Regionals Data Project 2018-2019 Season

[Truthiness 4,938]

It's that time of the year folks! Regionals Season starts this weekend. With the new season comes our annual data project, which attempts to collect lists and determine trends with the meta. It has gone a long way toward putting hard numbers behind problems. Throughout the season, you can go to this link and see what the data looks like:

https://docs.google.com/spreadsheets/d/14kjCf7_-KjTjAqMyUrXayZI8u9oBlqDyMLbYAMeaMxk/edit?usp=drivesdk

So what does that mean for you? First, I need lists. The more the better. No list, even last place, is unimportant because it all goes into the statistics. You can post lists to a thread, or message me, I don't care. The more detail the better. Second, I need to know what the community wants to look at. I'll be editing the formulas in the coming days to update from last season, but I only know what I think is important. I need to know what else to teach.

[AdmiralYor 205]

Is there any way you can read the wreckage of the defeated fleets to get a portent of Armada's future, or at least when the next wave will be announced?

[geek19 5,005]

5 hours ago, Truthiness said:

It has gone a long way toward putting hard numbers behind problems.

Do you have any data to support this?

[xanderf 5,991]

Not sure how easy it would be to do, but considering the overall event size is probably worth doing.

IE., it's a lot easier to get a list into the 'top 8' at a 12-person regional than it is a 32-person regional...

(Maybe instead of 'top 4' or 'top 8' groupings, have it at 'top 20%' or 'top 50%' or something like that?)

Edited November 29, 2018 by xanderf

[geek19 5,005]

Also, squad points and commanders seem obvious. I'm personally curious about "less-loved" squadrons, like basic e-wings, Lancers, or Scurrgs for example.

Same with ships. How many people are still using AFs? Does anyone fly Vic's anymore? How much are I-1s used?

[GiledPallaeon 2,465]

2 minutes ago, geek19 said:

 How much are I-1s used?

Working on making that not zero, maybe. I dunno, turbolaser, or DR and Ordnance? It's really hard to justify frequently.

[xanderf 5,991]

6 minutes ago, geek19 said:

Also, squad points and commanders seem obvious. I'm personally curious about "less-loved" squadrons, like basic e-wings, Lancers, or Scurrgs for example.

Squad points and commanders both seem to be there, already.

I'd be keen on a breakdown of 'scatter aces' vs the rest, though.  Even the brace aces seem slightly undercosted for their return on investment (vs the corresponding generic), but 'scatter aces' are crazy cheap compared to what they should be (relative to their corresponding generic) - I bet we'd see that in the numbers if it was tracked.

[GiledPallaeon 2,465]

@Crabbok You are welcome to send Orlando’s lists back with me if you want. It’s no big deal to me.

[Truthiness 4,938]

3 hours ago, xanderf said:

Not sure how easy it would be to do, but considering the overall event size is probably worth doing.

IE., it's a lot easier to get a list into the 'top 8' at a 12-person regional than it is a 32-person regional...

(Maybe instead of 'top 4' or 'top 8' groupings, have it at 'top 20%' or 'top 50%' or something like that?)

I might have an idea on how to do it.

[Crabbok 9,453]

If you're putting lists together for science... I really should run something ridiculous so we can go back and say "See... 0.002% of players ran Ackbar Nebulons at Regionals - it's a real number!"

3 hours ago, GiledPallaeon said:

@Crabbok You are welcome to send Orlando’s lists back with me if you want. It’s no big deal to me.

 

Ummm, sure.  I wasn't actually planning on collecting anyone's lists in the first place.  Like, I may do some video of like "Hi, wow look at all those ISDs!" but I'm not actually running the regional or anything, I'm just there to play and geek out with fellow geeks!

Edited November 29, 2018 by Crabbok

[Truthiness 4,938]

15 minutes ago, Truthiness said:

I might have an idea on how to do it.

Alright, I've got a solid concept, but it's going to involve some manual input. For my sanity, I'm also going to need it to be one of the earlier columns in the full lists tab. Also for my sanity that means I have to obliterate one of said early columns so that I don't have to re-do every single formula. Any objections to me replacing the MoV column? Tournament Points alone is probably more than enough for our curiosity.

[xanderf 5,991]

Wait, Google sheets doesn't automatically adjust all formulas when you do a column insert?

REALLY?!

Excel has been able to do that since version...geez, I dunno, 97 or so....

[Ginkapo 7,926]

16 minutes ago, xanderf said:

Wait, Google sheets doesn't automatically adjust all formulas when you do a column insert?

REALLY?!

Excel has been able to do that since version...geez, I dunno, 97 or so....

One is free and web based. The other is microsoft. 

So on paper a fair competition but in reality not. 

[xanderf 5,991]

8 minutes ago, Ginkapo said:

One is free and web based. The other is microsoft. 

So on paper a fair competition but in reality not. 

Ooof, I'd try copy+paste out into Excel and then back again or something.  That fries my brain, I just can't even imagine trying to work with that limitation...

[Thrindal 946]

I will be TO for Derry NH on January 12th so I will have all the lists.  I will post them for you.

[Truthiness 4,938]

2 hours ago, xanderf said:

Wait, Google sheets doesn't automatically adjust all formulas when you do a column insert?

REALLY?!

Excel has been able to do that since version...geez, I dunno, 97 or so....

The first page is pulling from a second sheet, which means you have to be very specific about what column you call from. Even Excel doesn't update when you do that. Sheets still updates formulas within the same sheet, just like Excel. The formulas I've created are labor intensive when I make them, but they run themselves as the season goes on. It's a massive improvement over the last incarnation of the data project.

Edited November 30, 2018 by Truthiness

[Truthiness 4,938]

The Regionals for the weekend appear to be the following:

Independence, Missouri

Orlando, Florila

Brisbane, Australia

Good luck to everyone playing!

[xanderf 5,991]

11 minutes ago, Truthiness said:

The first page is pulling from a second sheet, which means you have to be very specific about what column you call from. Even Excel doesn't update when you do that. Sheets still updates formulas within the same sheet, just like Excel. The formulas I've created are labor intensive when I make them, but they run themselves as the season goes on. It's a massive improvement over the last incarnation of the data project.

FWIW, Excel actually does still handle that cross-sheets.  Especially if the content in a sheet is defined as a "table", in which case the columns are identified by name rather than position anyway - and you can even move them across documents and it can keep the relationship.

Anyway, that's neither here nor there - this isn't Excel, and it just makes it all the impressive how much of a marvel of work you are doing for the community!  Thanks a million for this - the game unquestionably benefits from it!

(Uhhhmmm…any way to get a track line in for those scatter aces as a squadron type?  😉 )

[Truthiness 4,938]

27 minutes ago, xanderf said:

(Uhhhmmm…any way to get a track line in for those scatter aces as a squadron type?  😉 )

Right now I have the ability to pick out the Scatter Aces. I could compare the percentage of lists that include X number of Scatter Aces, just like with ship sizes. If you're looking at normal vs. Scatter numbers, that might be a little harder, but possibly doable.

Edited November 30, 2018 by Truthiness

[xanderf 5,991]

1 minute ago, Truthiness said:

Right now I have the ability to pick out the Scatter Aces. I could compare the percentage of lists that include X number of Scatter Aces, just like with ship sizes. If you're looking at normal vs. Scatter numbers, that might be a little harder, but possibly doable.

Hmmm.  Yeah, wonder what the most conclusive analysis would be.

Certainly the overall impression is that scatter aces are worth more than their cost lets on, compared to their generic counterparts.  But would the best way to look at that in a tournament format be just overall comparisons?  IE., lists with with max squads and 50% scatter aces place so high, where max squads and 10% scatter aces do something else?

That is - the question is probably best phrased as percent of squads taken as scatter aces?  Or...actually...maybe even better...

Percent of squad limit (134 pts) that is scatter aces?  THAT might be illuminating...  (after all, someone could bring just two squads and both are scatter aces - if that list does well or poorly, does it really tell us much?  Or...maybe it does, actually...hrm…  No, wait, maybe it works this way, too.  Say 36 pts in scatter aces is 27% of max squads...comparing a list with the same cost in squads but no scatter aces at 0% then, if they trend in sharply different directions...)

[geek19 5,005]

4 hours ago, xanderf said:

Hmmm.  Yeah, wonder what the most conclusive analysis would be.

Certainly the overall impression is that scatter aces are worth more than their cost lets on, compared to their generic counterparts.  But would the best way to look at that in a tournament format be just overall comparisons?  IE., lists with with max squads and 50% scatter aces place so high, where max squads and 10% scatter aces do something else?

That is - the question is probably best phrased as percent of squads taken as scatter aces?  Or...actually...maybe even better...

Percent of squad limit (134 pts) that is scatter aces?  THAT might be illuminating...  (after all, someone could bring just two squads and both are scatter aces - if that list does well or poorly, does it really tell us much?  Or...maybe it does, actually...hrm…  No, wait, maybe it works this way, too.  Say 36 pts in scatter aces is 27% of max squads...comparing a list with the same cost in squads but no scatter aces at 0% then, if they trend in sharply different directions...)

I assume this is primarily Imp focused? I can see Shara/Tycho, but that's a basic Rebel dump of squadrons that doesn't need much more support is all.

[geek19 5,005]

How hard would it be to add in a thing for each ship? Not just X large or whatever, but how many CR90s, how many Assault Frigates, how many Nebulons Yavari? While knowing the number of ships is interesting and number of larges/mediums, i think you'll get more traction out of THIS data personally.

[xanderf 5,991]

19 minutes ago, geek19 said:

I assume this is primarily Imp focused? I can see Shara/Tycho, but that's a basic Rebel dump of squadrons that doesn't need much more support is all.

At the moment, yeah, obviously - although don't knock Blount.

Still, that 'don't need much support is all' speaks to the point of the issue.  I mean, as that example - does anyone seriously think the Shara + Tycho is worth only 3 standard A-Wings?  Good grief, they'll easily take double the number of attacks to kill as the entire other group - and that's just the defensive value of the tokens, nevermind the pilot's ability.

Certainly, the same is true (and worse) for the Imperials - you sneeze at a TIE Fighter and it explodes, while it takes 3 powerful attacks to take out Howl (enough that you get at least a couple accuracies in there).  Yet she's only double the cost, nevermind her pilot ability.

The issue is less pronounced with the brace aces, as they don't take an average of as many attacks to kill, but scatter aces are quite the durability value for their price.  At least, this certainly seems anecdotally true of games, and a casual review of historical games, but I think we could pretty easily put hard numbers to it to quantify more clearly.

[Truthiness 4,938]

5 hours ago, geek19 said:

How hard would it be to add in a thing for each ship? Not just X large or whatever, but how many CR90s, how many Assault Frigates, how many Nebulons Yavari? While knowing the number of ships is interesting and number of larges/mediums, i think you'll get more traction out of THIS data personally.

It's easy to compare how many fleets included one or more of a ship. The formula for the ship sizes is done by checking against a list of ships, so making a formula that just counts one specific ship type is easy. How would this be compared? Are you looking for the raw number of each ship taken, or the percentage of fleets that take at least one of a ship type?

Edited November 30, 2018 by Truthiness

[Truthiness 4,938]

10 hours ago, xanderf said:

Hmmm.  Yeah, wonder what the most conclusive analysis would be.

Certainly the overall impression is that scatter aces are worth more than their cost lets on, compared to their generic counterparts.  But would the best way to look at that in a tournament format be just overall comparisons?  IE., lists with with max squads and 50% scatter aces place so high, where max squads and 10% scatter aces do something else?

That is - the question is probably best phrased as percent of squads taken as scatter aces?  Or...actually...maybe even better...

Percent of squad limit (134 pts) that is scatter aces?  THAT might be illuminating...  (after all, someone could bring just two squads and both are scatter aces - if that list does well or poorly, does it really tell us much?  Or...maybe it does, actually...hrm…  No, wait, maybe it works this way, too.  Say 36 pts in scatter aces is 27% of max squads...comparing a list with the same cost in squads but no scatter aces at 0% then, if they trend in sharply different directions...)

Ooof. That might be possible, but I'm not sure I'm skilled enough to pull it off.