Event rankings

Yvendros

How do event rankings work? All the players who achieved the same score, what determines their order?

First to achieve points? Past records? Random?

Kinda lame that I have the same score as everyone in the 25-50 range, but am in the 50-100 ranking...

Morphis

Same scores(ties)are ordered like a queue: the first to get that score will be placed first in the tie.
In the past there was a bug about it but has been officially fixed recently.
That's what patch notes reported but had not the chance to extensively test it.

On a preliminary not deep testing it looks like it was indeed fixed.

So basically who wins faster gets the better placement.

DuskPaladin

At this point, seeing how strong everyone is, and how heartbreaking it is to be tied with many other people and not get the same reward; maybe the best would be to get rewards based in total points, like:

350 points = Mythic + Fat pack,
345 points = Rare + Fat pack,
340 to 300 points = Rare + Booster pack,
etc.

You would be encouraged to play as well as possible, rather than it being a race against time. Leave that to Quick Battle, Nodes of Power should just be about how consistently you can win.

jimilinho_

DuskPaladin wrote:

At this point, seeing how strong everyone is, and how heartbreaking it is to be tied with many other people and not get the same reward; maybe the best would be to get rewards based in total points, like:

350 points = Mythic + Fat pack,
345 points = Rare + Fat pack,
340 to 300 points = Rare + Booster pack,
etc.

You would be encouraged to play as well as possible, rather than it being a race against time. Leave that to Quick Battle, Nodes of Power should just be about how consistently you can win.

I hit a perfect score and currently 19th! (Same bracket as a lot of you I imagine!)

Crazy!!

babar3355

Thumbs up for the Paladin.

The fun part about these PvP events is that you can take your time and work though battles rather than trying to blitz through them like in Quick Battle.

Yet now you have thrown all platinum players in the same bracket and make it a race to the finish line. A perfect score is barely going to get top 25 this competition.

Another option is get the AI to stop making insane plays like murdering their own creatures, always spamming first PW skill, or casting buffs on creatures with summoning sickness.

Yvendros

that's funny Paladin, i was just ranting about this to my wife (bless her for putting up with my games...) She suggested the same thing.

Or at the very least, it'd be nice to see progression rewards go further. maybe not as often as the current, but like every 50 points up to a perfect score. makes this sting a little more bearable...

shteev

13th place with a perfect score. Want to know what I was doing when my final nodes respawned? Was I in bed? At work? Maybe I was just playing really slowly? Nope.

https://scontent.xx.fbcdn.net/t31.0-8/f ... JpIjoidCJ9

This is absolutely unacceptable. Completely * unacceptable.

The very idea that the first 5 people to finish their last games as close to 1pm GMT will get the highest rewards is some of the tinniest * ... have ever inflicted on the player base. *** I am not at all happy with these small cats. Yeah, you heard me.

Are we the players all just going to stand for this? The problems is getting worse with every event, and it's completely unfair by any reasonable human definition of the word fair. Only an insane lunatic * would suggest otherwise.

*Admin edited for language.

Yvendros

shteev wrote:

tinykitty.

I used to think tinykitty was an auto-replace. i'm skeptical now...

Ohboy

shteev wrote:

13th place with a perfect score. Want to know what I was doing when my final nodes respawned? Was I in bed? At work? Maybe I was just playing really slowly? Nope.

https://scontent.xx.fbcdn.net/t31.0-8/f ... JpIjoidCJ9

This is absolutely unacceptable. Completely * unacceptable.

The very idea that the first 5 people to finish their last games as close to 1pm GMT will get the highest rewards is some of the tinniest * ... have ever inflicted on the player base. *** I am not at all happy with these small cats. Yeah, you heard me.

Are we the players all just going to stand for this? The problems is getting worse with every event, and it's completely unfair by any reasonable human definition of the word fair. Only an insane lunatic * would suggest otherwise.

*Admin edited for language.

I'm actually kind of OK with it. Even though timing is not the best for me, it makes sense that people who finish faster should be placed higher in the case of a tie.

BTW it looks like you're getting weak WiFi AND cell signal. Don't think it's the games's fault here.

yunnnn

Ohboy wrote:

it makes sense that people who finish faster should be placed higher in the case of a tie.

While this is true, the end result is that a game with a global community now requires a single time that all aspiring players need to be online (in order to win). This is incredibly tiresome and unhealthy for the game.

As it is, the current system requires me to wake up incredibly early in the morning so I can beat tiebreakers by finishing earlier. And I grudgingly do this, allowing me to finish in #1-5 rank every time.

This implementation needs to be changed because:
1. It's not fun to set the alarm for 5am. I'm sure it's also not fun to have the event end during work hours for people in Europe. The burnout is real. I could certainly stop, but this system isn't fun for anyone unless they have a schedule that is perfectly synced with this arbitrary event start time.
2. Other players who can't be online at a very specific time are at a ridiculous disadvantage.
3. Winning ASAP isn't supposed to be the goal of these events. I played a Hixus deck on my last node that took 15 minutes to grind out, and I dropped from 1st place to 4th due to tiebreakers (even though I still had perfect score).

Possible solution implementations:
1. Ties always get the same reward tier, rounded up. This might be hard to implement because NOP is often a 30-way tie.
2. Additional heuristics to determine rank. These can be: The rarity of cards used, turns taken, the number of matches, the overkill damage. I won't be the judge of what is fair here.
3. More secondaries that are challenging. NOP is a 30-way tie because its not hard to win every match. You're really just hoping the game doesn't crash in your matches.
4. Events are fired off in 4 different time brackets, and users can select which time bracket is more appropriate for them. PST players can pick something that suits them, and so forth. The ranking can still be in the same pool, but at least players can pick a better event start/end time.

Ohboy

yunnnn wrote:

Ohboy wrote:

it makes sense that people who finish faster should be placed higher in the case of a tie.

While this is true, the end result is that a game with a global community now requires a single time that all aspiring players need to be online (in order to win). This is incredibly tiresome and unhealthy for the game.

As it is, the current system requires me to wake up incredibly early in the morning so I can beat tiebreakers by finishing earlier. And I grudgingly do this, allowing me to finish in #1-5 rank every time.

This implementation needs to be changed because:
1. It's not fun to set the alarm for 5am. I'm sure it's also not fun to have the event end during work hours for people in Europe. The burnout is real. I could certainly stop, but this system isn't fun for anyone unless they have a schedule that is perfectly synced with this arbitrary event start time.
2. Other players who can't be online at a very specific time are at a ridiculous disadvantage.
3. Winning ASAP isn't supposed to be the goal of these events. I played a Hixus deck on my last node that took 15 minutes to grind out, and I dropped from 1st place to 4th due to tiebreakers (even though I still had perfect score).

Possible solution implementations:
1. Ties always get the same reward tier, rounded up. This might be hard to implement because NOP is often a 30-way tie.
2. Additional heuristics to determine rank. These can be: The rarity of cards used, turns taken, the number of matches, the overkill damage. I won't be the judge of what is fair here.
3. More secondaries that are challenging. NOP is a 30-way tie because its not hard to win every match. You're really just hoping the game doesn't crash in your matches.
4. Events are fired off in 4 different time brackets, and users can select which time bracket is more appropriate for them. PST players can pick something that suits them, and so forth. The ranking can still be in the same pool, but at least players can pick a better event start/end time.

A good solution really is to rotate event timings as someone suggested before, so everyone gets a shot at an optimum timing. Having this for quick battles would be nice too.

Of course then I guess people who had good timezones will complain about how it's harder now and not enjoyable and yadda yadda.

I think the best solution is the obvious one? Spend some time to seriously improve the AI or give us true pvp. Then we won't have 30 way ties and it'll be possible to award all tied winners. Or even better... A death match tiebreaker!

HunMike

I'm fully agreed with steev, including the language (btw. i placed 12th in the same bracket)

I finished at around 1:50 CET. This means, with the estimate of 6 min matches (loading screens included), I lost ~20 minutes to the perfect time (right at the beginning when using up the first charge on every node) and haven't got a chance. I looked at the leaderboard around 1:30 and the first 7 places were filled by then. So, the real time frame to being in the top5 I estimate less then 5 minutes this time. I find this very unsettling, to be polite. It's down to sheer luck, what opponent you got. With a more token-heavy ai, one doesn't have a chance to finish in time.

I've arranged that i had been able to play right at the start, and right when the last charges appeared on my nodes, so i can say i put in max effort and still doesn't had a chance. Not that it shook my world much, but still, i was very disappointed.

The game tends to crash at the start of an event (no wonder, see the above to get a picture about how crowded the server is in the first 30 mins), so some changes would be for the benefit of the devs on watch just as much as for us.

Several players suggested that we shouldn't start with full nodes. That alone would help avoiding a couple crashes at the starts. yunn was very polite :

2. Other players who can't be online at a very specific time are at a ridiculous disadvantage.

To make it more clear: they are shut out. (Lucky me with CET timezone...)
Not to mention that those in relation can easily imagine a better pastime right before going to sleep than an hour of obligatory MTGPQ

I find the extended progression rewards a great idea. Basically it would compensate the players who are highly competitive but had some dumb luck/missed a couple minutes. The last prog reward (for perfect score) should be some minor thing, like 3000 runes in order to avoid the uproar - let's face it, the ones with 340 points are more than capable to reach perfect score, they just got frozen 1-2 times (let it go, let it gooooo...), so we can say the game itself were biased against them- the reward for 325 (or whatever) should be the big one. For maintaining balance i suggest a small cut in leaderboard prizes in parallel with the extended prog rewards.

I know that my first suggestion would lead to an even heavier competition. The point is, we players aren't disappointed becouse of the harshness of the race, we are angry about the displayed chance of getting a mythic which isn't there in reality, despite the effort we had put in. Those who are on a level where they can optimize for timing in a scale close to seconds clearly don't need those rewards (Note that i'm not stating anywhere they dont deserve it for their additional effort). They have a deeper understanding on the game mechanics, which has nothing to do with how good they play (perfect) or how good their decks are (OP).

Another thought on that: Those, who put a minimum of net 6 hours (rounded down seriously) of their time per event, in which the first and the last half hours are tightly sceduled (rounded down again), achieve a perfect score, have a fair claim not to decide the distribution of the mythics on a less than 5 min difference. It's not like a marathon, where the result is based (more or less) purely on the runners themselves. Here, we have in-game freezes, server downtimes so it's near to impossible to plan ahead within this little margin of error. I wouldn't call that planning when i just hope for some hardcore players having a crash so i can squeeze myself in the top5 with my 35-50 min per 5 round performance (with the hope of luck avoiding those same freezes).

shteev

Ohboy wrote:

I think the best solution is the obvious one? Spend some time to seriously improve the AI or give us true pvp.

Those are good solutions but require a lot of coding.

If I might suggest a simpler one to implement straight away (which would combine with further fixes to the problem)... how about we only play opponent's decks that they have submitted for the Event itself? One reason we all get so many easy wins is that a large number of our games are played against untuned decks which our opponents may not have changed in months, or worse, are full of bad cards they have been mastering.

Ohboy

shteev wrote:

Ohboy wrote:

I think the best solution is the obvious one? Spend some time to seriously improve the AI or give us true pvp.

Those are good solutions but require a lot of coding.

If I might suggest a simpler one to implement straight away (which would combine with further fixes to the problem)... how about we only play opponent's decks that they have submitted for the Event itself? One reason we all get so many easy wins is that a large number of our games are played against untuned decks which our opponents may not have changed in months, or worse, are full of bad cards they have been mastering.

I don't think the wins is due to this. I didn't meet walkover decks last event, and everyone still got perfect scores. The AI is just incapable of making games hard enough to differentiate skill levels.

shteev

Ohboy wrote:

shteev wrote:

Ohboy wrote:

Spend some time to seriously improve the AI or give us true pvp.

If I might suggest a simpler solution to implement straight away... how about we only play opponent's decks that they have submitted for the Event itself?

I don't think the wins is due to this.

Well perhaps not many, but it would take 5 minutes to implement my solution, countless man hours to implement yours, and the two are not mutually exclusive.

shteev

yunnnn wrote:

2. Other players who can't be online at a very specific time are at a ridiculous disadvantage.

Actually now I think about it, it's two very specific times, isn't it? I have to play my first games at 5pm GMT as fast as possible to give me the opportunity to start my last games at 1pm GMT 2 days later, and then finish those as fast as possible too. Is that 5am and 1am for you? Doesn't seem convenient, or fair, to me.


This event the game has decided that one of my frequent opponents will be majincob. He's a pretty fearsome opponent as some of you will no doubt know (and all round great guy!). Frankly, THIS is why I play MPQ in the first place. Really hard matches against really good opponents. My first match this round was Ajani v Sorin, and it was EPIC. If MPQ was a sport (it's not, but this analogy seems to be gaining traction around here for some reason), then the spectators would have been on the edge of their seats. I mean yes, yes, of course I can outplay the AI, but it got some lucky cascades and gave me a really hard ride. Maybe his deck was better than mine? Who can say! And what was my reaction to completing this 20 minute marathon battle of the titans? Elation at my victory over the machine? No. It was to curse his name to the high heavens for dragging out my first game for so long that I'll never be able to get in the top 5 now even with a perfect. Tiny tiny tiny, kitty kitty kitty. If your ears are burning, majincob, that'll be why.

Corn_Noodles

yunnnn wrote:

1. Ties always get the same reward tier, rounded up. This might be hard to implement because NOP is often a 30-way tie.
2. Additional heuristics to determine rank. These can be: The rarity of cards used, turns taken, the number of matches, the overkill damage. I won't be the judge of what is fair here.
3. More secondaries that are challenging. NOP is a 30-way tie because its not hard to win every match. You're really just hoping the game doesn't crash in your matches.
4. Events are fired off in 4 different time brackets, and users can select which time bracket is more appropriate for them. PST players can pick something that suits them, and so forth. The ranking can still be in the same pool, but at least players can pick a better event start/end time.

#1 is completely right and should have been implemented from the start. They seem to be paranoid about giving out too many prizes.

#4 could work, but you'd need to account for coalitions as well. How would coalition ranking work with members in different time brackets? Perhaps the coalition leader would choose what time bracket all the players of their coalition use?

yunnnn

Corn Noodles wrote:

yunnnn wrote:

1. Ties always get the same reward tier, rounded up. This might be hard to implement because NOP is often a 30-way tie.
2. Additional heuristics to determine rank. These can be: The rarity of cards used, turns taken, the number of matches, the overkill damage. I won't be the judge of what is fair here.
3. More secondaries that are challenging. NOP is a 30-way tie because its not hard to win every match. You're really just hoping the game doesn't crash in your matches.
4. Events are fired off in 4 different time brackets, and users can select which time bracket is more appropriate for them. PST players can pick something that suits them, and so forth. The ranking can still be in the same pool, but at least players can pick a better event start/end time.

#1 is completely right and should have been implemented from the start. They seem to be paranoid about giving out too many prizes.

#4 could work, but you'd need to account for coalitions as well. How would coalition ranking work with members in different time brackets? Perhaps the coalition leader would choose what time bracket all the players of their coalition use?

The rankings are all still in the same pool; it's just that there are different start times. This means that rewards and points are all the same as before. At the end of all 4 time-groups, all points are collected into the same pool and the rankings are solved.

Another benefit to this model is that the event start time isn't going to overload the servers, as it should hopefully be distributed across different times.

This doesn't really solve the problem of tiebreakers, but it should solve the server load and convenience issues.

Corn_Noodles

yunnnn wrote:

The rankings are all still in the same pool; it's just that there are different start times. This means that rewards and points are all the same as before. At the end of all 4 time-groups, all points are collected into the same pool and the rankings are solved.

Another benefit to this model is that the event start time isn't going to overload the servers, as it should hopefully be distributed across different times.

The only way I can see this working without them redesigning and reprogramming events from the ground up, is for them to run 4 events with staggered start times in a group and if you join one of them, you're locked out of the other 3. The game servers right now can't seem to handle running two concurrent events smoothly so I doubt it's going to be able to run essentially 4 events concurrently without any problems. It seems to me there's some underlying design issues that would need to be addressed and that would take money and time. I'm all for having the staggered start times and it should have been done from the beginning, but I don't see them doing anything about it any time soon.

yunnnn

Corn Noodles wrote:

yunnnn wrote:

The rankings are all still in the same pool; it's just that there are different start times. This means that rewards and points are all the same as before. At the end of all 4 time-groups, all points are collected into the same pool and the rankings are solved.

Another benefit to this model is that the event start time isn't going to overload the servers, as it should hopefully be distributed across different times.

The only way I can see this working without them redesigning and reprogramming events from the ground up, is for them to run 4 events with staggered start times in a group and if you join one of them, you're locked out of the other 3. The game servers right now can't seem to handle running two concurrent events smoothly so I doubt it's going to be able to run essentially 4 events concurrently without any problems. It seems to me there's some underlying design issues that would need to be addressed and that would take money and time. I'm all for having the staggered start times and it should have been done from the beginning, but I don't see them doing anything about it any time soon.

Implementation:
Having 4 start times is not reprogramming it from the ground up. There is still a single event, but the users' startTime and endTimes are different, based on the start time that they selected when joining the event. I see it as a single new attribute, with some simple refactoring for existing methods to use this attribute, rather than the current hardcoded value (9am pst).

Introduce a new +hour offset field with 4 (or n) possible values; when the user joins the event, they pick one of the times, which the server persists as one of the +hour offset values. I picked 4 because it allows for flexibility, while not being so granular that it's too confusing to beginner users.

All the event code is exactly the same, except that the event is not startable until the user's startTime is reached, rather than the global startTime. Similarly, the eventEnd time needs to be also incremented with the +hour offset for the user. In other words, we're just refactoring in the start and end time-gates to use an additional persisted field.

Summary:
The real problem I see is that it might be confusing to users; however, with the convenience of better start times, it might still be worth implementing.

This is something I deal with in my actual line of work: Dealing with major usability issues while trying to keep a stable ecosystem of experienced, as well as beginner users. I agree that this is a hard problem to solve, but it can also be a profitable one. I don't really know the metrics of this game's IAP wrt Time Played (ie, do noobs spend more or do pros spend more), but I would suspect that the mid-high end players spend more money on this game (due to their desire to stay ahead), and making their lives less complicated will keep this user group from burning out and quitting.

yunnnn

yunnnn wrote:

Implementation:
Having 4 start times is not reprogramming it from the ground up. There is still a single event, but the users' startTime and endTimes are different, based on the start time that they selected when joining the event. I see it as a single new attribute, with some simple refactoring for existing methods to use this attribute, rather than the current hardcoded value (9am pst).

Introduce a new +hour offset field with 4 (or n) possible values; when the user joins the event, they pick one of the times, which the server persists as one of the +hour offset values. I picked 4 because it allows for flexibility, while not being so granular that it's too confusing to beginner users.

All the event code is exactly the same, except that the event is not startable until the user's startTime is reached, rather than the global startTime. Similarly, the eventEnd time needs to be also incremented with the +hour offset for the user. In other words, we're just refactoring in the start and end time-gates to use an additional persisted field.

Summary:
The real problem I see is that it might be confusing to users; however, with the convenience of better start times, it might still be worth implementing.

This is something I deal with in my actual line of work: Dealing with major usability issues while trying to keep a stable ecosystem of experienced, as well as beginner users. I agree that this is a hard problem to solve, but it can also be a profitable one. I don't really know the metrics of this game's IAP wrt Time Played (ie, do noobs spend more or do pros spend more), but I would suspect that the mid-high end players spend more money on this game (due to their desire to stay ahead), and making their lives less complicated will keep this user group from burning out and quitting.

Weird thing... I just started playing marvel pq and it looks like this is exactly how their events work; I joined one and it made me pick an end time.