Author Topic: Tokens - Deflation, distribution, and other issues  (Read 1256 times)

Arataka Reigen

  • The Greatest Psychic of the 21st Century
  • Core 'Shroom Staff
  • Star Spirit
  • ***
  • Gabumon
  • 100% Legit
    • View Profile
Re: Tokens - Deflation, distribution, and other issues
« Reply #20 on: March 17, 2017, 11:06:08 PM »
Very sorry about not making the meeting. For some reason I had it noted down for tomorrow, so I was asleep (it's 4 am here right now). I will try to explain why I believe worrying about price rates at this time is not productive.

The gist of it is that figuring out the economy and how much every prize will cost is the end goal, while curbing present and future inflation is an intermediary step TOWARDS that end goal. To determine the price rates, we need to know how many tokens will be in circulation this year. To figure out how many tokens will be in circulation, we need to know how many tokens will carry over from last year, as well as how many tournaments there will be that give out tokens.

Needless to say, we do not know either of these things right now. We're essentially trying to figure out an equasion that's composed of only variables. That's impossible. This is why I put forth that we focus on filling in one of these variables by making a decision about carry-over tokens and deflation. Once we have one constant in the equasion, we can begin figuring out the other variables and thus solve the entire thing.

Basically, a rudimentary timeline I am proposing is this. We would go through the timeline in order and move on to the next point once the previous one has been finalized and they become relevant:

1. Make a decision on how to handle token deflation
2. Gauge interest in hosting tournaments, map out a tentative list of likely tournaments from that, and try to determine how many tokens that will add to circulation
3. Figure out how many shroom tokens the Shroom will give out, and how many awards tokens we will allocate to that
4. Roughly determine the amount of prizes there will be (this may take a while)
5. Use information from 1,2, 3, and 4 to make an informed decision on how much every prize will cost

There are no definite dates because the exact dates depend on various external circumstances. The public would be officially informed about token deflation once we have finalized point 1, which ideally would be some time next week.

tl;dr version:
Finalizing token deflation is one step in figuring out how prizes will work. Deflation and price rates are NOT in opposition to each other, rather they are part of the same progression, with the latter actually being the end goal. As such, one should not be treated as an obstacle to the other. Let's go through the steps one by one, instead of taking all the steps all at once.

Anton

  • Froppy
  • #mwchat Admin
  • Star Spirit
  • *****
  • Hypnotoad
  • kero?
    • View Profile
Re: Tokens - Deflation, distribution, and other issues
« Reply #21 on: June 30, 2017, 12:04:33 AM »
As per a decision to tackle the problem of inflation plaguing our awards token economy, we've taken the first step in our restructuring of token worth, distribution, and prize rates.  The timeline we have for this is:

Quote from: Credit to Edo for spearheading this
1. Make a decision on how to handle token deflation
2. Gauge interest in hosting tournaments, map out a tentative list of likely tournaments from that, and try to determine how many tokens that will add to circulation
3. Figure out how many shroom tokens the Shroom will give out, and how many awards tokens we will allocate to that
4. Roughly determine the amount of prizes there will be (this may take a while)
5. Use information from 1,2, 3, and 4 to make an informed decision on how much every prize will cost

Step 1 was finalized in the meeting tonight, which is to deflate the token amounts, dividing by 10, and rounding to the nearest whole number.  We realize that at first this looks like we are confiscating tokens, but I assure you that the following steps will result in prize prices being decreased and fall within range for everyone to have a good chance of achieving without necessitating joining every single tournament and dominating.

If there are any questions, comments, or concerns please let us know!  We are doing this to make the whole process more fair to everyone, so we appreciate any input.  Feel free to check our public meeting logs and the Full List document for the rest of the decisions we have made so far.

2017 Token Sheet

So hey, we need to get on this soon.  Right now is Step 2 (and Step 3, really).

Spoiler: List of tournaments (click to show/hide)

afaik this is everything that is running and will run, potentially a pokemon doubles and ssb 3ds, but I'm going to doubt those at the moment.  How should we exactly go about determining how many tokens each tournament would potentially give out?  There isn't exactly a solid formula to this so just guessing won't work.  Base it off previous years / similar tournaments?  Just have the hosts do tokens as normal?

As for Step 4, I can see into the future and that being potentially few things being added this year to what already is left over, thanks to Steam being a pile of *bleep* and ruining this for us, leaving few options left such as praying Humble Bundle will have a summer sale and spending a ton on individual gift cards that are grabbed up very quickly.


Arataka Reigen

  • The Greatest Psychic of the 21st Century
  • Core 'Shroom Staff
  • Star Spirit
  • ***
  • Gabumon
  • 100% Legit
    • View Profile
Re: Tokens - Deflation, distribution, and other issues
« Reply #22 on: June 30, 2017, 12:36:37 PM »
afaik this is everything that is running and will run, potentially a pokemon doubles and ssb 3ds, but I'm going to doubt those at the moment.  How should we exactly go about determining how many tokens each tournament would potentially give out?  There isn't exactly a solid formula to this so just guessing won't work.  Base it off previous years / similar tournaments?  Just have the hosts do tokens as normal?

Wait, there isn't a set of guidelines that determine how many tokens are allocated to each tournament??

I... didn't really expect that. That's a rather huge oversight that should be addressed. I mean I get that hosts should be allowed to distribute the available tokens to their winners as they see fit, but the total amount of tokens available to them should really be controlled by a centralized entity. Otherwise there's not really anything stopping people from exploiting the system by hosting a tournament and giving themselves whatever amount of tokens they want.

I'll have a look at some data and get back to this.

Anton

  • Froppy
  • #mwchat Admin
  • Star Spirit
  • *****
  • Hypnotoad
  • kero?
    • View Profile
Re: Tokens - Deflation, distribution, and other issues
« Reply #23 on: June 30, 2017, 02:44:45 PM »
It's only ever really been based on what other people are doing, and kinda doing what everyone else is, but nothing formal has been set up because it wasn't exactly needed until icemario decided to exploit the system like how you said.  tbf, I don't think anyone really expected the token economy to become like...an economy.  This is really the first year of me having a part in Awards administration that there's even been time to address this as a topic on a substantial level because of how lax we did the poll discussion.  And tbh, I see it undergoing a substantial change anyways now that Steam is out of the picture.  If gift cards end up being all we can do, we might have to shift to giving prizes only to winners or Top 3 in tournaments because I don't have that kind of money and I doubt anyone else is going to step up to just dish out a few hundred dollars for prizes when in the past Steam allowed quickie $3 game grabs by the bundle.


Arataka Reigen

  • The Greatest Psychic of the 21st Century
  • Core 'Shroom Staff
  • Star Spirit
  • ***
  • Gabumon
  • 100% Legit
    • View Profile
Re: Tokens - Deflation, distribution, and other issues
« Reply #24 on: June 30, 2017, 02:50:07 PM »
So I thought this over for a while and here is the conclusion I came to. Off the top of my head, here's three ways tournament tokens could be handled:

1. Let the hosts set the token total themselves

The most straight-forward approach. Have tournament hosts walk up to the committee and be like "Hey, give me 600 tokens for my tournament". This is the way that requires the least amount of effort on our part, but also the most exploitable. There's really not much stopping you from hosting a tournament, thinking of how many tokens you would want to receive from it, and then manipulating token distribution to get that amount. Like, say I want 100 tokens. So I'm going to host a tournament that I can expect to place highly in, ask for 400 total, and distribute tokens in a way that'll ensure I get at least 100. But whoops, something unexpected happened and I placed lower than I anticipated. I now can't justify 100 for my position, because then there'd not be enough for the people who placed ahead of me. No problem though, I'll just ask for 600 tokens total instead. I won't have to justify the increase because nobody knows I was originally only going to ask for 400.

So I don't think that's a good idea.


2. Give every tournament a flat amount

This would be the easiest to calculate. We have 14 tournaments. Every tournament gets 400 tokens total. Boom, 5600 tokens we need to account for. However, this is also the most terrible way to go about this. Tournaments vary in scope. Some tournaments are just smaller or bigger than others, and it wouldn't be entirely fair to the really large ones to get the same token amount as everyone else. Which leads directly into the third point.


3. Make the total dependent on the number of participants

This way involves having a set base amount of tokens for tournaments, and then multiplying it by the number of people participating. With this approach, the token total increases proportionally to the size of the tournament. In essence, say the base token value is 30. 10 people sign up for a tournament, thus it gets 300 total. 15 sign up, it gets 450. Et cetera. This would ensure that the tournaments that attract the most interest are also the most rewarding. In theory, it should also incentivise people to sign up more, as the higher the amount of participants, the larger the payout if you win. This makes sense to me, since the more competitors you have to deal with, the bigger your prize should be.

I've ran this approach against some numbers I collected from last year (I apologize if there are any errors), and here's how the approach would play out (assuming a base value of 30 tokens per participant).

Art Contest
Participants: 21
Token total '16: 704
New total: 630

Pokémon
Participants: 16
Token total '16: 571
New total: 480

Mario Kart 7
Participants: 14
Token total '16: 324
New total: 420

Mario Kart 8
Participants: 16
Token total '16: 324
New total: 480

SSB4
Participants: 14
Token total '16: 500
New total: 420

Yu-Gi-Oh
Participants: 8
Token total '16: 265
New total: 240

I think a base value of 30 works out best because it's close to last year's figures, with a bit of subtraction to account for the deflation that will be going on. The final amount can be set later; for now this is just a proof of concept. Minecraft games is also not on this list, because that's something I'd have to get into later if we decide to go with this.



So in the end, which approach should we go for?

Personally, I think the most effective way to go about this would be a combination of 1. and 3.. Something along the lines of us making the calculations like in 3., and then giving the result to the hosts as a guideline. Like, send them a message like "We've calculated that the optimal total for your tournament is 480 tokens. Please plan your payout accordingly." This would give hosts the chance to reply to the effect of "With the way I am distributing tokens, it would be more convenient for me if I had 495.", to which we could answer with "That's reasonable, here's 495 instead.". It also reduces exploitability, since if a host decides to say "480 isn't enough, give me 700 instead", we could just be like "nope". It would ALSO give new hosts who have no idea what amount of tokens is appropriate for their tournament a suggestion to start with.

This approach would give us tangible numbers we can make plans with, while also being flexible enough to allow hosts to make reasonable adjustments as they see fit.

Anton

  • Froppy
  • #mwchat Admin
  • Star Spirit
  • *****
  • Hypnotoad
  • kero?
    • View Profile
Re: Tokens - Deflation, distribution, and other issues
« Reply #25 on: June 30, 2017, 03:06:29 PM »
if in your "combination of 1 and 3" it accounts for a tournament requiring heavier participation than a tournament with an equal number of participants, then I support it, because I feel that's an aspect that needs to be covered otherwise we're gonna be setting up a system where the most incentivized tournament will be a no-effort free-for-all

also, do presentation and shroom tokens fit into this, or are those two separate wild beasts?


Arataka Reigen

  • The Greatest Psychic of the 21st Century
  • Core 'Shroom Staff
  • Star Spirit
  • ***
  • Gabumon
  • 100% Legit
    • View Profile
Re: Tokens - Deflation, distribution, and other issues
« Reply #26 on: June 30, 2017, 03:16:13 PM »
We could modify the token base value depending on how much participation is required. Like, if you're hosting a standard tournament, you get 30 per person, but if you host one where the goal is to press a button once, you get 5.

I don't really know what constitutes a low-participation tournament though. Have we had something like this before? Do you have examples?

Shroom and presentation tokens aren't tournaments. There's no competition going on, they're more like bonuses. As such they would be separate entities.

Lord Bowser

  • Super High School Level Bowser
  • Core 'Shroom Staff
  • Power Star
  • ***
  • "It's funny because we're all living in a simulation and free will is a lie."
    • View Profile
Re: Tokens - Deflation, distribution, and other issues
« Reply #27 on: June 30, 2017, 03:21:36 PM »
participation is a really subjective thing imo, perhaps it could be something decided by the committee as a whole?

like we can look over all the tourneys and the committee can collectively argue "x tournament had a lot of participation but y tournament not so much", along with asking the hosts themselves for their judgement

stuff like awards mafia, art contest, etc. i would view as high participation, while smaller things with less players and less competition (perhaps something like the mk7 tournament? that's all i can really think of as low for this year) would be low participation



(credit to TPG)

(credit to Mr. Edo)

Arataka Reigen

  • The Greatest Psychic of the 21st Century
  • Core 'Shroom Staff
  • Star Spirit
  • ***
  • Gabumon
  • 100% Legit
    • View Profile
Re: Tokens - Deflation, distribution, and other issues
« Reply #28 on: June 30, 2017, 04:06:51 PM »
"Participation" in this context only means "effort that needs to be exerted to perform your function as a participant". Number of sign-ups does not factor into this, because that is already covered by the "x amount of tokens per head" clause. So I guess let's call this "low-effort" instead, to make it less confusing.

stuff like awards mafia, art contest, etc. i would view as high participation, while smaller things with less players and less competition (perhaps something like the mk7 tournament? that's all i can really think of as low for this year) would be low participation

What makes MK7 low-effort compared to everything else? It's still an active competition that I think is on-par with the other tournaments. The only difference I can think of is that Mario Kart generally has a higher number of players per "match", which means fewer matches total, and thus it doesn't take as long as other tournaments. But I'm not sure if that really means it's low-effort.

Like, if we set up different effort categories, one thing we should definitely be wary of is the mindset of "well, we have this low-effort category, so we have to put SOMETHING into it". That's not the case at all. If there's nothing that fits, nothing should be put in.

participation is a really subjective thing imo, perhaps it could be something decided by the committee as a whole?

like we can look over all the tourneys and the committee can collectively argue "x tournament had a lot of participation but y tournament not so much", along with asking the hosts themselves for their judgement

In my experience, while this would be an ideal setup in theory, in practice this sort of thing is very hard to coordinate and often just fails due to inactivity. Getting everyone to sit down and thoroughly, actively, and individually evaluate every tournament would be a huge hassle, and more likely than not just result in the head staff running around begging people for input, as it usually does. Throw having to wait for host input into the mix on top of that, and you've got yourself a situation.

In the long run, I feel like it would be more efficient to just have a standard setup that can be easily referenced, and then discuss the exceptions (if they exist) rather than each and every entry.

Lord Bowser

  • Super High School Level Bowser
  • Core 'Shroom Staff
  • Power Star
  • ***
  • "It's funny because we're all living in a simulation and free will is a lie."
    • View Profile
Re: Tokens - Deflation, distribution, and other issues
« Reply #29 on: June 30, 2017, 04:47:04 PM »
idk, mk7 was the one thing that came to mind for some reason, but yeah after further inspection it's pretty on-par

and yeah you make a good point re: effort categories, i didn't think about it that way. it would be best to avoid labeling things as low-effort for the sake of having a low-effort category

unfortunately i can't really think of any other possible solutions to this issue atm so i'm not really able to contribute further :/



(credit to TPG)

(credit to Mr. Edo)

Anton

  • Froppy
  • #mwchat Admin
  • Star Spirit
  • *****
  • Hypnotoad
  • kero?
    • View Profile
Re: Tokens - Deflation, distribution, and other issues
« Reply #30 on: July 10, 2017, 09:38:24 PM »
I'm kinda just waiting on humblebundle to do something before I finalize this, which hopefully happens sooner than later .. .. .. ....... . .. . . . ..... .. ..


Anton

  • Froppy
  • #mwchat Admin
  • Star Spirit
  • *****
  • Hypnotoad
  • kero?
    • View Profile
Re: Tokens - Deflation, distribution, and other issues
« Reply #31 on: July 16, 2017, 04:48:11 AM »
Here's some logs of me blabbering on with a tl;dr being:

*Edo's plan sounds gr8
*30 tokens per person in each tournament granted to the host to divvy up as they see fit
*Hosts can request wiggle room if they need more or less
*If things seem unfair, we can still stick our nose in
*Prize prices will remain the same, with an average goal being $0.12 USD/token, give or take, using that as a base combined with babby's first supply & demand chart

This is technically open to discussion but I'm pushing pretty hard for this to be a quick and easy decision to begin applying asap after poll tallying is done

Quote
20:31   Anton{Politoed}   what I take by your numbers is that
20:31   Anton{Politoed}   we're not wildly going to change rates or anything, but
20:31   Anton{Politoed}   just create a system that makes more sense
20:31   Anton{Politoed}   and is consistent
20:32   Anton{Politoed}   therefore, prize totals are remaining the same
20:32   Anton{Politoed}   which means I just gotta spin my dick around and collect every tournament and how many people participated/will receive tokens
20:32   Gabumon   ok
20:32   Anton{Politoed}   because hosting also awards tokens
20:32   Anton{Politoed}   and that didn't really have a system either beyond how much I felt they deserved for the work they put in
20:33   Anton{Politoed}   this shouldn't be too hard, just
20:33   Anton{Politoed}   gotta do it
20:33   Anton{Politoed}   and also deciding if your example of 30/person is what we're using
20:34   Anton{Politoed}    30 did create totals that were representative of what we tried to aim at for fairness in previous years
20:34   Anton{Politoed}   so I don't think we should spend much time debating whether 30/35/40/etc. will be better
20:34   Anton{Politoed}    30 works
20:35   Anton{Politoed}   "omething along the lines of us making the calculations like in 3., and then giving the result to the hosts as a guideline. Like, send them a message like "We've calculated that the optimal total for your tournament is 480 tokens. Please plan your payout accordingly.""
20:35   Anton{Politoed}   when you say that
20:35   Anton{Politoed}   does that imply AC staff will create a projected final tally, or
20:35   Anton{Politoed}   just the sum
20:36   Anton{Politoed}   'final tally' as in like
20:36   Anton{Politoed}   "here's what we feel 1st place should receive, 2nd place, 3rd, etc."
20:36   Anton{Politoed}   that sounds like too much micromanaging so if you did mean that i'm gonna go ehhhhh
20:36   Anton{Politoed}   but I don't think you did
20:37   Gabumon   no, I didnt mean that
20:37   Gabumon   we just give people a total with some wiggle room
20:37   Gabumon   how they divide it up is up to them
20:38   Anton{Politoed}   with the unspoken note that if their divisions are suspicious, that we retain the authority to change it
20:38   Anton{Politoed}   not gonna piss myself over someone in 1st place receiving 70 and going 'waaaah they should get 75'
20:39   Anton{Politoed}   but if it's like 1st gets 150 tokens and 2nd gets 35 then
20:39   Anton{Politoed}   _
20:39   Anton{Politoed}   I'll probably post these logs as a confirmation of what we're doing
20:40   Anton{Politoed}   in that thread
20:40   Anton{Politoed}   instead of just doing it immediately
20:41   Anton{Politoed}   re: low- vs. high-effort/participation tournaments
20:41   Anton{Politoed}   I think the difference between most, if not all, of them are within a standard deviation
20:42   Anton{Politoed}   so we could just take that up as per exception as they become obvious
20:42   Anton{Politoed}   instead of codifying numbers into this right now


Rintarou Okabe

  • his smile and optimism = gone
  • Donkey Kong
  • *******
  • freakworld
  • El Psy Kongroo
    • View Profile
Re: Tokens - Deflation, distribution, and other issues
« Reply #32 on: July 16, 2017, 05:13:19 AM »
I've stated before that I like this plan