After the Dugtrio ban in BW OU where a simple majority was used, I noticed that despite Smogon’s general attempts to keep stuff in line with precedent and have consistency, its strange that our suspect process is still heavily inconsistent. In different gens and different tiers, different requirements are used. A quick glance at the blind voting forum shows this:
1. OU – 60%
2. UU - 51%
3. NU - 51%
4. RU - 51%
5. LC - 60%
6. DOU - 60%
7. PU – 51% to ban, 60% to unban
8. BW OU - 10/19 No percentage specified in the OP
9. Monotype 60%
I believe it's ideal for this process to be standardized. There does not appear to be a reason for the variation across tiers and this system can potentially make suspect tests susceptible to outside biases based on the suspect in question.
The recent BW OU suspect test is a strong example of the issue of standardization. Since the last Excadrill suspect used a different process, I took a look at the initial Excadrill suspect test. That one used a 60% suspect vote while the current Dugtrio ban used a 51% suspect vote. Considering they were both for a relevant oldgen that is still heavily played in official tournaments, it is detrimental to the suspect process that two different requirements were used for the same tier.
The precedent is giving as much independence as possible to all sections of our site. Fixing small inconsistencies across similar areas isn't a top priority, because every area has different userbases, needs and preferences. Consistency can be neat, but neatness alone isn't enough justification to force everyone to do the same.
Why should OU, a tier in which the list of Pokemon available only changes when new games or the rare event Pokemon are released, have the same standards as UU / RU / NU / PU, tiers that can see major changes on a monthly basis. Why should older gens, that have very inactive ladders but very strong tournament presence, follow the same standards as current gens, where the least active ladder is more active than all ladders from RBY to BW put together.
Forced consistency isn't necessarily a good thing. There are some arguments in favor of keeping things, such as percentage of votes needed across tiers, consistent across tiers, but precedent isn't one of them.
Another example of this can be seen in LC, where the tests have seen significant inconsistencies:
1. Torchic - 60%
2. Vulpix - 60%
3. ORAS Drifloon - 60%
4. ORAS Diglett - 50%
5. ORAS Diglett + Gothita - 66%.
The complete lack of consistency actually mattered here, as Drifloon received 62% ban votes, which would have resulted in it staying in the tier if it was under the same requirements as the Diglett/Gothita test done before. In fact, LC here changed its requirements in 3 back to back suspect tests. Both these situations highlight the importance of a standardized system.
That's pretty much macle's fuckery in 2016. Why? who the fuck knows, but afaik it's an isolated incident and way beyond my time as tiering admin. That aside, I agree that kind of inconsistencies are terrible and I'll make sure that doesn't happen again. However, forcing tiers to be consistent with their % doesn't mean we need to force all tiers to have the same %.
Personally, I believe we should standardize the 60% vote requirement rather than 51% majority. In a hypothetical situation in which votes are tied at 20-20 in a field of 41, if the simple majority is attained, the resulting ban is not as definite as it would have with a 60% majority. It is possible this could help minimize future unbanning of controversial suspects.
It could also increase the chances of a controversial suspect not getting banned, which can force the majority of the voters to play in a tier with a Pokemon they expressed they didn't want in the tier and/or forcing tier leaders to do quick resuspects. 51% isn't as definite as 60%, but you don't risk going against the wishes of 59% of the voters. One option favors the status quo and potentially more stability, and the other the majority of the voters.
Both options have their flaws and merits. Tiers like OU will always favor the status quo, as the tier rarely changes outside of teambuilding trends. Tiers like UU are changing on a regular basis, for example Amoonguss went up last month and that alone suddenly left the tier without one of the best check / counters to some of the metagames biggest threats, so there's very little reason to go with the "pro status quo" option, so simply following the wishes of the majority of voters is preferable.
It's really not a coincidence that the 4 tiers that can see monthly changes are the ones using 50% + 1.
Anyway, as tier leader I always believed that 60% was the perfect balance between having a sizable number of people in agreement but not getting carried away with needing too many votes so that nothing gets done. Honestly, simple majority is crap. 50% + 1, 51%, however you want to look at it, is not a healthy way to be deciding major changes for a metagame. It's not that 51% is too small, it's that 49% is way too large of a group for you to ignore in a suspect when the "victorious" group isn't much larger than that. In some suspects, that can be fewer than 10 people.
There really isn't much else to say about this topic. Feel free to argue 60% vs 66% if you want, but 50% just doesn't cut it for me. Also, regardless of what we end up deciding for suspects, whether you're banning or unbanning something, the % used to make a change should be the same. I don't understand why they would be different.
I don't understand how "49% is too big to ignore" is an argument used against the 50% cutoff. With the other cutoffs you can ignore 59% and 65%, so by that metric alone they are worse. 50% has some flaws, but it's pretty clearly the better option in terms of not ignoring large number of voters.
