A simple scaling solution

Eddiemon

pasa_ wrote:

You seem just snip the points -- in hulk there was reset every X hours and nodes gave points many times. There is a substantial difference grinding-wize if a node can give you points 20 times or 10 or 5 or 2. If you actually think it's no difference elaborate.

Well I figured Jozier dealt very well with the 'everyone gets the same points from 2 attempts and everyone gets the same prize' issue. If you'd like me to repeat his argument again I can but it seems redundant.

Unknown

I like scaling. Now where did I put my Spidey?

Eddiemon

gamar wrote:

Why on Earth would it require everyone to complete the node two extra times? For people who just want to complete the nodes once or twice, they can, and for people who want to compete for covers, they already have to get way more points than you would recieve for doing that, so absolutely nothing would change

Player X and Player Y are competing. Player X has a better roster/is luckier/plays better/buys health packs and as a result can defeat level 100 enemies while Player Y can only defeat level 90 enemies.
Add in 2 full clears of the board before scaling starts.
Player X and Y aren't going to see that difference until they complete the two scaling free clears of the board AND all the battles to get their scaling up to the point where the difference between them shows.
Those 2 full clears don't serve any function in their competition, they are just clears that everyone has to get out of the way before they start to compete.

Or to put it another way, if those two full clears are worth 5000 points then everyone's final score will increase by those 5000 points. They won't stop where they would stop today because it's 'way more points', they will push as far as they can go. You've just made 'as far as they can go' 2 full clears of the board further.

Unknown

We shouldn't talk about scaling and rubberbanding as if the two things are the same. There are events without rubberbanding that have scaling (Heroic Oscorp) and there are also events without scaling but have rubberbanding (pretty much all other events prior to scaling being introduced in TaT) and both models are workable.

By itself, scaling does not discourage grinding. Without rubberbanding, it takes you the same number of wins as someone who grinded to catch up to him, so you'd face the same increase in difficulty as he did. Scaling is mostly to correct for over/underestimating player ability, as we see that most missions settled on 230X3 for the high end rosters because that's how hard something has to be to be challenging to a high end roster, but we also see a few mission actually go down because they turn out to be a lot of tougher than it looks (Devil Dino + Bagman being the most famous one).

The problem you have is that rubberbanding and scaling creates a unique environment where grinding is punished. If you have just scaling, the guy who grind is still ahead by grinding since his pursuers needs to grind just as much as he does to catch up. If you have just rubberbanding, the guy who grind is still ahead by grinding because he will just grind to #1 before the the last refresh, and that puts him in the best possible position to win since the guy who is #1 is the guy who can afford to play the latest. But, what doesn't work is when you put both together. That is, while you're grinding for a positional lead, your enemies are now all 230X3, and you can't possibly fend off a charge from a guy facing say, level 150 enemies with the same roster as you because that guy can clear missions twice as fast as you. Note that without scaling, both you and your pursuers would be clearing missions at the same pace, and thus if you started out at #1, you can either build a lead that nobody can ever catch up (this happened a lot in TaT and The Hunt when sub bracket award was 500 HP), or in the worst case you can at least know that your starting position of #1 at the last refresh affords you the most amount of time to wait before you have to start playing, which is obviously advantageous.

Unknown

Eddiemon wrote:

gamar wrote:

Why on Earth would it require everyone to complete the node two extra times? For people who just want to complete the nodes once or twice, they can, and for people who want to compete for covers, they already have to get way more points than you would recieve for doing that, so absolutely nothing would change

Player X and Player Y are competing. Player X has a better roster/is luckier/plays better/buys health packs and as a result can defeat level 100 enemies while Player Y can only defeat level 90 enemies.
Add in 2 full clears of the board before scaling starts.
Player X and Y aren't going to see that difference until they complete the two scaling free clears of the board AND all the battles to get their scaling up to the point where the difference between them shows.
Those 2 full clears don't serve any function in their competition, they are just clears that everyone has to get out of the way before they start to compete.

Or to put it another way, if those two full clears are worth 5000 points then everyone's final score will increase by those 5000 points. They won't stop where they would stop today because it's 'way more points', they will push as far as they can go. You've just made 'as far as they can go' 2 full clears of the board further.

Well, I didn't say the first two clears "wouldn't count towards scaling," I said you'd get them scaling free - the scaling multiplier wouldn't apply to them. In other words, your third play would be scaled the same as it is now.

But even that is irrelevant - even if those "first two battles" DIDN'T count towards scaling, you'd still be wrong, because the stack depletion and rubberbanding would be the same - you wouldn't get two full clears and then all your stacks refresh! It just means your first two battles on a node would be easier.

Unknown

gamar wrote:

Eddiemon wrote:

gamar wrote:

Why on Earth would it require everyone to complete the node two extra times? For people who just want to complete the nodes once or twice, they can, and for people who want to compete for covers, they already have to get way more points than you would recieve for doing that, so absolutely nothing would change

Player X and Player Y are competing. Player X has a better roster/is luckier/plays better/buys health packs and as a result can defeat level 100 enemies while Player Y can only defeat level 90 enemies.
Add in 2 full clears of the board before scaling starts.
Player X and Y aren't going to see that difference until they complete the two scaling free clears of the board AND all the battles to get their scaling up to the point where the difference between them shows.
Those 2 full clears don't serve any function in their competition, they are just clears that everyone has to get out of the way before they start to compete.

Or to put it another way, if those two full clears are worth 5000 points then everyone's final score will increase by those 5000 points. They won't stop where they would stop today because it's 'way more points', they will push as far as they can go. You've just made 'as far as they can go' 2 full clears of the board further.

Well, I didn't say the first two clears "wouldn't count towards scaling," I said you'd get them scaling free - the scaling multiplier wouldn't apply to them. In other words, your third play would be scaled the same as it is now.

But even that is irrelevant - even if those "first two battles" DIDN'T count towards scaling, you'd still be wrong, because the stack depletion and rubberbanding would be the same - you wouldn't get two full clears and then all your stacks refresh! It just means your first two battles on a node would be easier.

Right, the first two battles on a node are trivial so everyone does them, and then the competition starts at battle 3. So, add in an extra 2 battles of grinding per node.

Unknown

jozier wrote:

gamar wrote:

Well, I didn't say the first two clears "wouldn't count towards scaling," I said you'd get them scaling free - the scaling multiplier wouldn't apply to them. In other words, your third play would be scaled the same as it is now.

But even that is irrelevant - even if those "first two battles" DIDN'T count towards scaling, you'd still be wrong, because the stack depletion and rubberbanding would be the same - you wouldn't get two full clears and then all your stacks refresh! It just means your first two battles on a node would be easier.

Right, the first two battles on a node are trivial so everyone does them, and then the competition starts at battle 3. So, add in an extra 2 battles of grinding per node.

Consider Player X and Player Y

Player X is a newbie, or maybe he just doesn't have the time this round to go for the top spots - but he still wants to see the story, get some tokens and iso from nodes + progression rewards. Right now, he might not be able to because the community scales all the battles beyond his reach. Sucky! If he got a couple of completions without scaling kicking in, he'd still be able to play the game at his own pace.

Player Y is you or me, going for the gold. What am I doing - playing the "featured character" nodes 3 or 4 times each for the last couple of refreshes? Maybe completing the map once and then picking off the highest point missions a few times? I'm getting much more points doing that than I would replaying the ~300 point missions and removing scaling from two playthroughs of those 300 point missions wouldn't change that! So for Player Y, nothing at all changes.

Unknown

Scaling is there to provide a challenging, but not impossible experience for all players. How you do that across a wide range of rosters and play ability is very difficult. You also have to make it so it can't be manipulated easily. Hence, the community scaling component. Without scaling newer players would not be able to compete and would be discouraged from playing. No new players = no new cash for the dev team = no MPQ for anyone.

Rubberbanding is there to allow some flexibility in play length/style. Not everyone can or wants to play 4+ hrs a day so rubberbanding allows those players a way to still compete. It's really not that much different from PvP where players with lower points get more points battling higher point players as a way to catch up and MMR helps to keep matchups challenging.

The purpose of the game is to make money while providing a rewarding experience for players so they'll be encouraged to spend said money. I'm sure it's a balancing act that is constantly in motion, but they seem to be narrowing in on a model that works.

Unknown

gamar wrote:

jozier wrote:

gamar wrote:

Well, I didn't say the first two clears "wouldn't count towards scaling," I said you'd get them scaling free - the scaling multiplier wouldn't apply to them. In other words, your third play would be scaled the same as it is now.

But even that is irrelevant - even if those "first two battles" DIDN'T count towards scaling, you'd still be wrong, because the stack depletion and rubberbanding would be the same - you wouldn't get two full clears and then all your stacks refresh! It just means your first two battles on a node would be easier.

Right, the first two battles on a node are trivial so everyone does them, and then the competition starts at battle 3. So, add in an extra 2 battles of grinding per node.

Consider Player X and Player Y

Player X is a newbie, or maybe he just doesn't have the time this round to go for the top spots - but he still wants to see the story, get some tokens and iso from nodes + progression rewards. Right now, he might not be able to because the community scales all the battles beyond his reach. Sucky! If he got a couple of completions without scaling kicking in, he'd still be able to play the game at his own pace.

Player Y is you or me, going for the gold. What am I doing - playing the "featured character" nodes 3 or 4 times each for the last couple of refreshes? Maybe completing the map once and then picking off the highest point missions a few times? I'm getting much more points doing that than I would replaying the ~300 point missions and removing scaling from two playthroughs of those 300 point missions wouldn't change that! So for Player Y, nothing at all changes.

Newbies are faced with severely deflated personal scaling so that they can get through the missions and see the story amongst other things. So your new system doesn't improve on that.

I also really don't see how your system is any different from the current system at all. Especially the way you've just described it.

Eddiemon

gamar wrote:

Well, I didn't say the first two clears "wouldn't count towards scaling," I said you'd get them scaling free - the scaling multiplier wouldn't apply to them. In other words, your third play would be scaled the same as it is now.

I'm sorry that I cannot differentiate between 'doesn't count towards scaling' and 'scaling free'. My interpretation of the terms would have 'scaling free' neither being affected by nor counting towards scaling. I've tried searching for the original reference to see if I misconstrued it back then or if you had used different terminology but after 10 minutes of searching for the magical combination of words across all of your posts I've failed to find the original suggestion.

I have no issue with suspending the application of scaling for the first round 'so everyone can see the storyline' as long as the first round still contributes to scaling. My issue is when the outcome is mostly determined by 'time spent playing' as opposed to any other factor.

But even that is irrelevant - even if those "first two battles" DIDN'T count towards scaling, you'd still be wrong, because the stack depletion and rubberbanding would be the same - you wouldn't get two full clears and then all your stacks refresh! It just means your first two battles on a node would be easier.

This system would mean a node I could clear 10 times I can now clear 12 times. While I used a base score to be illustrative I agree you wouldn't get a fixed score improvement, that was just demonstrative. You would however still get jammed with 2 extra clears to compete, which was the point of the demonstration.

Unknown

There's a fundamental disconnect between what scaling needs to accomplish and what is fair. If we have scaling based on victory, it is safe to say winning 10 times means you should face a harder challenge than winning once.

But, due to rubberbanding, those two actions may result the same amount of points, and that is in no ways fair to the guy who won 10 times. Indeed, the only way to avoid this is if you base scaling on PvE points earned as opposed to victories. You'd then set it such that the scaling does not kick in until you have (base value of all story missions X 10), so that someone who started late can still get the starting difficulty even with the maximum rubberband modifier and see the story.

I think scaling based on PvE points which do not kick in until you have 10X(all story mission base value) could work. Let's say we decide that being 15K above this threshold means you should see enemies 50 levels higher than whatever your base level is (determined by roster), then it doesn't matter how you got your 15K points, after you got that much points you'll be fighting enemies that are 50 levels higher. That'd at least solve the problem of people getting unreasonably low or high scaling factor.

Unknown

If you wait long enough to take advantage of scaling, your levels go up due to community scaling. The grinder will have easier first battles.

Unknown

jozier wrote:

If you wait long enough to take advantage of scaling, your levels go up due to community scaling. The grinder will have easier first battles.

Right now the personal scaling component seems to dominate the overall difficulty that it's still a huge advantage to keep that low even if the community component is slightly higher for doing it late.

I recall the first Simulator Basic was the other way around and does behave like how you described, but shouldn't the guy who play first have an advantage?

You can simply have any scaling suspended before people reach an amount of points equal to whatever you think is the 'goal' of the tournament. For something with a story, that should be the sum of the story mission base points X 10. For something like Simulator Basic, I'd imagine that'd be equal to the (sum of the missions on the 1000 iso reward branch) X 10. The current weighing of personal/community rating would be fine in this system where scaling is determined by PvE points, not number of victories.

Unknown

Might be interesting to have a toggle.

Toggle on: Much the same as it is now, scaling etc. earn points as normal
Toggle off: Scaling turns off, matches tailored just to your personal PVE MMR, no event points for completion. Essentially becomes solo mode.

Unknown

Eddiemon wrote:

But even that is irrelevant - even if those "first two battles" DIDN'T count towards scaling, you'd still be wrong, because the stack depletion and rubberbanding would be the same - you wouldn't get two full clears and then all your stacks refresh! It just means your first two battles on a node would be easier.

This system would mean a node I could clear 10 times I can now clear 12 times. While I used a base score to be illustrative I agree you wouldn't get a fixed score improvement, that was just demonstrative. You would however still get jammed with 2 extra clears to compete, which was the point of the demonstration.

Maybe my experience is unusual, but I'm usually limited by stacks, not difficulty (and it's not like I have a top tier roster, so I'm assuming that's even more the case for people who do). So I'm not seeing situations where I clear a node 6 times and then the 7th is just too hard - I'm seeing that I can clear a node down to 1 and there's no point clearing it further for 1 point. Giving two "easy" clears wouldn't actually make me able to get more points.

And if you ARE facing nodes that end up scaling so high you can't beat them AND you're competing for top5 placement, I'm assuming at the least you have to have the capability to beat the top-value nodes a few times at the end of the sub - so there's not going to be a time where your "minimum capability" is fewer than four/five completes on a node, so if getting two "free" completes helps you increase your number of possible completes, those free ones will be earlier in the event and get wiped out by rubberbanding anyways, so they won't actually help you increase your points?

Unknown

Cryptobrancus wrote:

Might be interesting to have a toggle.

Toggle on: Much the same as it is now, scaling etc. earn points as normal
Toggle off: Scaling turns off, matches tailored just to your personal PVE MMR, no event points for completion. Essentially becomes solo mode.

What exactly is the point of that?

Unknown

jozier wrote:

Cryptobrancus wrote:

Might be interesting to have a toggle.

Toggle on: Much the same as it is now, scaling etc. earn points as normal
Toggle off: Scaling turns off, matches tailored just to your personal PVE MMR, no event points for completion. Essentially becomes solo mode.

What exactly is the point of that?

If the scaling is making it impossible to even play the game, it gives us something to do besides complain about scaling

Unknown

Scaling has gotten better with each event, so there's hope they'll get it working well enough.

I really don't see how else they can do it. D3 has a certain number of covers/rewards they want to give out each event so it has to be competitive. If it was progression based then it turns it into a grind fest to meet the progression in the allotted time and isn't as controllable over rewards given out (or it would be a race to get the rewards).

Transferring old content to the prologue and adding older covers to some of the nodes is a good idea long term and maybe D3 has that planned. We're only in episode 4 so far, but maybe after a certain point we'll see those earlier episodes added to the prologue or a new "Episode" tab for players to chip away at their leisure.