30 minutes ago, Cuz05 said:

 I like to fly what I like. I do like effective things but it's more about making effective decisions with whatever it is you've put on the table.

The secret is its possible to think logically about....meta, trends, good vs bad ships, good vs bad in game decisions, and whatever ships you enjoy. Thinking or talking about one doesn't prevent anyone from thinking about other stuff.

*Also Wedge is bad.

23 minutes ago, evcameron said:

So don't stop sharing your work @Tlfj200!

Please do...

Oh don't encourage his threats, he was going to be insufferable with stats no matter what

59 minutes ago, Brunas said:

Facts. We're so bad at stats have developers have started lying to us about their rng so we'll stop complaining it's broken

https://serenesforest.net/general/true-hit/

XCOM doesn't do this. How many times have you heard that xcom is broken because someone keeps missing 90%+ shots?

I used to play a lot of Blood Bowl. The computer/console version is a literal translation of the board game, complete with 6-sided dice. The game, however, displays success chances as percentages. Now, most people in here could probably figure out that a 66.7% chance of success is "roll a 3+" and an 83.3% chance is "roll a 2+," but apparently that's not true of people who played that game and there were a lot of people on the forums complaining that the odds calculator was off... ?

Just now, skotothalamos said:

I used to play a lot of Blood Bowl. The computer/console version is a literal translation of the board game, complete with 6-sided dice. The game, however, displays success chances as percentages. Now, most people in here could probably figure out that a 66.7% chance of success is "roll a 3+" and an 83.3% chance is "roll a 2+," but apparently that's not true of people who played that game and there were a lot of people on the forums complaining that the odds calculator was off... ?

ALL SHALL KNOW LIL SKITTLES AND HIS 2 SKULL INTO 2SKULL AURA

31 minutes ago, Cuz05 said:

Loving @jesper_h 's view. A page ago I was pondering how almost every game of 2.0 I've seen and played, was decided by decisions on the table, rather than ships and dice.

I have a lot of views about a lot of things. Many of them are hypocritical, at best. But thanks.

1 hour ago, Brunas said:

Facts. We're so bad at stats have developers have started lying to us about their rng so we'll stop complaining it's broken

https://serenesforest.net/general/true-hit/

XCOM doesn't do this. How many times have you heard that xcom is broken because someone keeps missing 90%+ shots?

https://www.reddit.com/r/XCOM2/comments/45u81x/yes_xcom_2s_rng_cheats_in_your_favor_heres_how/

If you're bad, it does. (This is also probably why so many people complain, because on lower difficulties they ace all those 65% shots and then up to commander/legendary and suddenly don't know why they only hit 65% shots 65% of the time)

1 minute ago, svelok said:

https://www.reddit.com/r/XCOM2/comments/45u81x/yes_xcom_2s_rng_cheats_in_your_favor_heres_how/

If you're bad, it does. (This is also probably why so many people complain, because on lower difficulties they ace all those 65% shots and then up to commander/legendary and suddenly don't know why they only hit 65% shots 65% of the time)

Oh man, I hadn't played/seen XCOM2 - this is amazing.

8 minutes ago, Brunas said:

Oh man, I hadn't played/seen XCOM2 - this is amazing.

I want to make alt rules now in X-Wing that incorporate streaks. Like a free focus token next turn if your shot misses.

2 hours ago, Kaptin Krunch said:

Say boomowl had lost to Holt at Coruscant (meme format aside), Holt advances to top 16 at minimum. ****, say he somehow wins his top 16, and his top8 is Chumby who concedes for the memes? Top 4 somehow

The dream...

2 hours ago, gennataos said:

I wish there was a meaningful way to discuss individual players, their approach/choices/etc, more in depth. I find that infinitely more interesting that list breakdowns.

Well we can't at the moment just due to low sample size. But assuming 1.0 levels of data collection there's no reason we couldn't (and further no reason we shouldn't) model in player performance. As a whole we've generally pruned the data by looking at cuts only, major events only, or both. The assumption being that if it made the cut it was flown by someone decent (or the list is just that good) and people are actually trying and maximizing logical self interest by playing the best list at major events. By narrowing our focus to just what lists are played and stipulating that all else is equal it simplifies things quite a bit. But I'm not sure those are good assumptions. Bunn or Duncan playing a particular list and losing badly with it is just as much if not a more valuable data point than -=darthjank69=- rolling natties all day and making the cut. Blaire is going to be pushed towards running Zs due to personal preference regardless beyond the degree the meta would indicate.

Who's flying a list and how good they are makes a massive difference but to this point it's largely been absent in any analysis I've seen

10 minutes ago, Makaze said:

Well we can't at the moment just due to low sample size. But assuming 1.0 levels of data collection there's no reason we couldn't (and further no reason we shouldn't) model in player performance. As a whole we've generally pruned the data by looking at cuts only, major events only, or both. The assumption being that if it made the cut it was flown by someone decent (or the list is just that good) and people are actually trying and maximizing logical self interest by playing the best list at major events. By narrowing our focus to just what lists are played and stipulating that all else is equal it simplifies things quite a bit. But I'm not sure those are good assumptions. Bunn or Duncan playing a particular list and losing badly with it is just as much if not a more valuable data point than -=darthjank69=- rolling natties all day and making the cut. Blaire is going to be pushed towards running Zs due to personal preference regardless beyond the degree the meta would indicate.

Who's flying a list and how good they are makes a massive difference but to this point it's largely been absent in any analysis I've seen

I swear, every couple days I go through the exact same thought process:

"Whoa Rebel Zs are killing it"

"...dammit Blair"

1 minute ago, Brunas said:

I swear, every couple days I go through the exact same thought process:

"Whoa Rebel Zs are killing it"

"...dammit Blair"

Rebel Zs are actually pretty solid.

Talas w/ Clusters are a nice swarm option

1 minute ago, Do I need a Username said:

Rebel Zs are actually pretty solid.

Talas w/ Clusters are a nice swarm option

15 minutes ago, Makaze said:

Well we can't at the moment just due to low sample size. But assuming 1.0 levels of data collection there's no reason we couldn't (and further no reason we shouldn't) model in player performance. As a whole we've generally pruned the data by looking at cuts only, major events only, or both. The assumption being that if it made the cut it was flown by someone decent (or the list is just that good) and people are actually trying and maximizing logical self interest by playing the best list at major events. By narrowing our focus to just what lists are played and stipulating that all else is equal it simplifies things quite a bit. But I'm not sure those are good assumptions. Bunn or Duncan playing a particular list and losing badly with it is just as much if not a more valuable data point than -=darthjank69=- rolling natties all day and making the cut. Blaire is going to be pushed towards running Zs due to personal preference regardless beyond the degree the meta would indicate.

Who's flying a list and how good they are makes a massive difference but to this point it's largely been absent in any analysis I've seen

Who is flying what is definitely a data point people can gather. That's not what I'm looking for, though. I'm looking for players talking through their decision process in a game. We see theory in occassional, brief FFG guest articles and @Biophysical's blog posts, but the closest I've seen are the videos @Musical Xeno was doing with Gold Squadron for a bit. Bring that back!

Just now, gennataos said:

Who is flying what is definitely a data point people can gather. That's not what I'm looking for, though. I'm looking for players talking through their decision process in a game. We see theory in occassional, brief FFG guest articles and @Biophysical's blog posts, but the closest I've seen are the videos @Musical Xeno

Well that takes a lot of effort, is difficult to quantify, and only a small fraction of games are streamed so not all that useful for doing stats with. Not that all that stuff isn't informative, interesting, and nifty but if it isn't going to result in a concrete numerical representation of how good/bad Wedge is then pffft why bother?

1 minute ago, gennataos said:

Who is flying what is definitely a data point people can gather. That's not what I'm looking for, though. I'm looking for players talking through their decision process in a game. We see theory in occassional, brief FFG guest articles and @Biophysical's blog posts, but the closest I've seen are the videos @Musical Xeno was doing with Gold Squadron for a bit. Bring that back!

I'm willing to do a pilot project if I have:

1. A video of a match to go over (I've spent a few minutes today thinking about re-capturing recorded stream games)
2. Somebody far more interesting and smarter than I. I'm rather dull and boring unless I have somebody to play off of, and further I'm mediocre at best on strategies. Not saying I have to remain silent and make everybody else do the dirty work, I just don't expect anybody would WANT to hear me babble through idiocy.

I'm willing to put time into a video, do post-work and cuts, and whatever, and post the final result on youtube.

9 minutes ago, Do I need a Username said:

Rebel Zs are actually pretty solid.

Talas w/ Clusters are a nice swarm option

Scum Zs vs. Rebel Zs... sure not that different

But Scum Zs plus Drea/Qauds vs Rebel Zs plus um... yeah... Leia...?

Just now, LagJanson said:

I'm willing to do a pilot project if I have:

1. A video of a match to go over (I've spent a few minutes today thinking about re-capturing recorded stream games)
2. Somebody far more interesting and smarter than I. I'm rather dull and boring unless I have somebody to play off of, and further I'm mediocre at best on strategies. Not saying I have to remain silent and make everybody else do the dirty work, I just don't expect anybody would WANT to hear me babble through idiocy.

I'm willing to put time into a video, do post-work and cuts, and whatever, and post the final result on youtube.

I'm full of spicy opinions if you need a color commentator.

If anything, the GSP thing where players live commentated their own games was helpful. I can do that.

8 minutes ago, Makaze said:

Well that takes a lot of effort, is difficult to quantify, and only a small fraction of games are streamed so not all that useful for doing stats with. Not that all that stuff isn't informative, interesting, and nifty but if it isn't going to result in a concrete numerical representation of how good/bad Wedge is then pffft why bother?

Hah, true dat.

I know it's a pipe dream, it's just what I find most interesting...how people actually play their list. Like, the stuff we go over with our opponents after/during a game. When we're actually practicing, our local group talks through our games and our decision process, sometimes during decisions, sometimes after the fact. We should probably do it more often than we do, certainly after the fact.

6 minutes ago, LagJanson said:

I'm willing to do a pilot project if I have:

1. A video of a match to go over (I've spent a few minutes today thinking about re-capturing recorded stream games)
2. Somebody far more interesting and smarter than I. I'm rather dull and boring unless I have somebody to play off of, and further I'm mediocre at best on strategies. Not saying I have to remain silent and make everybody else do the dirty work, I just don't expect anybody would WANT to hear me babble through idiocy.

I'm willing to put time into a video, do post-work and cuts, and whatever, and post the final result on youtube.

I think to arrive at what I'm talking about, you'd have to find the participants of a given streamed game and ask them to do an after-action commentary.

I went back to my 1.0 stuff because of this discussion, and now I'm a bit sad that I never finished the write-up.

Circle size corresponds to the ratio between number of ships in cut and swiss, while y-axis is the amount of a pilot in swiss (e.g. Quickdraw was played 290 times in total) and the x-axis is the overall win percentage (including of those that didn't make the cut). The red line marks the (weighted) mean win percentage of pilots that made the cut (as opposed to all used), which was at just below 58%. So pilots to the right were better than the average who made the cut, and ships to the left were worse. This should mean that the ships on the right advanced more often into the higher placements. Indeed, the 18 pilots that are found on the right placed 8th on average. And the two ships with the lowest win percentage also placed worst.

With this I'd have identified Jess Pava, Omicron and Kanan Jarrus as the top3 sleeper hits: (same circles and red line, but new y-axis which is now the placement in the cut). Yorr has a very high win percentage but also not such a high placement, same as Inquisitor.

With this analysis I could have identified and guessed that PalpAces with Yorr/Omicron and Inquisitor were actually a good choice at the end of 1.0 and if I remember correctly they really were making their comeback. Now of course plenty people realized this without going through this detour. But my point here is simply that enough data can actually be useful to identify pilots with high potential.

I'm super interested in, and if I ever quit my current job may get around to trying it, is doing some AR type things with ships. It would require some markers on the mat/ships but that would let you capture movement, firing arcs, heat maps, obstacle placement, ect from the streamed game

Maybe the Mynocks "special" stream will do something along these lines

A while ago I was doing some analysis of interesting X-Wing matches from Youtube:

https://xwingtube.wordpress.com/

The focus was on what was the sequence of playes, which were the key decisions, and what types of strategies and tactics were employed by the players.

I hope it could have been useful for newer players, although I primarily did it as an exercise in analysis and writing english language

5 minutes ago, gennataos said:

I think to arrive at what I'm talking about, you'd have to find the participants of a given streamed game and ask them to do an after-action commentary.

I think the difference is you're looking for a sports-type post-game analysis, where I'm thinking more documentary type discussion.

Just now, LagJanson said:

I think the difference is you're looking for a sports-type post-game analysis, where I'm thinking more documentary type discussion.

sports is a documentary.

1 minute ago, Kaptin Krunch said:

sports is a documentary.

It can be, yes.