So where were the guild players at adepticon?

[4thstringer]

1 minute ago, Franchute said:

We dont want these numbers to tell us something about the best players. They have to be useful for the average player.

I think top 10 at Nova is very good.

My friend just corrected me and said it wasn't top 10.  It was top third.

[RustAndTheCity]

I can see that some masters can have a beginner bias in the data. 

 

For example, Rasputina has a low win rate in logfaux. One of the best players in my area played Rasputina for the last several months. He only lost 2 and tied 3-5 games out of 35 games total. In the hands of an expert player she is very powerful. But she's perceived as a.good player for a new master, which may skew down her overall results. There's also the issue that he didn't log results. If he had, her results would be higher because of the small sample size.

 

I guess logfaux is an imperfect data set. There is bias in who reports games. There is also bias in player skill and who chooses what master. As the data set becomes larger the results will be more valid. 

 

With the raw data and a lot of time, a statistician could use the data to control for player skill if there was enough data.

[4thstringer]

32 minutes ago, RustAndTheCity said:

I can see that some masters can have a beginner bias in the data. 

 

For example, Rasputina has a low win rate in logfaux. One of the best players in my area played Rasputina for the last several months. He only lost 2 and tied 3-5 games out of 35 games total. In the hands of an expert player she is very powerful. But she's perceived as a.good player for a new master, which may skew down her overall results. There's also the issue that he didn't log results. If he had, her results would be higher because of the small sample size.

 

I guess logfaux is an imperfect data set. There is bias in who reports games. There is also bias in player skill and who chooses what master. As the data set becomes larger the results will be more valid. 

 

With the raw data and a lot of time, a statistician could use the data to control for player skill if there was enough data.

Even then, if you have players that play their weaker masters into less competitive players, that could still mess things up.

.

[RustAndTheCity]

10 minutes ago, 4thstringer said:

Even then, if you have players that play their weaker masters into less competitive players, that could still mess things up.

.

That's very true. I painted up lady J with the idea she'd be more beginner friendly for opponents than McMourning. So I'm 5-0 with her so far. I'm not sure if I would take her in bad matchups against strong opponents with her.

[iamfanboy]

Maybe we should try to make Logfaux a more Henchman motivated thing? Most Malifaux communities have one embedded, having good data to recommend Masters to people would be appreciated, and - okay, it would be more extra work on top of all the other stuff a Henchman is doing to get soulstones, but overall might be seen as worth it? Just taking notes at events a Henchman runs (Guild McMourning 2/1/1) and entering them later wouldn't be too hard.

I think.

[trikk]

That actually requires storing a ton of data. Selected masters, selected schemes etc.

[iamfanboy]

12 minutes ago, trikk said:

That actually requires storing a ton of data. Selected masters, selected schemes etc.

A fair point, but maybe logfaux tracks too much data if that's the case - do we want to know - "Okay this master loses if FFM and Spring the Trap are on the table" - or do we just want more general data - "Guild McMourning is up, Guild Sonnia is down"?

 

The idea of weaker/stronger players messing up the data, though, is harder. I know X-wing list juggler solves it by having two data entry points - between the 'entry' lists and the lists that actually make a cut in tournaments, Top 4/8/16 - so you can easily see what's rising to the top among a field of potential casuals. For example, the only T65 X-Wing that matters for top cuts is Biggs but lots of noobs bring Wedge, Luke, and other EU characters to the table.

But I think that it's just the noise that we'd have to accept if we want any data at all on what masters are being used successfully or not.

Or we could just, ya know, play whatever. I think defining what's good/bad is nearly impossible for a game like Malifaux because it's so situational and each player can choose NOT to bring a certain Master into a bad scheme/strat set. Compared to WM/H, where it's basically one game - "Capture the Flag and/or Kill the Enemy Boss" - Malifaux's games are all over the place and frankly, I think that data would be almost uselessly noisy in the general game. It's nearly unsolvable even if there were only 20 different models in the game TOTAL, let alone in the 200-300 range!

Henchman Hardcore would be different, because the number of optimal solutions is smaller, but even GG17 doesn't look easy to solve, and filtering the data would be just... useless IMHO.

EDIT: Example, there is GOOD, and there is BETTER, in a game like Malifaux (Gupps are best Scheme Runners in Neverborn, but Silurids/Terror Tots/Insidious Madnesses are also valid choices in certain pools). Sonnia might be bad AT certain Schemes, but that does not make her a BAD Master - and data which showed her down in successful usage would be noisy because of how variable the game itself is.

[trikk]

I think it would actually be better to use the UK/USA/POL rankings and add master/scheme selection to the uploaded CSV.

[4thstringer]

I like that logfaux isn't limited to top 4 of major tournaments.  I wanna know how they are doing across skill levels because the game balance isn't just for the tournament players.  It affects my casual group too. Balance issues really effect peoples enjoyment of the game at all levels. 

[trikk]

Yes, but unless it has a big csv upload for tournaments I doubt I`ll spend an evening to put the results in and I think they are a bit more valuable data than single games.

[tactician]

12 hours ago, Franchute said:

Anyway, I do think logfaux is a great tool for the community. We should all keep on uploading our games.

If necessary, I offer myself to work a bit on the data to clean up the numbers for possible "player fixed effects" once more data will be gathered (in case @tactician would be interested).

Yes, I would definitely be interested! As I've mentioned here, opening the data to the public is on my to-do list.

[Tris]

Guys, do you know that you can filter games in logfaux for tournament/casual games?

So, where's the problem with casual games skewing the data?

[Franchute]

4 hours ago, Tris said:

Guys, do you know that you can filter games in logfaux for tournament/casual games?

So, where's the problem with casual games skewing the data?

I've heard in some podcasts people complaining about the difficulty to upload tournament games. As a result, they simply upload them as casual games.

[Tris]

22 minutes ago, Franchute said:

I've heard in some podcasts people complaining about the difficulty to upload tournament games. As a result, they simply upload them as casual games.

Don`t know, is there any difference?

I upload all games from our monthly tournaments and have no issues - EDIT: in fact, I double checked it right now, there is literally no differnece between reporting a casual or tournament game.

 

So,if the data is added wrong it`s imo not right to complain.......

 

Any TOs around here having issues with logfaux?

If so, what are the problems you`re facing?

Anything @tactician should know?

 

It`s our responsibility to add the data, as tournament games are arguably more important for the meta, so we should try to do our best if we want the data to be useful

[RustAndTheCity]

I think the issue is that (as far as I know) the tournament organizer has to upload the tournament results (or someone else). I don't see at the moment that an individual can upload their games as a tournament result.

[tactician]

10 minutes ago, RustAndTheCity said:

I think the issue is that (as far as I know) the tournament organizer has to upload the tournament results (or someone else). I don't see at the moment that an individual can upload their games as a tournament result.

What you describe is the way it's encouraged to be. However, it's not the way it has to be.

You are free to create a placeholder tournament and add all of your games to it. In fact, if the TO of the events you attend will neither submit games to Logfaux nor share the data, just go ahead and do it.

[Fetid Strumpet]

My mastery of programming is limited to basic self taught skills in Dreamweaver, but given that bag o tools seems to be becoming more popular as a tournament software is there anyway to collaborate with the developers of that app to just give an option to automatically generate a report at the end of an event that can just be uploaded to logfaux to record the event results?

[tactician]

1 minute ago, Fetid Strumpet said:

My mastery of programming is limited to basic self taught skills in Dreamweaver, but given that bag o tools seems to be becoming more popular as a tournament software is there anyway to collaborate with the developers of that app to just give an option to automatically generate a report at the end of an event that can just be uploaded to logfaux to record the event results?

As much as I'd love that, it can't be done without a re-design in one of the tools. The reason is that Logfaux requires the Master to be specified and Bag-o-Tools doesn't record Masters.

[Tris]

So, the 'issues' are lazyness?

Then nobody should wonder why there's not more data on tournament games....

[Franchute]

39 minutes ago, Tris said:

So, the 'issues' are lazyness?

This is a fine excuse. Malifaux is a hobby, not a profession.

[RustAndTheCity]

1 hour ago, tactician said:

What you describe is the way it's encouraged to be. However, it's not the way it has to be.

You are free to create a placeholder tournament and add all of your games to it. In fact, if the TO of the events you attend will neither submit games to Logfaux nor share the data, just go ahead and do it.

Thanks for the reply. I've been logging all of my games with logfaux. I think it's a great tool. I really like how easily I can record my own games. I definitely don't want to seem critical of the service! I can just see a few things that could potentially allow for more data to be available.

As a tournament organizer, I admit that it is tough to take the time to upload results in another way after an event. 

The formats for results of different programs don't record specific points for strategies and schemes either. And so adding it afterwards is guesswork for the most part unless I record it all separately.

Maybe a way to get more tournament results would be a way for TOs to input only player, master, and score for each game instead of individually adding scores for each Strat and scheme.

 

For individual players, it would be nice to have a pulldown menu to select from tournament, casual or league game. It would also be nice if they could submit results for a tournament or league. I know that is more complicated though.

 

Again, I really like logfaux and appreciate all the work that must go in to it. Just trying to think of ways to make it easier to add more data into it!

[Tris]

35 minutes ago, Franchute said:

This is a fine excuse. Malifaux is a hobby, not a profession.

Sure, I know that TOing is a lot of work, not paid for and I'm glad that people actually do that

Imo, it's just no good to complain about logfaux when the issue is that people aren't submitting the results properly.

It's not that hard and as tactician said, every player could proxy a tournament where he submits all his/her tournament games (if the TO doesn't)

 

I'm fully aware that's submitting the results is an additional step of work, but either we want a better data base, or not

 

As to how to get the data at the tournament - I hand out spreadsheets at the event, with space for the chosen strat/schemes, played faction and master and sort through these usually the week after the event has taken place

[tactician]

5 minutes ago, RustAndTheCity said:

I definitely don't want to seem critical of the service!

You don't. Besides, "critical" is not a bad word in my book.

8 minutes ago, RustAndTheCity said:

As a tournament organizer, I admit that it is tough to take the time to upload results in another way after an event.

I know. I have issued an open call for tournament game data, new and old. Just send them to me and I'll take care of the hard work.

11 minutes ago, RustAndTheCity said:

For individual players, it would be nice to have a pulldown menu to select from tournament, casual or league game.

You are right, it would. But, as a designer, I would be painting myself in a corner doing that. The reason being that not all tournament games would be attached to a tournament. I would then have to differentiate between those that do and those that don't and expose the difference in the UI, too. Before I bore everyone any further, it doesn't fit the long-term design of Logfaux. Besides, the work-around I mentioned in the previous post requires minimal effort.

[tactician]

59 minutes ago, RustAndTheCity said:

Maybe a way to get more tournament results would be a way for TOs to input only player, master, and score for each game instead of individually adding scores for each Strat and scheme.

That's the way it works now! Strategy and scheme VPs are not required.