Guns of Icarus Online - Fantasy League

[Thomas]

The Idea
Piggy-backing off of the League play currently being developed discussed here:
https://gunsoficarus.com/community/forum/index.php/topic,3799.0.html

I think it might be fun to have a 'Fantasy League' to pair along with it. Similar to Football and Fantasy Football. With the highly organized and predictable nature of the proposed League, it should be pretty simple to get a Fantasy League running alongside it.

How This Works
It's based upon the same structure as any other fantasy team. We have a roster of players who will participate in the league, and those wishing to take part in the Fantasy League build their own 'team' from this list. The players receive points for their performance in League matches.

However, we are limited to actions that can be easily tracked, particularly from a streamed game. While it's possible to have players screen shot their end match performance stats and send it in, that's a bit of a hassle and also not an accurate representation of their performance.

For our purposes, we will divide players into two categories, 'Captains' and 'Crew'. The Crew players will earn points for any action that is reported by the combat log (upper left of the screen), which is parts destroyed and parts rebuilt.

Crew

• 1 Point per Gun Destruction/Rebuild
• 2 Points per Engine Destruction/Rebuild
• 3 Points per Balloon Destruction/Rebuild
• 4 Points per Hull Destruction (+1 if this is followed by the ship being destroyed)
• 5 Points per Hull Rebuild
• -2 Points per Disconnect (or divide total score in half if Fail to Return)

Captains will receive points based upon their crew performances, match outcome, and match performance.

Captains

• 50 Points per Ship Kill
• -20 Per Ship Death
• (Sum of Crew Score)/4 ((The average score of the crew))
• (Average Survival Time)/(Match Length)*50
• -10 per Disconnect (Divide Score by 4 if Fail to Return)
• Sum of End Game Score * 1.5 (if Victor)

How to Create a Roster

There's a few ways for participants in the Fantasy League to make their roster.

The easiest and probably most boring is for an All Pick, where you can compose your team of anyone, even if they've already been selected in someone else's team.

To keep it interesting and competitive, we can go with a Random Pick or Draft Pick. With a random pick, your team will be assigned to you, where with a Draft Pick, players take turns choosing from the Roster. Generally the turns will be based upon your performance in the previous Fantasy League. Also due to the time it would take, it's probably best to go with Random, at least for the initial Fantasy League.

Your Roster will be composed of 14 Players.
2- Captains
6- Crew
2- Replacement Captains
4- Replacement Crew

Only the score from 2 captains and 6 crew will be used each week. You can change your crew formation anytime before those weeks matches are played.

Replacements will be used when someone in your starting team does not participate in a given week.

For example if a team in Match 1 has one of your players, but that player couldn't make it. Your #1 replacement will have their score used instead. If your #1 replacement also didn't play, your #2 replacement will be used. If Match 4 happens and another one your picks is missing, your #3 replacement will be used in their stead. If you don't have enough replacements for the number of players missing, you will only get scored for the players that played. (ie: if only 7 of your 14 players played, you will only get points for those 7).


Trading

Part of the fun for Fantasy Teams is building up your team, which is where trading comes into play.

All trades must be completed before 1 AM Thursday (GMT) to be official. Both players must notify the organizer before this time (probably in a thread). Scores previously earned by the players are not transferred.

Once the Final Tournaments begin, there is no more trading allowed.

Prizes

Ideally there will be a prize for the player who earned the most Fantasy Points at the end of the season. This has yet to be discussed or determined (although I'm fairly certain we could get Muse to chip in costumes <3 ).


Discussion

The set up and proposed rules and scoring system are still in the drafting phases, so any input would be appreciated. What should change, what should stay, and why? Did we leave out anything important?

[Urz]

As an exercise, you should scrub through the first two weeks of Sky League VODs, transcribe the part logs, then decide if that's something you want to commit to doing every week indefinitely. Seems like it would be super tedious and not really the best use of anyone's time.

My recommendation would be to base it on K/D, or something equivalently quick to tally.

[Skrimskraw]

I agree with urz, you need to test this first.

the combat log lies, it always have. Sometimes I am credited for taking a hull down, even though it was my gunner etc.
also the combat log system will have your star players being those of the old playstyle, those who stand around and just repair stuff individually. Teams have moved so much and ship crews are now working in diffrent ways. - A great engineer that have become part of the meta is for example is the wrench buff chem who chemsprays and buffs your entire ship, but rarely rebuilds.

[Von Zincington]

To keep it interesting and competitive, we can go with a Random Pick or Draft Pick. With a random pick, your team will be assigned to you, where with a Draft Pick, players take turns choosing from the Roster. Generally the turns will be based upon your performance in the previous Fantasy League. Also due to the time it would take, it's probably best to go with Random, at least for the initial Fantasy League.

My one issue with doing it this way is if a person gets an absolute donkey of a team (they get a Blaine Gabbert, Trent Richardson and any of the Panther receiving corps) whereas another gets the bees knees (randomly given Tom Brady, Calvin Johnson and Adrian Peterson). If at the very least people get to make a 1st round pick and then get randomly distributed players through the rest of the team selection process.

This way you feel like you have some control and that you made a wise initial choice.

[redria]

Hmmm....
What I was sort of thinking was... do we all have to enter into the same fantasy league? Now, admittedly fantasy football has a bigger following than we ever will, but would there be a situation where a front office (so to speak) releases fantasy points to each player and we could have our own leagues running? Say if a clan wanted to play amongst themselves.
Just a thought, and we probably wouldn't have enough interested people to make it necessary anyway.

As far as how to collect data...
It all depends on what Muse can do to help. If we are independent, then anything more extensive than kills is going to be pretty difficult to track. We have only the ticker and the end report screen to go by, and only the players get to see their end report. It makes for a difficult task unless we have a lot of people working to collect data. It could certainly be done, just not easily.
The other alternative is if we get support from Muse. The easiest support would be if they had some sort of logs that they could send us from the matches which would basically give the ticker data in text format. More interesting would be some sort of in game support/tracker for the fantasy points a player gets, similar to how the bounty system works. Naturally this would take a lot more work and would take away time from introducing new content.

[Thomas]

Well the original idea was to have myself watch the videos and record the points. With the hard cap on the time limit and only having best of one matches, this should be a lot easier than trying to track matches with overtime or best of three's.

The biggest downsides that I foresee is that matches need to be recorded, and of a relatively decent quality. If it's super pixelated, it could be impossible to read the ticker. And if it's not recorded, there's no way to obtain the information. Disconnects would be harder to track if the casters hide the chat, but it shouldn't be too hard to ask them to bring up the scoreboard or such when this happens to show who disconnected.

The goal is to make it complicated enough not to be boring, and to differentiate the players from the ship as a whole. If it was just based upon the K/D, you'd be getting at least 8 players with the exact same point values every single week.



Von Zincington and Skrimskraw bring up some great points.
I love Zincington's compromise of having participants doing a partial pick and getting the remainder at random.
For the combat log, we'll just have to roll with it. As far as I'm aware it doesn't happen to often. For the points, they're easily adjusted.

A dry run is another great idea, and will help normalize the point system.

With the Sky League, there's about 18 teams that participated. Which means 9 matches, or up to 4.5 hours of videos. With a week between matches, that should be plenty of time to observe and update the points.


For the league, we could easily make the stats public for players to run their own fantasy leagues if they chose. For data collection we'll just be using the combat log in the upper left. However, getting Muse to help would be a great idea. I know they've been looking into more data collection, but had to put in a pin in that because of all the work being done for co-op and matchmaking. By the time the League is ready we might be able to work something out with Muse.

[N-Sunderland]

Honestly, I'd say the log is wrong about who destroyed components at least 20% of the time.

[redria]

Honestly, I'd say the log is wrong about who destroyed components at least 20% of the time.

I feel like that is mostly due to the odd way things work.
If I destroy my own engines with kerosene, the part destroy goes to the last person who actively damaged it -> either an enemy, nobody, or an ally with mines or tar.
Same goes for guns, hull, and balloon.
If you hit a wall and something is destroyed, the same goes into effect. An ally ramming you into the wall does not actively damage you, so whoever actively damaged you last gets the destroy.

I don't know that I have seen the log be wrong outside of that peculiarity.

[Imagine]

Honestly, I'd say the log is wrong about who destroyed components at least 20% of the time.

I feel like that is mostly due to the odd way things work.
If I destroy my own engines with kerosene, the part destroy goes to the last person who actively damaged it -> either an enemy, nobody, or an ally with mines or tar.
Same goes for guns, hull, and balloon.
If you hit a wall and something is destroyed, the same goes into effect. An ally ramming you into the wall does not actively damage you, so whoever actively damaged you last gets the destroy.

I don't know that I have seen the log be wrong outside of that peculiarity.

I had done testing for some of this stuff using mercs to knock out components, and sometimes it's just blatantly wrong. Like, I wouldn't fire a single shot (at most I'd be the one that loaded in ammo for it), and yet the log would still say I knocked out something even though someone else was shooting... but not always. It's rather bizarre.

[Skrimskraw]

how does pilots get points?

[Thomas]

Pilots are counted as Captains in the OP, since they're sometimes engineers and what not.


Captains

• 50 Points per Ship Kill
• -20 Per Ship Death
• (Sum of Crew Score)/4 ((The average score of the crew))
• (Average Survival Time)/(Match Length)*50
• -10 per Disconnect (Divide Score by 4 if Fail to Return)
• Sum of End Game Score * 1.5 (if Victor)

Really I'm just spitballing, especially for this part. We want to reward them for kills, punish them for deaths; but that doesn't give a whole lot. I know there's a lot of technical and strategic maneuvers a captain can pull, but these take some judgement to call, and have a decent chance of being missed during streaming. (Say they're watching two of the ships duke it out and miss an epic goomba stomp by the other ship).

Essentially we give them an average of their crew scores (crew does well, the captain does well). Then they'll get supplement points based upon their k/d and survival time. Chances are this will absolutely skew the point system to favor captains over crew, so we'll have to adjust it. Ideally I'd like to remove the average crew score from it, and just do it based upon the other factors, but we'll have to see how the dry run goes first.