Survivor General Discussion Thread/Statistics

[ajhockeystar]

Hey everyone, I updated the standings spreadsheet now that S9 is over! Lemme know if anything is wrong: https://docs.google.com/spreadsheet...v0Xuh5gCbsxyv-QLWR7hPP21d8/edit#gid=979432923

Here are your new fun facts:
-VigilanteVigoroth is the first person to win in their first game since HeaLnDeaL all the way back in S3. We've had 6 seasons in between then and now where the winner had played at least one game before their win!
-RADicate has joined an exclusive club with Hannahh where he has gotten into the top 5 on 4 different occasions! As an added bonus, neither of them have ever placed below that (for Hannahh, her lowest is 4th).
-Drookez tied the record for most individual immunity wins in a season (with 6), and tied the record for most consecutive individual immunity wins (with 5). Both of these records are shared with HeaLnDeaL.
-After being a hydra in this past game as a twist (they ended up splitting at merge!), Umbreon and Takes Azelfie are now tied in the rankings. What are the odds?
-JALMONT joins Marjane, Hannahh, and ajhockeystar as the only people to make FTC multiple times.
-Season 9 had 12 newcomers, which is the most since CAP S2 (which also had 12).

In terms of times played,
-As of Season 9, we now have a total of 118 unique Smogon Survivor players, and 50 multi-timers.
-50 users have played in 2 or more seasons (24 have played twice exactly)
-26 users have played in 3 or more seasons (11 have played thrice exactly)
-15 users have played in 4 or more seasons (Hannahh/RADicate/ajhockeystar/rssp1/zorbees/Mikaav/Drookez/MK007 have played in exactly 4 seasons)
-7 users have played in 5 or more seasons (Flyhn/Whydon/RODAN/Toni XY have played in exactly 5 seasons)
-3 users have played in 6 or more seasons (Vooper and JALMONT have played in exactly 6 seasons)
-Trace has played in the most seasons (7 to be exact). Of the 11 seasons that have been hosted so far, he has only missed out on Circus S2, CAP S2, Circus S8, and Circus S9.

I'll probably get to updating the immunities and the total votes sections at some point too, so stay tuned for those!

 

[HeaLnDeaL]

JALMONT joins Hannahh and ajhockeystar as the only people to make FTC multiple times.

You forgot the first player to make FTC multiple times, which was Marjane.

 

[HeaLnDeaL]

Also, just to represent, EpicUmbreon29 is the only multi-timer to find idols in all seasons that he's played I think :D

As of this season, multi-time players who have never been eliminated before the merge are:

2 seasons:
Marjane
sparktrain
Blazade
Animus
DLE
apricity
Haruno
spiderz

3 seasons:
Umbreon
Zelfie

4 seasons:
Radicate
Hannah

6 seasons:
vooper PANDA KING


I might be missing people, if I am feel free to holler at me :)

 

[rssp1]

Also, just to represent, EpicUmbreon29 is the only multi-timer to find idols in all seasons that he's played I think :D

If you don't count s1, which had no idols, I also meet this criteria

 

[ajhockeystar]

Hey! With the conclusion of S10, here is the updated stats sheet! Let me know if anything needs to be fixed: https://docs.google.com/spreadsheet...v0Xuh5gCbsxyv-QLWR7hPP21d8/edit#gid=979432923

 

[heuryalone]

Pretty cool that no one has ever placed last in their first game then never played again.

 

[ajhockeystar]

Pretty cool that no one has ever placed last in their first game then never played again.

I think you might be looking at the multi-timers tab! These players fit your criteria: Hitmonleet, Nomark, THE_IRON_KENYAN, Sundar, KrazyCake, Floss

However I do agree that it’s good that the other players that placed last all played again - since there’s been 10 seasons + 2 CAP seasons, that means half of the people who placed last played again!

 

[Da Letter El]

technically it's more than half, since s3 had two elims simultaneously, so there's been 13 people to ever get "last place" in a smogon survivor out of the 12 games

 

[apricity]

Sharing this script to generate 1-1 chats between users in discord. Will create the channel, set permissions, all it needs from you are user ids, role ids, and server ids. Populate the users array, then update the server id as well as the role ids to set specific permissions.


let users = [
['Pash','DISCORD USER ID'],
['Whydon','DISCORD USER ID'],
['alimdia','DISCORD USER ID'],
['Inferno','DISCORD USER ID'],
];

bot.on('message',async message => {
if(message.content=='init'){

var everyoneRole = bot.guilds.get('SERVER ID').roles.find('name', '@everyone');
for (i=0;i<4;i++){
for(j=0;j<4;j++){
if(i<j){
var nametemp = users[j][0].concat("-");
var name = nametemp.concat(users[0]);
await message.guild.createChannel(name, 'text')
.then(r => {
r.overwritePermissions (bot.id, { VIEW_CHANNEL: true });
r.overwritePermissions (users[j][1],{ VIEW_CHANNEL: true })
r.overwritePermissions (users[1],{ VIEW_CHANNEL: true })
r.overwritePermissions ('ROLE ID',{ VIEW_CHANNEL: true,SEND_MESSAGES: false }) //specs
r.overwritePermissions ('ROLE ID',{ VIEW_CHANNEL: true }) //bot
r.overwritePermissions ('ROLE ID',{ VIEW_CHANNEL: true,SEND_MESSAGES: false }) //eliminated
r.overwritePermissions(everyoneRole, { VIEW_CHANNEL: false });
r.setParent('CATEGORY ID');//

})
.catch(console.error);}

}
}

}
}
);

 

[heuryalone]

Looking forward to the standings spreadhseet after the 80 player survivor is over.

 

[HeaLnDeaL]

I think the new data sheets are uh really bad, and the changes seem weird and made on a whim. The whole points earned is entirely subjective and doesn't remotely measure skill and the decision to randomly include non survivor games into the points is highly questionable. Like if you want some new circus leaderboard, then make one separately in a new doc instead of overwriting the survivor data.

 

[apricity]

New sheets look great! Thanks AJ

https://docs.google.com/spreadsheets/d/1ioCwnMNuJ3FdT9Ksnv0Xuh5gCbsxyv-QLWR7hPP21d8/edit?usp=sharing

 

[HeaLnDeaL]

Let's go over all the data that was specifically LOST from the update.

-Challenge performance data
-Idol/advantage data
-separations between multi timers and single time players, which is important because the skill comparisons are quite different between categories
-voting data (which AJ may or may not be compiling to be better than it was before, but it's entirely lacking now at least)

And what was added in their place?
-Non-survivor games
-A point system that directly misrepresents survivor scores and greatly benefits multi timers regardless if they flopped in one or more of their seasons.

In the new sheets, you get more points by playing, not by doing well. An average system, like we had before, shows a more accurate representation of skill and achievement. You could play 5 times and get 10th each time under the current system but still get more points that someone who played twice and got 5th both times.

I get the desire to have some point system to measure more current activity levels, but each new season individually already does that. The only way that I think a point system works is if it's implemented in a seasonal (3 month or so) period like a temporary leaderboard or temporary ladder across a variety of circus games. Having a seasonal leaderboard could encourage current activity while still allowing the normal stats to be kept in a separate tab without being distorted by "did you play the most?" bias. Furthermore, the benefit of seasonal leaderboards is that they could actually chart and show progress across time when you compare a player's seasonal ranking from one season to the next. The current system doesn't really show that and instead arbitrarily rewards points based hugely on participation over placement.

 

[ajhockeystar]

What was deleted was wholly unfinished, without any indication of what seasons were finished or not. What I am working on after these current parts is adding in sheets for ranking:

individual immunities
individual immunities, 1 season
tribal immunities
tribal immunities, 1 season
tribal immunities above/below expected
tribal immunities above/below expected, 1 season
voting correctly % (this I might not do because it's hard to know with split votes and stuff)
total correct votes (same applies)
total correct votes, 1 season (same applies)
tribal councils attended
tribal councils attended, 1 season
votes cast against
votes cast against, 1 season
lowest votes cast against / tribal councils ratio (quits not counted)
lowest votes cast against / tribal councils ratio (quits not counted), 1 season

What has been done is about 2-3 days' worth of work. I haven't made any announcements about it because it is unfinished. I'd appreciate if I wasn't criticized about what's missing after working on all this and adding all this in.

You mention that non-survivor games are added, as if it's a bad thing. They were literally added as a separate section. If you want a sheet for survivor games only, there is one for you. I am unsure of what you are complaining about.

I understand your criticisms about the point system, and I understand and agree that it isn't perfect. It rewards playing more than once, and it uses a large time period to do so. I am fully open to criticism and suggestion from anyone. I believe that this new system is better than the old one since the old one rewards people for playing once, having a good performance, then never playing again. Although the new system does favour playing a lot, it encourages playing more. Once again, if anyone can think of a better method of measuring this, I am open to it.

 

[JALMONT]

I’ll let everyone in on a secret...

do NOT share if you value ur life

Your ranking on an extraordinary small corner of the internet that’s on a game that’s not especially competitive has absolutely zero meaning at all to who you are as a person in real life or even online life and in fact it’s probably quite unhealthy to view it any differently so there’s no need to get worked up about it :)

 

[SlimShadow1027]

Let's go over all the data that was specifically LOST from the update.

-Challenge performance data
-Idol/advantage data
-separations between multi timers and single time players, which is important because the skill comparisons are quite different between categories
-voting data (which AJ may or may not be compiling to be better than it was before, but it's entirely lacking now at least)

And what was added in their place?
-Non-survivor games
-A point system that directly misrepresents survivor scores and greatly benefits multi timers regardless if they flopped in one or more of their seasons.

In the new sheets, you get more points by playing, not by doing well. An average system, like we had before, shows a more accurate representation of skill and achievement. You could play 5 times and get 10th each time under the current system but still get more points that someone who played twice and got 5th both times.

I get the desire to have some point system to measure more current activity levels, but each new season individually already does that. The only way that I think a point system works is if it's implemented in a seasonal (3 month or so) period like a temporary leaderboard or temporary ladder across a variety of circus games. Having a seasonal leaderboard could encourage current activity while still allowing the normal stats to be kept in a separate tab without being distorted by "did you play the most?" bias. Furthermore, the benefit of seasonal leaderboards is that they could actually chart and show progress across time when you compare a player's seasonal ranking from one season to the next. The current system doesn't really show that and instead arbitrarily rewards points based hugely on participation over placement.

 

[HeaLnDeaL]

I understand your criticisms about the point system, and I understand and agree that it isn't perfect. It rewards playing more than once, and it uses a large time period to do so. I am fully open to criticism and suggestion from anyone.

But then...

I'd appreciate if I wasn't criticized about what's missing after working on all this and adding all this in.

The point isn't that you want to add new stuff and that adding new stuff is bad, the point is that you deleted old stuff that was actually helpful. You seem to recognize that the point system isn't a good system either, which is definitely a major part of the critique, but could also be easily alleviated with not deleting the distinction between multi timer and single timer tabs and by using averages across both tabs. Adding stuff is fine, I don't see how the non survivor games are relevant in a survivor data sheet but their presence alone does not hurt the survivor data, it's a net neutral in my book. Deleting some of the previously existing survivor data and distinctions is the net negative.

The conversation isn't blaming anyone for wanting to make proactive updates, the question is why were these particular changes made (removing the multitimer and single timer tabs, adding points that only benefit super multi timers) and what kind of systems do we actually want in rankings?

I believe that this new system is better than the old one since the old one rewards people for playing once, having a good performance, then never playing again. Although the new system does favour playing a lot, it encourages playing more. Once again, if anyone can think of a better method of measuring this, I am open to it.

The old system did not discourage repeat players or encourage people to stop playing. With the old system, we still repeatedly had good players keep playing again (I think a survivor series that encourages super super repeat players is a problem by itself though, but that's a topic for later). The old system literally had two different tabs, one for multi timers, one for single timers. Having two tabs met both goals. The new system removes half of it, and skews so far into super multi timer direction that it's more important to keep playing and survive only a few votes than it is to actually do well.

 

[apricity]

The new system removes half of it, and skews so far into super multi timer direction that it's more important to keep playing and survive only a few votes than it is to actually do well.

In the below image, aura guardian was the merge boot.

 

[HeaLnDeaL]

In the below image, aura guardian was the merge boot.

View attachment 270445

You do realize that +2 points is... MORE points than before? And that if you get 5th in one season, and then merge boot in the next, you still end up with a higher score despite doing worse? The problem with a total points system is that there's no average weight, there's no downside to doing poorly, there's literally only upside to playing again. It's a system that rewards one time high placements (and multi time high placements) while saying poor placements don't mean anything, effectively erasing the poor placements from history.

 

[ajhockeystar]

You do realize that +2 points is... MORE points than before? And that if you get 5th in one season, and then merge boot in the next, you still end up with a higher score despite doing worse? The problem with a total points system is that there's no average weight, there's no downside to doing poorly, there's literally only upside to playing again. It's a system that rewards one time high placements (and multi time high placements) while saying poor placements don't mean anything, effectively erasing the poor placements from history.

This is a good point! So what if we used the same formula, except we subtracted (# of players/4) from everyone's score in each game? This would change the screenshot hal posted above to:

This would make it so:
-The person who achieved the middle placement gets 0 points
-The person in 1st place gets a larger gain relative to the rest of the game than before (therefore rewarding wins)
-There is, in general, a net positive in points gain (ie not punishing people extremely hard for one bad game)
-There is a potential downside to joining games, instead of having bad games be "erased"

 

[JALMONT]

Who cares?

I don't mean that in a rude way. I'm dead serious; is there a group of people out there that's somehow benefiting or losing out by these rankings? As far as I know, I have not benefited in any way from these rankings and I continue to not benefit in any tangible way (actually I might be disadvantaged if anyone in hal's current game sees this...). If I'm supposed to receive a check in the mail or added internet fame, let me know because I haven't gotten any of those things yet.

What exactly is the public interest here in having "accurate" "scores"? Is there a group of people out there trying to figure out who the greatest Smogon Survivor players are? I'm genuinely curious as to why this project, which someone took their time to do presumably for free, is "really bad". Maybe heal can make his own separate sheet?

 

[apricity]

I'm genuinely curious as to why this project, which someone took their time to do presumably for free, is "really bad". Maybe heal can make his own separate sheet?

this is mostly my takeaway. The utility of the sheet is mainly in the other tabs where I can scroll through and get a sense of how the game went when it’s up beside postgame recaps. In that sense it’s a substantial improvement over what we had before, even in its current not-finished state. I also think it’s kind of fun to easily see how people I know perform in different games.

The concern, the urgency of the criticism, and the personal hurt that we seem to have caused Heal is comically mismatched with the level of importance of what AJ took on as a passion project over the weekend.

 

[HeaLnDeaL]

There is no personal hurt or urgency, don't give yourself that much power to hurt me with your oh so mean words :rolling_eyes: Maybe, just maaaaaaaybe, I'm critiquing the point system because the point system actually has logical flaws to it?

While I certainly don't want to be saying that AJ's work free time work is unhelpful, we also have multiple people before AJ that started collecting the data, and a lot of that past data is going to waste or being deleted. You seem to care about AJ's work, but not about anyone's else work to get that earlier data or that it is deleted and their hard work is gone. AJ's collection of the voting data is certainly helpful and the ongoing presentation of that is certainly being improved, but it seems to be mostly a copy/paste of existing voting data, so it also feels sort of disingenuous to credit AJ for doing this all on his own. There's certainly holes, there's certainly investigation he's doing and he's trying to find out more, and that's good, but the truth is that the majority of that is still just copy/pasting, which is a simple button click. And likewise, with a simple button click, a lot of the other data sheets that others spent time doing original work to compile just went poof when certain sheets got deleted. I mean, I have backup copies of it to a certain point, but not for everything though.

If your entire argument comes down to "who cares about rankings?" then why are you defending AJ's rankings? Your arguments seem to be increasingly ad hominem to support someone and discredit someone else based on the notion of it doesn't even matter. Again, you are trying to bypass the arguments that the only-addition system was flawed and not engage with that content at all.

If you're going to have a ranking tab, and put it in front of the season tabs at that, then people are going to look at it, people are going to make comparisons. We sort of have a responsibility then to make those comparisons actually mean something and to be decipherable and logical to the viewer. I don't know the exact formula AJ is using for the points, I'd guess the minus system is an improvement on the always positive system, but I still don't know the formula or what it is exactly trying to do. I do know what exactly averages tries to do though. Again, maybe the minus system fixes things, maybe if I look in the formula bar the universe makes sense now. But it certainly doesn't make sense to reward a player who got 20th in one season and 5th in the other with all the benefits of one and none of the detriments of the other.

AJ, from your perspective, what exactly are the positives of having your point system? I don't really buy the argument that it encourages people to play again, and I also don't see it as necessary for players to even play again in the first point. *Why* does the point system exist, what problems does it solve over a multitimer average system?

 

[Lrr]

I’ll let everyone in on a secret...

do NOT share if you value ur life

Your ranking on an extraordinary small corner of the internet that’s on a game that’s not especially competitive has absolutely zero meaning at all to who you are as a person in real life or even online life and in fact it’s probably quite unhealthy to view it any differently so there’s no need to get worked up about it :)

false, I base my entire identity on that one game of survivor I played

e: also stats like this are mostly misinterpreted anyway since I don't think any of us have statistics degrees or equivalent experience

 

[JALMONT]

There is no personal hurt or urgency, don't give yourself that much power to hurt me with your oh so mean words :rolling_eyes: Maybe, just maaaaaaaybe, I'm critiquing the point system because the point system actually has logical flaws to it?

While I certainly don't want to be saying that AJ's work free time work is unhelpful, we also have multiple people before AJ that started collecting the data, and a lot of that past data is going to waste or being deleted. You seem to care about AJ's work, but not about anyone's else work to get that earlier data or that it is deleted and their hard work is gone. AJ's collection of the voting data is certainly helpful and the ongoing presentation of that is certainly being improved, but it seems to be mostly a copy/paste of existing voting data, so it also feels sort of disingenuous to credit AJ for doing this all on his own. There's certainly holes, there's certainly investigation he's doing and he's trying to find out more, and that's good, but the truth is that the majority of that is still just copy/pasting, which is a simple button click. And likewise, with a simple button click, a lot of the other data sheets that others spent time doing original work to compile just went poof when certain sheets got deleted. I mean, I have backup copies of it to a certain point, but not for everything though.

If your entire argument comes down to "who cares about rankings?" then why are you defending AJ's rankings? Your arguments seem to be increasingly ad hominem to support someone and discredit someone else based on the notion of it doesn't even matter. Again, you are trying to bypass the arguments that the only-addition system was flawed and not engage with that content at all.

If you're going to have a ranking tab, and put it in front of the season tabs at that, then people are going to look at it, people are going to make comparisons. We sort of have a responsibility then to make those comparisons actually mean something and to be decipherable and logical to the viewer. I don't know the exact formula AJ is using for the points, I'd guess the minus system is an improvement on the always positive system, but I still don't know the formula or what it is exactly trying to do. I do know what exactly averages tries to do though. Again, maybe the minus system fixes things, maybe if I look in the formula bar the universe makes sense now. But it certainly doesn't make sense to reward a player who got 20th in one season and 5th in the other with all the benefits of one and none of the detriments of the other.

AJ, from your perspective, what exactly are the positives of having your point system? I don't really buy the argument that it encourages people to play again, and I also don't see it as necessary for players to even play again in the first point. *Why* does the point system exist, what problems does it solve over a multitimer average system?

I applaud you for not clicking my super secret spoiler about survivor ORGs but if you had read it then it should've been abundantly clear that I don't care in the slightest about aj's rankings or the criteria he's using or anything else you're complaining about. I'm not even attacking you! I asked some pretty good questions I thought! i think the overarching confusion I have comes from your bizarre response to someone who ostensibly is doing something positive here (although i suppose from your pov, aj's done you a big insult i guess? that's what the best i can figure). it's not quite clear to me why aj's system is so awful.

anyways, the bolded is the closest you came to answering any of my questions - why exactly do "we" have this responsibility again?

As far as I know, we are not a government organization, multimillion dollar corporation, a non-profit, an academic group or researcher, or any sort of group reporting a board of directors or other overseers. We aren't even a club or any kind of defined social organization. There's no reputation to protect here. So that takes us back to my original question...who cares? Who is using this information to drive their decision making process that they might be so affected by a possibly incomplete ranking system? Is anyone claiming this to be the end all be all rankings of smogon orgs? Who actually cares about that sort of thing? Is there an actual audience that cares about that, and if they exist, to what purpose could they possibly be using this "illegitimate" system for? Look I'd get what you're saying if aj was making up a ranking system to make it look like smoking doesn't cause cancer or something like that. but this can't possibly affect the average viewer's worldview in a drastic and meaningful manner? (can it?)

Again, I think the solution here if we aren't willing to accept that this just isn't that important is to create your own stats sheet. This seems to be something very important to you and luckily, since we are not a structured organization, everyone is free to do their own thing! It shouldn't be too hard, it's just copy and paste right? Data gets lost all the time sadly (although it's not quite clear how something from a google sheet can be permanently deleted + the average rank stat that you want is still there?) and that's why you have to backup up what's important (although again I'm not sure why we should be so mournful when according to aj what was deleted was unfinished and unorganized - a great loss and insult to the authors nonetheless). Hopefully we can all use this as a learning experience regarding how to handle data that is important to us so that actually serious projects that matter in life don't meet the same fate as our incomplete statistics.