Community > Community Events

Competitive Concept - Single Ship Teams

<< < (2/5) > >>

Skrimskraw:
https://gunsoficarus.com/community/forum/index.php/topic,5667.0.html

? ;)

ZnC:
@Thomas: The group stage round robin I'm talking about is, as you mentioned, similar to a bracket. Those that advance the group, get into a group with others that advance their groups. Those that don't are eliminated. I still think 2v2s are the way to go, as they are much more playable than 4v4 chaos.

To compensate for odd-numbered participants, revive a ship when a group is lacking one, or advance 3 ships (instead of 2) when there's only one ship in the next group.

Thomas:
@Skrim: That was a tournament with teams of 8? It had a lot of it's own issues, particularly with low interest from the community, too short and spaced too far apart. A lot more teams than expected fell off the radar before the Summer tournament rolled around. The Summer tournament itself had a lot of scheduling issues, as a lot of other organizers wanted to organize events around that time... none of which seemed to happen. Kind of disappointing.

@Zanc: I think it's a great idea for a tournament.

Agent Sylka:
You have brought together a lot of different concepts for this that play well together. BoCA will support you in any way that you need if you move forward with this league, from organizing help to publicity. This is definitely something we are interested in getting involved with but our involvement will depend on you and or your organizing teams desire for our level of involvement. 

We like this idea as far as the general set up with 4-player teams, yet still competitive.

8 weeks sounds like a decent amount of time for a league like this. It gives all the teams enough time to gain/lose points to have variation, enough time for the newer teams to have played with more experienced teams that they improve over this duration. But not too long that it becomes too fatigued. An idea for "bye" weeks could be that on week 4 and week 5 (seeing as week 1-3 regular season, week 4&5 "bye" weeks and week 6-8 regular season with week 9 being the FINALE!) half the teams would be off the first week while the other half plays and vice versa the next week. Each team would get 1 week off in the middle of the league.

It maybe a bit hard to add a team during the league for the simple reason they could have a disadvantage or advantage on points over the teams that have been in the league for the whole duration. Or maybe if there are 8 weeks the team starts with either 400 or 800 points (50 or 100 points for each week) and if a team is added in on say week 3 they would start with 300 or 600 points (400-100 or 800-200) so it removes the number of points for the weeks they weren't playing in? Other wise maybe have a "sub" list and if a team doesn't show up or bows out of the competition the "sub" ship that comes in would have points averaged to all the teams in the league?

As for the blind pick thing, Ayetach had suggested a pretty interesting idea on the SCS forums, that may work out well for this idea. It allows for shorter lobby time, refs not to have too much to watch for, teams to be able to "specialize" in ship builds, and for a slight amount of randomization to take into account for enemy ship builds/maps.

--- Quote ---When players were mentioning a draft system, why not have each team suggest 2 or 3 sets of ships and loadouts to the ref a day before for instance? and during the lobby they can choose to change from their first preferred set to the next but can’t return to the first one again. That way they have flexibility of changing loadouts but will take a chance by moving forward to the second set with a ‘no return’ policy. That way you have a drafting of 2-3 ‘strings’ of ships that teams pick out. Its worth trying that out as well since these ideas are being tested for the SCS
--- End quote ---

This is the OP

Thomas:
Alright, so I've put together a Mock Season for how an event like this might go down. For this mock season, we're having 8 teams (A-H). Each of these pretend teams has a different skill level. Team A has the highest with 80, all the way down to H with 10. When a team is formed, the team with the higher combined skill level will win. So if there's team BF vs DE, team BF will win (100 vs 90). In the event of a tie, the team with the highest individual skill level will win. This is a pretty good example of how it might work in a real tournament, although in reality teams won't be quite so spread out in a skill, there would be random factors involved (such as how well teams work together obviously).

During the first week, teams are assigned randomly. After week one, they're paired from highest to lowest scores. Since we're doing more than one match a week, you can't set them up in a win/loss fashion like I had initially hoped. Otherwise we could do two matches a week and have teams shuffle after every single match. But that would be a mess. Whoever loses has 10% of their points taken, combined, then distributed evenly to the winners. So if team AB loses, and team A has 700, team B has 500; team A loses 70 points, team B loses 50 points. The teams that won against them get 60 points each. Every team starts with 500 points. This makes week 1 points pretty easy to distribute. Each team will play 3 rounds each week, changing partners at the end of the week. This will go for 8 weeks, with half the teams taking a break week 4, the other half on break week 5. Points will be rounded to whole numbers. It should be noted that this would be the minimum number of teams. It actually works out nice for a consistent round robin. With more teams it wouldn't be possible for a full round robin every week, so opponents would be assigned randomly.

TeamSkill LevelStarting PointsA80500B70500C60500D50500E40500F30500G20500H10500




Week 1:
Paired TeamsCombined SkillA G100B D120C H70E F70
At the end of week 1, the scores and ranking are as follows:

TeamScoreB645D645A546G546C445H445E364F364

This means that for week 2:
Paired TeamsCombined SkillB F100D E90A H90G C80
 The scores get a bit messier as we go into week 2, but that's totally fine, as they should separate the teams more and make it easier to rank them. At the end of week 2 and the start of week 3, the scores are as follows:

TeamScoreB790A586D563F509H495G397E336C324
Making the teams for week 3:

Paired TeamsCombined SkillB C130A E120D G70F H40
Things are still quite messy in the rankings. By the end of the league, what we'd like to see is the teams being relatively close to alphabetical order, as that's how their skill is based. It's looking kind of iffy, but we'll keep going. At the end of week 3 and into week 4 we have the following scores and teams:


TeamScoreB930A621D497C464E395F371G361H361

Paired TeamsCombined SkillB H130A G120D F70C E40
Half the teams are taking a break, and half continue. How we'll do this by pairing the teams up, having half of those combined teams taking a break, and the other half compete in their matches. The just reverse that the next week. For this example, and ideally for a more complex tournament, this actually affects nothing. So we're going to combine weeks 4 and 5 to give us a full week. The scores at the end of week 5 and going into week 6 (and the new teams for week 6) are:

TeamScoreB919A767G507C410H408D363E355G271

Paired TeamsCombined SkillB H130A G120D F70C E40
Then of course we come to the end of week 6 and starting week 7. Here's some more tables (we're almost done!).

TeamScoreB907A906E494G371C367H365F325D265

Paired TeamsCombined SkillB D120A F110E H70G C40
So it's still quite messy. A and B are neck and neck, which is pretty good, still worrying to see C and especially D that low. Places seem to change a lot from week to week for everyone but the top two teams. In a real league, this might be typical, but in a real league the skill level of teams wouldn't be nearly as easy to categorize.  As we finish week 7 and come into the final league week of 8. More tables.


TeamScoreB1045A891D403F368E360G336C332H265

Paired TeamsCombined SkillB H80A C120D G70E F70

So then we finally come to week 8 where we wrap up all the points as follows. The top 4 teams will move ahead to the finals where they will compete in a double round robin. Whoever has the most wins at the end wins. If there's a tie, we will use the points from the league to determine the winner. In the event those are tied as well, we will engage in a GoIO themed jeopardy round hosted by Howard (probably not though). Anyway, the final points come out to:


TeamScoreA1026B1011C467D356H309G301F269E261
So we have our actual top 4 teams going into the finals, which is pretty cool. The bottom half isn't as clean of course, but I feel it would rearrange itself properly if it's run long enough. Considering that real team skill isn't that clear cut, that real matches suffer a lot of random factors, and that teams are more likely to be closer in skill level than these fake teams, I think it turned out pretty well. As for the finals, we have:

A with 6 wins, B with 2 wins, C with 2 wins, and D with 2 wins, making A the champions~




=========================================================

Some things I noticed while running this mock season is that the early random matches set up some teams for an early lead that was difficult to overcome. There was also a huge split in the score near the end. To keep from going crazy with decimals, I had to do some shady rounding at times to keep whole numbers. In an actual tournament, we may have to round better and the total available points between each team might drift up or down accordingly. I ensured the total available points stayed the same, hence the shadiness.


Overall I feel that it would function better in a realistic setting. I understand there are practical difficulties with putting two groups of strangers together and telling them to compete, but where GoIO is right now it's more realistic than trying to keep the 8 man team system going.

Navigation

[0] Message Index

[#] Next page

[*] Previous page