Ranking players by power is necessary for multiple things, especially to create balanced teams. The biggest problem with that is finding a metric that actually succeeds in measuring a players power.
Ranks (private - marshal) only correlate with amount of games, barely with skill, so they are not a reliable metric.
Here is my very easy to code solution: Use the average amount of medals a player got in his last 100 games (giving a player his avg medal score starting after 25 games completed, otherwise too low n. Only completed games count).
Why is this a good method?
easy & fast to code
reliably depicts the player being best at something
further gives high scores for high impact players, eg players who blow up many tanks, build rallies, cap many points, use different soldier classes etc
assigns low scores to one dimensional players (eg someone only farming score by shooting down bombers but not contributing another way to the game will reliably get some medals, but only few)
one sided stomp games give out few medals, as tresholds are usually not met, therefore people farming wins in easy games will have lower avg scores (especially stacks)
additional stack balance: since stacks usually cannibalize all medals among themselves, players who mostly play in stacks will have low to medium medal scores, players playing alone and carrying their teams to wins with disciplined work will have many medals → perfect depiction of player impact and game balance
EDIT: The obv big advantage is that using this medal average score for matchmaking will actively create balanced teams; much better than other metrics. As every team will try to be assigned to get high medal players who build rallies etc and then filled up with the lazy randoms.
Also this is a soft rule, not a hard rule. So it doesnt slow down match making at all, it just improves it by creating more balanced games.
EDIT2: This system would need a fixed timer to gather a pool of queuing player to combine into teams. 20-30 sec would be fine. In addition a deviation value can be set that defines well balanced teams; in times of high player numbers games wouldnt need to wait but start instantly once the balance level according to the deviation value is reached.
there are plenty stuff that was easy and fast to code and they didnt do it fast. e.g. mm fix that enabled 13v10 matches required 5 minutes to code… they fixed it 3 months after acknowledging it. merge the way they implemented it required max week or two for single programmer to implement, they did it in ~10 months.
it could maybe even work if MM wasnt already extremely complicated. but when you have playerbase divided across 6 servers and further across 4 nations and across 3-4 BR ranges, then it leaves no room for SBMM except maybe in peak 2-3 hours for crossplay on.
this is a soft rule not a hard rule so it doesnt make matchmaking any harder or slower. its a strict improvement over what is used atm as it gives additional data to matchmaking to make it more precise.
i made the post as i know DF doesnt play its own game and wont come up with a good matchmaking solution anyway.
this is straight impossible to implement for crossplay off and it is questionable on how effective it would be on crossplay on.
it would slow down MM, cause you can either have FIFO queue that gives you match as soon as 20 players get together or wait till your MM tries to match similarly skilled players which requires at least 40 players to even start shuffling.
you still dont understand: a soft rule means the mm algorithm tries to team up players as good as possible according to the medal scores with the currently queuing players. the algorithm tries to do it with as high precision as the current amount of players allows. this has NOTHING to do with longer mm times as calculating it takes less than a second and if perfect mm isnt possible the games will still be created. this system just increases the average balance of matches created. i recommend the usual posters in this forum to spend 10 minutes thinking about something before forming opinions.
if there was faction queue where hundreds of players spent minute or two before getting match your idea would be viable at 0 cost. but you dont have that. you have FIFO queue that creates match whenever you get 10 players from both team. so any MM that doesnt give match as soon as 20 players are ready is already slower MM.
if thats how enlisted matches are truly made, it explains how horribly imbalanced the matches are and also that there needs to be a better matchmaking system with 10 - 30 sec waiting time. in that case my system becomes the best proposed so far for enlisted.
that is how queues work. first in first out (aka FIFO). SBMM wouldnt be hard to implement if enlisted didnt have complex MM. but cause enlisted has complex MM, SBMM is almost impossible to implement without either forced join any MM or any vs any MM.
i will do oversimplified napkin math for number of players enlisted has available:
at peak time enlisted has ~10k players, off peak ~3k players. cause enlisted has 20% desertion rate actual number of players can be anywhere between 8-10k in peak hours or 2.4-3k in off peak hours.
enlisted has 3 servers and lets assume perfect distribution(which it will never have), you get 2666-3333 players in peak and 800-1000 players in off peak hours.
average match lasts ~20 minutes, so it will get 133.3-166.66 players in peak and 40-50 players in off peak hours waiting in queue every minute.
enlisted has 4 factions and if we assume perfect distribution you will get 33.3-41.65 players in peak and 10-12.5 players in off peak hours per faction every minute.
for simplicity of calculations lets assume that enlisted only has 3 BR MM ranges (BR 1-2, BR 3-4 and BR 5) and you will get 11.1-13.83 players in peak and 3.33-4.17 players in off peak hours per faction/BR MM queue every minute.
do you see reason why any kind of SBMM is hardly possible without significantly increasing queue times or changing current MM rules?
My squadmate has been complaining about this for years now. Player ability needs to be taken into account.
Now I agree that ranks are not a good indication of skill, on the other hand Im sure you understand our frustration when the enemy team leaderboard the top 5 are marshals with hundreds of kills and our team is all low ranks with our bolt action rifles.
I believe a soft rule for your idea should work. My only query would be for example my own squad:
Myself, Im a reasonably competent player, unlocked everything in Allied tech tree, unlocked to BR III in Germany (but play II), unlocked and play BR II Japan.
My squadmate is a much slower, more long range, less aggressive player. You mentioned shooting down bombers, thats literally one of the things he likes to do and hes very good at it (excellent shot too), but hopeless with closer range SMG/LMG stuff.
30 kills in a match would be a really good day for him (and hes okay with that, like this guy had never played a computer game in his life before War Thunder and then Enlisted).
So to get to the point, how would you deal with our squad? Im a LOT more skilled/advanced, than my squadmate. So where would we rank?
Would we be prevented from playing together given the ability discrepancy? Or would it be like War Thunder where our abilities are sort of averaged and we could be ‘uptiered’ thanks to my ability or ‘downtiered’ due to his?
Nice suggestion indeed. This and stacks versus stacks matchmaking are the need of the hour. However, I feel like this system can best be implemented when 1. We have more players in-game. 2. When all the BRs are made separate. In the meanwhile they can test out this new system though as a pilot project.
the mm would wait for 20 sec (a fixed time can be set then games start) then create as balanced as possible teams based on average medal score of the team. due to playerbase stats this wouldnt be always perfect, but definitely more balanced than the current lack of any mm at all.
Logically, it would be in contradiction to the BR soft rule. So in effect, both would become completely ineffective.
This soft rule argument is just so ridiculous.
You simple can’t have infinite number of soft rules and expect them to influence something.
This game can’t even manage to do stack vs stack on consistent bases. It’s extremely naive his idea (if being implemented) would change something in positive way. Quite the opposite, it would only parasite on currently existente soft rules and effectively weakening them.
there is simply not enough players to ensure faction vs faction and BR MM and SBMM. it barely functions without SBMM, not to mention with.
also any SBMM would need somewhat equal number of same skilled players on both sides, which is impossible to guarantee in faction vs faction MM. you could have 100 veteran in queue in one faction and 100 newbies in other faction and no SBMM can make such matchup good.
like i said before any SBMM would need to see either removal of faction vs faction MM aka. any vs any(like in WT, WoT, WoWS) or it would need to force join any on all players.
