it should. Regardless I lose the next turn when they use ice beam again; however, because of your blanket rule I can call for a rematch since a glitch occurred, and I can support my case by saying maybe other calcs were affected in the match, even if it wasn't apparent that there were. And because it's a team tournament there's no way i'm going to agree to the match being valid.

This is why including some sort of oversight to determine the extent it affected the match is needed, people will abuse a rule like that. Maybe not to the extent presented in my hypothetical but ya.
Okay, I'll give you that. My blanket rule went to far. However, I still think that we should be way faster to have games like this replayed, and at the very least change the tournament rules and guidelines to reflect the actual current ruling.
 

Marty

Always more to find
is a Site Content Manageris a Battle Simulator Administratoris a Programmeris a Member of Senior Staffis a Community Contributoris a Top Researcheris a Top Tiering Contributor
Research Leader
Well that's a pretty huge flaw when we've been playing official tours for years with inaccurate damage formulas lmao
Yeah, I'm aware, and it's an embarrassment. Four years ago when whoever was in charge of such things asked if Gen 4 was ready for tournament play on PS instead of PO, I said hell no, the damage formula isn't even right. Then I left for a year and a half and suddenly every gen's being played on SmogTours regardless of its completeness and no one seems to care anymore, so I prioritized working on ORAS and fixing the dozens of bugs Gen 6 already had when I came back. I'm not here to throw people under the bus though, so if you want to blame me for this then sure.

Anyway this is totally off topic so here's my two cents on the point of this thread:
If it was the game decider, the simulator should be fixed and the game recreated or the match replayed with new teams, depending on when and how the glitch occurred.
The first part might be unrealistic to expect to have done in any reasonable amount of time depending on the bug, such as the one in the OP. It would be up to the TD to decide what's considered reasonable I guess, but I agree with the rest of the post.
 

Bughouse

Like ships in the night, you're passing me by
is a Site Content Manageris a Forum Moderator Alumnusis a CAP Contributor Alumnusis a Tiering Contributor Alumnusis a Contributor Alumnus
Not throwing anyone under the bus. I think you're an incredible contributor to Smogon and PS... just wish we actually knew it had been wrong. Thank you for all you do.

Can we prioritize getting ADV and DPP damage calcs perfected since we do really, really know how those work for sure? Damage calc fundamentals like this are way more important imo than esoteric bugs about how Sheer Force Dragon Tail interacts with hazards, etc.
 

Marty

Always more to find
is a Site Content Manageris a Battle Simulator Administratoris a Programmeris a Member of Senior Staffis a Community Contributoris a Top Researcheris a Top Tiering Contributor
Research Leader
Not throwing anyone under the bus. I think you're an incredible contributor to Smogon and PS... just wish we actually knew it had been wrong. Thank you for all you do.

Can we prioritize getting ADV and DPP damage calcs perfected since we do really, really know how those work for sure? Damage calc fundamentals like this are way more important imo than esoteric bugs about how Sheer Force Dragon Tail interacts with hazards, etc.
Just to be clear, I meant I wasn't going to throw the people who were supposed to have been working on it after I quit under the bus. So yeah, although I was replying to your post, the "you" was referring to whoever was reading.

Thanks for the kind words! I'll try to work on it soon, but I have limited time in the foreseeable future and even if I hardcode literally every part of it, it'll probably take a while. I'm only one person, after all. :(
Outside of me not understanding at all why we've been using the wrong damage formula for years, there's also still something wrong:

196 Atk Life Orb Carvanha Aqua Jet vs. 156 HP / 76+ Def Gligar in Sun: 3-8 (12 - 32%) -- 0.1% chance to 4HKO
(3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 8)

That's the BW calc. The damage dealt was 2. Outside of both players confirming their spreads, I can guarantee that the spreads used in my recreation were correct. As such, there's still a bug there.
Forgot to reply to this earlier. Basically what's happening here is I implemented Gen 4 Life Orb's no-recoil-against-substitutes thing long before Gen 5 even had its damage formula correctly done, so it's using Life Orb code that's five years old that no one thought of fixing when they changed the current damage formula. So now Life Orb's modifier is done in the wrong place and gets floored earlier in the process and results in 1-2 HP lower rolls sometimes, which becomes more obvious at lower levels.


Anyway, no more derailing from me; please VM me if there's anything else!
 
Situations like these will be handled on a case by case basis. This is obviously a bug we were not in the know about, and the LCPL match directly referenced here falls outside my jurisdiction. I made my suggestion to the host and left it at that - if it doesn't take place in the tournaments subforum I'm not going to take action unless I feel its a particularly egregious break of what we expect in a tournament on Smogon. This event does not justify TD intervention in my opinion.

I appreciate the attempts to create blanket solutions but they simply won't work. I trust the general reasoning capabilities of my team more than I do any attempt to make such a sweeping policy for something that comes up so incredibly rarely, and as such, would prefer the flexibility to rule as we see fit when the situations themselves do happen.
 

Users Who Are Viewing This Thread (Users: 1, Guests: 0)

Top