FFA league - concept and discussion

Discussions about the various leagues
Post Reply
User avatar
Zehra
Private First Class
Private First Class
Posts: 914
Joined: Sun Oct 18, 2015 3:36 pm
Location: Within the BZFS API and Beyond it
Contact:

FFA league - concept and discussion

Post by Zehra »

Greetings all,

Something I've noticed has not been done in quite a while, is a Free For All league. I think while CTF is really enjoyable and allows for team vs team rankings and some statistics on players overall, it does not offer something very much like a leaderboard ranking to take place. Since it would be a league type format, nobody has to worry about having stats logged during casual game play sessions, but only when matches do take place.

If there is enough interest in a FFA league, I do not mind backing this venture fully, and indeed I am quite interested in seeing a FFA league exist and am open to suggestions and ideas on what may take place. Even a rabbit hunt/chase type of league too, if so desired.

There is several ideas I have relating to scoring formats and stats, such as kills per minute, kill to death ratios, best flags used for stats. Scoring ranges from strict leaderboards used for overall scoring to team types ranging from standard team selection to various team type "factions" or alliances which may be joined.

What is everyone's thoughts on this? Is a FFA league even wanted? Would a small scale demonstration be appropriate? Any specific considerations I should give? What would you like to see?

-Zehra
Those who are critical of me, I'll likely be the same of them. ~Zehra
The decisions we make are the ones we look forward too and the ones we regret. ~Zehra
There's a difference between knowing my name and knowing me, one shows respect to my name and the other is to who I am. ~Zehra

See where I've last been active at Strayers.
Visit BZList.net for a modern HTML5 server stats site.

Click here to view the 101 Leaderboard & Score Summaries Last updated 2021-01-12 (YYYY-MM-DD)
Latest 101 thread
User avatar
macsforme
General
General
Posts: 2069
Joined: Wed Mar 01, 2006 5:43 am

Re: FFA league - concept and discussion

Post by macsforme »

Individual competitive/ranked/ladder play is a viable concept in most games, in tandem with the more formal league play with organized teams. I imagine it would work just as well in BZFlag. However, I do not believe that BZFlag's current ailments would be solved by the addition of such an option at this time. In my view, competitive play is built upon the foundation of a healthy game and community, not the other way around.

Edit: Sorry, I just realized that I misunderstood your premise (I thought you were suggesting solo competitive play with random teams). I believe that competitive FFA would quickly deteriorate into everyone targeting the weakest players, and you might have trouble recruiting and retaining new participants.
button
Private First Class
Private First Class
Posts: 7
Joined: Mon Mar 14, 2011 5:02 pm

Re: FFA league - concept and discussion

Post by button »

The second level of the hix map would quickly turn into skid row with the amount of unsolicited camping there would be. Tent prices would skyrocket and I don’t know if the bzflag economy would be able to keep up with such a drastic increase of homelessness…
User avatar
Zehra
Private First Class
Private First Class
Posts: 914
Joined: Sun Oct 18, 2015 3:36 pm
Location: Within the BZFS API and Beyond it
Contact:

Re: FFA league - concept and discussion

Post by Zehra »

macsforme wrote: Tue Apr 11, 2023 6:31 am Individual competitive/ranked/ladder play is a viable concept in most games, in tandem with the more formal league play with organized teams. I imagine it would work just as well in BZFlag. However, I do not believe that BZFlag's current ailments would be solved by the addition of such an option at this time. In my view, competitive play is built upon the foundation of a healthy game and community, not the other way around.

Edit: Sorry, I just realized that I misunderstood your premise (I thought you were suggesting solo competitive play with random teams). I believe that competitive FFA would quickly deteriorate into everyone targeting the weakest players, and you might have trouble recruiting and retaining new participants.
TL:DR; (summary)
I've seen a few examples of ladders in games, I agree with you on it working well in BZFlag, it not addressing/solving BZFlag's current ailments. Competitive play is difficult to factor in, but I don't disagree with you. My predictions/calculations do line up in terms of issues of targeting weaker players, but I believe some solutions may address this.

-

I will admit I'm not that familiar with individual competitive/ladder/ranked play, although I've seen a few examples of ladders and think it would go well with BZFlag. I agree with you that it doesn't address or solve BZFlag's current ailments, especially at this time. (Currently CTF maps dominate the field, unlike even a year or two ago when Urban Jungle and similar would be of noted activity on the list.)

In terms of competitive play, one aspect which I've mostly now focused on, is interactions of player groups. The basis of this, is that much of player groups and preferences can be understood in a neatly summarized way and what would ideally work. (I'll explain in more details on this further below as a general note.)

You're right on the premise of weaker players being targeted, I do have some difficulties in attempting to prevent this type of abuse from occurring. My idea to mitigate this to some extent is finding a better scoring system which would grant less weight on stronger vs weaker players in terms of score adjustment. One of my main difficulties is finding a system which could take into account both joint scoring and individual scoring, while at the same time being meaningful and detailed enough and also properly factoring in updates. (I'll get into more details on this as a general note at the end of my post.)

button wrote: Tue Apr 11, 2023 6:56 pm The second level of the hix map would quickly turn into skid row with the amount of unsolicited camping there would be. Tent prices would skyrocket and I don’t know if the bzflag economy would be able to keep up with such a drastic increase of homelessness…
Lol. I can understand camping turning the map into a noob hunting session. Or creating a wide gap between those who have tents and those who don't. :p

I'm looking at a few possibilities from using map data/analytics to either assign a new camping metric which is different from the main scoring or assigning a lower weight to camping techniques in scoring. Either by itself or in combination with adjusted scoring, I do believe a better spawn algorithm would wind up reducing the unfortunate crisis.

---

General note:

TL:DR; (summary)

In terms of scoring, I haven't found an ideal scoring solution, but several bits of inspiration are somewhat close and far from it at the same time. (I leave an important side note as to why some areas may be confusing.) I see spawning as a very much needed improvement. I believe the overall player dynamics will change in several areas. I see player groups as very important focus. My vision of FFA league. Endeavours similar to a FFA league and their difficulties.

(Keep in mind, this is all aggregation of data, not direct word by word quotes from the threads!!!)

-

Originally, FFA league would simply be a basic ranking/leaderboard and not strictly a league itself. PlayerRanking is considered the original ranking: a new idea - player ranking

But some issues would be present in terms of scoring: Player rankings, ratios and more... tend to become stuck/fixed over time. For example: A player with a million to half a million in K/D ratio would be on top of the leaderboard and virtually never have their K/D ratio adjust ever again. This would present accuracy problems in terms of recentness and balancing, as newbies would see their improvement rate less than it actually is and pros would have difficulties in improving their rate at all.

This is a simple system, yet already it faces that issue/problem to begin with. 1vs1 league has a similar issue in terms with its scoring: A new year, a new beginning.

The scoring becomes inaccurate and gives too much weight in some areas and the leaderboard winds up being dated. (Nobody wants to play when the risk so much to possibly gain so little.)

When this occurs, other newer data becomes preferable to that extent: Spring 2019 Babel Triples Challenge

On a side note:
One aspect which adds perhaps confusion to the whole concept of a FFA league is the following:
What is "Team Death Match" in other games, is called "Free For All" in BZFlag. "Free For All" in other games, is known as "Open Free For All" in BZFlag.

For FFA league, I'd explore both concepts as possibilities: In TDM/FFA, it would be a "Joint Scoring" option", while in FFA/OFFA would be an "Individual Scoring" aspect.

But I do admit, it is tricky to figure out how to make "Joint Scoring" meaningful.

In terms of joint scoring and teams, making the scores meaningful is very tricky. In a similar scenario, issues on how to make the scoring/added teams meaningful are brought up: A new matching variant for GU league

Joint Scoring/Team(s) scorings has some difficulties in terms of giving weight to cooperation with teammates, similar to the following thread: Public Team Score Contribution Value

(Map analytics may help address/resolve this.)

From current observations, it seems that we've slowly began to shift away from team based matches, as well as limitations within map selections, this may be seen in the following thread: Thoughts on competitive/league play

Overall, I do envision something similar to the following: General Ranking Between Registered Players

But in a more comprehensive form and perhaps with more weight given to time as well.

-

I see spawning as something which needs major overhaul and enhancements to make games work better in any setting. Spawning into teammates bullets, spawning into bullets, spawning facing walls, or spawned "safely" in an entirely unpractical location do not make sense. (You can even spawn in locked up sections of the map without no way out.)

The luck factor of good/bad spawns is nice for adding a bit of randomness to the game, but it can make the game very painful to play. Spawns should fall in semi-good or semi-bad range overall and no impractical spawns or bad spawns.

(I may expand on this as its own thread for further discussion/reference.)

-

In BZ - What's the future for it?, I did cover some predictions regarding cycles and player groups, while some of the more underlying principles are perhaps better summarized in Long term observations - What to possibly consider.

Previously, with player groups being much larger, there was less need to factor in player groups. (You would easily have 10+ active players per active group.)

In my opinion, as player counts decreased, the overall dynamics began to switch up in game. Which could be seen in types of maps played: List of most popular/classic 1.x maps

This later resulted in a wide variety of various "issues" which arose and were commented on in 2011 to 2013: I will point out something interesting from it all, this partially involves player groups and their dynamics/perspectives on some matters. A more recent comparison could be made, but it wouldn't be as comprehensive as this one.

This lines up with how some of the friction which has occurred over the years, but now we see it within a much smaller scale. Which is partly why I expect another change in the whole dynamic of the game to occur within a year or two.

Relating to the subject is how only certain map types will only attract certain types of players: Campy Players, Not Maps (Case Study: "Two Tanks")

This may be seen as why Two Tanks is no longer a map which is actively played as the player groups for it, are no longer active at a level sufficient to generate activity on the map. Which falls in line with how less player groups have generated new player dynamics/cycles.

I'll use the following threads for reference: Agatha's Thoughts on High-Level Gameplay, Ducati style overhaul, and "Two Tanks" Pyramids-to-Jumpers Shots Chart

One common misconception is difficulty vs competitive games. (Jumping Skills Impossible vs HiX.)

Currently with player groups, there is very specific core interests they each have. While some maps have players overlap, it tends to be because players fit into more than one group vs the other way around. (Which is partly why when player groups decrease overall, the variation in maps which are seen online tend to decrease or rather be less random than they normally are.)

Related to this is now the increasing skill gap between players, which means players will wind up on the side of more skill, but pros will still lead overall. (As long as their skill increases at the same rate or more.)

One of several reasons why this remains the case, is simply the fact that basic tactics, while extremely effective/enticing, do not do much to affect the overall game in terms of strategic play. This can be easily explained by that fact that despite it offering better potential in several cases, simply what strategy can be studied/researched, is not strictly desirable or wanted by the majority of players. (Which is why don't I see a shot chart/guide for HiX, when it is a simpler map to calculate ricochets, especially when it would be of great help to new league players.)

(I may later release parts of my rico rotation tutorial type of map.)

Adding to this is the player types, which while before were more distributed among player groups, now player groups tend to contain very specific player types.

-
My vision: Something which proves challenging enough with light-hearted competition. Ideally, a map would be created which has elements which can play in a rock-paper-scissor type of manner in terms of player group preferences.

The match stats offer some really good insights/advice for players on where to improve and to enhance their overall enjoyment of the game.

Players can play maps they enjoy, get to see their progress and are rewarded for their efforts.

-

Related to endeavours similar to FFA league:

I've also considered a league for offering novelty game modes at a rate of 1 per month or week or perhaps bi-weekly.

I do realize the main difficulties are the activity, the management, and the starting of a league.

Currently, the activity rate is difficult, but I do believe anyone who can use the cycle of player dynamics changing, will be successful to some extent. For management, it is understanding the player groups. For starting of one, it's either the ability or the connections with people who have the ability to make it happen, same as their support.


-Zehra

p.s. Does anyone find the TL:DR; sections helpful? If so, any suggestions for improvements on them.
Those who are critical of me, I'll likely be the same of them. ~Zehra
The decisions we make are the ones we look forward too and the ones we regret. ~Zehra
There's a difference between knowing my name and knowing me, one shows respect to my name and the other is to who I am. ~Zehra

See where I've last been active at Strayers.
Visit BZList.net for a modern HTML5 server stats site.

Click here to view the 101 Leaderboard & Score Summaries Last updated 2021-01-12 (YYYY-MM-DD)
Latest 101 thread
User avatar
Zehra
Private First Class
Private First Class
Posts: 914
Joined: Sun Oct 18, 2015 3:36 pm
Location: Within the BZFS API and Beyond it
Contact:

Update: Possible scoring system to be used for FFA league:

Post by Zehra »

TL;DR
Previously, I did cover some of the issues faced with scorings and similar. I believe I have worked out a simple metric which is sufficient for the rankings to be used. The metric is K/D ratio, but with additional factoring for a "strength"/"accuracy" metric. To make it easy enough to sort out, strength, accuracy and type metrics are part of it. Main listing is an average of averages of other scores sorted by map type.

Long version:
There is a lot of issues with attempting to solely rewarding activity or solely "believed" strength in scorings. Both wind up with inaccuracies, as they provide a "wrong" metric and don't reward accomplishments well.

To avoid this, the following can be adapted as a notion:
* Activity == increased accuracy in a strength rating.
* Type == Map type tells of what kind of skill a player has on a given map.
* PlayerStrength == K/D ratio.

So we wind up with the following considerations:

PlayerStrength = Player(K/D)
PlayerExperience = Activity(Type)
PlayerRank = (PlayerExperience, PlayerStrength)

So who is predicted to win based on the metrics used:
If a player has the same PlayerStrength, whoever has the higher PlayerExperience rating, is predicted to win. (Thus a good indication of strength, even if the K/D ratios are the same.)

But how do we know when an experience level changes based on activity?
Upon reaching the following in K+Ds: 100, 500, 1000, 5000, 10000

Which gives the following consideration:
Level 0: (K+D < 100)
Level 1: (K+D < 500)
Level 2: (K+D < 1000)
Level 3: (K+D < 5000)
Level 4: (K+D < 10000)
Level 5: (K+D < 50000)

Or we can use the following notion, which is simpler to understand/provides more accuracy:
Upon reaching: 100, 500, 1000
Level[1000][500][100]

So a player with a sum of 2745 K+D, would be level[2][1][2].

New players are either given a grace period or a form of "modified" initial rankings to prevent abuse. (60 days or 20 matches which may be used to set initial levels, but not affect the K/D ratios. This can be averaged out.)


On top of this, since maps gain a type classification, there is an average "overall" player ranking and a map type specific player ranking. (Hence someone can have a 5-4 K/D ratio overall on average, but depending on map type, may for instance have a ratio of 7-5 on some maps, while 5-5 or 5-6 on other maps.)

In the long run, it provides a simple and easy way of knowing skills of players with higher accuracy as more matches increase.

...

Note: Statistics is considered a separate item, same as analytics, which will be covered at a later time period.
Those who are critical of me, I'll likely be the same of them. ~Zehra
The decisions we make are the ones we look forward too and the ones we regret. ~Zehra
There's a difference between knowing my name and knowing me, one shows respect to my name and the other is to who I am. ~Zehra

See where I've last been active at Strayers.
Visit BZList.net for a modern HTML5 server stats site.

Click here to view the 101 Leaderboard & Score Summaries Last updated 2021-01-12 (YYYY-MM-DD)
Latest 101 thread
Post Reply