The thing about game balance.

Unknown

Here's the thing about game balance. No matter what you do, someone is on top. Given your game style (forced featured character with choice of 2 other characters) and your game's potential choices (which characters to bring) the player base is ALWAYS going to settle on the "best" combo and use that in high end game play 95% of the time. It's just a fact, an immutable law of game design. Someone is on top.

For MPQ that was Thor/XF. This is not a bad thing, someone has to be on top why not these guys.

We can all agree that infinite combos are bad, that pay to win boosted insta wins are bad, and that some chance however small should exist that you get your butt whopped by the cascading board of doom between turns. I'm pretty sure we can also agree that a 50% loss rate would be unacceptable. That's what a "balanced" game looks like. You win half the time you lose half the time. So the game is already slanted towards the players.

Now, you are a game designer. You want to shake up the current meta for whatever reason. Here is what you do: Release an obvious counter to a selected character that dethrones them as top dog. This fosters competition and healthy gameplay as combos are figured out to counter other combos. Here is what you don't do: Take that character who many have put time and effort and money into and rip it to shreds. This just builds resentment in your playerbase for not delivering on false promises.

C.Mags - Worthy of the nerf bat. Infinite combo (bad). Came out the other side looking ok, not great, but playable
Spidey - Worthy of the nerf bat. Infinite combo (bad). Double nerfed by true healing so totally unplayable now
Sentry - Worthy of the nerf bat. Boosted 2 turn victory (bad). Crippled so hard as to be unplayable now.
Thor - not worthy of the nerf bat. Seriously, what did she ever do to you besides be a fan favorite and above average character? No infinte loops, no boosted 2 turn wins, at best you knock out one character after 3 turns which plenty of people can do if you bring 12 AP to bear, no reason to lower her damage rate in any way.

Now, you want to change charge tiles and how they are weighted? That's great, and you should definitely do that in the future, but HANDS OFF MY THOR. Or at the very least Smite needs a ridiculous ridiculous damage boost to compensate for the damage lost.

grifterx777

Gosh, I couldn't agree more with everything you just said. I could handle the other changes but beating down my 4hor set me over the edge. She's my favorite character and now look at her. I'm so upset that I haven't played the game since I read the changes. And that's saying something for me. Smh... Oi ve. Give me my money back!

Unknown

Since MPQ as a game obviously cannot work if it was balanced (50% win rate) I think the best you can do is like A beats B beats C beats A. There's an additional complication in MPQ where beating really means 'convince someone to skip you', because no one really wins a defensive match unless it's a grossly overpowered opponent or an incredible mismatch (e.g. you're running Thor + X Force against Loki and The Hood, and for some reason you didn't boost green which would decide the fight by turn 5 and instead spent the next 20 minutes doing match 3s). When factoring in the psychology of players X Force/Thor pretty much *beats* everyone, even though it's surprising weak against The Hood or Loki, but it doesn't matter because most players with Loki/The Hood wouldn't dare to attack and you'd always use boosts when attacking either of the AP drainers unless you really love pain.

Right now Iron Fist is probably the only character that has some kind of inequality. He obviously dominates The Hood/Loki, but is not necessarily dominant (though also not weak) against the slower guys with a lot of HP, while The Hood/Loki are great against slow guys and get chewed up by Iron Fist. This is mostly because Iron Fist can create unacceptable attrition for the opponent even if he doesn't win and while the way he does this is pretty heavy-handed, it's the right idea. There should be some matchups that'll create unacceptable attrition for the attacker even if they can still win, since any team that can win defensively in MPQ would have to be unbelieveably overpowered and whoever can do that would by definition not be balanced. Another way to do this would be to create unacceptable length of time battles that hopefully isn't a variant of the Prehistoric Arms defense.

Mawtful

Phantron wrote:

Since MPQ as a game obviously cannot work if it was balanced (50% win rate) I think the best you can do is like A beats B beats C beats A. There's an additional complication in MPQ where beating really means 'convince someone to skip you', because no one really wins a defensive match unless it's a grossly overpowered opponent or an incredible mismatch (e.g. you're running Thor + X Force against Loki and The Hood, and for some reason you didn't boost green which would decide the fight by turn 5 and instead spent the next 20 minutes doing match 3s). When factoring in the psychology of players X Force/Thor pretty much *beats* everyone, even though it's surprising weak against The Hood or Loki, but it doesn't matter because most players with Loki/The Hood wouldn't dare to attack and you'd always use boosts when attacking either of the AP drainers unless you really love pain.

Right now Iron Fist is probably the only character that has some kind of inequality. He obviously dominates The Hood/Loki, but is not necessarily dominant (though also not weak) against the slower guys with a lot of HP, while The Hood/Loki are great against slow guys and get chewed up by Iron Fist. This is mostly because Iron Fist can create unacceptable attrition for the opponent even if he doesn't win and while the way he does this is pretty heavy-handed, it's the right idea. There should be some matchups that'll create unacceptable attrition for the attacker even if they can still win, since any team that can win defensively in MPQ would have to be unbelieveably overpowered and whoever can do that would by definition not be balanced. Another way to do this would be to create unacceptable length of time battles that hopefully isn't a variant of the Prehistoric Arms defense.

This may not be the place to discuss it, but since you were talking about being skipped =/= defensive victory, I started thinking about making defense mean something.

What if non-retaliation skips cost points instead of ISO? (-1 point from your score to skip a fight)

Yes, there are some other changes that I think would need to accompany something like that; more than 3 fights would need to be presented, the queue would probably need to refresh after every fight so that you don't get stuck with a node that's had the same opponent listed from the beginning of an event, skipping could never result in rerolling the same opponent, etc. Certainly it would make a good defensive team mean something, and I feel it would play more into the SPR triangle of team strengths you describe. It may even influence people to swap between teams more regularly as they climb.

It could also be horrible, but I just had the thought and wondered if you'd already theorycrafted that sort of environment.

GrumpySmurf1002

Thor had a potential infinite loop that got squashed by turning Twin Pistols into a finishing move instead of an AP generator. They in theory could've addressed the charged tiles out of the gate, but chose Hood instead. I presume that's because they had no actual data on charged tiles, but knew what Hood could do.

As far as she's concerned, I suspect beyond the charge tiles, there's a metric that says her battles were significantly faster than other characters, which is the same metric that took out Sentry. That's why the Smite damage also got nerfed rather than just removing some charge tiles. Even with that, I have no idea where yellow comes into play, but that's for the other discussions.

I also suspect that's why XForce has avoiding a tinkering. Few people in PvP are using underleveled Thor (at least that I've experienced), but lots are using underleveled XForce. Those underleveled XForce take appreciably longer to win a match because they either don't have the full board explosion (5 green) or the full benefit of Surgical (5 black). Even for maxed XForce, taking out XForce-4Thor with just XForce-3* (I think) takes appreciably longer than a mirror match, where 4Thor can speed things up.

So taken on the whole of the player base, XForce's speed probably seems in line with other characters, so there's no reason to revisit.

If XForce-Fist is as fast as people theorize even after Fist has been addressed, then you may finally see XForce addressed. But I wouldn't hold my breath.

simonsez

GrumpySmurf1002 wrote:

They in theory could've addressed the charged tiles out of the gate, but chose Hood instead. I presume that's because they had no actual data on charged tiles

Yes, god forbid they actually playtest and gather that sort of data...

simonsez

Mawtful wrote:

What if non-retaliation skips cost points instead of ISO? (-1 point from your score to skip a fight)

Awful idea. Don't penalize players for the **** matchmaking algorithm.

Unknown

As long as there are shields the majority of opponents you wish you could be matched against will be unavailable and it will be forced to offer up suboptimal matches. That is not the algorithms fault, it is a side effect of giving players shields. So tell me honestly, do you want them to make the matchmaking better but remove shields or keep it as is and be able to protect your score?

GrumpySmurf1002

simonsez wrote:

GrumpySmurf1002 wrote:

They in theory could've addressed the charged tiles out of the gate, but chose Hood instead. I presume that's because they had no actual data on charged tiles

Yes, god forbid they actually playtest and gather that sort of data...

Well, the other thing, which I've seen mentioned in other threads, is that they don't play the game like I might play the game, and I don't play like a top-10 alliance plays the game. Unless they decide to get some people to beta test a character, no amount of play testing is going to find exploitations like the top end player base does.

The general point is the same though, they've made it really clear with words and actions that they're a data driven company. One can kick, yell, scream all they want about various elements of the game, but changes only happen when there (seems to be) data to support it.

Unknown

Cryptobrancus wrote:

As long as there are shields the majority of opponents you wish you could be matched against will be unavailable and it will be forced to offer up suboptimal matches. That is not the algorithms fault, it is a side effect of giving players shields. So tell me honestly, do you want them to make the matchmaking better but remove shields or keep it as is and be able to protect your score?

Bull ****. I climb to around 500 and have my page on the leaderboard is all unshielded competitors around my score. Yet no matter how much I skip, I am offered 18-20 point matches. I can see 9 other people in my bracket alone that would be worth 24-26 points. Additionally as I climb 100 points, I can still see plenty of unshielded players around my score. Those would have been worth 26-33 points at the initial climb. Yet no matter how much you skip, they stay in the 18-20 range (22 is a big win). It's not shielding that is the issue - the MMR algorithm sucks. At higher scores, shielding may be a bigger effect, but at lower scores it is obvious the MMR is broken

Unknown

Call me odd but I have climbed many a ladder off 17-23 point wins, and see nothing wrong with that. If it started offering me single digits all the time then that is a different story. Especially since as you climb the nodes you already have lined up drop in value so if you lined them up at a 25 (spot on same as you) you take the 35 point node you found instead and then the 25 you had is now only a 22, stupid system we all want 30+ point nodes all day but that is mathematically impossible so lets blame the algorithm.

There might be forumers with enough experience with MMR algorithms to offer constructive advice on a system incorporating varied levels and shields but calls for "fix it, fix it, fix it" don't seem to want a balanced system that works optimally for the most people, they want it to give them preference so they get the matches they want all the time.

Unknown

papa07 wrote:

Cryptobrancus wrote:

As long as there are shields the majority of opponents you wish you could be matched against will be unavailable and it will be forced to offer up suboptimal matches. That is not the algorithms fault, it is a side effect of giving players shields. So tell me honestly, do you want them to make the matchmaking better but remove shields or keep it as is and be able to protect your score?

Bull tinykitty. I climb to around 500 and have my page on the leaderboard is all unshielded competitors around my score. Yet no matter how much I skip, I am offered 18-20 point matches. I can see 9 other people in my bracket alone that would be worth 24-26 points. Additionally as I climb 100 points, I can still see plenty of unshielded players around my score. Those would have been worth 26-33 points at the initial climb. Yet no matter how much you skip, they stay in the 18-20 range (22 is a big win). It's not shielding that is the issue - the MMR algorithm sucks. At higher scores, shielding may be a bigger effect, but at lower scores it is obvious the MMR is broken

Until you reach a certain point matches are based off your personal MMR. So it doesn't matter if there are scores around yours if they aren't in your MMR bracket. The MMR calculation doesn't seem to work amazingly i'll admit but that's why you cannot find them and just oppening on score earlier means ppl will see lots of completely inappropriate teams all the time.

Of course when you reach a certain point it switches to a score based matchmaking system and you would expect to see anyone around your score who is unshielded

bit off topic though

simonsez

Cryptobrancus wrote:

As long as there are shields the majority of opponents you wish you could be matched against will be unavailable and it will be forced to offer up suboptimal matches.

This is complete nonsense. At any given point in time, I can cycle through the same 5 or 6 people over and over again. Don't tell me that with 20,000 daily users, there are only half a dozen in my shard who aren't shielded.

kensterr

You know that their data showed 2* Wolverine has dealt the most damage since Christmas.

NERF HIM NOW!

Unknown

Cryptobrancus wrote:

Call me odd but I have climbed many a ladder off 17-23 point wins, and see nothing wrong with that. If it started offering me single digits all the time then that is a different story. Especially since as you climb the nodes you already have lined up drop in value so if you lined them up at a 25 (spot on same as you) you take the 35 point node you found instead and then the 25 you had is now only a 22, stupid system we all want 30+ point nodes all day but that is mathematically impossible so lets blame the algorithm.

There might be forumers with enough experience with MMR algorithms to offer constructive advice on a system incorporating varied levels and shields but calls for "fix it, fix it, fix it" don't seem to want a balanced system that works optimally for the most people, they want it to give them preference so they get the matches they want all the time.

I climb on the under 20 point matches out of necessity too, but the problem is that the retaliations are -26 or -30, so for each one that retals, I have to win 2 new ones. And I sure as hell can't get my points back from the retal because that is worth much less than they took from me and they will just come back again.

I don't want 30+ point nodes all day (it would be nice though) but 23-27 would be great (as well as mathematically possible) and make retaliations less costly.

I have said it many times, that they should scrub the MMR and give people match options between 150% and 75% of their current score. People will still get to their equilibrium level, but they will have options to climb that will offer points, and in theory, more points for tougher teams (those teams would be able to climb higher)