15 minutes ago, Makaze said:

It's a fairly basic chart with a very clear statement about where the data came from and SotL has been very open with answering questions about his methodology. If you don't like the dataset then tough, it's the best we have available. If you don't like the methodology then suggest a better one. But quibbling about the semantic difference between "ranking" vs. "popularity in top 8 cuts" and demanding massive and nuanced disclaimers is just ridiculous.

I also get annoyed when people use shoddy methods like forum polls to back up spurious claims. But that's not what happened here at all, he didn't proceed to go on a rant about the J5K or demand change to the game based on it. Rather he just said here's some data that I've summarized, sorted, and colored for your reading convenience. Which by the way I found very helpful and informative (thank you for that) even if it's not 100% ironclad statistically accurate and published in a tier 1 peer reviewed journal.

4 minutes ago, Makaze said:

R2-D2 is certainly good but only defines that particular Norra or Coran build, not the majority of lists for an entire faction like Palp did. If R2-D2 disappeared tomorrow then both ships could easily be replaced by another Rebel regener with similar points cost.

Biggs though... yeah

The R2-D2 part of my post was more theoretical then actually looking to seem him nerfed. I wanted to see if Blodvargarna actually believed the argument he was using or was just using it for his convenience. Our brief conversation here has dispelled my thoughts on the latter option.

I added Biggs in later for precisely your reaction. There were arguments to be made on R2-D2's case. Less with Biggs.

1 hour ago, Stay On The Leader said:

Haha, I have no sway that I'm aware of!

Really? Because you just did this:

1 hour ago, Lampyridae said:

Wow. I knew I was handicapping myself with them but not this badly.

5 minutes ago, haslo said:

...

Hello sir, one hell of tool, that is. Can I have some few simple questions?
- does the model calculate weight based on tournament rank? Winning regio/open series will give ships more points than winning store champs?
- does the model calculates the number of players? ie, getting in top3 on 100 people tournament will give those ships more points than top3 on 16 people tournament?
- how is a mediocre to low result affecting the ships final grade, generally speaking? In a said example, list placed on 60th position on a 120 people tournament, will actually increase the ships final grade or reduce it? Example:
16 people on the tournamen
TOP1: list with Corran and miranda
TOP8: list with Corran
No more lists with those ships (or no data). Will Miranda be higher than Corran?

Based on looking at meta-wing analyzer (amazing by the way) R3-A2 is the most popular astromech, then M9-G8, then R4-D6, then R2-D2.

If nothing else comes out of this than that more people use Meta-Wing then the post was worthwhile!

I'll happily use whatever 'sway' I have to that end.

Just now, BlodVargarna said:

Based on looking at meta-wing analyzer (amazing by the way) R3-A2 is the most popular astromech, then M9-G8, then R4-D6, then R2-D2.

Yet I havent heard a lot of "NERF R3-A2" shouts.

It's more compliated than that. Stresbot makes it more difficult for you to play in a short term: you get a stress or two, try to shake it, prioritise the stressY, hopefully kill it. Okay, problem solved,

While R2D2 will sometimes mean: you put 4 damage on corran horn, you are almost there, he runs away, regenerates... and you feel like a hard-worked victory just slipped out of your hands. I dont want to say "negative play experience", but I know people who would say that in similar situation.

2 minutes ago, haslo said:

I do talk with sozin he's a great guy, showed me the way to the JSON API of Lists Juggler for example, which made importing much easier

The way this is currently done: The percentile is multiplied with the logarithm of the number of players in the tournament for the "magic" value. This way, larger tournaments will have a much larger impact. You can configure that away with the dropdown box that says "Large tournaments = more Magic" if you want to see the difference.

Good to hear. I do similar in my player rankings. I normalize them though afterwards so the converse is not true as well (one bad showing at a large tournament destroying a ships ranking etc.)

35 minutes ago, BlodVargarna said:

The fact that it is no longer the go-to for Imperial lists IS a good thing. Like, diversify your meta duuuude.

My lists are fine and diverse. I´m just pointing out Palpatine´s nerf wasn´t needed. He wasn´t in 99% of the tournament lists. I didn´t even play that much with him. I´m just pissed off he got nerfed when there wasn´t a need for that.

Just now, Voitek said:

Yet I havent heard a lot of "NERF R3-A2" shouts.

Apart from me ;-)

R3-A2 is so much better at dealing out stress than the other options that when stress becomes a key part of the metagame it warps hugely around a couple of Rebel ships that can double-tap with R3-A2.

I see lots of abstract criticism in this thread but a suprising lack of example on why you are criticizing.

SotL used meta wing for the chart and meta wing pulls data from listjuggler.

Since the FAQ, we only had 3 major event (major= 50+ people attending) afaik: 2 sos and 1 regional.

Of those event, listjuggler still miss the biggest one, tatooine.

This means that in the chart, there are many small events that, this I admit being just a supposition of mine, weighted similar to 2 events present in listjuggler.

We however have data from tatooine: I posted the top 8 and majorjuggler collected it in his topic.

So, how would the chart be if we only consider the top 8 from those 3 events (this is very limitating, I would much rather prefer to use 7-3 from tatooine, 6-3 from hoth and 4-2 from the regional, but unfortunately I don't have them)?

I suck at both maths and graphic, but I think that jumps would still be #1, Ks #2 and protectorate #3.

But Biggs (xwing) would definitely be #4 rather than lancers (which would be lower, with just 1 entry), followed by the ghost and the shuttle.

Defenders would be the "best" imperial ship, closely followed Tie/sf.

As I already said, I love SotL work. It might just be a bit too early to compile something like this chart thought

3 minutes ago, Voitek said:

Yet I havent heard a lot of "NERF R3-A2" shouts.

It's more compliated than that. Stresbot makes it more difficult for you to play in a short term: you get a stress or two, try to shake it, prioritise the stressY, hopefully kill it. Okay, problem solved,

While R2D2 will sometimes mean: you put 4 damage on corran horn, you are almost there, he runs away, regenerates... and you feel like a hard-worked victory just slipped out of your hands. I dont want to say "negative play experience", but I know people who would say that in similar situation.

Oh, believe me. I would love to lead a nerf campaign against the Stressbot. I just think they do more harm than good here.

Is the X-Wing at 5 only because of Bigg? Y-Wing's up there too. The closest imperial is 9. huh...

Just now, Stay On The Leader said:

Apart from me ;-)

R3-A2 is so much better at dealing out stress than the other options that when stress becomes a key part of the metagame it warps hugely around a couple of Rebel ships that can double-tap with R3-A2.

Correct, but I know you are generally a big fan of nerfhammer flying around like Mjolnir in Thor's hand

I have heard more complains about rebel regen in general (with no distinction to R2D2, R5P9, or Miranda), than about stressbots. Granted, Miranda recently became the dominant rebel ship, so that might be a factor as well.

1 minute ago, Ken at Sunrise said:

Is the X-Wing at 5 only because of Bigg? Y-Wing's up there too. The closest imperial is 9. huh...

Biggs is ~81% of all X-Wings recorded since the nerf.

And I still don´t see those who have claimed Imperials are OP in the recent threads to back their words up now!

Just now, Stay On The Leader said:

Biggs is ~81% of all X-Wings recorded since the nerf.

I wondered about that. So without Biggs where might the X-Wing rank? Either way it does see table time.

Just now, Sunitsa said:

Since the FAQ, we only had 3 major event (major= 50+ people attending) afaik: 2 sos and 1 regional.

Of those event, listjuggler still miss the biggest one, tatooine.

Tatooine is on Juggler last I checked.

20 minutes ago, Sunitsa said:

I see lots of abstract criticism in this thread but a suprising lack of example on why you are criticizing.

SotL used meta wing for the chart and meta wing pulls data from listjuggler.

Since the FAQ, we only had 3 major event (major= 50+ people attending) afaik: 2 sos and 1 regional.

Of those event, listjuggler still miss the biggest one, tatooine.

This means that in the chart, there are many small events that, this I admit being just a supposition of mine, weighted similar to 2 events present in listjuggler.

We however have data from tatooine: I posted the top 8 and majorjuggler collected it in his topic.

So, how would the chart be if we only consider the top 8 from those 3 events (this is very limitating, I would much rather prefer to use 7-3 from tatooine, 6-3 from hoth and 4-2 from the regional, but unfortunately I don't have them)?

I suck at both maths and graphic, but I think that jumps would still be #1, Ks #2 and protectorate #3.

But Biggs (xwing) would definitely be #4 rather than lancers (which would be lower, with just 1 entry), followed by the ghost and the shuttle.

Defenders would be the "best" imperial ship, closely followed Tie/sf.

As I already said, I love SotL work. It might just be a bit too early to compile something like this chart thought

+1

though tatooine is there with about 50% of the lists (which is more than naboo or Hoth)

I just think this was posted without any real thought towards the data behind it. It is very incomplete at best.

1 minute ago, Voitek said:

Hello sir, one hell of tool, that is. Can I have some few simple questions?
- does the model calculate weight based on tournament rank? Winning regio/open series will give ships more points than winning store champs?
- does the model calculates the number of players? ie, getting in top3 on 100 people tournament will give those ships more points than top3 on 16 people tournament?
- how is a mediocre to low result affecting the ships final grade, generally speaking? In a said example, list placed on 60th position on a 120 people tournament, will actually increase the ships final grade or reduce it? Example:
16 people on the tournamen
TOP1: list with Corran and miranda
TOP8: list with Corran
No more lists with those ships (or no data). Will Miranda be higher than Corran?

Certainly

• The model currently does not take tournament rank into account. You can filter by rank if you want, but there's no impact beyond that.
• Yes, number of players is taken into account by means of a multiplier of log(number of players) for each tournament's percentile rating of a squad in the magic value
• Yes, a mediocre result is affecting the ship's final grade in two ways:
  • It'll drag down the average percentile of the ship
  • It'll increase the multiplier for number of times the ship was used + 1 (because log(1) is 0, and a ship that was used once should still show up)
  • It'll thus increase or decrease the magic ranking

So in your example:

Corran has a percentile of 75% (once 100%, once 50%), and shows up twice. So its magic will be 0.75*log(2+1)=0.36
Miranda has a percentile of 100% and shows up once. So its magic will be 1*log(1+1)=0.30

So Corran will be slightly higher than Miranda.

It's all a bit fuzzy, thus the resulting value is called "magic". But when I tried out the idea and saw the results, they looked rather reasonable to me. The magic value is thus:

avg(percentile * log(tournament size))*log(number of times used + 1)

...and you can configure away both of these with the filter settings, and will then have a list ranked by percentile (which will be dominated by one-off lucky punches).

1 minute ago, Timathius said:

+1

thought tatooine is there with just the top 8 (which is more than hoth)

I just think this was posted without any real thought towards the data behind it. It is very incomplete at best.

Wow, I don't know what when into it but do you really think "this was posted without any real thought towards the data behind it". I thought this only represented the data. Perhaps it is a smaller subset of data than might be otherwise more useful. But still. That was rather personal don't you think?

I applaud the use of whatever data we have available, and this is our best and most current data. It may be incomplete, but the minimal and scattered information and rampant confirmation bias that will result from each individual's own observations are far more likely to lead to inaccuracy. We can always (and should) repeat this analysis in a month or three when the meta has settled. Until then, this is the absolute best we can do.

My only question is whether the ranking is of the popularity, percentile, or 'magic' figures.

Edit: Sorry, that was a bit sharply-worded.

Just now, Voitek said:

I have heard more complains about rebel regen in general (with no distinction to R2D2, R5P9, or Miranda), than about stressbots. Granted, Miranda recently became the dominant rebel ship, so that might be a factor as well.

I don't think there's much to be worried about with regen, people are misplacing their concerns. There's still so much alpha/ordnance stuff in the meta that Biggs meat shield is your best defence not regen - there's no Poe, little Corran, very little R2-D2...

There's not much 'Rebel Regen' - what there is, is Miranda. Miranda is good for many different reasons and the regen is obviously a huge part of it, but so is TLT and so is ordnance. The suite of nerfs I like (shared elsewhere) addressed that as well as R3-A2. It's a topic for another day, but yes I think R3-A2 is a problem for a healthy diverse metagame - we need more effective stress options not one super-effective stress option.

Just now, Stay On The Leader said:

I don't think there's much to be worried about with regen, people are misplacing their concerns. There's still so much alpha/ordnance stuff in the meta that Biggs meat shield is your best defence not regen - there's no Poe, little Corran, very little R2-D2...

For me at least, the alpha isn't the most fun to play. But when discussing stats and tournaments fun isn't as much a priority I suppose.

1 minute ago, Ken at Sunrise said:

Wow, I don't know what when into it but do you really think "this was posted without any real thought towards the data behind it". I thought this only represented the data. Perhaps it is a smaller subset of data than might be otherwise more useful. But still. That was rather personal don't you think?

It may have come off that way, but I think that if you look through the thread I have been very level headed in my criticism of how this has been presented. SoTL please don't take that comment personally, was not meant as a slight at all.