Perfect scores should be less possible perhaps?

[Deleted User]

I was just thinking about how perfect scores should maybe be less possible. This may build upon my "Dispersing Wealth" discussion found here.

The presence of a high number of perfect scores creates a high number of people getting scores that fill up the top end of the leaderboards, who are usually the same people (myself included at times). I don't hate perfect scores but I feel they can sometimes create unwanted stress or frustration. I hear for platinum (as I am Gold), a loss in 1 point drops you down like more than 30 ranks and it really frustrates a lot of people. Perhaps it has a lot to do with the amount of effort/time spent vs the rewards they receive? And I am sure there are positive aspects but anyhow..

I remember when the AI (Greg) was super OP omnipotent and everyone hate it, as did I - but I did adapt.. then it was super weird when Greg become less than average Greg. The fact of the matter was how excessive he was. However, what if for a particular event your last 5 games, Greg was super OP and hard to beat, dependent on how many total points you had for the event thus far. For instance, if there are a total of 30 games for an event (probably a PvP one) and each game is worth 9 points (like RTO, I think?), if you have played 20 games and have a score range of 171 (1 loss) - 180 (perfect), your next 5 games results in facing a Greg with a higher difficulty. Then dependent upon your score after those 5 games, your next 5 games could be facing the omnipotent OP Greg. Perhaps this will reduce the number of perfect scores such that achieving a perfect score is actually a feat worth bragging about.

Ultimately, this may cause some more fluctuation of the top end of leaderboards, provide a challenge (for some), create less perfect scores so that maybe reward opportunities can be provided to others, possibly reduce some frustration/stress as perfect scores will be harder to achieve and become more of the norm with less people tie-ing for the same score so there are less total ranking drops because 50 people are tied for 1st place (although the opposite could be true as well when facing the harder Greg; but, initially I was thinking it would be like an "Oh, well. It's not so bad" since the negative connotations associated with not getting a perfect will be less severe, considering its not an isolated one-time lucky Greg I faced; but one, almost everyone at this level faces), etc etc.

I was wondering what people maybe thought of this? Honestly, I was writing this and thought... is this really a good idea? I like getting top tier prizes... hmmm.. not sure I will be happy with this change; but, perhaps, just perhaps, its worth it and may make the game more fun and reasonably competitive with more flux.

DBJones

I think this could be good for the game. Though, should they decide to do something similar in PvE, the score that triggers it should absolutely be at or above personal progression. Actually, that might be interesting in the one player PvE, get a perfect score to unlock the nuts OP greg version. I'd take on that challenge from time to time.

Quantius

If and only if they do something about Greg's cascades. It's great if they make it smarter, but not if the casting cost of cards is irrelevant.

khurram

Players seem to forget the cascades that they get themselves.

[Deleted User]

khurram said:

Players seem to forget the cascades that they get themselves.

Oh I remember.. I still see them now randomly. And trust me, I have rage fits... but don't worry my phone is still in one piece! However, I have noticed Greg starts matches usually making terrible matches (unless a lucky cascade falls); but as the match progresses, he seems to start making slightly smarter matches.

Quantius said:

If and only if they do something about Greg's cascades. It's great if they make it smarter, but not if the casting cost of cards is irrelevant.

True, but feels hard to implement as then they would have to code individual exceptions and specific decision trees based on individual cards in relation to others for Greg, which is a HUGE amount of work. The tremendous cascades are quite a problem but I guess it would end up being the point for the last few games to really filter out the best of the best or the luckiest - though I admit I am scared of this.

And one thing is certain, PWs like Koth, elspeth, saheeli, Huatli2, i.e. any with HUGE mana gains should not have the same super OP AI setting as say Gideon1 etc etc.

Sirchombli

I definitely cascade. I don't really think it's by design, but it seems fairly rare that Greg ever gets just one match in a round, though. I have also noticed a difference in behavior when I destroy gems vs when Greg does. Koth is a good example. In the koth mirror, Greg using his first turns into cascade city . I couldn't tell you how many times I've popped it and had nothing happen. Same story with the ever ubiquitous volcanic rambler. This is manageable with the derpy ai . When the ai was super cascades hurt a lot ,and they were constant. I did adapt, and I'd rather the game be too hard than too easy . I ,personally, feel like the ai difficulty should scale up with your mastery . There's really no reason that bronze and platinum should be playing the same Greg. 

jtwood

Bring back OP Greg! He was a blast!

Thuran

More than anything, I suspect the exsistence of Cycling contributes in no small part to the problem of too many perfect scores. Therefore, we should at least give them the benefit of the doubt, and wait to see if the trend continues after next week, when cycling and other very potent linear strategies go the way of smugglers copter. 

That said, the problem is also largely inherent to the nature of secondary objectives:

- Either you make them too luck based, such as HoD with "take 10 or less" which you straight up lose if koth fires a turn 1 inferno jet or something like that. 

This has the huge problem of making the game incredibly frustrating, as your placement becomes too luck-based and you can play flawlessly with a perfect deck, and yet lose purely due to an initial cascade. Or you get caught up in impossible objectives, simply because you have to take damage from a deck unable to do damage (met a cycling DB whose ONLY SOURCE OF DAMAGE was drake haven, so i was unable to take damage at all from him), or the classic example of meeting a creatureless deck when you need enemy creatures to kill. k

- Otherwise, you make the objectives manageable, which however leaves the door open for anyone to achieve those objectives consistently if they play well and build a good deck for them.

This has the advantage of being a way better play experience, as luck is less of a factor, and if you and your decks are good enough, you will get the kind of reward you deserve. However, this also means that anyone not good enough to get flawless, is pushed almost out of any significant prices, even if they miss a single objective, because the competition is as tight as it is. 

That said, i feel like RTO has hit a solid balance, with objectives that you should aim to achieve, but are also forgiving enough to not become frustrating. Some objectives are dangerous, and the "win with X or less life" is really horrible, but overall the event feels fairly enjoyable and fair.

The real solution, is probably to rethink the objectives and objective system itself, as well as improve the skill level of greg.

Let's face it, getting a 95%+ winrate is incredibly easy, and it is rare for the AI to provide any sort of real challenge, meaning that the base challenge of simply winning the game is almost trivial, even if the decks should in theory be of comparable strength.

So, rather than change objectives, which seems almost impossible to get perfect, we need the base game to become a lot harder at platinum/Diamond.

Frankly, they should probably just straight up give greg something like +1 to all mana gains, or some other advantage, to offset it's inherent disadvantage and try to make winning outright hard.


If you go almost flawless at platinum, it should be an achievement, not the norm.

By comparison, the top professional magic players typically have a winrate of about 60-65%.

If the game is providing an apropriate challenge, you should expect to only achieve a winrate of about 50%, 95% is just too darned high! 

Tilwin90

I for one am looking forward to cycling going away. Mark my words, after that happens, I wish everyone good luck in RtO. Those green and black nodes? If you complete them flawlessly you're either super lucky with your matchmaking and initial setup, or have one adamant collection & deckbuilding skill. 

wereotter

So here's the issue:

Not everyone wants a super hard AI. Some people want to play for fun and relaxation and don't want to have to fight for every win. Others want a challenge and want to feel like they've accomplished something by beating a super intelligent AI.

I think the devs of this game have decided you're more likely to run people off by having a highly tuned AI than by having a simple one. People like to win, and if they're winning more matches, especially when they start playing, they're more likely to stick with the game than if they're losing a lot.

One solution is to allow each player to set the difficulty they would like to face, but that comes with its own set of issues, namely that you couldn't ever do another ranked event as those who chose a lower AI difficulty would have an inherent advantage over those who didn't, and you couldn't divide by what difficulty you chose for tiers as new players might be paired off against established players who just want a more chill experience.

Additionally the reward structure right now is geared toward players winning the majority of their matches to make progression. If the AI were bumped up to make it more a 50/50 win-loss ratio, then they would also have to completely revise that setup.

I sympathize for people who miss the challenge from the previous AI, but I would have to say that the dumbing down is better for the long-term health of the game, and not just because I prefer lobotomized Greg.

sjechua

wereotter said:

One solution is to allow each player to set the difficulty they would like to face, but that comes with its own set of issues, namely that you couldn't ever do another ranked event as those who chose a lower AI difficulty would have an inherent advantage over those who didn't, and you couldn't divide by what difficulty you chose for tiers as new players might be paired off against established players who just want a more chill experience.

AI difficulty setting of smart / smarter / smartest could allow for a point bonus / multiplier. The smarter the AI difficulty setting, the greater the bonus ...

Tilwin90

sjechua said:

wereotter said:

One solution is to allow each player to set the difficulty they would like to face, but that comes with its own set of issues, namely that you couldn't ever do another ranked event as those who chose a lower AI difficulty would have an inherent advantage over those who didn't, and you couldn't divide by what difficulty you chose for tiers as new players might be paired off against established players who just want a more chill experience.

AI difficulty setting of smart / smarter / smartest could allow for a point bonus / multiplier. The smarter the AI difficulty setting, the greater the bonus ...

If by multiplying you mean give the current rewards to the smartest ai matches and a multiplier of -50% and -75% to the lower difficulties, sure!!! Let's not give them any more ideas to chip away the already selective and skewed rewards... 

wereotter

Tilwin90 said:

sjechua said:

wereotter said:

One solution is to allow each player to set the difficulty they would like to face, but that comes with its own set of issues, namely that you couldn't ever do another ranked event as those who chose a lower AI difficulty would have an inherent advantage over those who didn't, and you couldn't divide by what difficulty you chose for tiers as new players might be paired off against established players who just want a more chill experience.

AI difficulty setting of smart / smarter / smartest could allow for a point bonus / multiplier. The smarter the AI difficulty setting, the greater the bonus ...

If by multiplying you mean give the current rewards to the smartest ai matches and a multiplier of -50% and -75% to the lower difficulties, sure!!! Let's not give them any more ideas to chip away the already selective and skewed rewards... 

You can’t really change rewards based on AI settings unless it’s a one-time choice and then you’re only ever matched against people who chose your same difficulty level else you’ll run into potential deck imbalances. Additionally that would lead to issues regarding PvE events. Hence why I said that implementing that kind of system would have to kill all ranked events by design. 

Tilwin90

Not necessarily... When you enter an event you choose your difficulty and based on that you get your rewards adjusted... AI isn't everything.
Sure it would skew things unless each difficulty got its brackets but my point was this is an extra opportunity to further handicap rewards or up difficulty for no gain... 

span_argoman

You could scale points gained from the win and secondary objectives based on the difficulty chosen. That allows you to change AI difficulty even during the event.

And/or have certain Progression awards for winning X matches in <insert difficulty> during an event.

Or have the AI difficulty increased only after full individual Progression is met.

Or have the AI gain a bonus (or multiple bonuses) depending on your current total score in the event.