Replacing Tiers / Matchmaking

bk1234

In TG yesterday, I played 3 N1 decks -- I think only one had a mythic in them. I was using my Aj1 "FaceSmasher" deck (it's just as bad as it sounds -- Goggles -- HUF - Deploy - big dinos). 

Playing low mythic / original PW decks has become a norm in Platinum -- as players tier up too fast and find themselves in an environment where it must be very hard to compete in a game where there is still a huge veteran base. 

When it was introduced 2 years ago, the mastery and tier system was a good way to determine who the most competitive players were. It worked for awhile, but last summer when AER and Amonkhet hit, we had a couple of mass exodus' of top tier players. The current top player base is newer to the game (we have on on my top 10 team who has only been playing a few months) and they don't have the resources to compete in Platinum. In addition, many veteran casual players have slowly made their way to Platinum and also do not have the same resources that the veteran players have. 

This system no longer works. It is not fair. 

I spend a lot of time talking to players of all levels, deckbuilding, decision making, etc -- and one thing really pops out at me as a factor of how someone can compete at certain levels. 

Maxed Planeswalkers. 

A top tier player with 150+ mythics is more likely to have most or all of their PW maxed. 

Top players have the crystals -- we have or can quickly get the runes. I currently have 35/36 PW maxed and am maxing Karn. I put 20K runes into him yesterday, because I have a top-tier Koth deck that can grind that much in Story Mode with little effort. 

At other levels, players don't have the decks to be able to acquire runes at a fast pace. 

In addition, a newer or mid level player has to balance crystals and pick and choose which PW to buy while also having the opportunity to acquire new cards to be more competitive. Top tier players don't have to pick and choose which PW to pick up -- we have been around a long time and most of us picked them all up when they came out. We aren't catching up -- so our crystals are going to new cards versus old PW. 

I'm not saying this is the solution, however if we are going to discuss how to fix matchmaking, this needs to be a key point, as it completely differentiates how competitive a person can be in PvP. 

Mburn7

I've been saying we need a higher tier for probably close to a year now.  Platinum was full when Kaladesh came out, let alone now.  There are too many cards out for the same level of mastery.

Either add another tier, or increase the mastery needed to go up (or both).  The current scaling was put in at Innistrad levels, there's no way it was designed for the current collection

Bil

    There would be less platinum players with small colections if we had to master the 5 colors to get access to the next tier.
   It's not a big deal to get to platinum in a single color, it is way more complicated to master the 5 colors.
   I'm not saying we don't need some other changes to make matchmaking more eficient but that would at least prevent players from reaching platinum without a decent collection.

Mburn7

Bil said:

    There would be less platinum players with small colections if we had to master the 5 colors to get access to the next tier.
   It's not a big deal to get to platinum in a single color, it is way more complicated to master the 5 colors.
   I'm not saying we don't need some other changes to make matchmaking more eficient but that would at least prevent players from reaching platinum without a decent collection.

I dunno if that's really an issue, since platinum in one color only qualifies you for platinum in events with that color.  True a lot of events now are 5-color, but I don't think its THAT big of an issue.  

The issue (as I see it) is there is no difference between someone who has been playing 2 months hitting platinum and someone who has been playing 2 years and has been platinum since Innistrad.

There should be another tier for the top-top players, or just longer progression between the tiers.  The system can stay

bk1234

Mburn7 said:

Bil said:

    There would be less platinum players with small colections if we had to master the 5 colors to get access to the next tier.
   It's not a big deal to get to platinum in a single color, it is way more complicated to master the 5 colors.
   I'm not saying we don't need some other changes to make matchmaking more eficient but that would at least prevent players from reaching platinum without a decent collection.

I dunno if that's really an issue, since platinum in one color only qualifies you for platinum in events with that color.  True a lot of events now are 5-color, but I don't think its THAT big of an issue.  

The issue (as I see it) is there is no difference between someone who has been playing 2 months hitting platinum and someone who has been playing 2 years and has been platinum since Innistrad.

There should be another tier for the top-top players, or just longer progression between the tiers.  The system can stay

A new tier based on the current system is not the solution. 

Mastery should not determine that tier.

I have a HUGE collection with less than half of my cards mastered.

Tiers should be based on indicators of ability, mastery is simply and indicator of grinding. It does nothing to predict how well someone can do -- maybe total leveled PW and some kind of combination of collection size. That would indicate ability to perform well. 

This problem also still goes the opposite way as well. There are players in Gold who should be in Platinum. They get an unfair advantage -- if there were an ability based matchmaking system, this wouldn't be a problem. 

GrizzoMtGPQ

What if we had a system that was opt-in that got you to a higher level. A test of sorts, that pitted you against the type of decks that you'd be likely to face, that go you into the higher level. It would definitely feel like you achieved something if you could pull it off and it would keep the unworthy out.

Mburn7

bken1234 said:

Mburn7 said:

Bil said:

    There would be less platinum players with small colections if we had to master the 5 colors to get access to the next tier.
   It's not a big deal to get to platinum in a single color, it is way more complicated to master the 5 colors.
   I'm not saying we don't need some other changes to make matchmaking more eficient but that would at least prevent players from reaching platinum without a decent collection.

I dunno if that's really an issue, since platinum in one color only qualifies you for platinum in events with that color.  True a lot of events now are 5-color, but I don't think its THAT big of an issue.  

The issue (as I see it) is there is no difference between someone who has been playing 2 months hitting platinum and someone who has been playing 2 years and has been platinum since Innistrad.

There should be another tier for the top-top players, or just longer progression between the tiers.  The system can stay

A new tier based on the current system is not the solution. 

Mastery should not determine that tier.

I have a HUGE collection with less than half of my cards mastered.

Tiers should be based on indicators of ability, mastery is simply and indicator of grinding. It does nothing to predict how well someone can do -- maybe total leveled PW and some kind of combination of collection size. That would indicate ability to perform well. 

This problem also still goes the opposite way as well. There are players in Gold who should be in Platinum. They get an unfair advantage -- if there were an ability based matchmaking system, this wouldn't be a problem. 

Obviously some sort of ability-based system would be best, but would probably be difficult to implement.  Would it only be based on previous events?  Would it be like a golf handicap, where its based on your last 10 results?  Something else?

I don't think the current system is really THAT bad, remember the purpose was to encourage people to use different cards and not spam the same few combos all the time.  And with your huge collection you probably have a ton of mastered and almost-mastered cards just because of the number of different decks you've run over the years.

I'm not saying its a perfect solution, but it would definitely make everything much better and give Oktagon time to come up with something more permanent.

bk1234

GrizzoMtGPQ said:

What if we had a system that was opt-in that got you to a higher level. A test of sorts, that pitted you against the type of decks that you'd be likely to face, that go you into the higher level. It would definitely feel like you achieved something if you could pull it off and it would keep the unworthy out.

Wouldn't that also give people the option of staying in lower tiers though?

I don't think that would be great for the game. 

However this brings up a good point -- win % is also an indicator of ability and should be included somehow. 

Furks

Get rid of tiers. It's just a bad system on every possible level.

It makes the rich richer, by handing out the best progression rewards to people who need it the least. It doesn't make sense that a player of higher tier would get a better reward for doing the same event. It's especially painful for events with an entry fee where its a net loss to play them at a low tier.

Tiers also don't reflect actual skill, they merely show whether you grind a lot (or have spent money on the game) win/loss ratio is a much better metric representing skill. (like in every other game ever) 

babar3355

Seems like I have seen this movie before.

But we keep bringing up the same things, because you guys keep refusing to do anything about it.

We want an ELO system.  Card mastery is a great concept but should be utilized for a different purpose in game (e.g. leveling PW, earning additional rewards, etc)

Rhasget

Mastery is unrelevant as standard changes.
A player who had mostly good KAL/AER decks now might have a standard collection that would be better off in a lower tier (unlikely since AKH and HOU dragged on but you get the point).

The most fair is to incorporate an Elo-system. Then you will always face players in your current level (whether skill or cards or the like).

nexus13

Does ELO work when you aren't really head to head? 

Enygma6

Here’s an idea: separate the rankings between Standard and Legacy.  Keeps most of the current card mastery based systems intact, while encouraging use of newer card sets.  
Of course there would be a fair bit of overhead work: Determine what’s an appropriate value for each tier based on the current Standard card set. Having a process to recalculate all players’ mastery points when Standard rotates in/out sets.  
  
Lets old-school players who have been away a while or don’t care for new sets keep Platinum status in Legacy events, and allows those with more modest Standard collections have a chance to get a feel for new sets without immediately being thrown into the jaws of too many Mythic dinosaurs.

FindingHeart8

bken1234 said:

GrizzoMtGPQ said:

What if we had a system that was opt-in that got you to a higher level. A test of sorts, that pitted you against the type of decks that you'd be likely to face, that go you into the higher level. It would definitely feel like you achieved something if you could pull it off and it would keep the unworthy out.

Wouldn't that also give people the option of staying in lower tiers though?

I don't think that would be great for the game. 

However this brings up a good point -- win % is also an indicator of ability and should be included somehow. 

eeeeeeh I don't like this idea of win % being factored in.

Before quick battle was removed I had a ~95% win ratio, and that's even including 500+ defeats as I slowly gained the cards I needed to become competitive.

I would rather there be some form of randomness in the level of opponent I'm facing, rather than always only face players with 95% win ratios or higher.

While one could counter-argue, that your win percentage will go down as you lose more, and then you'll face easier opponents, to that I would respond that I don't want to have to go on a mandatory losing streak to face players who aren't running God combos.

As a strong advocate for beginner players, I'd be supportive of an extra tier or two to give newer players a buffer; but if you've invested enoguh time to get all your colors to platinum or beyond...well...you're not a beginner player anymore.

Gabrosin

The fundamental question is: what is the purpose of the tier system?  The only sensible purpose is to protect newer players from being paired against players with cards that the newer player will have a tough time winning against.  It's a set of training wheels, combined with an incentive to eventually graduate by mastering cards in order to achieve maximum rewards.

In light of this, it doesn't make sense to move to an ELO system, or add another mastery tier, or anything like that.  Players should simply be matched based on the relative size of their collections in the relevant format.

I've been playing the game for about four months now, and I've collected a little over 900 cards.  I've gotten most of the commons and uncommons in standard but I still have very few rares and mythics, because I've been spending my crystals on accumulating planeswalkers and I have put very little real cash into the game.  I don't have many legacy commons and uncommons because the freebie packs don't provide them and there are few opportunities to open older packs.  So when getting matched in standard events I'd probably get matched against other players who have been in the game for a few months; brand new players wouldn't have to face me because they own only a few hundred cards, and I wouldn't be stuck against high-resource or money-paying players because their regular investment into packs from the vault would lead them to have more cards than me.  As my collection grows and I fill out more of the Origins set, and I start having spare crystals for packs rather than PWs, I would start to pair against those players more often.  Meanwhile in legacy events, I would almost never see the players with tons of old mythics at their disposal.  Conversely, an old player trying to get back into the game wouldn't be stuck at a disadvantage in standard, but wouldn't be able to run up the store on players in legacy.

Card quantity is the most reliable metric for doing pairings.  There's a natural incentive to collect more cards for everyone; in fact it's almost impossible to avoid, so there would be no gaming of the tier system for easier matches.  It's trivially easy to calculate and maintain, and would require almost no new code for the devs, unlike instituting an ELO system and trying to maintain it across different events/formats.

Do away with the tiers.  Give everyone the same rewards for the same events.  Pair players based on collection size.

Sirchombli

By your definition of top tier, I'm floating in limbo. I have over 100 mythics . 30 of my 32 pws are maxed, but anybody who knows me knows I'm terrible at resource management. I tiered up too early, so I became very good at playing sub optimal decks. I feel that I have more top tier problems than mid level and I play to the level of a top tier .

I think I'm walking, talking evidence that platinum is wonky . I shouldn't have been in platinum when I got here and I've been noticing a lot of decks that appear to have the same issue II did when I jumped up too soon. I like to test janky original pw builds in tg , but my janky builds are stuffed with mythics and have synergy . I don't care about your legacy mythics . Huf me. Deploy me . I'll be fine , but that's because I spent so much time just losing to people who had better collections than me . I sometimes forget, but I didn't have much fun when I was in gold just getting wrecked by Olivia. I now understand there was more to it than that , but it sure felt like it at the time . 

I'm not entirely sure that the tier system really works , but there probably needs to be another tier . Another issue with platinum is that at this level, being f2p can be a liability. I've managed to find a balance, but the way the last 2 sets released was rough and I wound up spending more money than I would usually. I feel like the tier past platinum could be based on percentage of total cards rather than mastery . The mastery system is a bad metric because when you're new ,you want to master all of your cards . Then you end up in platinum when you're barely even ready for gold .

DBJones

The problem with matching by collection size is how much better certain cards are than others. For instance, a new player could get Omniscience from a daily reward HOU pack while their total collection is only a few hundred. That would be an extreme example, but the difference in card power, even within rarities, is huge in this game. Also, besides protecting new players, providing actual battles rather than slaughters to higher level players is another important part of matchmaking.

Personally, I think player's collections would make a good baseline (separate rankings for standard and legacy), modified by your performance in the last 5-10 relevant events you've battled at least one opponent in (specifically, I think the modifier should be your points divided by the total possible for the number of battles you actually fought.)

Furks

I'm sorry but the argument that win/loss isn't a good metric because you don't want to 'arbitrarily' lose games is like a chess master saying he only wants to play against beginners because he doesn't want to lose.

In every sport, game or videogame you win some and you lose some. You get matched against people of similar skill level because that's what's fair for both sides.

Obviously if the game moved to a system like that, rewards may need to be adjusted. On the other hand, you are still facing Greg and even though you might get matched to someone with a 95% win rate, Greg remains stupid and you will have an advantage over him.

The training grounds would also need to be kept track of separately because otherwise people could abuse intentionally loosing in there to make other events easier.

That said, if we can't have a system like that, collection size would be better than the current system, provided that the tiers and their different rewards go away. 

Sirchombli

Furks said

I'm sorry but the argument that win/loss isn't a good metric because you don't want to 'arbitrarily' lose games is like a chess master saying he only wants to play against beginners because he doesn't want to lose.

I didn't say win/loss wasn't a good metric . Mastery level has zero to do with your win/loss percentage . The mastery is about how many times you play with each card . Considering that story mode awards mastery points at a rate higher than pvp makes it inherently flawed . If they decided to base mastery levels on win percentage, I'd be all for it . It would generally be fair. At least as fair as it can be . That sounds like an acceptable solution and I wish I thought of it myself .

hawkyh1

if they based mastery points on win percentage at the
time of winning it may be a better representative of
mastery. a player who wins on off 50% with a fixed
deck is not as good as someone who can win
consistently with the same deck. it might also help limit
mastery by lucky cascades.

HH

Gabrosin

DBJones said:

The problem with matching by collection size is how much better certain cards are than others. For instance, a new player could get Omniscience from a daily reward HOU pack while their total collection is only a few hundred. That would be an extreme example, but the difference in card power, even within rarities, is huge in this game. Also, besides protecting new players, providing actual battles rather than slaughters to higher level players is another important part of matchmaking.

Personally, I think player's collections would make a good baseline (separate rankings for standard and legacy), modified by your performance in the last 5-10 relevant events you've battled at least one opponent in (specifically, I think the modifier should be your points divided by the total possible for the number of battles you actually fought.)

Sure, this is a thing that will wind up happening, but it will be the exception, not the norm.  New players will naturally be drawn to the strongest card they happen to open because they don't have many options.  I played the hell out of Majestic Myriarch because it was my only mythic and it let me take over a game if things went just right for it.  Will some lucky souls get a hold of masterpieces and ride them in every event?  You bet.  And if they're successful, they'll accumulate packs, leading them to open more cards, leading them to drift towards the appropriate levels of competition.

There are people now who sandbag their mastery level to stay in a lower tier.  They know enough to identify only the very best cards and play those over and over.  Pairing on collection size, the only way to sandbag and stay low would be to stop opening new cards, and what's the point of the game then?  And if you even out the rewards, there won't be any reason to do it.  You could make things a little easier if you could game the system to stay low, enabling you to win top rewards more frequently, which will result in... more cards.

At best, a player could simply refuse to open packs of cards and try only to accumulate jewels to let them open Rare+ and Mythic+ packs, keeping their collection ultra-small while still having a trickle of mythics coming in.  That won't necessarily let them have good decks; some commons and uncommons are as crucial as those mythics in order to stay balanced.  But if that turns out to be a concern, the concept could be modified to look for players with high percentages of mythics and masterpieces and bump them up the rankings a bit.  Perhaps keep it similar to current mastery, where each common you have is worth 2 points, uncommons 4, rares 8, mythics 16.  Regardless, the key is that you don't let players choose whether or not to grow their competition level as their resources expand.