Tournament and leaderboard integrity
The tournament structure does not reward the best player. It rewards the player with the luckiest matchmaking. At first I thought this was *purely* a matchmaking problem, but it goes beyond that into how points are awarded for the tournament.
The way you can earn promotion, and placement on the leaderboard is determined by your tournament score. The tournament score is determined by:
1. Participation (setting defense is 720 points)
1. Win a battle (1k points)
2. Feats (minimal)
3. Battle score (0-1.5ish k points)
The main problem here is obviously the variable one: battle score can eclipse the points from a win. And battle score is the most influenced by matchmaking (feats are too, but don't give many points). This means that matchmaking has a *huge* impact on your score, and thus both promotions, and the placement on the leaderboard.
This should, imho, be changed. The simplest is of course, to just remove the points for battle score: just give a fixed point amount to wins. Let the score from battles be a tie breaker. This is like how in most sports championships, the most important thing is to WIN the match. Running up the score is only important for tiebreakers.
However a fairer system is to use some form of ELO ranking, and tie the win score to that. You can still use battle score as the tie breaker, but ties should rapidly be far less common. The reason fixed points for a win work in most sports is because things are run in either a tournament format, or a full round-robin league format. If you have a league format (e.g leaderboard), but can't guarantee everybody plays everybody, having fixed amounts for wins is unfair. One player may simply get easier matchups than another, despite them being "equally" good at the game. This should be reflected by their score. Winning a match vs an equally skilled opponent should give more points than beating a lesser skilled one. And beating a *more* skilled opponent should reward more points still. This is how the point system works in chess, on the Starcraft 2 ladder, and, in fact, most ladders in all eSports games. The points aren't allocated by some complex weighting of your roster: they are allocated by who you beat. Everybody will start with some placement matches (random opponents: this can of course be weighted by GP or some other measure of roster power), and be placed somewhere in the global ladder. After that, every match awards points based on the difference between your place on the ladder and your opponent's. There are different ways of tweaking this to better adjust to the specifics of each game, and how that should be done for SWGOH I don't know. But it seems by *far* the fairest way to create a leaderboard.
It's also worth mentioning that of course matchmaking will still play a large part, and in fact, how matchmaking is done can be almost entirely decoupled from how the ELO-esque system awards points for wins (and removes points for losses), or can be almost entirely based on this same score.
Finally, a little bit about the matchmaking algorithm this GA. I actually think it is very good and has created a very fun GA (at least for most of the people I've talked to). It can probably use some further tweaks to reduce the huge discrepancies in total GP between opponents, but in general looks to be a vast improvement in creating fun and exciting matches. That said, right now it doesn't award clever investment in your roster at all. A large of the game is resource allocation: choosing what toons to invest in. If you have 3m GP but it's all invested in toons that are relevant in the current GA meta, you could probably beat someone who has 5m GP. The GP is *only* a (rough) reflection of the total resources invested. Not whether those resources were invested "wisely" (I use this term very loosely). This choice in how to allocate resources is entirely ignored by the new matchmaking system (and was probably taken too far in the other direction previously). That means it doesn't matter if you are incredibly good or incredibly bad at this part of the game, as only your development of the top teams is even looked at right now, and you are matched with people who have invested similarly to you in their top teams. That does create intense and exciting battles. But it also cheapens the experience of investing cleverly in your roster. I say investing cleverly, but the same can be achieved by spending money. The matchmaking right now *also* reduces the reward for spending money on your pvp teams (you will simply be matched with other players who did the same... and won't be given the opportunity to pulverize f2p noobs). That seems directly counter to any goals this game has in terms of monetization.
I apologize for the long post. TLDR: tournament points need to be "decoupled" from the points you earn in a battle, and preferrably replaced with an ELO-like system for the ranking. Matchmaking needs to create interesting matches for this to be fun. Right now it's in a good spot, but probably disregards the resource allocation part of the game too much.
The way you can earn promotion, and placement on the leaderboard is determined by your tournament score. The tournament score is determined by:
1. Participation (setting defense is 720 points)
1. Win a battle (1k points)
2. Feats (minimal)
3. Battle score (0-1.5ish k points)
The main problem here is obviously the variable one: battle score can eclipse the points from a win. And battle score is the most influenced by matchmaking (feats are too, but don't give many points). This means that matchmaking has a *huge* impact on your score, and thus both promotions, and the placement on the leaderboard.
This should, imho, be changed. The simplest is of course, to just remove the points for battle score: just give a fixed point amount to wins. Let the score from battles be a tie breaker. This is like how in most sports championships, the most important thing is to WIN the match. Running up the score is only important for tiebreakers.
However a fairer system is to use some form of ELO ranking, and tie the win score to that. You can still use battle score as the tie breaker, but ties should rapidly be far less common. The reason fixed points for a win work in most sports is because things are run in either a tournament format, or a full round-robin league format. If you have a league format (e.g leaderboard), but can't guarantee everybody plays everybody, having fixed amounts for wins is unfair. One player may simply get easier matchups than another, despite them being "equally" good at the game. This should be reflected by their score. Winning a match vs an equally skilled opponent should give more points than beating a lesser skilled one. And beating a *more* skilled opponent should reward more points still. This is how the point system works in chess, on the Starcraft 2 ladder, and, in fact, most ladders in all eSports games. The points aren't allocated by some complex weighting of your roster: they are allocated by who you beat. Everybody will start with some placement matches (random opponents: this can of course be weighted by GP or some other measure of roster power), and be placed somewhere in the global ladder. After that, every match awards points based on the difference between your place on the ladder and your opponent's. There are different ways of tweaking this to better adjust to the specifics of each game, and how that should be done for SWGOH I don't know. But it seems by *far* the fairest way to create a leaderboard.
It's also worth mentioning that of course matchmaking will still play a large part, and in fact, how matchmaking is done can be almost entirely decoupled from how the ELO-esque system awards points for wins (and removes points for losses), or can be almost entirely based on this same score.
Finally, a little bit about the matchmaking algorithm this GA. I actually think it is very good and has created a very fun GA (at least for most of the people I've talked to). It can probably use some further tweaks to reduce the huge discrepancies in total GP between opponents, but in general looks to be a vast improvement in creating fun and exciting matches. That said, right now it doesn't award clever investment in your roster at all. A large of the game is resource allocation: choosing what toons to invest in. If you have 3m GP but it's all invested in toons that are relevant in the current GA meta, you could probably beat someone who has 5m GP. The GP is *only* a (rough) reflection of the total resources invested. Not whether those resources were invested "wisely" (I use this term very loosely). This choice in how to allocate resources is entirely ignored by the new matchmaking system (and was probably taken too far in the other direction previously). That means it doesn't matter if you are incredibly good or incredibly bad at this part of the game, as only your development of the top teams is even looked at right now, and you are matched with people who have invested similarly to you in their top teams. That does create intense and exciting battles. But it also cheapens the experience of investing cleverly in your roster. I say investing cleverly, but the same can be achieved by spending money. The matchmaking right now *also* reduces the reward for spending money on your pvp teams (you will simply be matched with other players who did the same... and won't be given the opportunity to pulverize f2p noobs). That seems directly counter to any goals this game has in terms of monetization.
I apologize for the long post. TLDR: tournament points need to be "decoupled" from the points you earn in a battle, and preferrably replaced with an ELO-like system for the ranking. Matchmaking needs to create interesting matches for this to be fun. Right now it's in a good spot, but probably disregards the resource allocation part of the game too much.
3
Replies
But devs seems to be adamant for doing matchmaking purely based on roster instead of separate PvP rating for reasons they don't want to explain.