Forum Settings
Forums
New
Reply Disabled for Non-Club Members
Pages (5) « 1 2 [3] 4 5 »
Jan 2, 2018 9:09 AM
noiz cancelling

Offline
Jan 2008
1012
Kineta said:
@Numi I understand that TriZen's comment on "eagerness" was misguided, and would have hit a sore point with me if I were currently a FAL staff member (or recently retired). That being said, I do think he has two valid points.

I really disappointed with the fact that there isn't any notification or any form of communication from the admin/staffs regarding the cancellation of FAL Fall 2017.
Judging from the club comments, I think many were disappointed about this. Players were waiting until the very last minute for a thread to appear. The two comments made by staff in September (and since) were asking for people to apply (last minute). That the season may not happen was never mentioned. I think asking for a notification saying, "Sorry guys, we couldn't make Fall 2017, but we hope to be back strong for Spring 2018! <application information>" is a small request.

If there weren't enough staffs that can handle FAL, also please let the users know at least 1 month before.
You said, "FAL has been seeking this replacement since March." Luna's post is from 2015, and she is not currently a FAL staff member. A new thread was not created. The existing thread was not shared with the club (via MAL's notifications system). There is no request for members to apply in the Final Results/Week 14 thread. Nothing was posted in the club comments asking for new staff members between Jul 10 to Sep 23.

I'm the last person on MAL to ever negatively criticise or blame about things being late, and I know all the struggles that come with manning a group of volunteers when you have a real life and a bunch of other projects on the go. I'm casting zero blame here. When I realised the FAL Fall season wasn't going to happen, I thought, "Hopefully this gives them enough time to get a solid team together for Spring 2018". So zero anger, blame, or negativity from me. I just think TriZen has two valid points, and I hope that the staff members who read his post are aware of it :)


I fully agree that both of those points are valid. It is because I did not see myself as having any place to comment on the remaining staff's choice in actions since I stepped down in July that I wasn't comfortable leaving remarks on that in particular. I just wanted to make it clear that hey, we'd been seeking a replacement since March. The intention was to train the new person/people in during Spring 2017's season so they would know what they were doing well by the time I retired. It didn't go that way, but it wasn't something that could make me push back stepping down by a season. And since then I do not have the knowledge of the remaining staff's actions, communications, etc regarding the topic beyond what's visible in the club public spaces. Not enough insight on why they did or didn't take certain actions, no longer someone who should speak for them regarding seasons that would start after I stepped down, therefore it's not something I'm going to comment on further than I already have.

All I'm going to ask is that in the future people not make assumptions on the staff's intentions and keep it to concrete things such as actual actions/ideas on how to improve. As I said, I do agree that those two points are valid.

In any case, I'll be removing myself from the conversation after I post this comment and just keep an eye out for what responses the staff I left behind choose to make.
Jan 2, 2018 12:47 PM
Lead Admin
Faerie Queen

Offline
Aug 2007
6303
Numi said:
I fully agree that both of those points are valid. It is because I did not see myself as having any place to comment on the remaining staff's choice in actions since I stepped down in July that I wasn't comfortable leaving remarks on that in particular.
Understood, thanks for clearing that up :)

I was a bit concerned that after your post, the rest of TriZen's points would be disregarded. So I just wanted to draw attention to them. Hopefully it didn't bother you (or anyone else) any.
Sep 12, 2020 5:05 AM
Nostalgia Addict

Offline
Feb 2018
243
is this still active?
Sep 13, 2020 2:37 AM

Offline
Oct 2018
461
WIIXAnd said:
is this still active?
Yes.
Sep 13, 2020 7:27 AM
Nostalgia Addict

Offline
Feb 2018
243
So i think we should raise nuke value
its useless and nobody use it
Sep 26, 2020 3:47 AM

Offline
Feb 2007
2910
Can we get an update on the banners for the Top 3 from the last four FAL seasons? I know it's probably a low priority but it has been a while since the 2019 seasons and I would like to have those banners so I can display them on my profile. They don't have to be done yet, I'm happy with just an ETA. (I wasn't around for the Fall 2018 season but I'm not sure if the Top 3 got their banners either).
Mar 29, 2021 4:12 AM

Offline
Jul 2012
1009
Don't you think it's time to raise the points for aces and wildcards? I feel like gaining 3k or sacrificing 1.5k for nukes (if you're nuking someone not on the same rank) isn't much.
MAL has gained a lot of users in the past couple of years so I feel like the game is pretty much set by just selecting a good team and that's it.

I've also got a suggestion for an extra wildcard, and that is having an extra anime in bench. But the catch is you have to use up points to swap it in and you have to choose it before the season starts. The anime you swap it with can't be brought back either.
Just like the other wildcards, you can't use it until the 9th week. It's possible that you won't ever need to use it, since you could already have a good team or the reserved anime is not doing well, but it might add extra competition later on. It will also act as an extra trump card that can let you gain a few ranks.

During team registration it could look like this:
active anime 1
active anime 2
active anime 3
active anime 4
active anime 5

benched anime 1
benched anime 2

reserved anime
SamiiMar 31, 2021 8:26 AM
Apr 16, 2021 11:41 PM

Offline
Aug 2018
24
I don't know if it just me but I have been having a number of issues trying to see my team and my benched anime because for some reason the fmal website hasn't been working on my computer over the course of the last two weeks. Is there something wrong?
Sep 29, 2021 2:51 AM
Offline
Apr 2015
2
Guys, little help pls, how do I pick my team? When I click on FAL it only shows me a countdown
Sep 29, 2021 9:11 AM
Nostalgia Addict

Offline
Feb 2018
243
urosh said:
Guys, little help pls, how do I pick my team? When I click on FAL it only shows me a countdown


Registration has closed, you are late :(
Sep 30, 2021 3:21 AM
Offline
Apr 2015
2
WIIXAnd said:
urosh said:
Guys, little help pls, how do I pick my team? When I click on FAL it only shows me a countdown


Registration has closed, you are late :(


Yeah, I figured, thanks.. Ill be ready for next season xD
Oct 2, 2021 12:57 PM

Offline
Apr 2017
144
Small request on the (apparently?) new design: Can we get a direct link for the different entries? It's a bit annoying to search the titles every time I want to check up on their current stats. A direct link would really be convenient for this.
Oct 3, 2021 1:14 AM

Offline
Apr 2020
43
RobbiRobb said:
Small request on the (apparently?) new design: Can we get a direct link for the different entries? It's a bit annoying to search the titles every time I want to check up on their current stats. A direct link would really be convenient for this.
Yeah I was thinking the same, It's really tedious to manually search for titles.




anime list favourites | slipfunc - stay the night
‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎ ‎‎ ‎ ‎krytm - away

Oct 3, 2021 5:51 PM

Offline
Dec 2019
2861
Question about the ace.
If I aced something, will it have to beat my previous aces?
I aced Blue Period last week, and I'm not confident that my other anime can beat Blue Period this week.
Oct 3, 2021 6:28 PM
Offline
Nov 2010
1162
fluffycow17 said:
Question about the ace.
If I aced something, will it have to beat my previous aces?
I aced Blue Period last week, and I'm not confident that my other anime can beat Blue Period this week.


Yes, unless that previous ace has more than 85,000 watching/completed.
Oct 6, 2021 6:03 AM

Offline
Jul 2019
348
eplipswich said:
fluffycow17 said:
Question about the ace.
If I aced something, will it have to beat my previous aces?
I aced Blue Period last week, and I'm not confident that my other anime can beat Blue Period this week.


Yes, unless that previous ace has more than 85,000 watching/completed.


so, it means total pts of week1+week2 count for ace?

Oct 6, 2021 7:25 AM
Offline
Nov 2010
1162
_48454449_ said:
eplipswich said:


Yes, unless that previous ace has more than 85,000 watching/completed.


so, it means total pts of week1+week2 count for ace?


No, no, only for the week. Anyway, you can see your anime's score for the week on https://fal.myanimelist.net/
Oct 6, 2021 7:33 AM

Offline
Jul 2019
348
eplipswich said:
_48454449_ said:


so, it means total pts of week1+week2 count for ace?



No, no, only for the week. Anyway, you can see your anime's score for the week on https://fal.myanimelist.net/


so we have to calculate the points of this week total pts- previous weeks pts

Oct 6, 2021 7:38 AM
Offline
Nov 2010
1162
_48454449_ said:
eplipswich said:



No, no, only for the week. Anyway, you can see your anime's score for the week on https://fal.myanimelist.net/


so we have to calculate the points of this week total pts- previous weeks pts


??? You don't have to calculate. The points of the week you see on https://fal.myanimelist.net/ are already calculated for you... Just look at your anime in your team. The points below each anime are the points for the week.

Just base your ace on those points.
eplipswichOct 6, 2021 7:41 AM
Oct 6, 2021 7:44 AM

Offline
Jul 2019
348
eplipswich said:
_48454449_ said:


so we have to calculate the points of this week total pts- previous weeks pts


??? You don't have to calculate. The points of the week you see on https://fal.myanimelist.net/ are already calculated for you... Just look at your anime in your team. The points below each anime are the points for the week.

Just base your ace on those points.

thank you, I thought it's pts of both weeks

Oct 7, 2021 11:34 PM

Offline
Jan 2021
23
The score that is showing for that week? Is that too easy? I was able to figure out which one would be Ace. If this is the case, why is it -5000 when the answer is wrong?

Oct 8, 2021 12:37 AM

Offline
Jul 2019
348
i can't do 2 swaps this week please fix that, there's no rule that i can't do this
A player may swap a maximum of four (4) times during the 14 weeks of the FAL season
_xxXHediolic_Oct 8, 2021 12:42 AM

Oct 10, 2021 2:26 PM
Offline
Jan 2018
7
Reshirom said:
The score that is showing for that week? Is that too easy? I was able to figure out which one would be Ace. If this is the case, why is it -5000 when the answer is wrong?

The score is changing each hour, so the anime with the highest score in your team can change during the week. In some cases it may be problematic, but I also think that usually it is very easy to choose ace correctly
Oct 10, 2021 11:35 PM

Offline
Sep 2011
9879
_48454449_ said:
i can't do 2 swaps this week please fix that, there's no rule that i can't do this
A player may swap a maximum of four (4) times during the 14 weeks of the FAL season
If you read the wildcare section, it says "Extra swap: At a fixed cost of points, immediately swap one active anime to bench and one bench anime to active. "
Oct 14, 2021 5:46 AM
Offline
Sep 2017
11
Just a suggestion, but I think it would be cool if you could add another table for friend ranking in the ranking section so that we could see where we place compared to our friends. Maybe something you can think about for the next anime league season.
Oct 24, 2021 6:07 AM
Offline
Jul 2018
564084
Well I was going to wait until the season was over before making a post about all the things I hate about this new setup by aww man do they keep getting worse and worse each week. So I guess i'll go ahead and drop it all here.

First of all, not having a place somewhere where you can see how each anime got the points it did each week sucks. The point breakdowns are helpful to new players so they can understand the rules and correct any misinterpretations they might have about how scoring works.
Also not being able to see how many discussion points are counted every 2 weeks is extremely annoying to me. If i'm trying to plan ahead for future weeks I like to see how the discussion points have been trending and it's harder to make a guess about what might an anime might score in 2 weeks if I don't have the discussion point numbers. For bench swapping this can be seriously important.
Could I just go in each discussion thread and find the number of posts that were posted before the cut off time? Sure but that's going to get increasingly annoying as there becomes more threads to look through. Plus I don't want to see or read spoilers for shows I want to watch at some point.

The weekly results threads are also sorely missed. Not just because of the points breakdown I mentioned above but it also kills a lot of the community aspect of the game as well. It was interesting to read what people's reactions were to the shifting scores and leaderboards each week. That's just completely gone now. Maybe there is still some of it happening on Discord by not everyone (including me) uses Discord. Killing the MAL-side weekly community discussions doesn't make sense if the goal of this season was get new players....

...and oh god were there a lot of new players. So many that even now, 3 weeks later, you can't scroll up and down the Team Ranking page without the little scroll box breaking. I'm lucky i'm in 24th place so it's not that hard to scroll up to see what the top score is. But if I wanted to find another player who was much further down, it would be pretty much impossible. There should just a "Find User" search option where you just type in a name and see what their score and position are.
Feb 24, 2022 7:42 PM
Offline
Apr 2020
28
For Spring 2022 Season could the badge you get change based off of how well you did at the end? Like if you get #1 you get winners badge, top 100 or top 1000 a different badge, or maybe like top 1% top 10% finisher top 50% or something like you get at least a different badge if you do well.
Mar 12, 2022 10:59 PM

Offline
Feb 2007
2910
Time for the part of the FAL discussion cycle again where the relevance/fairness of nukes is brought up again. Maybe it is time to finally remove them?

They can be useful in breaking ties (as shown last season <_<) and overtaking players close to you, but ties at the end can also be decided by the points you get from your bench anime, which goes back to the core part of FAL; picking the best team, and that should be more important.

With the team size increasing to 8 anime last season, the chance of multiple top players having the same bench anime has become smaller, so nukes might be irrelevant now for breaking ties. There is the scenario however where someone tied with you has a stronger bench, so nuking is the only option to get ahead of them. Then it comes back to the topic of whether that's strategically fair or not.

---

Whether nukes are removed, I did come up with a Wildcard idea, since wildcards give players something to do at the end of the season. So how about a chance to Ace an anime you missed out on earlier in the season? Usually due to a popular anime passing the Ace threshold but not being the top scorer for the weeks it was eligible.

Extra Ace: At a fixed cost of points, you are able to Ace an anime that has passed the Watching + Completed Members threshold. It has to be an anime you have previously not declared as an Ace.

That last part might be a sticking point. I'm sure players that failed an ace would like to get it right later in the season, but it would cost them even more points to do it. But at least this way it reduces complexity and makes this Wildcard option easier to understand/implement. This Wildcard is an alternative to players that don't need to make an extra swap, or it can be useful to players who missed out on their chance to Ace something earlier.

I'm sure there's something I'm not taking into account that makes this idea dumb, but guess we'll see. :p
Apr 2, 2022 5:53 PM
Offline
May 2021
44
I don't know if this is a thing yet, so I'll put it there just in case: somewhere that declares exactly what week it is so I don't have to keep track would be nice. Like, Weeks 7 and 13 have a score point booster and I want to make sure I have certain shows active, so I'd be convenient to make sure I don't swap a week early/late and not get the points I was planning around.
YO!
Apr 17, 2022 6:50 AM

Offline
Sep 2016
108
@kineta
Does it take a while for the FAL Badge to show up in our profiles?
I've seen it in some profiles, but some of us are still missing the badge.
Apr 17, 2022 7:47 AM
Lead Admin
Faerie Queen

Offline
Aug 2007
6303
Zacko said:
@kineta
Does it take a while for the FAL Badge to show up in our profiles?
I've seen it in some profiles, but some of us are still missing the badge.
As far as I know, they haven't been given out yet.
Apr 17, 2022 2:23 PM

Offline
Sep 2016
108
Kineta said:
Zacko said:
@kineta
Does it take a while for the FAL Badge to show up in our profiles?
I've seen it in some profiles, but some of us are still missing the badge.
As far as I know, they haven't been given out yet.


Ooooh, I've been looking at the 2021 badge! Thanks :)

Follow up question:
If I've aced an anime this week and it just went over 85k viewers, do I need to cancel my ace and pick one with less than 85k or can I keep it since I aced it before it crossed the limit? EDIT: Got a response from Discord telling me to cancel the Ace when it goes above 85k so I'll do that.
ZackoApr 17, 2022 2:33 PM
Apr 22, 2022 4:07 PM
Go read Medalist
Offline
Apr 2007
285
Some suggestions to maybe help balance the scoring system a bit. I know it's already been greatly overhauled compared to 2021, but I think there's still work to do there.

1. Tone down the score for watching+completed number to 0.4 for odd and 0.2 for even weeks respectively. Right now the scoring is still massively skewed towards watching+completed numbers which are a metric of quantity—as opposed to rating or favorites which are metrics of quality. Even the rules/FAQ page estimates the score gain from that alone to be at around 50% of the total for the entire season which is still quite accurate despite the changes from last year when viewership was even more prominent. This makes picking a "good" team a lot easier even if you know next to nothing about the shows in question just by looking at the numbers of people having them on their lists and flailing your way into top-200 like I did last year in my first FAL. Additionally, the episode discussion score is more or less also a derivative of active viewership number because bigger audience always generates more discussion even if there is less substance to discuss per se. So realistically, closer to 2/3 of each particular show's points total depend on that one number alone. Crazy!

I understand that part of it is because MAL seeks to encourage people to engage in the fandom and such in general, and viewership is the main metric thereof, but I think for the purposes of the contest it would be more interesting to try to also guess which shows will be most well-received and not simply most watched. For instance, Kingdom S4 outright destroys the entire board rating-wise, but by the end of the season it won't even crack the top 10 under the current scoring system due to its puny viewership. Ya Boy Kongming has shot up to overtake even Komi S2 (imagine that!) in rating and is the clear sleeper hit of the season, but as it stands, even with a massive rating advantage it has no chance catching up to Aharen-san, the ResidentSleeper hit of the season, on non-rating weeks. If the scoring were more even, people who picked their teams based on numbers alone and without researching into the potential hits would be entirely blindsided by this. I think it works towards FAL's goals to encourage them to find out more information on the upcoming shows before they finalize their teams and discover hidden gems ahead of time, which is always cool.

2. Discount points for pre-favoriting and pre-dropping. In other words, only count series as favorited or dropped if they have been favorited or dropped AFTER they had started airing, and only begin counting these actions after that time. It's both in common sense and fair that you can only make a decision to declare a series favorite or drop it if you had a chance to watch any of it at all. I don't think it would be particularly difficult technically, either, since it's a simple deduction of an already-known value at an already-known time. You only need to do it once for every show on the list and then apply this number as a correction to the formula.

Why is this important? Well, consider for a moment that some of the most anticipated shows of the season easily gain over a thousand favorites before the first episode. You can check it right now with Kakkou no Iinazuke which will only start airing in two days. Another good example is Shikimori-san and Date a Live S4 which both similarly started with over a thousand favorites but only gain a small fraction of that amount per week.

Now I don't think there is necessarily foul play involved (although there easily can be), and it's a relatively common thing that people "reserve" a favorite spot for something they've already read and loved the source for (Chainsaw Man is already at OVER NINE THOUSAND!!!11111, for instance, and it's still months away), but it still has a potential to significantly skew the balance of the contest itself as the popular franchises and shows with popular source materials gain the advantage all too easily without having to "work" for it during the season itself, so to speak.

3. Consider also counting favorites and drops as a ratio to watching+complete value rather than absolute numbers. Right now both of these metrics mainly scale with raw viewership yet again, making shows with higher base values have respectively better or worse showing than they represent statistically. I think for a more adequate representation you could award/penalize shows that aren't just favorited/dropped more, but rather favorited/dropped more often, similar to how you award and penalize shows for meeting the 6.0 rating threshold on respective weeks.

For instance, at the end of week 2, Shikimori-san led with a massive 3054 drops, which sounds horrible on paper, but by that point the ratio of drops to active watchers was ~1:39. But look at Kunoichi Tsubaki or Mahoutsukai Reimeiki's performance in the same period, and you'll see ratios of ~1:17 and ~1:20 respectively. They had fewer absolute drops but were dropped almost twice as often which is a more telling metric. (No, their situation hasn't improved much since then.)

It's similar with favorites, but is harder to assess properly because of the aforementioned insane pre-favoriting bias, and unfortunately I've failed to note down pre-airing values to have a clearer picture. But just week to week it's easy to tell that some shows are managing to keep a relatively steady ratio (e.g. Tomodachi Game, Love Live S2) with one particular title climbing from ~1:211 week 1 to ~1:121 week 2 to 1:116 as of writing this line (yep, Ya Boy Kongming again), showing a rare example of initial values not polluted by pre-favoriting as well as a breakaway performance in terms of quality.

4. Make Aces a less-trivial, higher-stakes point source. Right now the hardest part about Aces is understanding how they work (as illustrated by the fact that there are multiple explanations available everywhere around, and people still ask the questions). Once you understand, it becomes all too trivial to just set up a reminder and declare your Ace at the eleventh hour when there's zero guesswork or risk involved. Barring very rare circumstances when two shows go neck-and-neck for points and it really is a toss as to which one of them gets it (which isn't a big deal, either, since you can typically just skip this week and try again on the next), it's never hard to tell which one you should Ace at any time because you can do it right before the final score calculation for the week. The numbers you see are almost identical to those that will end up on the scoreboard in an hour and a half, so the extent of player's effort there is simply not forgetting to click the button.

A clear solution to this would be forcing the player to declare an Ace at least 48 hours before the week's end. In return, you can make each successful guess worth 30k points instead of 20k since failure is an actual possibility that way even for people who know what they're doing. This will encourage planning ahead and make it a lot more engaging to keep track of point gain for your team and make more tactical bench swaps for juicing Aces mid-season.

5. Give more recognition to non-winners who end up near the top. Right now somebody who finished at, say, #15 receives the same badge as somebody who finished at #1500 or #15000. I think it would encourage more people to try harder if, say, finishing in the top 200 or so would receive some kind of a different badge compared to a regular participant. I think something similar has already been suggested before.

This is to recognize those who care about their participation and/or achieve meaningful results. Would be even better if one could choose to display their badges in place of the graphic signature or somewhere else visible on the forums, kind of like the Easter Egg Hunt badges you have right now next to your avatars.

6. Rework the nuke wildcards. The point difference between the top contenders is often so tiny and the scoring fluctuations between weeks so large that the penalty cost of 5000 points will move you back a few places all by itself and achieve a result opposite to the intended, not to mention it takes incredible effort to precisely calculate the exact point gain for everyone who could potentially overtake you on the last week: in the previous FAL, the #10 spot went to someone who jumped up 131 places! Who could see something like that coming? It's silly that the scoring system puts people so close together that you could do something so nuts, and equally silly to expect a mileage out of this feature that would be on par with the boost or extra swap.

7. Speaking of jumping up a hundred places: the current scoring system is too rigid in general. The most apparent way to tell is looking at how many ties remain near the top even after the contest is over and how many players are tied at the same value. That way an increase in score of just a couple thousand points can move you up a dozen places at a time when you're already in the millions, or even hundreds of places if you're further from the top. This is a symptom of the lack of agency on the players' part, so even if you want to put in more effort, there's very little you could do once the registration is over and teams are finalized. I think this point of criticism has already been brought up earlier by BlueYoshi and others, and I think it's also possible to mitigate.

Perhaps the easiest solution would be to award or subtract extra points depending on how late in the week existing voluntary actions are made, e.g. acing or swapping. For Aces in particular—and this ties into my suggestion above—perhaps you could do away with the point penalty for the wrong guess entirely (having wasted an Ace is a penalty on its own since you can't do it for the same show again when it's actually eligible) but scale the reward down instead depending on the day the Ace was declared: 30k if declared on Monday, 25k Tuesday, 20k Wednesday, 15k Thursday, 10k Friday. You could similarly scale the booster wildcard depending on the week it's used. With swaps, you could incur a penalty scaling from –1k on Wednesday to –5k on Sunday.

This way you introduce both higher variance for players with similar teams as well as risk-reward avenues that get them to consider their options more carefully and go the extra mile to ensure they aren't mistaken. They would obviously favor players that collect information and plan ahead, and hence encourage to engage stronger with the anime choices and community both prior and during the contest weeks, check out stats for the shows, etc. This way it actually feels more like a contest where you participate yourself and less like a horse race gamble where you mainly bet on audience behavior.
moozoohMay 23, 2022 11:45 PM
Apr 23, 2022 12:49 PM
Go read Medalist
Offline
Apr 2007
285
Oh, and two more things.

First, I've noticed from reading the prep thread that people think increasing the threshold will lead to a higher total number of Aces per week. This is actually backwards, and comes from a fundamental misunderstanding of what determines the potential number of Aces on an optimized team.

The way acing works is that your 2–4 top performing series during the first few weeks are also likely to be your strongest performers for the rest of the season, essentially preventing any other show from being aced for as long as they remain on the active team. There are only two ways to circumvent that restriction:
1) swapping away your top performers you've already aced before to make room for another Ace, or strategically waiting before bringing them onto an active team in order to juice the rest of the shows for Ace points (not a good tradeoff in most cases, last week being the notable exception);
2) waiting for them to—you've guessed it—hit the active viewership threshold so that they stop gatekeeping underperformers.

Lowering the threshold makes rotating Aces easier, not the other way around! Up to seven shows is possible to ace this season at 85k, and it would be even easier in the future if the threshold were to be lowered, but you will find it extremely hard to ever ace more than five if it were raised instead.

To illustrate: this season, assuming a highly optimized team, everyone aced Aharen-san first week. Tomodachi Game became available for acing immediately on the second week since Aharen became ineligible due to crossing 85k active viewers while Komi-san S2 and especially Shikimori-san never had a chance. Third week those who had Summertime Render on their bench were able to swap it in for another easy 20k points. That is three Aces in three weeks if you've played your cards right. The next opportunity will come as early as week 5 and then week 7, although I haven't seen anyone on the scoreboard with the particular team composition that would enable taking advantage of all five (but it was possible!).

Now consider if the threshold were, say, 100k: those who picked Shikimori-san for their choose-one would've aced nothing on the second and third weeks because Shikimori san-itself still shot up past the threshold by the end of week 2 while both Tomodachi Game and Summertime Render would've ended up gatekept by Aharen-san still being the better performer during that time. Those who picked Komi-san S2 would've been able to ace it on week 3 but would've still ended up with only two 40k Ace points after the first three weeks. Fourth week there's a good chance for Tomodachi Game to pull ahead of Aharen-san for a short while, so Komi-san pickers will be only one week behind on having three Aces while Shikimori-san pickers will have to wait even longer to get their third. If only two non-mutually-exclusive shows go over the 100k active viewers threshold by the end of the season—which is extremely likely—nobody will have more than five Aces. So how does this help anyone if they want more? The answer is it doesn't.

Personally, I think it's a very cool mechanic that introduces a good deal of player agency to the contest, so lowering the threshold to 80k for the next game feels like a much better idea to me.

Second, the ideal point distribution should not take Aces—or wildcards, for that matter—into account because these are deliberate player actions and aren't part of the show's own performance. They should be considered separately and balanced around how much agency you want to give to the player. If you want to give them more agency, you can increase the point rewards without touching the show scoring formulae.

As for the shows proper, in my opinion, ideal point distribution should look closer to this:
40% watching+completed
35% rating
20% discussion
20% favorites
–15% drops

As I've mentioned before, awarding of favorites and penalizing of drops could be decoupled from the raw viewership numbers by treating them as ratios/percentages rather than absolute values. I wouldn't be opposed to making rating even more prominent than it is in this setup, but let's try out a more conservative increase first and see if it proves itself.
moozoohMay 23, 2022 11:52 PM
Jun 16, 2022 7:27 AM
Offline
Sep 2021
2
Change how Ace works
I think Ace should work like this:
* +20,000 points if the Aced anime collected the most points among animes that haven't passed the 85,000 members threshold in "Watching" + "Completed" for that week.
* −5,000 points if the Ace does not become the anime with the most points for that week, or has passed the 85,000 members threshold in "Watching" + "Completed".
Jul 3, 2022 4:33 PM
Reviewer

Offline
Apr 2012
201
I don't think bomber up/down should be a thing.

It just makes somebody else lose rather than making you win. I came in ranked 9 with only 20k points separating me from winning and being rank 11. If the player above or below me decided to ruin my chances, they could've just bombed me and I would've lost despite having great analysis and picks for the season, which I think is entirely unfair. I think it should be purely self-determined how well you do in the season, rather than hoping the people around you don't decide to be cruel.

Thankfully, this didn't happen and the people around me all used it on swaps and boosters
3xTrippleJul 3, 2022 4:38 PM

Jul 6, 2022 5:26 AM
A sandwich

Offline
Jul 2020
1947
Hi, I've just joined to say three things:
1) This was a season were I performed way higher than I had initially anticipated. I think what saved me were not only the anime choices but the use of aces distantly throughout the weeks, saving up one for the very end, so regarding how they work I've got no complaints. Maybe the score values need some tinkering though.

2) I would however change the way the other power ups work, mainly the bomber up and down. They are not only used to make another person lose while hindering your own score (just as @3xTripple said earlier) but unless you're at the top 3 by a wide margin it's just not wise to use them. I'd like to suggest a better option but I haven't come up with one as of yet.

3) Regarding the prizes for the top 10, reading the rules it says:

"Potential winners are the person with the top 10 ranking at the end of the event on July 3, 2022. Sponsor will email the potential winners at the email address registered at time of entry with further information and any additional requirements to accept the prizes. If a potential winner cannot be contacted within 72 hours, fails to respond to any attempted contact, is otherwise in non-compliance with these Official Event Rules, or declines to or cannot accept the prize to be awarded, such potential winner will be disqualified, his/her entry will be declared null and void and the Sponsor reserves the right, in its sole and absolute discretion, to select at random another eligible entry, in which event these provisions shall apply to such other eligible entry"

I came in 8th yet, as of now, I've received no email or notification about any of that. Should I contact someone about it? I apologise if this is not the place to ask about this.

Anyway, this was one hell of a ride I had with other people from my group who also participated but didn't come as close. We're looking forwarding to participating in the next one.
⠓⠥⠍⠁⠝⠊⠞⠽ ⠞⠓⠗⠕⠥⠛⠓ ⠎⠕⠍⠑ ⠎⠞⠥⠏⠊⠙⠊⠞⠽⠂ ⠎⠕⠍⠑ ⠺⠊⠇⠇ ⠎⠅⠊⠏ ⠺⠁⠎⠓⠊⠝⠛ ⠞⠓⠑⠊⠗ ⠓⠁⠝⠙⠎ ⠁⠝⠙ ⠑⠝⠙ ⠥⠏ ⠗⠥⠊⠝⠊⠝⠛ ⠊
⠞ ⠋⠕⠗ ⠑⠧⠑⠗⠽⠕⠝⠑
⠎⠓⠊⠊⠅⠕ ⠁⠕⠎⠕⠊⠂ ⠼⠃⠼⠚⠼⠁⠼⠓
Jul 8, 2022 4:29 AM
A sandwich

Offline
Jul 2020
1947
rasterman7 said:
I came in 8th yet, as of now, I've received no email or notification about any of that. Should I contact someone about it? I apologise if this is not the place to ask about this.
There it is.
⠓⠥⠍⠁⠝⠊⠞⠽ ⠞⠓⠗⠕⠥⠛⠓ ⠎⠕⠍⠑ ⠎⠞⠥⠏⠊⠙⠊⠞⠽⠂ ⠎⠕⠍⠑ ⠺⠊⠇⠇ ⠎⠅⠊⠏ ⠺⠁⠎⠓⠊⠝⠛ ⠞⠓⠑⠊⠗ ⠓⠁⠝⠙⠎ ⠁⠝⠙ ⠑⠝⠙ ⠥⠏ ⠗⠥⠊⠝⠊⠝⠛ ⠊
⠞ ⠋⠕⠗ ⠑⠧⠑⠗⠽⠕⠝⠑
⠎⠓⠊⠊⠅⠕ ⠁⠕⠎⠕⠊⠂ ⠼⠃⠼⠚⠼⠁⠼⠓
Sep 11, 2022 7:51 PM
Go read Medalist
Offline
Apr 2007
285
The next season is soon upon us, and we still haven't had any good discussions on the scoring system and stuff. I'm partly to blame because I promised to make this post back in early August to give everyone statistics to discuss but only recently realized that I actually do have all the data I need to prepare it—so I have. It's just that it was a lot of data to go through.

First off, here are some seasonal statistics for the top 10 shows on the resulting ladder (e.g. from Komi-san S2 through Yuusha Yamemasu). It's just 10 because fitting more numbers on graphs made them look a lot less readable.

Week-to-week progression of audience numbers:


(Full size on click.)

Points of interest
* Steep decline of Shikimori after week 2 due to poor reception and scheduling issues, allowing Komi to confidently catch up.
* Summertime suddenly gaining a lot of momentum on week 8 and even reverting the normal downward trend a mere 1/3 into its run. I have no explanation for this; perhaps the show had a lot more untapped potential for growth and was still being slept on due to the D+ jail (there were no major events that helped it otherwise, to my knowledge).
* Paripi Koumei getting a lot of new audience on week 5, presumably thanks to getting glowing endorsement from major anitubers' doing the rounds on seasonal shows.
* Otome Game's surprising performance in the second half of the season, considering the pits it started from.
* As usual and expected, all shows that ended during the season got a significant boost on the respective final weeks from people who didn't set them as watching or had them as completed while they were still airing. This indirectly helped all the remaining shows as starving seasonal anime watchers started picking up whatever was left.

Week-to-week progression of drops:


(Full size on click.)

Points of interest
* Contrary to what some Discorders have said and estimated early in the season, drop progression isn't linear—at least not for everyone, as best illustrated by week-to-week changes. For most shows, it follows a U-shape, peaking around week 3 or 4 (the "three episode" rule), then going on a decline, and peaking again close to the end of the season, with week 10 being almost a universal drop in drops (sorry). I'm not sure what causes the second peak, but likely reasons are fatigue and secondary/tertiary picks from seasonal viewers who finish their main ones and are starving for more content.
* Komi's performance was weirdly inconsistent, with significant changes week over week; I don't have anything in mind to attribute this to.
* Again, Otome Game defied all expectations and finished very respectably despite a disastrous start. It was also the only series that peaked on the first week and had no bumps of any sort on the following few.

Week-to-week progression of favorites:


(Full size on click. Also note that it starts on week 2. The reason is I didn't record the state of favorites on week 0, so week 1 data had all the pre-favorites added in, sending values into the stratosphere. So I cut them.)

Points of interest
* This is a mess. There is nothing resembling a universal trend (other than the usual end-of-season bump) or linear progression.
* Shikimori, DAL, and Aharen (the only show that lost favorites on week 2—not seen here, unfortunately) fell pretty hard from the lofty heights they started with.
* Komi, Gaikotsu, and Summertime had the only progression resembling linearity, with Summer's being the most erratic of the three but staying within consistent bounds.
* Paripi Koumei, Tomodachi, and, again, Otome Game showed everyone else how it's done.

Finally, rating (it's not as busy as the other graphs so I added a few more shows):


(Full size on click.)

Points of interest
* Several shows very quickly established a very narrow band and stayed there until the end-of-season bump: Paripi Koumei, Honzuki, KoiSeka, Gaikotsu, and to a lesser extent Summertime and Yuusha.
* For Paripi Koumei, Komi S2, and DAL, the end-of-season bump (typically a result of culling early droppers with low scores) didn't even come—they received nothing extra.
* There were both new and returning shows that took many weeks to climb to their potential, among them Kingdom S4, Ao Ashi, and particularly Tomodachi and Otome Game—the latter managing to climb almost a full point since the start of the season and outperforming many hyped shows in the process.
* DAL, Komi, Aharen, and Kakkou had their best rating at the start and then slowly declined. But they've got nothing on the disappointments of the season: Shikimori and Shijou.

All of this data and more can be found in this spreadsheet, which you can use as a template for your own data dumps and become a spreadsheet warrior like myself.
moozoohSep 22, 2022 5:56 PM
Sep 11, 2022 7:53 PM
Go read Medalist
Offline
Apr 2007
285
On to the more pressing concerns, let's discuss what this taught us about the new scoring system and where we might be going next.

Here's the table of the Spring 2022 season top 20 shows by overall score. Note these names and how they are positioned relative to each other.



Here are the same 20 shows with their score broken down by metrics.


(Full size on click.)

The average roughly corresponds to 51% of show's total score coming from audience numbers, 41% rating, 14% favorites, 6% discussion, and -12% drops. This proportion is relatively stable despite the small sample size, and doesn't change too much even if you start culling outliers. This is already a big success and certainly better than the one we had before, but it can (and should) be improved further; mainly because the numbers in the proportion are deceptive and don't exactly represent what you might think they do.

Two major points of criticism immediately stand out.

I. The scoring system is still massively skewed towards popularity over audience response.

* Popularity translates into viewership points directly and then scales two more quantity metrics—favorites and discussion—all at once, letting shows with lukewarm response from a big enough audience dwarf the opposite cases.
* Main ad-hoc audience response indicators in the current system—rating and drops—are relatively weak in comparison. The rating formula is solid but isn't penalizing enough, and drops need further tweaking to feel good.
* Pre-favoriting skews scoring by giving popular shows (Shikimori, Aharen, Kakkou) and shows announced far in advance (Date a Live IV) a massive headstart while the others have to work for it from the ground up. I have seen people on Discord raising this concern as well many times. Seems like nobody is really happy about this being a thing.
* Unlike rating, favorites are an absolute metric, so a show with a drastically bigger audience but same ratio of favorites to viewers will have much more points for it. While the same also applies to drops, the latter are decided during the season and not in advance, making it a more fair counterbalance to raw viewership.

All of this is illustrated by the fact that shows like Shikimori-san, Aharen-san, Yuusha Yamemasu, etc., were able to steal the march on those received much better and planting themselves firmly in top 10 early on. Shikimori and Aharen even ended up #3 and #6 respectively, meaning they were among the main breadwinners for any team that had them despite mediocre ratings and massive drop figures. Three of the seasonal shows that finished above 8.00 didn't even crack top-10 while the #10 place was held by a 7.08 title. This looks like a problem to me; the contest urges participants to "predict the anime of the season" but gives unfair and evident advantage to quantity over quality. By doing so it also gives unfair advantage to spreadsheet warriors like myself over a naive player who actually intends to predict the anime of the season and not one that best fits a mathematical model.

How this can be addressed:

1. Reduce watching+completed point rewards. Regardless of any other step, this is immediately beneficial for competition and is worth doing for the fall season.

2. Raise the rating threshold but make the reward bigger to compensate. Ratings for modern shows are massively inflated, to the point where anything under ~7.00 isn't "above average" anymore, but decidedly bad in almost every case. As the site-wide average is currently 6.79 and shows no signs of decreasing, it makes sense to balance around that number (make it 6.80 for a pretty decimal). Edit: This may have been inaccurate. A DB admin confirmed to me that the actual average is 6.37 at the moment, but thankfully, for our case it doesn't matter too much because it takes into account a lot more content than we need to balance around. For the record, week 13 average for the past season's top 20 shows is 7.66, comfortably above 6.80, so I still think that's a good threshold value—I'd dare say it's even on the conservative side, you can make it work just as well all the way up to 7.00.

3. Combat pre-favoriting. This can be done multiple ways.

* As a site-wide feature: simply make not-yet-aired titles ineligible for favoriting and purge their current favorites. Optionally send a system message to affected users saying what shows have been purged from their lists of favorites and explaining that they will become available for adding back when they start airing.
Pros: Neatly solves the problem in its entirety by putting every show on an even playing field.
Cons: Goes beyond the scope of FAL and causes (a minor) inconvenience to some users if admins think pre-favoriting is what they should be able to do.

* As a naive FAL-only feature: subtract the number of favorites at the hour the show starts airing from any further calculations (so e.g. if a show had 1000 favorites right before the first episode aired, week 5 will award 15 points per each favorite above 1000 for this show).
Pros: Entirely removes massive advantage from shows that benefit the most from pre-favoriting.
Cons: Also unfairly penalizes them by discounting a big chunk of legitimate favorites they would've gotten anyway (but happened to be delivered in advance), raises questions from the apparent discrepancy ("why does my show have 2000 favorites but scores so poorly this week?"), and introduces arbitrary numbers into scoring which users need to know to make their own calculations.

* A hybrid feature: FAL-eligible shows are forcibly removed from users' favorites lists at the season's start with an accompanying system message telling them they can add them back if they wish so.
Pros: Does not discount more than is necessary, keeps the formula clean and simple.
Cons: Disrupts general user experience without actually solving the main issue with pre-favorites, since users can just add the show back immediately. Even if they are removed and locked out for the entire duration of team selection, nobody prevents spreadsheet-savvy users from collecting the data in advance.

* Keep pre-favoriting in but combine it with the next step (making favorites count a percentage/ratio of active audience).
Pros: Partially solves the problem at zero risk and low investment.
Cons: Lazy, only solves one part of the issue.

Personally, I am in favor of the first option (remove the ability to pre-favorite entirely) but I don't know if it gets consensus among the admins. Failing that, the last option seems least controversial.

4. Count favorites and drops as percentages to take audience scaling out of the picture. The concept is simple: a show with 100k viewers and 10k drops, and a show with 10k viewers and 1k drops both have one drop per ten viewers, so they should be treated equally. Hence the values we should be working with are, respectively (number of drops)/(number of watching+completed)×100%, likewise with favorites. So a show with 18205 viewers and 308 favorites would be expressed as 308/18205×100%=1.69% favorited.

I suggest scaling the formula the same way it is done with rating: a threshold with a score multiplier that rewards/penalizes the show depending on which side of the threshold it lies on. Anomalies like DAL and shows with very small audience where each favorite matters a lot can be mitigated by capping maximum bonuses/penalties (see further below on examples).

5. Do something about the discussion points. This metric is currently in an awkward state where increasing the point value of posts is a potential abuse vector with players inflating post count with low-content one-liners and reaction responses, but without any sort of increase it barely adds anything. And in addition to that it still suffers from being closely coupled with audience size, so basically it's just free bonus points for popular shows.

Perhaps this can be partially solved by introducing a scale-down multiplier in a vein similar to the drops/favorites suggestion above, thereby decoupling this metric from popularity, then cautiously increasing post value.

But I think increasing last-week post value right now is very warranted, since the total reward is already naturally low due to only having one week's worth of posts and discussion dying down for most finished series (Shijou, Gaikotsu, and Paripi Koumei had posts in the single digits for the final week, and these weren't unpopular shows).

To help illustrate most of these suggestions, I created a sandbox on the second tab of the spreadsheet I linked above. It is fully interactive and contains an example scoring model I came up with. It achieves a distribution of 41/41/12/8/-2 and feels a lot better for shows with smaller audience numbers (for instance, it puts Kingdom, Ao Ashi, and Honzuki S3 in the top-10 which is more intuitive given their high rating). I personally think it still requires more effort to make discussion points more prominent and some more creative capping to work around pre-favoriting for shows like DAL.

Feel free to make a local copy and poke numbers until you feel like you've arrived at a good result, then share it here.


II. Players lack meaningful agency to make risky gambles or offset developments beyond their control.

In one of my previous posts (see points #4 and #7 there), I suggested a risk/reward system for aces, boosters, and swaps that rewards early (risky) decisions and penalizes late (safe) ones. I stand by that suggestion and would like to invite other players to discuss this, as I believe it would make things a lot more interesting and resolve stale ties that tend to happen when players pick heavily optimized teams. I think the earlier something like this is implemented, the more all of us benefit.

I'd also like to echo one of the suggestions for a wildcard (preferably in place of a nuke since it's bad in both concept and function) that I remember someone else made on Discord (?), but I can't find it.

In short, it's the ability to permanently swap a show with any other show from outside your team, at a massive score penalty; I think something on the order of at least 50k would be fair, but needs further experimentation. The point of this is to give those who made unfortunate team-picking decisions (like Shikimori in past season) a chance to catch up somewhat, and it's also a late-game, very high-risk gamble for those who miss one specific pick or want to hyperoptimize for extra aces and stuff. I think it's a lot more exciting than making another player's life worse for what isn't their fault.

I'll add to this post if I remember anything else I wanted to say.
moozoohSep 12, 2022 1:23 PM
Sep 18, 2022 2:58 AM
Offline
Dec 2020
6
this good
Sep 20, 2022 3:51 AM

Offline
Jun 2017
3172
AveryFB said:
For Spring 2022 Season could the badge you get change based off of how well you did at the end? Like if you get #1 you get winners badge, top 100 or top 1000 a different badge, or maybe like top 1% top 10% finisher top 50% or something like you get at least a different badge if you do well.
moozooh said:

5. Give more recognition to non-winners who end up near the top. Right now somebody who finished at, say, #15 receives the same badge as somebody who finished at #1500 or #15000. I think it would encourage more people to try harder if, say, finishing in the top 200 or so would receive some kind of a different badge compared to a regular participant. I think something similar has already been suggested before.

This is to recognize those who care about their participation and/or achieve meaningful results. Would be even better if one could choose to display their badges in place of the graphic signature or somewhere else visible on the forums, kind of like the Easter Egg Hunt badges you have right now next to your avatars.

I'm late on this and it probably can't be applied to Fall 2022 season either but please do something about this, it's frustrating that users who put some effort during the FAL season get the same badge as someone who just spent some minutes to make their team and then literally that's all they done for FAL. (well if also someone has a really good draft that for picking the right titles alone and they're in top 500 or something they deserve the different badge too)

Also to add something for risky picks/decisions,I'm saying this for the third time I think but one of the things I loved about old FAL was when an anime got picked by less than 3% of active participants, that anime would get double points, I like to see something "similar" to this in new system, for example the percentage be less than 1% instead (or even lower) of 3 because of lots of new sign ups in new system (22k sign ups in new system vs. at best 2k sign ups in old system) and the double points would be ×1.5 total points or something like that if double points seem to be unfair. With the point system FAL has, most of the unpopular titles won't have good points anyway even with double points (maybe some sequels like Idolish 7 S3 part 2 for Fall 2022 have a chance with this suggestion to be among tops but normally I doubt it gets in top 15 which makes it being a useless pick). It's sad that these unpopular titles can potentially bring new competitions and this potential was used in old system but not now
mhkrSep 20, 2022 3:56 AM
Sep 20, 2022 4:09 AM

Offline
Jan 2008
2990
Is there an actual register thing to click or does just choosing the anime count? (I did that part where I have my active and bench anime at the top)

It says "Submit your team before the timer runs out!" but there is no actual submit button that I can see
Sep 20, 2022 8:22 AM
Go read Medalist
Offline
Apr 2007
285
UberBat said:
It says "Submit your team before the timer runs out!" but there is no actual submit button that I can see

It's this button at the bottom of the page:



For me it reads "Change" because I've already submitted my team once, but if you haven't yet it will read "Submit" for you.
Sep 20, 2022 1:34 PM

Offline
Jan 2008
2990
moozooh said:
UberBat said:
It says "Submit your team before the timer runs out!" but there is no actual submit button that I can see

It's this button at the bottom of the page:



For me it reads "Change" because I've already submitted my team once, but if you haven't yet it will read "Submit" for you.


ok might have found the specific thing you showed me
UberBatSep 20, 2022 1:39 PM
Sep 28, 2022 12:09 AM
Offline
May 2019
2
It could be interesting to have an utility to export the team.
Oct 2, 2022 8:29 PM

Offline
Oct 2020
1638
yo how does ace work? is ace only available once or per week?
Oct 2, 2022 8:38 PM
Go read Medalist
Offline
Apr 2007
285
Sethispr said:
yo how does ace work? is ace only available once or per week?

It's once per week, and it's limited by the number of shows you have, so at most 8.

The conditions for acing can be a little complicated if written out in text, so here's a flowchart to guide you:

Oct 2, 2022 9:48 PM

Offline
Oct 2020
1638
moozooh said:
Sethispr said:
yo how does ace work? is ace only available once or per week?

It's once per week, and it's limited by the number of shows you have, so at most 8.

The conditions for acing can be a little complicated if written out in text, so here's a flowchart to guide you:


ayy alright tysm very helpful
Oct 5, 2022 2:11 PM
Offline
Nov 2020
6
"Fantasy Anime League (FAL) is a fantasy sport game with an anime twist." Admins should really consider changing the name as this game doesn't really resemble fantasy football other than there are teams and points.
Reply Disabled for Non-Club Members
Pages (5) « 1 2 [3] 4 5 »

More topics from this board

» Fall 2024 Registrations Now Open!

Kineta - Sep 15

26 by Hattyson »»
Oct 29, 2:33 AM

» A comprehensive guide to Fantasy Anime League

moozooh - Mar 25, 2023

12 by moozooh »»
Oct 8, 6:04 AM

» Ideas for solving the discussion points problem

moozooh - Jul 21, 2023

31 by Samii »»
Aug 18, 8:21 AM

» Spring 2024 Week 13: Ace/Swap Reminder & Discussion (final week)

eplipswich - Jun 23

3 by badabass »»
Jul 1, 6:52 AM

» Spring 2024 Week 12: Ace/Swap Reminder & Discussion

eplipswich - Jun 16

6 by kernel99 »»
Jun 27, 5:33 AM
It’s time to ditch the text file.
Keep track of your anime easily by creating your own list.
Sign Up Login