Jump to content
SaltMaster 5000

Carolina Krayts is the best X-Wing podcast

Recommended Posts

2 hours ago, svelok said:

Both formats, timeline from released of wave 4 to now, filtered to events with >=25 players and >= 75% of lists filled.

That's 3,262 total lists across 59 events (worlds/nova duplicate events removed, but that's it for pruning). Of those, 477 are Rebels.

Blue Squadron Escort is used in 17 lists.

For context:

Wedge Antilles: 233 lists.
Braylen Stramm: 147 lists.
Arvel Crynyd: 56 lists.

Either 97% of rebel players are sleeping on T65 generics or they really just aren't good, I promise

Just on MetaWing without as much attention paid to filtering/pruning, both the Blue X-Wing and B-Wing were around 40-50 lists (Wedge ~360, Braylen ~190, for comparison), and only a few slots apart from each other around rank 80.

http://meta.listfortress.com/pilots?ranking_start=2019-07-10&ranking_end=2019-11-11&large_tournament_multiplier=true&widespread_use_multiplier=true&use_ranking_data=all&tournament_type=&format_id=&

That's not great, but it's like 75-80% of the way from the bottom.  Not very good.  Not panic mode.  In American letter grades, maybe a C+.  Win percentiles for them aren't obviously terrible.  

If there's a C+ ship that folks just aren't super interested in flying, can't we just be fine with letting folks decide not to fly it?

That's just how folks choose stuff.  Most folks don't really look at marginal cost of energy (too much) when picking between Electric and Gas for a kitchen stovetop--they just like how either Gas or Electric behaves.  Folks don't pick transportation modes purely on price--they'll often choose car or bus based on what they find convenient, access to parking, and so forth.  Also, some folks just want to drive the red one. 

X-Wing is the same way.  Many folks find "words" more fun than "no words."  It's not purely win rates that'll lead to folks flying high initiative.  I think things should be better balanced, but we'll never get to a place where Wedge and Blue Escorts are equally common, and that's OK.

Share this post


Link to post
Share on other sites
28 minutes ago, theBitterFig said:

Just on MetaWing without as much attention paid to filtering/pruning, both the Blue X-Wing and B-Wing were around 40-50 lists (Wedge ~360, Braylen ~190, for comparison), and only a few slots apart from each other around rank 80.

http://meta.listfortress.com/pilots?ranking_start=2019-07-10&ranking_end=2019-11-11&large_tournament_multiplier=true&widespread_use_multiplier=true&use_ranking_data=all&tournament_type=&format_id=&

That's not great, but it's like 75-80% of the way from the bottom.  Not very good.  Not panic mode.  In American letter grades, maybe a C+.  Win percentiles for them aren't obviously terrible.  

so, arguments like this are the reason I made ATC.  I can't find any rhyme or reason to weighted scores vs performance in the wild.  What does the ranking order of things on meta wing even mean?  I honestly have absolutely no idea, and I suspect I've put a lot more work into trying to find out than the majority of people casually opening the site.

 

For example, Captain Sear is currently the highest "weighted score" on Meta-Wing.  His average percentile is... 44.21%?  But wait, metawing does percentiles the other direction: 

I'm ranking things based on one primary attribute: Ranking percentiles. For this, I calculate the percentile of each squadron's result in each tournament, already when importing the List Fortress data. Like that, I can then do all the heavy lifting in the database, which results in reasonable performance even for complex filters in queries.

Examples (slightly simplified): 3rd of 12 => 75th percentile, 9th of 90 => 90th percentile - and after the cut: 2nd of 16 in cut of 80 => 87.5th percentile.

The ranking then adds these percentiles for the chosen rankings (Swiss or cut or both), which results in a value I call "Weighted Score". Two multipliers are used on this Weighted Score value, if used for the filter: log(number of players in tournament) in order to reward heavy competition, and log(number of squadrons who used this thing, total) in order to bring things to the top that are prevalent in the meta.

If you want more precise information, then you can always Use the Source. Maybe I'll change the algorithm to something more founded in statistics some time, but for now, it appears to bring the squads one would expect to the top.

Except, Captain Sear's avg Percentile is 44.21%.  ...What? That's dramatically lower than what you would expect (in ATC for a somewhat different dataset he's at what would be 63%, basically.  And higher is beta on metawing.  His weighted score is... higher than everyone elses, but I have no idea what the weighted score is supposed to even be measuring.

TL;DR I can't find any rhyme or reason for how metawing works (I haven't looked through the source) for making arguments for "X is good" vs "X is bad".  It's extremely useful for seeing what things people play - if you just open the list archetypes view it gives a good representation of what people are playing in the wild near the top.  As for how those things are actually performing, I am extremely skeptical that the value add of "avg percentiles" and "weighted scores" for these lists is high.  Honestly, I think it might be negative - it would be better to sort by number of squadrons and just leave it at that.

 

To answer your question, you need these numbers:

VcSNy9L.png

Not "ATC is way better than meta-wing" but "please don't use meta-wing to make arguments like that".  Unfortunately, these ships are played so little we don't have real data to draw any kind of useful conclusions.

Share this post


Link to post
Share on other sites
24 minutes ago, Brunas said:

so, arguments like this are the reason I made ATC.  I can't find any rhyme or reason to weighted scores vs performance in the wild.  What does the ranking order of things on meta wing even mean?  I honestly have absolutely no idea, and I suspect I've put a lot more work into trying to find out than the majority of people casually opening the site.

 

For example, Captain Sear is currently the highest "weighted score" on Meta-Wing.  His average percentile is... 44.21%?  But wait, metawing does percentiles the other direction: 


I'm ranking things based on one primary attribute: Ranking percentiles. For this, I calculate the percentile of each squadron's result in each tournament, already when importing the List Fortress data. Like that, I can then do all the heavy lifting in the database, which results in reasonable performance even for complex filters in queries.

Examples (slightly simplified): 3rd of 12 => 75th percentile, 9th of 90 => 90th percentile - and after the cut: 2nd of 16 in cut of 80 => 87.5th percentile.

The ranking then adds these percentiles for the chosen rankings (Swiss or cut or both), which results in a value I call "Weighted Score". Two multipliers are used on this Weighted Score value, if used for the filter: log(number of players in tournament) in order to reward heavy competition, and log(number of squadrons who used this thing, total) in order to bring things to the top that are prevalent in the meta.

If you want more precise information, then you can always Use the Source. Maybe I'll change the algorithm to something more founded in statistics some time, but for now, it appears to bring the squads one would expect to the top.

Except, Captain Sear's avg Percentile is 44.21%.  ...What? That's dramatically lower than what you would expect (in ATC for a somewhat different dataset he's at what would be 63%, basically.  And higher is beta on metawing.  His weighted score is... higher than everyone elses, but I have no idea what the weighted score is supposed to even be measuring.

TL;DR I can't find any rhyme or reason for how metawing works (I haven't looked through the source) for making arguments for "X is good" vs "X is bad".  It's extremely useful for seeing what things people play - if you just open the list archetypes view it gives a good representation of what people are playing in the wild near the top.  As for how those things are actually performing, I am extremely skeptical that the value add of "avg percentiles" and "weighted scores" for these lists is high.  Honestly, I think it might be negative - it would be better to sort by number of squadrons and just leave it at that.

 

To answer your question, you need these numbers:

VcSNy9L.png

Not "ATC is way better than meta-wing" but "please don't use meta-wing to make arguments like that".  Unfortunately, these ships are played so little we don't have real data to draw any kind of useful conclusions.

I don't understand. Could you just give me a magic number. Call it the will to win please

Share this post


Link to post
Share on other sites
37 minutes ago, Brunas said:

so, arguments like this are the reason I made ATC.  I can't find any rhyme or reason to weighted scores vs performance in the wild.  What does the ranking order of things on meta wing even mean?  I honestly have absolutely no idea, and I suspect I've put a lot more work into trying to find out than the majority of people casually opening the site.

[...]

Not "ATC is way better than meta-wing" but "please don't use meta-wing to make arguments like that".  Unfortunately, these ships are played so little we don't have real data to draw any kind of useful conclusions.

All good.

I've been away a bit, might have missed when ATC became a thing.

Is there a link to it somewhere?

 

Share this post


Link to post
Share on other sites
3 minutes ago, theBitterFig said:

All good.

I've been away a bit, might have missed when ATC became a thing.

Is there a link to it somewhere?

 

Oh, no worries, it was officially "done" enough for public use... 22 hours ago?  I can't believe you, out of the loop for so long :p

and sure - http://advancedtargeting.computer/

 

Share this post


Link to post
Share on other sites
Just now, Brunas said:

Oh, no worries, it was officially "done" enough for public use... 22 hours ago?  I can't believe you, out of the loop for so long 😛

and sure - http://advancedtargeting.computer/

 

That's the thing about being away for a week: you don't know how much you don't know, and there are unknown unknowns.

48 minutes ago, Brunas said:

To answer your question, you need these numbers:VcSNy9L.png

I don't know that I'd draw conclusions, but there are some things I know I can't say.

  • 1. I can't say an X-Wing or a B-Wing is obviously better than the other.  Really similar performance (such as it is), not super high sample size (but better than many ships).
  • 2. I can't say they're obviously great.  I mean.  Duh.
  • 3. I can't say they're obvious trash.  A lot of ships do appear obviously bad in their results.  26% cut rate is respectable enough.  20% cut winrate--almost surely 1 out of 5 games :P--is still 1 game more than a lot of ships.

But in terms of broad strokes balance philosophy, maybe that's enough.  Seems like they win enough that, if someone really wants to play them, it's not hopeless.

Share this post


Link to post
Share on other sites
4 hours ago, svelok said:

Blue Squadron Escort is used in 17 lists.

I think part of the issue goes back to how Hyperspace got butchered. Back when it was only four ships for rebels, the escort worked well as a filler because it was one of the only options.

Share this post


Link to post
Share on other sites
2 hours ago, theBitterFig said:

That's the thing about being away for a week: you don't know how much you don't know, and there are unknown unknowns.

I don't know that I'd draw conclusions, but there are some things I know I can't say.

  • 1. I can't say an X-Wing or a B-Wing is obviously better than the other.  Really similar performance (such as it is), not super high sample size (but better than many ships).
  • 2. I can't say they're obviously great.  I mean.  Duh.
  • 3. I can't say they're obvious trash.  A lot of ships do appear obviously bad in their results.  26% cut rate is respectable enough.  20% cut winrate--almost surely 1 out of 5 games :P--is still 1 game more than a lot of ships.

But in terms of broad strokes balance philosophy, maybe that's enough.  Seems like they win enough that, if someone really wants to play them, it's not hopeless.

Oh, I don't even think you can say 2 and 3.  23 people have played them in the past... 4-5 months, across huge amounts of events.  23/26 lists doesn't pass the "noise" threshold at all, unfortunately.  Basically, "insufficient data for any useful conclusion"

Share this post


Link to post
Share on other sites
3 minutes ago, RynoZero said:

considering FFA xwings/Ywings were a list even at the end of 1.0 id say it has potential. 

Heh, I think last big 1.0 tournament was Polish Nationals won by 4Xt65. FFA were strong but ability to choose either to do 3 hard or 3 tallon with PS10 Wedge was just crazy strong. 

In 1.0 they fixed things by making other more broken things :)

Share this post


Link to post
Share on other sites
8 hours ago, theBitterFig said:

That's the thing about being away for a week: you don't know how much you don't know, and there are unknown unknowns.

I don't know that I'd draw conclusions, but there are some things I know I can't say.

  • 1. I can't say an X-Wing or a B-Wing is obviously better than the other.  Really similar performance (such as it is), not super high sample size (but better than many ships).
  • 2. I can't say they're obviously great.  I mean.  Duh.
  • 3. I can't say they're obvious trash.  A lot of ships do appear obviously bad in their results.  26% cut rate is respectable enough.  20% cut winrate--almost surely 1 out of 5 games :P--is still 1 game more than a lot of ships.

But in terms of broad strokes balance philosophy, maybe that's enough.  Seems like they win enough that, if someone really wants to play them, it's not hopeless.

A ship has built in linked focus into barrel, another has a gimmick way to get a linkable boost, but at the price of one red die and needing to commit to it before executing a manouver.

On a stroke of bad luck, one dies by two shots, the other by three (granted, one also could work in the other end of variance, but usually it's way more game compromising blanking out on first 2 attacks than rolling 4 evades there).

One ship is faster, that's true, but how often that's relevant in a ps 2 ship? Especially since, it's worth repeating, the slower ship got focus into barrel which is actually good in low ps ship too.

Then if we want to speak about results, ps 2 b-wings won a SoS.

 

But ok, if you think it's just personal preference I don't want to change your mind. It won't be the first time someone, me included, was wrong in this thread

Share this post


Link to post
Share on other sites
14 hours ago, Brunas said:

 

VcSNy9L.png

Not "ATC is way better than meta-wing" but "please don't use meta-wing to make arguments like that".  Unfortunately, these ships are played so little we don't have real data to draw any kind of useful conclusions.

My first instinct looking at that chart is that theres one person playing both in the same list and we're looking at thier tournament data through 23 games and also someone went 0-3 with the b wing.

Share this post


Link to post
Share on other sites

I find the intensity of the fight between 

8 minutes ago, catachanninja said:

My first instinct looking at that chart is that theres one person playing both in the same list and we're looking at thier tournament data through 23 games and also someone went 0-3 with the b wing.

No! Give me concrete meaning with those 23 data points over many months! I needs it!

Share this post


Link to post
Share on other sites
9 hours ago, Brunas said:

Oh, I don't even think you can say 2 and 3.  23 people have played them in the past... 4-5 months, across huge amounts of events.  23/26 lists doesn't pass the "noise" threshold at all, unfortunately.  Basically, "insufficient data for any useful conclusion"

Ah, but what about the not-so-useful conclusions? :P Like, there's some level of absurdly high or absurdly low winrate where we can kinds say "that ain't it."

3 hours ago, Sunitsa said:

A ship has built in linked focus into barrel, another has a gimmick way to get a linkable boost, but at the price of one red die and needing to commit to it before executing a manouver.

On a stroke of bad luck, one dies by two shots, the other by three (granted, one also could work in the other end of variance, but usually it's way more game compromising blanking out on first 2 attacks than rolling 4 evades there).

One ship is faster, that's true, but how often that's relevant in a ps 2 ship? Especially since, it's worth repeating, the slower ship got focus into barrel which is actually good in low ps ship too.

Then if we want to speak about results, ps 2 b-wings won a SoS.

 

But ok, if you think it's just personal preference I don't want to change your mind. It won't be the first time someone, me included, was wrong in this thread

I'm not saying they're exact, but... close enough?  But like, they're played about as much as each other anyhow.  It isn't like there are three or four times as many B-Wings as X-Wings out there.  We can see there's a clear gap between those and, say, a Tempest TIE/x1 or A. Cartel. Marauder.

As to low-init speed, it does open up deployment options.  B-Wings can't really fan out too well, since they're so slow and easily isolated.  X-Wing gives a few more options; it's easier to stick one on the flank, since 4-straight boost and 3-bank-boost is a lot more speed.  Speed also helps close out games.  B-Wing that gets lost in space towards the end is going to stay lost.  Boosts help an X-Wing out of the fight get back into the fight.  Out-of-combat speed was a massive part of why FFA was so good on low-Init generics at the end of 1e.

Is that worth the loss of a linked roll (it's good!) and slightly better defensive statline?  Well, probably depends a lot on the player, the rest of the list, etc etc.

Share this post


Link to post
Share on other sites
10 minutes ago, catachanninja said:

My first instinct looking at that chart is that theres one person playing both in the same list and we're looking at thier tournament data through 23 games and also someone went 0-3 with the b wing.

source.gif

 

 

Lowkey suggests they're close enough to each other, if the player is like, "I'm playing both in the same list."

Share this post


Link to post
Share on other sites
15 minutes ago, catachanninja said:

My first instinct looking at that chart is that theres one person playing both in the same list and we're looking at thier tournament data through 23 games and also someone went 0-3 with the b wing.

 

1 minute ago, theBitterFig said:

source.gif

 

 

Lowkey suggests they're close enough to each other, if the player is like, "I'm playing both in the same list."

Jokes aside... that's partially the Franke Forte data point (he unironically loves '4bz' where he splits Bs and 65s)

Share this post


Link to post
Share on other sites
5 minutes ago, Tlfj200 said:

Give me concrete meaning

Step 0: Select Ships/List You Like
Step 1: Review Listfortress, ATC, Metawing, Pinkbrain
Step 2: Adjust Some Things
Step 3: Test Against Logical Things 
Step 4: Find Concrete Meaning 
Step 5: Ignore Step 1 - 4, realize ensnare/jedi exist and nothing matters
Step 6: Live Free, Dont Join. run what you want man

Share this post


Link to post
Share on other sites
3 minutes ago, Boom Owl said:

Step 0: Select Ships/List You Like
Step 1: Review Listfortress, ATC, Metawing, Pinkbrain
Step 2: Adjust Some Things
Step 3: Test Against Logical Things 
Step 4: Find Concrete Meaning 
 

giphy.gif?cid=790b761114ca949eb2601389ff

 

1 minute ago, Boom Owl said:


Step 5: Ignore Step 1 - 4, realize ensnare/jedi exist and nothing matters
 

469.gif

2 minutes ago, Boom Owl said:


Step 6: Live Free, Dont Join. run what you want man

giphy.gif

Share this post


Link to post
Share on other sites
15 hours ago, theBitterFig said:

Just on MetaWing without as much attention paid to filtering/pruning, both the Blue X-Wing and B-Wing were around 40-50 lists (Wedge ~360, Braylen ~190, for comparison), and only a few slots apart from each other around rank 80.

http://meta.listfortress.com/pilots?ranking_start=2019-07-10&ranking_end=2019-11-11&large_tournament_multiplier=true&widespread_use_multiplier=true&use_ranking_data=all&tournament_type=&format_id=&

That's not great, but it's like 75-80% of the way from the bottom.  Not very good.  Not panic mode.  In American letter grades, maybe a C+.  Win percentiles for them aren't obviously terrible.  

If there's a C+ ship that folks just aren't super interested in flying, can't we just be fine with letting folks decide not to fly it?

That's just how folks choose stuff.  Most folks don't really look at marginal cost of energy (too much) when picking between Electric and Gas for a kitchen stovetop--they just like how either Gas or Electric behaves.  Folks don't pick transportation modes purely on price--they'll often choose car or bus based on what they find convenient, access to parking, and so forth.  Also, some folks just want to drive the red one. 

X-Wing is the same way.  Many folks find "words" more fun than "no words."  It's not purely win rates that'll lead to folks flying high initiative.  I think things should be better balanced, but we'll never get to a place where Wedge and Blue Escorts are equally common, and that's OK.

I think it's super fine that some things exist in that kind of low end not good but not awful range. People have this weird obsession/entitlement where they feel like they should be able to pull any random ship and spam it and do well, or that they shouldn't be punished for having some weird list building restriction. "I should be able to run every pilot on a chassis" "I should be able to run ace with two generics that are identically equipped and win"  I don't think a game can be balanced. 

 

There will always be somethings that come out more or less efficient, if the end game is still fun, it doesn't really matter. 

Edited by catachanninja

Share this post


Link to post
Share on other sites
9 hours ago, Oldpara said:

Boost is super strong. Especially linked one. 

Not tested it, but I am little bit afraid of 5 Xt65. But maybe I am wrong?

I think 5X probably crosses the line against the 2-Ship archetype.

I think Ghost+1 and Falcon lists or Double-Ace lists become... possibly Meme status? ...in a 5X meta. 

5Yions got tested and slapped. 

5 Cartels ... We'll see if they survive the nerf-bat. Obviously, Aces cut through a 5 Cartel meta at Worlds, so they seem fine? 

 

Edited by Bucknife

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...