General Weapon Refinery discussion/speculation/creation thread

[Ice Dragon]

8 minutes ago, Etheus said:

Which is why units do not exist in a vacuum. You might (MIGHT) have a point if this game was a 1v1 dueling game. It is a team-based strategy game involving four units. And that ranged unit who has lower stat points may not be able to safely engage on a specific enemy unit in a certain situation, but they more than make up for that by taking out other units on the enemy team, supporting allied units who CAN safely engage the problem unit, and reaching the enemy backline with greater range and movement. We cannot score the value of that in a meaningful way and we should not be trying to.

A unit that cannot kill a specific enemy unit that they should be able to handle (for example, green units should be able to handle all blues) is clearly worse than a unit that can. Being able to run to the back lines to snipe something is not significantly more useful than being able to straight up kill the thing right in front of you.

Furthermore, if you need support to perform at the same level as another unit, that's effectively equivalent to two units on your team being worth the same as the one other unit, and it makes sense to rate the unit that needed support as less than the one unit that didn't.

We can score this in a meaningful way.

 

19 minutes ago, Etheus said:

The flaw in your argument is that to maintain whales, one must also maintain minnows. If there are no counter-balancing mechanics in place to keep the lesser sea creatures playing, there is no game and nothing for those whales to compare themselves to. 

There is no such flaw. I'm well aware of the fact that whales need small fish to exist, which is exactly why @bottlegnomes's assessment that the Arena may as well just be a pool for whales is a terrible assessment.

If you make a pool for whales and only whales, they will just leave because there are not enough other whales to provide sufficient content and to fill up the lower rankings.

 

21 minutes ago, Etheus said:

That is highly reductive. No one is saying "burn it all." We're saying "this works, and this doesn't. Let's replace what doesn't work with something that might."

This idea of maintaining the system as is without critique, which basically boils down to "it exists as is therefore it should exist as is" is how we end up with most of the world's problems, large and small.

You're burning down the use of stat totals to the ground despite the fact that the ship has already sailed with the Duel skills.

Unless you have a way of resolving the fate of the existing Duel skills, you cannot get rid of the system that they apply their effect towards.

I'm not arguing this to stick up for the status quo. I'm arguing this because changing the status quo requires consideration of all of the related features and downstream effects. Even small changes in an isolated area are not safe to make without understanding the downstream effects.

[Corrobin]

15 minutes ago, Ice Dragon said:

Furthermore, if you need support to perform at the same level as another unit, that's effectively equivalent to two units on your team being worth the same as the one other unit, and it makes sense to rate the unit that needed support as less than the one unit that didn't.

What constitutes as "support"? Rallies and movement assists, sure, but does having someone stand next to the unit to activate a Bond, using Owl tomes, or Spurs and Drives count?

[Fabulously Olivier]

6 minutes ago, Ice Dragon said:

A unit that cannot kill a specific enemy unit that they should be able to handle (for example, green units should be able to handle all blues) is clearly worse than a unit that can. Being able to run to the back lines to snipe something is not significantly more useful than being able to straight up kill the thing right in front of you.

Except that it very much is. Matches can and often are won off the back of proper use of Gravity Kiting, Pain & Dance combos, and other means of slowing, strategically eliminating, or debuffing the backline.

Again, it is not the matchmaking system's job to tell us how to build our team in a teambuilding game. It is our job to do so in a manner that is effective based on the current metagame.

6 minutes ago, Ice Dragon said:

Furthermore, if you need support to perform at the same level as another unit, that's effectively equivalent to two units on your team being worth the same as the one other unit, and it makes sense to rate the unit that needed support as less than the one unit that didn't.

We can score this in a meaningful way.

So Bladetomes & cavalry buffs are not powerful because they require multiple units to deliver great effect? What utter nonsense.

6 minutes ago, Ice Dragon said:

 

There is no such flaw. I'm well aware of the fact that whales need small fish to exist, which is exactly why @bottlegnomes's assessment that the Arena may as well just be a pool for whales is a terrible assessment.

If you make a pool for whales and only whales, they will just leave because there are not enough other whales to provide sufficient content and to fill up the lower rankings.

 

You're burning down the use of stat totals to the ground despite the fact that the ship has already sailed with the Duel skills.

Unless you have a way of resolving the fate of the existing Duel skills, you cannot get rid of the system that they apply their effect towards.

I'm not arguing this to stick up for the status quo. I'm arguing this because changing the status quo requires consideration of all of the related features and downstream effects. Even small changes in an isolated area are not safe to make without understanding the downstream effects.

Now here is an actual argument. The significant misstep that was Duel skills throws an actual non-arbitrary obstacle against any attempt at rework.

For this, we must look into a means to maintain the spirit of these skills. As these skills are directly tied to arena score, we have to come up with a comparable functionality. I have some rough ideas for a system without BST scoring.

- HP +5. If unit dies, arena score will not be affected. 

- HP +5. Unit constitutes a bonus unit for the purpose of arena scoring.

- HP +5. If unit's merge level is below 10, unit constitutes a +10 merged unit for the purpose of arena scoring.

- Some combination of the above.

[Ice Dragon]

1 minute ago, Corrobin said:

What constitutes as "support"? Rallies and movement assists, sure, but does having someone stand next to the unit to activate a Bond, using Owl tomes, or Spurs and Drives count?

Standing next to a unit at the beginning of the turn for passive buffs is typically rudimentary support and should simply be expected within reason. For example, Hone Cavalry is rudimentary, but both Hone Cavalry and Fortify Cavalry is borderline. Two stacks of Litrowl on enemy phase is rudimentary, but only one stack on player phase. For 2-range effects, two stacks is rudimentary for buffs that activate at the beginning of the turn or combat buffs on enemy phase.

The kind of support I'm referring to that are above and beyond an expected level of unit independence are things like Pain to chip a unit's HP low enough to kill, Candlelight to attack a unit without risk of counterattack, or Dance to give the unit an extra turn to finish a weakened unit off.

 

1 minute ago, Etheus said:

Except that it very much is. Matches can and often are won off the back of proper use of Gravity Kiting, Pain & Dance combos, and other means of slowing, strategically eliminating, or debuffing the backline.

Again, it is not the matchmaking system's job to tell us how to build our team in a teambuilding game. It is our job to do so in a manner that is effective based on the current metagame.

Matches that are won with Gravity and Pain are typically also winnable (in fewer turns, mind you) by using units with superior stats that can flat-out kill the enemy. Both are viable for the end result of winning, but one doesn't rely as heavily on synergistic effects with other team members to achieve that result.

I'm not saying that those strategies aren't viable. I'm saying they're not more efficient.

 

5 minutes ago, Etheus said:

So Bladetomes & cavalry buffs are not powerful because they require multiple units to deliver great effect? What utter nonsense.

Refer to the response to Corrobin earlier this post.

Also, Litrblade is no longer the overwhelming powerhouse it once was. It's still good, but too many units now have enough bulk to deal with it more effectively. For example, Fjorm is nearly impossible to kill with a green tome when Ice Mirror is charged and will kill on the counterattack due to the amount of mitigated damage added.

 

18 minutes ago, Etheus said:

Now here is an actual argument. The significant misstep that was Duel skills throws an actual non-arbitrary obstacle against any attempt at rework.

Sorry about being so roundabout. I thought I mentioned it earlier, but I must have either missed mentioning it or deleted it from an earlier post during editing.

 

21 minutes ago, Etheus said:

For this, we must look into a means to maintain the spirit of these skills. As these skills are directly tied to arena score, we have to come up with a comparable functionality. I have some rough ideas for a system without BST scoring.

- HP +5. If unit dies, arena score will not be affected. 

- HP +5. Unit constitutes a bonus unit for the purpose of arena scoring.

- HP +5. If unit's merge level is below 10, unit constitutes a +10 merged unit for the purpose of arena scoring.

- Some combination of the above.

Of these the only one that is at all comparable in value to the current effect in the current format is the second. However, they all still have some issues:

1. Most players already play with the expectation that their units don't die. While this would be insurance against it, it's more valuable for "worse" players than experienced ones.
2. The biggest problem with this is that it has the same problem as removing the Arena bonus unit system. This effect ensures that you always have a bonus unit on your team, ensures that you can choose a unit that is combat-capable to be a bonus unit for kills, and removes the need to pull for new units.
3. This one is weird. It's most valuable for players with fewer resources rather than players with more resources, but players with more resources can use this to bypass the need to pull for +10 merges of new characters. It also has a similar problem as (2) where you can put the Duel skills on Alfonse, Sharena, Anna, and Fjorm and get away with never needing to pull for new units.

And none of them are viable for use on an Arena defense team.

[XRay]

2 hours ago, Hilda said:

That might be true for Arena Assault, but your ponies wont go against 4x Armors in normal Arena because of BST scoring soooo. The whole problem with Arena is the BST bonus Armorers get. I am fine with the SP skill being calculated into points but base BST?! Come on get rid of that fuckery.

I want to get rid of SP scoring because Firesweep players like me are being penalized for it. Firesweep Weapons cannot be refined and Life and Death does not score as high as others. I guess most units got the same issue with Aether too, but it is especially terrible for Player Phase units that want to be in Desperation range and the healing is entirely pointless for Firesweep nukes that almost never take damage in the first place.

Firesweep healer players who do not have Atk/Spd Push are also being unfairly punished as Attack +3/Speed +3 score ridiculously low.

2 hours ago, Etheus said:

If players wish to not use an entire unit on their team, that is their prerogative and the ranking system should not be penalizing that. They are only gimping themselves by not properly building around their bonus unit and relying on a core of three. That's a self imposed challenge, and if they can maintain peak arena performance despite effectively being down a character, all the power to them.

Running competent three man cores with a random Bonus unit tacked on has been the standard for quite a while now. I personally do not see it as much of a self imposed challenge since running a three man team while maintaining peak performance is pretty doable.

For Firesweep teams, two man teams are very much viable against most team compositions besides pony teams. Against pony teams, you often need a third unit to help speed up the killing since enemy ponies can still rush and swarm you easily on pre-trench Arena maps if you do not take them out quickly enough.

1 hour ago, Ice Dragon said:

Actually, it is to a reasonable degree. Ranged units have a positioning advantage, but they lose out on combat performance due to simply not having enough points in stats. Being able to out-position an opponent doesn't help if you simply cannot safely engage them in combat.

 

48 minutes ago, Ice Dragon said:

Furthermore, if you need support to perform at the same level as another unit, that's effectively equivalent to two units on your team being worth the same as the one other unit, and it makes sense to rate the unit that needed support as less than the one unit that didn't.

In my opinion, it is the combination of ranged combat, Firesweep Weapons, Dance/Sing Reposition combo, and ideally flier movement together that makes a Player Phase team so good. Losing out on some combat performance is definitely worth being able to engage any enemy safely no matter what their skill set is.

While not all regular ranged nukes are Firesweepers, regular nukes are still good enough to kill non Distant Counter units with Brave Bows and Blade tomes, and most have enough bulk to survive other ranged units' counter attack.

Alone, nukes and Dancers/Dancers are each worth less than a armor unit combat wise, but the combination of a nuke and Dancer/Singer together is worth about the same as two armor units in my opinion, and easily worth three armor units or more if the nuke is of the Firesweep variety. The nuke and Dancer/Singer combo as a whole is worth more than the sum of its parts.

[Fabulously Olivier]

13 minutes ago, Ice Dragon said:

Standing next to a unit at the beginning of the turn for passive buffs is typically rudimentary support and should simply be expected within reason. For example, Hone Cavalry is rudimentary, but both Hone Cavalry and Fortify Cavalry is borderline. Two stacks of Litrowl on enemy phase is rudimentary, but only one stack on player phase. For 2-range effects, two stacks is rudimentary for buffs that activate at the beginning of the turn or combat buffs on enemy phase.

The kind of support I'm referring to that are above and beyond an expected level of unit independence are things like Pain to chip a unit's HP low enough to kill, Candlelight to attack a unit without risk of counterattack, or Dance to give the unit an extra turn to finish a weakened unit off.

 

Matches that are won with Gravity and Pain are typically also winnable (in fewer turns, mind you) by using units with superior stats that can flat-out kill the enemy. Both are viable for the end result of winning, but one doesn't rely as heavily on synergistic effects with other team members to achieve that result.

I'm not saying that those strategies aren't viable. I'm saying they're not more efficient.

 

Refer to the response to Corrobin earlier this post.

Also, Litrblade is no longer the overwhelming powerhouse it once was. It's still good, but too many units now have enough bulk to deal with it more effectively. For example, Fjorm is nearly impossible to kill with a green tome when Ice Mirror is charged and will kill on the counterattack due to the amount of mitigated damage added.

 

Sorry about being so roundabout. I thought I mentioned it earlier, but I must have either missed mentioning it or deleted it from an earlier post during editing.

 

Of these the only one that is at all comparable in value to the current effect in the current format is the second. However, they all still have some issues:

1. Most players already play with the expectation that their units don't die. While this would be insurance against it, it's more valuable for "worse" players than experienced ones.
2. The biggest problem with this is that it has the same problem as removing the Arena bonus unit system. This effect ensures that you always have a bonus unit on your team, ensures that you can choose a unit that is combat-capable to be a bonus unit for kills, and removes the need to pull for new units.
3. This one is weird. It's most valuable for players with fewer resources rather than players with more resources, but players with more resources can use this to bypass the need to pull for +10 merges of new characters. It also has a similar problem as (2) where you can put the Duel skills on Alfonse, Sharena, Anna, and Fjorm and get away with never needing to pull for new units.

And none of them are viable for use on an Arena defense team.

How is this?

HP +5. The arena scoring value of unit will match that of the highest scoring unit on your team, if unit is not the highest scoring unit on team.

Also, since when was arena defense scoring relevant? Any defense win with a bonus unit is as valuable as another, regardless of score.

[bottlegnomes]

Well, screw it, I don't feel like figuring out how to get this to work across two pages, so responses in bold.

1 hour ago, Ice Dragon said:

No, you don't. You need a team that can babysit. As long as your bonus unit can deal at least 1 damage to anything, you're good.

I can maintain Tier 21 without too much difficulty with even low-merge unoptimized bonus units.

Fair enough. I am going to make an assumption here, which is not meant to be taken as a rebuttal to this particular point, but does affect my subsequent statements. That being, from everything I've heard, it's impractical to be able to maintain T21 with just F2P units, i.e. 3 170 BST armors with a refined weapon, 400 SP assist, Aether or Galeforce, 3 240 SP skills, and a 240 SP seal.

That is only one facet of it. It's like saying college exists for the purpose of getting people higher-paying jobs. Yes, it does do that, but it does a lot more than that.

The number of people who make that argument on a daily basis is astounding. I literally can't count the number of times I've heard someone say, "If you didn't want debt you should've picked a different major." That said, I don't agree with them, and...

The Arena serves two primary purposes from my perspective. The first is to be a means of encouraging spending through the use of bonus units and relatively lucrative rewards. The second is to be a goalpost for players to set their sights on. The latter purpose is why there are always free units in the list of bonus units every week. If the game mode were only a means of encouraging spending, free units would never appear as bonus units as that would be counterproductive.

I never said it was the only reason. I said they might as well explicitly acknowledge that that's it's purpose: 

"[IS could] Just come out and say that arena is meant to encourage whaling instead of adding more and more features that tilt it that way without actively acknowledging it."

Notice I never said sole purpose and that I said encourage, not necessitate. Brackets added to clarify my sentiment.

Anyway that goalpost exists to encourage investment and spending. It's basically saying, sure you can get here, but you could stay here if you spent. The free unit is basically a carrot that serves two purposes. One, give F2P people a taste of where they could get with spending, re. the previous sentences. Two, keep the ones who won't be swayed into spending around to make whales happy.

Because "toss everything in a fire and start all over" is not constructive in this situation.

I never said toss everything. Rarity and merges would still matter.

Being serious, neither is saying, "If you can't think of a better way, you shouldn't voice your dislike":

"I see a lot of people saying what parts of the matchmaking rating should be removed or what should be considered, but no one has yet come up and actually given a concrete 'better' means of calculating a matchmaking rating without handwaving away the 'figure out how "strong" a unit is' part, which is kind of the most important part of the calculation. Have any of you actually tried tackling coming up with the calculation to determine how 'strong' a unit is?"

If I'm misinterpreting that quote from your post, my apologies, though I would like some clarification on what you meant.

Arena clearly has flaws, and IS is just appears to be compounding those flaws which is where most people take issue. Yeah, there's not an easy way to fix it that will satisfy everyone, but not acknowledging that or poo-pooing expressions of what should be changed doesn't help anything (again, my apologies if I misinterpreted your statement). I personally would get a kick out of IS being that bald-faced, but if we're talking an actual approach, I don't think there's any gameplay issue with @Etheus's approach of just go based on team strength. If I've put enough dedication into my team of Clarise, vanilla F!Robin, Tobin, and Saias that it can beat a team of optimized B!Hector, V!Hector, H!Myrrh, and H!Jakob, what's it matter that the unit strength in a vacuum is lopsided? Yeah, changing a unit would alter that, but if I'm thinking of changing a unit, I should either factor in alterations or have to suffer a slip in standing for a bit. That's no different than someone in a fighting game needing to practice with a new character outside competitive play or slip in rankings until they get used to the character. Maybe it wouldn't encourage spending enough, but eh, there can still be other factors.

 

37 minutes ago, Ice Dragon said:

which is exactly why @bottlegnomes's assessment that the Arena may as well just be a pool for whales is a terrible assessment.

Again, for the record, I never said just. My hypothetical formula would probably place a 5* +5 3*-base somewhere around a +1/2  or so 5* exclusive. Setting aside the potential bad taste it would leave in people's mouths, what's the difference between my +10 Reinhardt not being enough to get into T21 because his BST isn't high enough and not being enough because his base rarity isn't high enough? Either way, I have to trade unit selection for security in staying at the top. For players like me personally who wouldn't want to invest in units just because they score well, not much changes. For players who are perfectly okay with dumping a ton into arena without budgetary concerns, they get a different, most likely larger, unit pool. The only people who really suffer are the F2Ps who dumped a ton into very specific units for the sole purpose of scoring well as most of those units would have their value drop somewhat significantly since they're generally 3* bases. The top investors stay at the top, those willing to play to the system remain just outside, and those who are apathetic fall into whatever place they do. Nothing really changes other than who falls into what group.

All that said, again, I never actually said it was a good idea. I said it was a possibility if they wanted to be incredibly bald-faced. In general, I'm actually fairly happy with FEH. It's a bit stagnant in types of content, but outside of arena, most of those content types are enjoyable and accessible without locking significant rewards behind major hurdles. Heck, even arena, I have a bit of a gripe with, but my bigger gripe is the previously mentioned favoritism for armors. Armors have been the dominant movement type for almost a year at this point and IS has done nothing to mitigate that. A week ago I would've been more irritated about it, but the introduction of the guard weapons seems like a sign that they might be intending to. And yeah, I'm fully aware that a not insignificant portion of my feelings there is due to my cavalry bias, but again, cavs and fliers have 1 cooldown speed-up skill, which is basically what the current meta revolves around.

[bottlegnomes]

Addendum because I can't edit: The issue I personally take with bonus units being prioritized is that it serves more as making players babysit or bend over backwards to make them viable rather than what it seems like it was initially intended for: to encourage spending and to encourage unit variety. Boosting score encourages people to pull for the units, and altering them encourages people to see how different units fit into what should theoretically be their favorite team. If one week, the bonus unit sits in the back because the player doesn't have a use for them, then so be it. That's not a unit they'd necessarily like. Another week a different bonus unit might play a major role and the player might be encouraged to add another unit to their rotation.

@Ice Dragon Regarding the cap on bonus unit points, I like the suggestion, but even 10 kills seems a bit much. That's still half of all the kills, which means that for full rewards, they have to be at minimum tied for most kills instead of just pulling their own weight. If it were meant to solely have them pull their own weight, 5 seems fair, one per unit per mach, but that seems a little low and doesn't really give the unit much of a spotlight, which is what it seems like bonus unit status is meant to do. Maybe like 7 or 8? That's about 1.5 per match and means they definitely get some shine and players can't just let them rot in the back. I know that probably seems like an almost trivial difference, but that means that they don't have to be at minimum one of the top two KOing units for the week.

Edited October 10, 2018 by bottlegnomes

[Ice Dragon]

2 minutes ago, Etheus said:

How is this?

HP +5. The arena scoring value of unit will match that of the highest scoring unit on your team, if unit is not the highest scoring unit on team.

Much better, actually. I may worry it's a bit too good, but that's not a complaint I care to argue at the moment.

 

3 minutes ago, Etheus said:

Also, since when was arena defense scoring relevant? Any defense win with a bonus unit is as valuable as another, regardless of score.

It isn't, but in the current system, Duel skills on your defense team aren't there to raise your defense score, they're there to introduce a unit to your defense team that offense teams in your score range are not used to having to deal with.

For example, I've recently seen Summer Tana, Celica, and Julia in the 750+ range.

 

12 minutes ago, bottlegnomes said:

That being, from everything I've heard, it's impractical to be able to maintain T21 with just F2P units, i.e. 3 170 BST armors with a refined weapon, 400 SP assist, Aether or Galeforce, 3 240 SP skills, and a 240 SP seal.

I managed to comfortably maintain Tier 21 two weeks ago with a team of

• +2 Laegjarn: 170 bucket with Duel, 1,700 SP bucket, 1 blessing stack
• +10 Hardin: 170 bucket, 2,100 SP bucket, 1 blessing stack
• +10 Legendary Hector: 170 bucket, 2,100 SP bucket
• +10 Legendary Robin: 160 bucket, 2,000 SP bucket

with only 14 out of 20 bonus kills.

I haven't bothered to optimize recently, considering Hardin is the only unit on my team running an Assist skill worth more than 150 SP.

My Hardin and Hector both have superbanes that keep them at 174 total stat points since I'd rather take a bit of performance over a bit of score. I'm not hurting for score, though I probably worry about cutting it close to the cutoff way farther from the cutoff than most people would.

 

23 minutes ago, bottlegnomes said:

If I'm misinterpreting that quote from your post, my apologies, though I would like some clarification on what you meant.

My biggest problem is that most people who say they want the Arena scoring system to be based on performance instead of just unit stats don't provide a concrete suggestion for how to design a scoring system that is based on performance. I think most of the time it's suggested, people don't realize how truly difficult it is to programmatically determine how well a team performs. And it doesn't help that it's at least partly recursive with the state of the meta.

Ladders sidestep this by experimentally determining how well a team performs, but Arena reward cycles are too short for a ladder format to work.

[bottlegnomes]

4 minutes ago, Ice Dragon said:

I managed to comfortably maintain Tier 21 two weeks ago with a team of

• +2 Laegjarn: 170 bucket with Duel, 1,700 SP bucket, 1 blessing stack
• +10 Hardin: 170 bucket, 2,100 SP bucket, 1 blessing stack
• +10 Legendary Hector: 170 bucket, 2,100 SP bucket
• +10 Legendary Robin: 160 bucket, 2,000 SP bucket

with only 14 out of 20 bonus kills.

I haven't bothered to optimize recently, considering Hardin is the only unit on my team running an Assist skill worth more than 150 SP.

My Hardin and Hector both have superbanes that keep them at 174 total stat points since I'd rather take a bit of performance over a bit of score. I'm not hurting for score, though I probably worry about cutting it close to the cutoff way farther from the cutoff than most people would.

Huh, go figure. The way F2P T21 people talk, it seems like a struggle to stay there at all. Basic breakdown, according to that, you should need about:

+8, 167 BST, a 350 SP refined weapon, a 300 SP assist (Tailtiu), a 500 SP special (Chrom and Cordelia), 1 240 SP skill (e.g. QR), two 200 SP skills (almost literally anyone), and a 240 SP seal (a nonfactor) for a total of 2,030 SP. Obviously 167 needs to be rounded up, and you can dump the bonus unit to a +2, but those can be mitigated with more 240 SP skills, but still, that's actually not too bad.

 

34 minutes ago, Ice Dragon said:

Ladders sidestep this by experimentally determining how well a team performs, but Arena reward cycles are too short for a ladder format to work.

This I still kind of disagree on. I'll return to my fighting game analogy. If a professionally competitive SSBB player can routinely get the to final round of tournaments with Ganondorf, I don't see any reason to make them change by effectively banning Ganondorf from top-level tournaments. As far as rotating team comp goes, the way it works out in FEH would be about equivalent to them dropping to making it to the semi-finals in top-level play while adjusting to the play style, but being allowed to participate at all*.

My standing should suffer for failings in my abilities and a failure to invest as much in the way of resources as my opponent, time in the case of SSBB and time and money in the case of FEH. It's my refusal to change with the times, a failing in investment (and arguably skill), that causes me to have issues facing armored teams on the desert map. Someone who's taken the time to optimize the same set of characters shouldn't be forced into the same reward tier as me just because we're running the same units. Deaths do that to some extent, but with how short arena runs are and how it's basically free to run, those are a pretty minor hurdle.

Fighting games dodge the unit quality issue by banning certain units in certain tournaments. An equivalent, taking BST as an actual indicator of performance, would be to have a set of tiers for people in each score range with all sets of tiers giving the same rewards—tier 20 for 680-700 score range would give the same as tier 20 for 720-740—but that seems even more prone to destroying unit variety an making a "whales only" arena, i.e. an all around worse idea.

*I know per your example that's not 100% equivalent, but that's the philosophical issue people take.

 

31 minutes ago, Ice Dragon said:

people don't realize how truly difficult it is to programmatically determine how well a team performs. And it doesn't help that it's at least partly recursive with the state of the meta.

That I'll definitely agree with, though. Seems like sometimes people have the expectation that computers can do anything.

 

That was a lot of words for what essentially amounts to I get where you're coming from on the complexity of rating unit quality, but I don't agree.

[Ice Dragon]

25 minutes ago, bottlegnomes said:

This I still kind of disagree on. I'll return to my fighting game analogy. If a professionally competitive SSBB player can routinely get the to final round of tournaments with Ganondorf, I don't see any reason to make them change by effectively banning Ganondorf from top-level tournaments.

No, because that's not how you experimentally determine what's good and what's not. A single professional player getting to the top with Ganondorf doesn't say that Ganondorf is good. It says that that player is good to be able to mitigate Ganondorf's weaknesses that well.

If everybody at the top is playing Ganondorf, then you can make an argument that Ganondorf is too good and probably should be banned.

I have already addressed how fighting game matchmaking differs substantially from "team building" game matchmaking in a ladder format, and a large part of it comes from the difference in what is being measured on a ladder for the two different games. In a fighting game, player skill is what is being measured, not the characters themselves (but it lets you see patterns in character usage). In a team building game, it's largely the team composition's effectiveness being measured and only some part the player's ability to execute the team's function.

[XRay]

2 hours ago, Ice Dragon said:

My biggest problem is that most people who say they want the Arena scoring system to be based on performance instead of just unit stats

I think most players want the scoring system to not have arbitrary scoring criteria. BST, SP, Bonus unit, and Bonus kills are arbitrary and it makes players build scoring optimized teams instead of building performance optimized teams. I do not think players would mind if the scoring system is based on performance by only considering the number of surviving units and and maybe the amount of merges.

I think it would be best to have a third Arena mode so players who already invest a lot in regular Arena would not feel like all their investment has gone to waste, and make Arena 2.0 be based purely on how well your team performs against a variety of enemies.

[bottlegnomes]

1 hour ago, Ice Dragon said:

No, because that's not how you experimentally determine what's good and what's not. A single professional player getting to the top with Ganondorf doesn't say that Ganondorf is good. It says that that player is good to be able to mitigate Ganondorf's weaknesses that well.

If everybody at the top is playing Ganondorf, then you can make an argument that Ganondorf is too good and probably should be banned.

You missed my point entirely. I'm saying the characters used should be irrelevant to ranking. I specifically used Ganondorf for the italicized reason you stated. Ganondorf being banned for being too bad is dumb for the same reason someone running mounted ponies being essentially banned from the top of the top is dumb.

Edited October 11, 2018 by bottlegnomes

[Ice Dragon]

16 minutes ago, XRay said:

I do not think players would mind if the scoring system is based on performance by only considering the number of surviving units and and maybe the amount of merges.

But that doesn't address matchmaking at all.

 

3 minutes ago, bottlegnomes said:

You missed my point entirely. I'm saying the characters used should be irrelevant to ranking. I specifically used Ganondorf for the italicized reason you stated.

This is a team building game, not a fighting game. The characters used in the team should be taken into account just as much as, if not more than, player skill.

[bottlegnomes]

4 minutes ago, Ice Dragon said:

This is a team building game, not a fighting game. The characters used in the team should be taken into account just as much as, if not more than, player skill.

What reason is there for a player who can beat teams of top tier units consistently with a team of low tier units to be theoretically excluded from the highest ranks? Again, gacha mechanics aside.

Edited October 11, 2018 by bottlegnomes

[Ice Dragon]

1 minute ago, bottlegnomes said:

What reason is there for a player who can beat teams of top tier units consistently with a team of low tier units to be theoretically excluded from the highest ranks? Again, gacha mechanics aside.

A player who can beat teams of top-tier units consistently with a team of low-scoring units should be excluded from the highest ranks. There's a huge difference between "low-tier" and "low-scoring".

For example, a team of two +0 Gravity+ staffies and two +0 Reposition bots can theoretically defeat just about any team in the game, but should certainly not score as high as a +10 version of the exact same team.

[bottlegnomes]

1 minute ago, Ice Dragon said:

A player who can beat teams of top-tier units consistently with a team of low-scoring units should be excluded from the highest ranks. There's a huge difference between "low-tier" and "low-scoring".

For example, a team of two +0 Gravity+ staffies and two +0 Reposition bots can theoretically defeat just about any team in the game, but should certainly not score as high as a +10 version of the exact same team.

That relates to BST how?

[XRay]

33 minutes ago, Ice Dragon said:

But that doesn't address matchmaking at all.

Players can match with any other player as long as their total merge levels are relatively similar I guess?

[Ice Dragon]

9 minutes ago, bottlegnomes said:

That relates to BST how?

Oh, that was what you were getting at. I'm sorry, I can't follow your fighting game analogy at all. It just doesn't work the way you say it does.

 

Fighting games typically use a ladder system to determine which opponents to match a player against. This is matchmaking by experiment. You pit players against opponents of various strengths until you determine a close enough position on the ladder to match their skill level to, then they fight opponents nearby and move up or down until their win-loss ratio becomes 1:1. This works for several reasons. (1) Player skill typically does not change quickly over time, meaning once a player has settled somewhere on the ladder, any movement from that position up or down will be gradual, and this results in always being matched up against opponents of similar skill. (2) Players are discouraged from changing characters to characters they are less familiar with due to the fact that they only have one ranked position. This means that until a player is comfortable playing with a character, they typically will not use the new character in ranked matches and therefore still results in the player always fighting against opponents of similar skill.

This doesn't work in Heroes. Seasons cycle too quickly and with too few matches played per season to properly place a player and have them settle down on the ladder. Furthermore, because performance is largely determined by team composition and less so by player skill, a player's performance can jump wildly simply by changing the team, requiring them to be re-placed on the ladder. This adds further inefficiencies to the ladder format, making it impractical for the short seasons. This is further complicated by the fact that the Arena format doesn't have players playing directly against other players, meaning (1) you don't have the ability to get players into a 1:1 win-loss ratio equilibrium and (2) we still haven't addressed methods of "ranking" the AI-controlled teams.

Matchmaking by experiment simply can't work with the restrictions required by the format. Therefore, the only way to reasonably provide matchmaking is to use a calculated metric from the players' and opponents' teams. The simplest way to do this is the assign points to various factors that affect how strong a unit is, and stat total happens to be one. As I have mentioned previously, units with fewer points in stats in general have a harder time fighting on their own with just rudimentary support. This I believe to be a fair enough reason to rate said units lower for matchmaking as weaker teams should be matched against weaker opponents.

In essence, this is exactly the same as the fighting game ladder except for the fact that instead of having to place yourself on the ladder over a period of time, the ladder places you on it based on the perceived strength of your team. Either way, both in the fighting game scenario and in Heroes, you will only fight against opponents ranked similarly as yourself, and scoring is simply a reflection of that placement.

 

 

I've apparently lost track of which thread this argument was being held in. How did we get here in this thread again?

[bottlegnomes]

@Ice Dragon on my phone, so excuse the not quoting.

The crux of what I'm getting at is that the current system favors perceived unit strength, not player skill. Note, I'm defining player skill as team-building ability, which also doesn't drastically change quickly.

Your analogy about a fighting game player not taking a new unit to competition is also a key point in my intended analogy. If I swap a unit and my score drops, that's my fault for not practicing with the new team comp, just as its a fighter's fault for not practicing with the new character. Theoretically, a player getting a better character is due to investment, which is one of two things that should be considered IMO, so their rank jumping is okay. I mean, yeah, luck and all, but that's always a factor and someone who's spent lots of blood, sweat, and tears (AKA BTS, I mean BST) on this game should statistically speaking have more good characters than someone who has just started.

I'm also saying we expand the window of standing to all of the player's time. The calcs would have to change, to maybe a percentage-based approach, but setting aside potential technical limitations of comparing that much data, it's possible. Maybe do something like average the win percentage of everyone in the tier, not just for the week but for all time, and bump up/down whatever percentile is deemed appropriate relative to that aveage. Someone showing up in tier 1 and winning all 7 matches with no losses is basically guaranteed to go up, but that's no different than someone showing up in tier 1 with a bunch of 5* +10s. The calculations obviously would need tuning to prevent wild jumps, like someone going from tier 19 to tier 15 over the course of a month due to bad luck, but that's the basic gist I think people are getting at. As far as def teams, similar percentages/percentile could be used. People in a certain tier could be matched against defense teams with a certain success percentage. Also would need tuning to not fluctuate wildly, but again, this is idealized.

Essentially, what I think people feel tiers should measure is two things: 1) player aptitude and 2) player investment. The current system kind of works.

Rarity - investment. The more time and money you've spent on the game, the higher rarity your team should be.

Level - investment. Same.

Merges - investment. Same.

SP - investment. Same.

Survival - player aptitude. It's a commonly accepted fire emblem metric, even if people have varying feelings on it. In this regard TV isn't hugely different. It's a reasonable if not perfect measure that FE fandom has long accepted.

Bonus units - some investment, but basically just an ad. Now with kills, they do lean a little toward skill, but I think ebecam agree they went a bit overboard.

The thing is BST is the odd one out, or at least was. Initially, it was meant to balance units, and it did a fairly good job at that. It had no relation to player aptitude or investment early on. I'm pretty sure they picked it just because they didn't know what else to do and held to the KISS approach. Sheena scored similarly to Hector. Gwendolyn scored higher than Olwen.

Now, it still has no bearing on player aptitude, and the only reason it's come to be a measure of investment is because IS has intentionally swayed things so only the really hard to get units have maxed out BST, meaning you need to invest a lot in either money or time and being thrifty with orbs. And even then, there's a disparity in that. SM Eirika doesn't specifically necessitate more investment to get and optimize than Hardin--both are the same rarity and have been out for a similar amount of time--but the perceived scoring and subsequently rank disparity is massive because if something like a 30 BST difference. (Being honest, I'm not that sure how much it actually affects, but that's the perception.) The issue isn't that people want to have a wild west where any random F2P can rank above the whales or people who've dedicated time and effort to building their teams. The issue, at least to me and as far as I can tell others, is that players are explicitly being handcuffed to a certain rank ceiling due to their character and play style preferences, not their personal ability and investment. It's one thing to cap out at tier 20 because I can't build a good enough team to beat the people above me or because the people above me have spent a lot more time and money on the game. It's another thing to cap out at tier 20 because the units I like to use are effectively excluded for basically arbitrary reasons or the claim that it's making the game fairer by pitting me against opponents of similar skill when the characters I use aren't intrinsically indicative of skill, hence my Ganondorf analogy.

Just to recap, and hopefully clarify, people want BST removed because they want rank to measure investment and player skill (team-building), not investment and unit value.

I'm fully aware my and others' proposed solutions have flows as well, but the thinking is that they can be tested and tuned. As is, they keep chugging with a system featuring am element that even a lot of whales don't seem to like. They probably can't just yank it because of the upset it'd cause to people who did play into the system, but they don't appear to be doing more than band aiding it at the moment which is another source of frustration.

 

And lastly, I'm not actually sure how we got on this either anymore.

[bottlegnomes]

@Ice Dragon Also, forgot to mention, those other metrics aren't perfect representations, such as SP, but they at least make some intrinsic sense. It makes sense that stronger skills are rarer and than strong skills also cost more, so someone having higher SP implies they have stronger and rarer skills which again, is usually a result of investment.

[JSND Alter Dragon Boner]

Using performance for gauging shit can get weird at times

 

Lets use an example from another nintendo team building game

 

Which one would you ban

Groudon and Kyogre 

 

Or Incineroar

 

 

You would say Groudon and Kyogre right? Anyone would

 

But if you refer to performance any sane person would say Incineroar. The total recorded performance of Incineroar beats Kyogre and Groudon combined

 

 

I believe FEH had a simmilar issue in this with Healers vs Armors

[redlight]

Couldn't we just impose a limit to how much one can score with BST and skills? And a duel skill would be the only way to go further than that limit

Also. Could we increase the bonus unit limit to 15?

Edited October 11, 2018 by silveraura25

[Ice Dragon]

10 minutes ago, silveraura25 said:

And a duel skill would be the only way to go further than that limit

Money grubbing much?

[Jotari]

Wasn't this thread meant to be about weapon refinery at some point, as interesting as it is to hear about people's opinion on the arena scoring.