Can someone please explain how rankings work?

JanksMcJank

I played every match possible (due to slow node refresh).
I lost a battle, but did score a few Secondaries. I didn't optimize for Secondary Objectives though.

So everyone on my page of the leaderboard in bottom rank has 61 points.
I look at the top of the leader board and there are a lot of ties there as well.

So it seems that primarily it's about points. I need to prepare and master these secondary objectives to rank higher than bottom tier.
But bracket 1 is 1-5. 3-11+ (can't see past the first screen) have the same score.

How are ties settled? There must be a secondary metric in place as to why ZW2007 is in the #1 tier and FoxyBro is in the #2 tier, even though they have the same score.

Nitymp

Ties are settled in the worst way imaginable - first come, first served.

losdamianos

If you happen to be lucky and live in the timezone where normal people dont sleep/work during the time LAST charge start and can finish it quickly You have a fair chance of

Dont get me wrong I will always be cynical and will have to admit that increasing the number of nodes and starting with not fully charged is a big PLUS but we should not stop there and fix it fully by rewarding equally everyone who happens to get same points #STRONGER TOGETHER (sounds like super glue advert)

ZW2007-

I'm in 5th place because I woke up at 4 a.m. to play my last three games before going back to sleep. I just happened to score my final 21 points slightly faster than FoxyBro. It's a terrible tie breaker system.

ZW2007-

And now I am rank 6 because someone with a perfect score just finished off their nodes. I'm never getting a mythic reward again because I'm never even attempting to get up at 4 a.m. again.

bk1234

All these changes were supposed to make this better --- because some dev who "hypothetically" doesn't play the game thought "If we make more objectives, less people will tie". In actuality -- the best players are the best because you can throw out anything and they'll meet the objectives.

I don't know a single competitive player who doesn't optimize for objectives -- and in platinum -- perfect games are prized. I didn't lose a single game in the last event, I dropped a couple +1s and a +2 -- I think I lost 4 points total and I did not rank well.

But to your original question, to get the top placings you do have to optimize your decks to BOTH objectives and as PP said you have to be the first to get there. In Platinum and Gold, nothing short of a perfect score will get you the top prizes.

hawkyh1

ZW2007- wrote:

And now I am rank 6 because someone with a perfect score just finished off their nodes. I'm never getting a mythic reward again because I'm never even attempting to get up at 4 a.m. again.

the bigger objective is to aim for a perfect score instead of
getting up at 4am. I'm not saying that the current tie breaker
system is not flawed. but you could get your mythic reward yet.

HH

James13

Opted not to wake at 4 am.

Woke at 5:30 and finished with a perfect score.

Ranked 6th.

babar3355

hawkyh1 wrote:

ZW2007- wrote:

And now I am rank 6 because someone with a perfect score just finished off their nodes. I'm never getting a mythic reward again because I'm never even attempting to get up at 4 a.m. again.

the bigger objective is to aim for a perfect score instead of
getting up at 4am. I'm not saying that the current tie breaker
system is not flawed. but you could get your mythic reward yet.

HH

Wrong.. I got a perfect score in 5 straight events. Finished, 11th, 26th, 4th, 16th, 20th. 1 mythic and that's before they changed the time to 4:00 AM and I could actually race

madwren

hawkyh1 wrote:

ZW2007- wrote:

And now I am rank 6 because someone with a perfect score just finished off their nodes. I'm never getting a mythic reward again because I'm never even attempting to get up at 4 a.m. again.

the bigger objective is to aim for a perfect score instead of
getting up at 4am. I'm not saying that the current tie breaker
system is not flawed. but you could get your mythic reward yet.

HH

My perfect score netted me 18th in platinum, and I finished my nodes 48 minutes after the final refresh (at 1:48am).

ZW2007-

hawkyh1 wrote:

ZW2007- wrote:

And now I am rank 6 because someone with a perfect score just finished off their nodes. I'm never getting a mythic reward again because I'm never even attempting to get up at 4 a.m. again.

the bigger objective is to aim for a perfect score instead of
getting up at 4am. I'm not saying that the current tie breaker
system is not flawed. but you could get your mythic reward yet.

HH

I'd still need to get up at 4 a.m. to have a good chance. It just so happened that lots of players missed one point in gold this time around so I think there were only 3 perfect scores in my bracket. Terrors had a lot of perfect scores. I might cave and attempt getting up at 4 a.m. if I have a perfect score going into the final refresh but then again that's something I shouldn't have to do in the first place. The real kicker of it all is I lost 1 point on the very first game of the event when my human that had turned into a werewolf proceeded to turn into a frog.

nexus13

I finished 16 minutes after the last refresh and that placed me in 10th. It is literally luck of the draw, the fact that I didn't happen to have the objective cards in my opening hands meant my games lasted a few more turns and took a little longer dropped me a few places. I fortunately have a relatively better time zone than others. Last refresh for me was at 1am, I'm normally up to 12:30 or later anyway so it isn't a huge burden to play then but it's still a race to see how fast you can play the last set of nodes.

I like the other changes, 1 day events, nodes not starting full, ok with 8 hour timer. Those changes are positive but the tie issue is probably worsened by it and its still a race at the end at a set time that isn't particularly fair to large swaths of the player base. The nodes of power objectives will probably shake up the ties a bit but I think its because some of them are bad objectives and not appropriate for PvP instead of PvE.

Irgy

They really should just make it random. I can see why they might not want to just give the mythics out to everyone, but time is the least convenient tiebreaker (well ok, reverse-time, i.e. last in best dressed, would be far worse, though also hilarious). If it was random everyone could just relax and get the points when they have time. Then you get a mythic or you don't based on random chance and it just folds into the existing duplicate chance and general jank chance. It's not satisfying but it's no less unsatisfying, and much more convenient.

The current system is kind of random anyway. It's random, plus playing fast (which I guess is kind of a skill, but people play turn-based games specifically to avoid needing to care about this), plus being able to be available at a particular specific time. Though at least it's now one specific time (at the end) you need to be available now instead of two (at the start and the end previously).

It's not even that hard to code. Each time you do a score update for a player, you insert the player in a random position relative to the other people on the same score (note that if there's x players there already then there's x+1 positions to consider). It's not immediately obvious this is fair, but you can prove it by induction with this as the inductive step:

Assume each player has a 1/n chance of being in each rank among n currently tied players. Insert a new player. Obviously the new player has a 1/(n+1) chance of being in each rank because that's how you insert them. For the other players, consider the chance of them being in position x after the insertion. There's two possibilities:
A) They're in position (x-1) now, and the new player is inserted ahead of them. This has a (1/n) * ((x-1)/(n+1)) chance of happening (the first term being the chance of them being in that rank, the second term the chance of insertion ahead of them given that rank).
B) They're in position x now, and the new player is inserted after them. This has a (1/n) * ((n+1-x)/(n+1)) chance (for similar reasons).
The total adds up to:
(1/n) * ((x-1)/(n+1)) + (1/n) * ((n+1-x)/(n+1))
= (1/n) * ((x-1)/(n+1) + (n + 1 - x)/(n+1))
= (1/n) * ((x-1+n+1-x)/(n+1))
= (1/n) * (n/(n+1))
= 1/(n+1)

James13

Irgy wrote:

It's not even that hard to code. Each time you do a score update for a player, you insert the player in a random position relative to the other people on the same score (note that if there's x players there already then there's x+1 positions to consider). It's not immediately obvious this is fair, but you can prove it by induction with this as the inductive step:

I don't really like that. I think a better solution is to just randomly sort the list of ties at event close.

Everyone has an equal shot at the ranking and it's simply random who gets what. It'll be fair and impartial.

Your way has a slight unintended effect of serving to make it a little (and only a little) bit of a problem if you are early to finish you'll have a chance each time of moving down and never a chance of moving up. It'll actually reward finishing as late as possible in that regard.

hawkyh1

I'm thinking tie breaks could be ranked by who was using
the weakest deck/planeswalker. hence who had the most
skill by completing the same task with the most handicap.

HH

jetnoctis

The current tie breaker system has other negative side effects. I live in Australia so the only way for me to even have a chance to get top 5 before the latest update was to wake up at 4am for the start and end of every event. Since the update, the start time is no longer a problem (thank you so much for the partially filled nodes) and for the first time, the end time is better for me (8pm).

However, a lot of the US player base now has to contend with waking up at 4am regularly to try to win. This is not fun and the disrupted sleep schedule can give you health problems, such as (obviously) tiredness, memory issues, lethargy, being 'on edge' and generally lowered immunity.

I recommend being considerate of those who work really hard to get perfect against such difficult, often luck-dependent objectives. They can be allowed to get their sleep back by, at least in the interim, either awarding all perfects with the top prizes, or at least randomising the perfects' placement, as per the precious post's suggestion. Following on from the latter suggestion, players wouldn't have to get up at 4am for events and could complete the nodes at healthier times and everyone is equally likely to win. I really pity that others have to go through what we in other parts of the world have had to go through for so long and I hope that this change occurs for the benefit of all.

Ohboy

jetnoctis wrote:

They can be allowed to get their sleep back by, at least in the interim, either awarding all perfects with the top prizes, or at least randomising the perfects' placement, as per the precious post's suggestion. Following on from the latter suggestion, players wouldn't have to get up at 4am for events and could complete the nodes at healthier times and everyone is equally likely to win.

I remember when this bug was a feature and people complained like crazy.

genapp

Is it normal for there to be a large disparity in the top rankings?

For example ranks 1 & 2 are sitting at or just under 2000 points with the next closest ranks all being under 500.

Is that normal? Am I going to have to aspire to putting up those high of scores to get in the top spots?

Alve

genapp wrote:

Is it normal for there to be a large disparity in the top rankings?

For example ranks 1 & 2 are sitting at or just under 2000 points with the next closest ranks all being under 500.

Is that normal? Am I going to have to aspire to putting up those high of scores to get in the top spots?

It is quite normal. When someone gets a good head start, others usually don't try to catch up. If #1 has a 200 pts advantage, they most likely will notice when someone else starts chasing them and can put up more points to keep their spot. You'd need to catch them off-guard to make up the difference. So #1 tries to increase their lead as much as possible to feel safer (so that they can, you know, get some sleep without losing their place) and the lower places don't feel like racing after them. Then, if you have 2 or more people in the lead, they will constantly try to outscore each other, putting up even more points. Players who see they have no chance to get the first prize or don't really care about it will just make sure to stay in top 10, which usually leads to a huge difference between the top spots and the rest of top 10. You should still pay attention and if someone's scores look very suspicious, take screenshots and dm them to me

genapp

Alve wrote:

genapp wrote:

Is it normal for there to be a large disparity in the top rankings?

For example ranks 1 & 2 are sitting at or just under 2000 points with the next closest ranks all being under 500.

Is that normal? Am I going to have to aspire to putting up those high of scores to get in the top spots?

It is quite normal. When someone gets a good head start, others usually don't try to catch up. If #1 has a 200 pts advantage, they most likely will notice when someone else starts chasing them and can put up more points to keep their spot. You'd need to catch them off-guard to make up the difference. So #1 tries to increase their lead as much as possible to feel safer (so that they can, you know, get some sleep without losing their place) and the lower places don't feel like racing after them. Then, if you have 2 or more people in the lead, they will constantly try to outscore each other, putting up even more points. Players who see they have no chance to get the first prize or don't really care about it will just make sure to stay in top 10, which usually leads to a huge difference between the top spots and the rest of top 10. You should still pay attention and if someone's scores look very suspicious, take screenshots and dm them to me

Thank you for the great response. That makes perfect sense. I think I'll settle for sanity over higher rankings for the time being.