TL;DR AT THE BOTTOM

I want to start by saying that the inclusion of Crystal Arena this season was great. As a start, the unlimited rewards system and vanity drops were cool and I think the fact that more people got to experience gameplay closer to what a Crystal 5v5 might be was awesome. However, I think there were a lot of aspects poorly thought out or some decisions that seem...dubious? At best.

Solo Queuing and it's expectations

I tracked several stats for 101 solo queue games that I played. Of these 101 games, 92 were in Gold 1. Of these 92, 56 were wins and 36 were losses. That's about a 60.8% winrate and 20 wins above .500. Regardless, from these 92 games alone I lost a total of 112 RP. Even with the most charitable analysis I couldn't see this as anything other than a bug or misguided expectation. The reason for this massive RP loss is that starting at Gold 1, penalties for losing start increasing and RP rewards for winning start to decrease. I averaged about -16 for every loss and only +8 for every win. Admittedly, I have noticed that this issue would be assuaged (slightly) if there were more players in these higher ranks queuing. However, because the current RP system actively punishes you for queuing at ranks Gold 1 and above in any situation where you can't guarantee an extremely high winrate, many players in this rank rarely queue with less than 5 players. These lopsided punishments don't make sense because: 1) the lobbies where you lose 16 RP seem almost identical to the ones where you win and gain 8 RP AND 2) expecting an exceptionally high winrate to be maintained is not congruent with the assumption that matchmaking strives to create teams where both teams have an equal chance to win. (If you will just take me at my word for these two points you can skip the next 3 paragraphs)

To expand on point 1, I have tried my best to figure out the rhyme or reason behind these big penalties when you reach Gold rank (compared to lower ranks), but after tracking my matches by hand, I can come to no other conclusion other than that this forced difficulty is put in by SBI (or is a bug). What leads me to this conclusion is that even though the lobbies that the game would create when I was/am Silver 2 are largely the same as the ones when I'm Gold 1, at silver 2 the RP +/- for wins and losses are very similar as opposed to Gold 1 where the each loss is worth significantly more than each win. And I am specifying the badge of Silver vs Gold because even if your rank number is the same, your +/- RP is much more heavily influenced by your badge (i.e. a loss at 2950 RP and silver rank is less punishing than a loss at 2950 RP and gold rank). My solution to this is simple, just make wins and losses worth very close to the same amount at all ranks, at least for solo queuing (yes, even at low ranks, I'll expand more on that later). In all honesty, I don't mind getting lower ranked teammates as my rank goes up, but it does not make sense to have progressively harder games AND getting punished harder for losing as well.

As far as point 2, it's fairly simple. The way I see it, it can't be two ways. Let me explain. Let's say, for arguments sake, that the higher loss penalties for higher ranks are legitimate because the lobbies that are made for these players are ones where the high rank players are EXPECTED and will most likely win. In that case, I can accept the higher penalty for losing because you are constructing a lobby where the high rank player has an easy time winning. However, if this is true and this is what your algorithm thinks is true, how can you justify putting the players who are playing against the high rank player in the lobby? Do you think they would feel good about you putting them in a game where they are expected to lose? Ok, now lets look at it from another point of view. Let's say that you believe the algorithm is creating two teams where both have about an equal chance to win. Which is good, I think that's where everyone would like it to be. If this is the case, and if SBI believes this to be the case with their current implementation of the algorithm, then how can you further justify the higher penalties at higher ranks? If both sides have about equal chances of losing and winning, shouldn't a win and a loss be worth the same amount of Rank Points, regardless of your current rank?

The win:loss ratio SBI is trying to force you to achieve is not reasonable in any sense because even in other games, most high rank players don't go above a 55%-60% winrate. For example, if we take a look at the top LP players in Korea, Europe West, and NA for League of Legends, very few players reach a 60% winrate at the highest level, and even fewer go very far above that. In contrast, some of the best and/or most accomplished Dota 2 players like Sumail, Miracle, Yatoro, Ceb/7ckingmad, and w33 don't manage to maintain a 60% winrate over thousands of ranked matches. Even in the all time chess elo rankings most of the top players hover around a 60% winrate and the only outlier is Garry Kasparov, whom many believe is the best chess player of all time. This idea that once you reach a certain rank you need an absurdly high winrate to climb when queuing solo is misguided. It may work for 5 stack queues, due to the removal of +0 games, but for solo queuing it's very nearly impossible. If the matchmaking system is working to create lobbies that are "fair", and we assume both teams have an equal chance of winning, should a player that has a positive winrate not just continue to climb? If anyone would still argue that as a player reaches Gold 1/Gold 2/Crystal, their winrate should be sustained above 66%/75%/etc, why should this be the case for a solo player? What makes 5v5 in Albion more special than games like Dota or League of Legends where this expectation is not in place?

Overall, I don't understand a lot of the expectations that seem to be carried over from 5 stacking into solo players' games. My suggestion would be to have solo queue games be more volatile (i.e. you gain and lose significantly more RP than the current implementation, like +- 20/25/30), but the wins and losses should be valued largely the same for the vast majority of games, assuming SBI has confidence that their matchmaking is making a balanced lobby. Additionally, I think SBI should consider increasing the rewards for solo queuing as opposed to queuing in a party. Doesn't have to be a lot, maybe 10%-20%, but I think encouraging this will lead to faster queue times and an easier time for the matchmaking to create fair lobbies.

Teammates
^{Losing RP is good, actually}

If you browse the forums or sometimes even comments in the reddit sub, people sometimes will complain about their teammates being bad, the lobbies being unbalanced, or something along those lines. What I found is that generally, the matchmaking algorithm is pretty good at having either team being within 300-400 RP of each other most (78%) of the time. The actual issue with "bad" teammates has nothing to do with the matchmaking in my opinion and more to do with the fact that the +/- RP calculations for players at the lower end of the rankings is also severely imbalanced. Generally, I don't see the result of this issue very often, since I mostly get matched with Silver players and occasionally a few bronze players. However, despite this I have encountered what I will call "anomalous players". These are players who consistently affect the games they take part in negatively for their own team, but because of how RP is handed out, they maintain a (most likely) undeserved rank for quite a while.

Lets say that we have this silver player , what I would consider to generally be about average (maybe slightly above average) stats for a Silver 1 player. He has 2129 RP, 127 total games played, and a winrate of about 59%. For the most part, a lot of silver 1 players will have stats similar to this and have about the same impact as this player. However, occasionally you will run into an anomalous player. This silver player has the same RP, more experience, yet their winrate is significantly lower than the aforementioned silver player (41%). These players are by no means are anywhere near each other in the impact they have in a match. Yet by all metrics, they will probably be treated as the same level of player by the matchmaking system. See, the issue at the lower end of the rankings has the same issue as the high end, except in the other direction. In lower ranks, wins net players much more positive RP than losses give negative RP (in some cases 3 times more). This starts to level off a bit once you hit silver but the reality is I've seen quite a few players able to maintain a ranking of Silver 1, even with a winrate as low as 40%, and able to maintain Bronze 2 with a winrate in the 30's%. So what is the solution at this end of the rankings? Same as above, just make wins and losses worth the same RP. Players who play above their rank will influence the game in a positive manner consistently and will eventually climb out of that rank. Yes, occasionally any player will get games that they were never going to win, "impossible games". But, if you play consistently above the rank you are currently at, you will eventually climb. Good players don't need the extra boost to get to the rank they "belong" at. More importantly, the "anomalous players" who maintain their rank despite more consistently affecting a game in a negative way will eventually fall to an appropriate rank. So in this sense, losing RP is good, actually. Especially when you lose enough games to deserve it.

Healers

I don't have much to say here. Contrary to what I've seen quite a few people say, I don't think healers alone determine the outcome of a match. For the most part my experience with solo queuing with random healers has been mostly positive. They're not all perfect players, but you don't need a perfect healer to win most matches. However, if you're going to continue to let people queue as healer in 5s with spells like Revitalize, Sacred Pulse, Holy Beam, or Holy Blessing, they should really be more useful in 5v5, otherwise just don't let people equip these spells in crystal arena. Moreover, at least for healers, when they enter the crystal arena instance they should have all spells unlocked for them. Additionally, while this has become MUCH less common than before the most recent crystal arena matchmaking change, I really don't think a healer should ever be the lowest rank player on either team, almost is never a good experience for the healer or the team. ALSO PLEASE MAKE A MUCH MORE VISIBLE AND OBVIOUS INDICATOR FOR THE DOUBLE HEALER PENALTY YOU HAVE PUT IN, YOU CHANGED IT WITH JUST A SMALL FOOTNOTE OF INFO AND NOW MOST PEOPLE DONT REALIZE ITS IN THE GAME. Make it like the message that a territory is low on food or something, anything really. It just needs to be much more obvious.

QOL

Include total elapsed match time and final score somewhere on the final stats screen, please.

TL;DR

Solo Queue

• Out of 92 games in Gold 1 solo queue I went 56:36 Win:Loss. I lost a total of 112 RP from this.
  
• Avg loss was -16 avg win was +8
  
• Expectations of high winrate from 5 stacking carried over to solo queue are unreasonable
  
  • LoL Ranked Ladder leaders rarely cross a 60% winrate
    
  • Dota 2 pros with thousands of matches usually are pretty far off from reaching a winrate of 60%
    
• What makes Albion 5v5 special to where we can put an unusually high expectation on every player as opposed to other 5v5 games?
  

Teammates

• Matchmaking algorithm is pretty good at having either team being within 300-400 RP of each other most (78%) of the time, contrary to popular belief.
  
• "Bad" teammates are mostly players who have benefitted from the fact that at lower ranks 1 win is worth multiple losses in terms of equivalent RP
  
  • This creates situations where a 41% winrate player ends up getting weighed the same as one who has 55%+ winrate
    

Both the above issues can be solved if wins and losses just give the same RP across all ranks, at least for solo queuing

Healers

• Most healers play well, at least good enough to win (from what I can tell solo queuing)
  
• If you are going to let people use Revitalize, Holy Beam, Sacred Pulse, and Holy Blessing in Crystal Arena, make these spells at least as useful as others in the same tree. Otherwise restrict their use just like double healer is (basically) restricted.
  
  • MAKE THE INDICATOR FOR THE DOUBLE HEALER PENALTY MORE OBVIOUS
    
  • Consider having all spells unlocked for healer in Crystal Arena
    

Yes, seeing players play in such obtuse ways that negatively affect matches in extreme ways is frustrating. However, I truly believe if the value of wins and losses were equalized, these players' ranks would fall and higher rank players would overall play with them much less. So that will continue to be my prescription on that issue for now. Yes, getting them on your team at all sucks, but I don't think it's entirely their fault since the only reason they are able to maintain their rank is due to how RP is handed out under ~2000.


The build I'm liking most for solo pubbing is 1h Arcane, Mage Cowl, Cleric Robe, Mistcaller, Leather Shoes on refreshing sprint, and Thetford Cape. Mana potions or res pots if there is a druid cowl on your team that is able to sustain you. I usually try to play by my healer and use the 1h Arcane E primarily to purge hellion jackets of enemy melee players so that they can't just sit on my healer. I've seen other people use offhands like Muisak with scholar cowl or druid cowl which is probably fine but I just personally really value having my cooldowns faster.

While I understand the intention behind this system, I think there should be less focus on the actual winrate. If RP +/- is equalized, I think most of the issues should fix themselves. The matchmaking already does a fairly good job of trying to make a "fair" lobby RP-wise, and if your winrate is above 50% playing against a team that this mostly within the RP range of your given team, you should just continue to climb if you win more often than you lose.

I think we mostly agree here but I just feel that a heavy focus on winrate has led to the situation we have now where most of the top ranked players play almost exclusively as a 5 stack and about half their games are against opponents that never had a chance to begin with.

Majority of the games you get when 5 stacking with other top players are just uncompetitive. Forgive me if I don't want to spend half my time sitting in front of the tent of a team with 3 silvers and 2 bronze players wondering how they managed to end up in a lobby against a team of all Gold+ players. That's how you see these winrates of 70%/80%/90%+. In no situation where you are playing in a "balanced" lobby should you expect a winrate that is that much above 60%. But, because you are 5 stacking, you're basically forcing the matchmaking algorithm to give you the "best case" scenario out of the other 5 stack teams that are currently queuing. Not to mention, even if I were to 5 stack just to raise my rank number, now I'm stuck in a situation where I am locked into only playing with 5 other players because solo queuing will almost certainly always result in a net-negative result.

If SBI wants players to 5 stack, ok cool, I can actually understand that. That would make a lot of sense and I wouldn't necessarily disagree. But if that's the case, just remove being able to queue with less than 5 players. If everything is going to be balanced around the 5 stack queue, don't leave in the afterthought of a queue that is basically a farce.