Level spread in League tiers

[Urist]

OK, so the "how you doing in "league" thread got a bit off topic, and the participants were rightfully chastised for it.

Anyway. It has been a few months since leagues were started. Nevertheless, an equilibrium hasn't been reached yet. There have been many people, including myself, who have predicted models for league populations, and an equilibrium isn't expected until late January or early February. I've mentioned this before, but the whole situation could have been avoided if, for the very first league, the top 20,000 or so highest level players were put directly into W2. But it's too late now, we just have to wait it out. At level 271, I am tied for 48th highest level in my S1 league. And I've been purposefully avoiding promotion for a while now. I don't know if that counts as "tanking" as I've never demoted, but there it is. Judge me as you will.

League rewards HAVE been modified since leagues were implemented. Perhaps not enough, but it may be a work in progress. And although yes, W1 and W2 rewards for 1st are better than lower placing rewards in higher leagues, you still have to account for the loss of kobans the following week if you're purposefully demoting. If you're averaging 120 kobans per week (240 for 1st place in W2, 0 the following week, repeat ad infinitum), you're still well below the koban income of someone who is taking 31st-45th in S1 consistently, week after week. Which, frankly, shouldn't be that hard.

 

Quote

Points x Level for wins: skews higher level heroes to the top for POSITIVE points.  Variability will still arise directly from relative advantages in attack power, ego, defense and harmony and indirectly from differences in luck regarding equipment/drops and individual strategy. 

I don't see how that helps anything. People taking 0 points now would still be taking 0 points.

Quote

Negative Points doesn't require skewing to distribute players more or less by level.  But it should be scaled so it is comparable to the positive scoring.  Points x (Highest_Level_In_League - Level) would push lower level heroes farther down the ladder for losses than relatively higher level players.

Funny thing, prior to leagues being implemented, I and a bunch of other people would purposefully weaken ourselves by swapping teams to help others in arena battles. Check out the thread about it here:

Of course, I no longer do this, but it's actually amazing how much your power drops simply by swapping your alpha. Check this out. Before:

After:

 

A loss of 34k ego, 9k damage and 2k defense. And that's just swapping alphas. People who REALLY want to get a low score could swap their entire teams, along with their equipment. Then they'll get a negative score, probably a lower negative score than an actual low level player would get from winning some, losing some.

Anyway, I guess what I'm saying is to just endure. Being low level sucks, but you won't be low level forever (what level are you, by the way?). And if you actually know what you're doing in terms of PvP team and equipment setup, you can consistently beat people many levels higher than you. I do all the time:

 

[snowman22]

There is a thread in the suggestions forum section about this. My take is that the xp problem is even worse than the koban problem.

• S1: 20th (a bit of tanking): 114 kobans and 90% max xp
• S2: 20th last (maximum effort): 0 kobans and 25% max xp.

My 1 week in S2 cost me about 1.2 levels.

How about:

• kobans for league N 20th - 5th the same as league N+1 85th - 70th?
• More xp for a win in a higher league?
• More spread in xp for a win depending on your opponents level? Against a level 200 / 300 I (level 246) currently I get about 150 / 200 xp . Maybe that should be 50 / 400 xp?

[[CNO] BananaJoy]

i would really prefer higher xp income but they nerfed it shortly after league was released. dont think kinkoid consider to raise it again.

[_shal_]

Meh, strong players are already getting tons more XP than they used to get before leagues existed. That's pretty much the least important area to "fix" anything. And the nerfing amounted to all of 14%, wheee.

Edited December 3, 2018 by _shal_

[windia]

1 hour ago, BananaJoy said:

i would really prefer higher xp income but they nerfed it shortly after league was released. dont think kinkoid consider to raise it again.

That was also because of the fact that you collected 5 tokens in 5 hours. the hard ones did not sleep 6-8 hours...

2 minutes ago, _shal_ said:

Meh, strong players are already getting tons more XP than they used to get before leagues existed. That's pretty much the least important area to "fix" anything. And the nerfing amounted to all of 14%, wheee.

True. Chthugha referenced highest player MadMax on nutaku having 3.5 million XP at beginning of league. Now there are 6.5 Million for this payer. From 10 of lvl 400.

[ZackMacKenzie]

Edit:  I did NOT nest the spoilers.  I have NO idea why they appear this way, and EDIT doesn't seem to present a means of changing them.

Welp, I have good news and bad news.  (Hopefully the use of spoilers will make the post more palatable than before. YMMV.)

Last Week & This Week's Results So Far

 

Through some miracle and a lot of work, I managed to eke out 45th in W3.  Given the decay of my ranking I had figured to finish around the 60th/61st cusp.  I guess I hadn't accounted for the number of people who play less on the weekends, and perhaps for the number of people running out of battles to win to grow as large as it did.

That was the good news.

The bad news is that this week's League is even more top-heavy than the last one.  While that might be a hot description for a woman (or man, if that's your thing) it ain't for a League.  This time around there were only 6 players of lower level, 10 of higher level that I knew I could beat, and 5 more that I thought I could maybe beat before the League closed.  My rank was 1st slightly into Day 2 but I knew full well it wasn't sustainable.  It has been precipitously falling ever since.  Of the 5 higher-ups that I thought were possible wins, I managed 2 of them.  I thought I'd managed a 3rd but the RNG somehow resolved 2 of the 3 battles as losses.  The last 2 (L107 & L110) have been active enough that they've stayed out of reach.  Perhaps I was too optimistic about them in the first place.

Demographics: 10x L250+,  14 more @L200+,  19 more @L150+,  34 more @ L100+ so 77 of 99 were at least 50 levels ahead.  I managed to go 9-for-9 against 1 each of L92, L98 & L108 players.  I also start off seeded 46th-50th in contests, so I know I'm competitive for my Mojo/level/veterancy, but that can only take you so far.  As of this moment, I'm 50th and falling with > 2 days left to play and nothing but losers left for me to cash in.

So, a week later into the development of towards equilibrium and we're seemingly still headed backwards, not forwards.  At least that's how it looks from the pools I've been placed into.

Response To Counter-Examples:

 

While I agree with much of the argument against my position

in spirit, I disagree in scope.  I was, and am still, disappointed about being grouped with a majority of players who are at least 100% higher in level.  Most of the "well you have to be able to punch up, I beat player X at level Y" are citing victories against players less than 50% higher than they were at the time.  That is an enormous difference.  For comparison, this week the 'toughest' wins I scored were against a player 68% higher, and hoped to be able to beat another one 72% higher which matches the intended counter-examples put forth by others.  The problem is in the proportions.

Power discrepancy grows with time/level.  On day 1 at L1 there isn't much difference between a good player and a bad player, or a lucky player and an unlucky one.  Here, a bad unlucky player can still beat a lucky strategic player from time to time.  More importantly, even the good lucky player hasn't had the opportunity to amass the resources necessary to punch up in any significant way.  This is one of the reasons you must wait to join League play until you've got some levels behind you. (Fear of accelerating the already fast low-level advancement is another.)

But from that point on, the differences in luck and ability accumulate to make the potential gap much wider.  It becomes increasingly common and significant for players riding near the top of the curve to punch up.  As per all things, the bulk of the population is somewhere near the middle, but the spread grows ever larger.

So, not only is the comparison not good because of the spread of levels between the player and his defeated opponent, it's also not a good one because it compares players in vastly different positions in life-cycle.

Why Fixing / Changing League Structure Is Important:

 

Another reason why it is important to fix this issue is that it has such far-reaching implications.  Being set up with a preponderance of can't-possibly-win pairings hurts you in so many ways beyond your League rewards (which are themselves quite significant).

• You receive 20% as much XP for losses.  This hurts your overall progression, hurts your performance in the weekly ToF XP chase, and hurts you in every Contest based on XP.
• Your ability to rack up PvP wins can be as much as halved by a highly unbalanced League.
• Your slower progression keeps your level down, which adversely affects your ability to advance your haremettes, accumulate Mojo, and accumulate Stat upgrades, all of which affect weekly or daily contests.  Further, they feedback to further lower your performance in the Arena.  To a lesser degree they also limit your overall harem level which feeds back into all PvP performance, though this effect decreases with level due to it being based on a square root.

The converse of these are also true — you get many spillover advantages for getting placed into a group made predominantly of newer players.

Again, I'm not asking to level the playing field between the very best high level players and the worst low level players.  I'm saying that a player shouldn't be competing against players that are so far advanced that spending kobans until you have 4 Legendary boosters & 6 Single-class Legendary items of your current class and level STILL leaves you no chance of winning against the majority of the field!  As near as I can tell, I could do both those things, AND spend kobans until I have ALL the Pachinko girls, and would STILL have no chance.

Why Didn't You Just Tank Like Everybody Else?

 

First of all, not everyone is tanking.  Many wouldn't gain any advantage by tanking so don't consider it.  Some never thought of it before.  Others consciously choose not to.

When I see a problem I work to find a solution.  When the problem is regarding something over which I have control, I just do it.  In all other cases, I consult with others and eventually leadership in whatever form it takes.

Only after every other option has been explored and exhausted will I choose to be part of the problem.  It's something I learned in kindergarten that stuck with me: "Right is right even if nobody does it, wrong is wrong even if everybody does it."  Some will agree, some won't.  Overwhelmingly, I find that choosing the local/personal, easiest, most advantageous method, i.e. being part of the problem, leads to terrible global results and eventually these spillover to hurt the local as well.  But, that's me.  I'll do me, you do you.  Hopefully, reality will meet somewhere in the middle.

Similarly, when I join a community I try to contribute to the best of my ability.  Some communities embrace new blood, some do not.  Since communities are made up of individuals, none are entirely one way or the other.  I haven't been here long enough to know which side we tend towards here at HH.

New Proposal

 

Perhaps the easiest, least churn solution (temporary workaround?) is changing how the groups of roughly 100 are formed.  They currently appear to be totally random — anyone who is currently in that League category (e.g. Wanker 2, Sexpert 3, etc) has an equal chance of being selected for the next spot to be filled when building the subgroups.  This can result in the lowest and highest ranking members of the same category appearing in the same sub-league.

Perhaps these would be better built using probabilistic ranking and weighting based on level, veterancy and/or Mojo?  This wouldn't exclude the possibility of the highest ranking member being dumped into a kiddie pool, nor would it eliminate the chance of a minnow being placed among sharks.  It would make it less likely, and tend to produce 'flatter' subleagues.  This would not only encourage participation, but it would introduce a new form of strategy around picking your battles.  (It might only enhance one that already exists in a limited way.)  Now, players will have to more closely monitor which battles to fight when, as differences in opponent strength will be more subtle.  Should I try punching up at this guy now or later.  Or, am I more likely to get stronger faster than he is and I should hold off for a day or two (or six).  On the flipside, do I take out this guy now or wait and give him a chance to catch up or pass me later?

While I've suggested probabilistic placement, a deterministic placement would flatten out the subleagues even more.  In the scope of that subleague for that week, it would certainly be 'more fair' but I suspect that the bigger picture might look worse. Having all the very biggest fish in the same pool means only 15 of them can advance in any given week. That's not terrible in and of itself but paired with the fact that 15 of the very smallest fish (for that League) will advance could eventually cause problems.  Regardless, it would prolong the pre-equilibrium state as much as it makes the week-by-week measure 'fairer'.  With probabilistic seeding it is still far more likely that the most experienced will advance from each pool, even in the last pool to be built that should be relatively bottom-heavy with newbs.  The main difference is going to be in the mid-ranks of each pool, with less of a chance of widespread advantage and disadvantage across many contests and ToF ranks.

Logistically, this would mean:

1. Determine the number of players in the League
2. Determine how many subleagues (COUNT) / pools to create and in so doing determine how many to place into each pool (SIZE)
3. Assign weights by strength measure (presumably level, veterancy, Mojo, or some combination of them)
4. Build data structure and logical operators for selection
5. Fill a pool by randomly selecting a player, one at a time, until SIZE or SIZE + 1 as needed
6. Start another pool and fill it as in Step 5
7. Continue until COUNT pools have been filled

If there's a tricky part it is probably the construction (and deconstruction) of the data and logic as players are removed by placing them into a pool.

What do the rest of you think?

Is this better than changing the scoring system within leagues as proposed earlier?

Do you agree that deterministic selection is worse than the current randomness?

Is probabilistic selection better than randomness?

Will it solve or at least reduce the problems at the core of the current system, or is changing the rewards hierarchy the only way to get it done?

I'm trying to come up with alternatives to changing the rewards system for 2 reasons.  First, I suspect that there would be HUGE resistance to any perceived reduction in rewards.  Second, I think the original concept for the Leagues hypothesized that in the long run, players that get promoted would finish roughly 31st-45th, which consistently compares (reasonably) in rewards to the previous level's 15th place.  If that is the long term goal/plan it probably shouldn't change.  It's fairly sound reasoning, before you account for gamesmanship on the part of the players.  The problem is that advancing all too often means finishing far lower, and that (or because of) people (are) tanking to demote themselves into easier and better rewards.

Thanks for taking the time to read whichever parts you read.

Edited December 10, 2018 by ZackMacKenzie
comment on strangely nested spoilers

[_shal_]

Second, I think the original concept for the Leagues hypothesized that in the long run, players that get promoted would finish roughly 31st-45th, which consistently compares (reasonably) in rewards to the previous level's 15th place.

Nobody ever said that, and how could that possibly work exactly, anyway, outside of some Lake Wobegon world? In any tiered competition, freshly promoted participants are very likely worse than almost all of those who already made it to that higher level at an earlier date. In the long run, most players will settle into exactly one level of league play, while a significant number of others will become elevator players who are too strong for one level and too weak for the next.

[ZackMacKenzie]

You are right, of course.  I think it's an issue of language.  Perhaps I should have stated "medium run"?  I'm not entirely sure that's a real term.

I'm used to short run, long run and very long run, but that's in the context of college math, economics and computing courses.  I don't know if these are universally used or a UK/US difference, or something else entirely.  Regardless, since this forum is none of those places, I should have realized that a more accurate way of describing it in English would have been a better choice.

My bad.  I'll try to remember to be more conversational and less technical.

Since I'm here already...

FWIW, people here initially guessed that the very long run equilibrium would be achieved in January or February. 

Modeling level spread is either beyond my level of mathematical understanding, requires computer simulations, or is easier than I think and I'm just blanking out on how to do it.  I suspect its #2, multiple simulations over a set of assumptions (player dropout rate, new player rate, rate of power growth relative to veterancy, random vacations from the game, etc.).  The Leagues may reach a relatively stable stratification — which is what we've been talking about — before they reach an equilibrium of proportion, they might not.  I'm not at all sure.  Proportional equilibrium means that the number of participants in each League remains the stable.  Stratification means that players' relative positions to one another has become stable, irrespective of League.

Forecasting an equilibrium of proportion is much easier and can be done with a spreadsheet, so I've done it.  (If I can figure out how to attach one, it'll be here for anyone to look at or critique.)  This forecast may not be 100% accurate since it only builds a proper pool for each full 100 players.  A League that has less than 100 players doesn't promote or demote anyone, until it has reached that critical mass of 100.  That may or may not be how it has been working.  One would have had to be in the promotes every week category to see how the leading edge works (or a dev), and that sure isn't me.

• Static player base of 100,000, with all of them starting in Wanker 1
• 15 of each full 100 promote to the League above.
• 15 of each full 100 demote to the League below.
• Wanker 1 has no League below it.
• There is no growth or shrinking of the player base.  This is to see if an equilibrium is even possible.
• Dicktator 2 is hypothesized in one model to see/show how it differs.

In the current 7-League model proportional equilibrium is achieved after 258 Seasons, otherwise known as just under 5 years.  Each League stabilizes at roughly 1/7th the size of the player base, with the high and low Leagues being slightly smaller.  The system gets close enough to equilibrium a month or two before the start of Year 3, so a bit more than half the time needed to get absolutely stable.

• Wanker 1: 14.290
• Wanker 2: 14,295
• Wanker 3: 14,295
• Sexpert 1: 14,295
• Sexpert 2: 14,295
• Sexpert 3: 14,295
• Dicktator 1: 14,235

In the 100, 000 is wildly high, then things will be a bit faster for absolute stability but dropping the population doesn't seem to affect when relative proportional stability happens.  The initial population size seems to only affect the amount of time it takes to progress from relatively stable to completely stable.

If another League is added, proportional equilibrium is achieved almost a year-and-a-half later, 331 weeks, AKA 6+ years!  Also, curiously, it forms a 2-week cycle of stability, flip-flopping forever in a binary state with each League holding roughly 1/8th of the player base.  We get within 10% of proportional equilibrium in less than half the time, just 2½ years.

Finally, the the adoption rate of new players affects stratification and proportion in opposite ways.  Stratification happens faster the more new players join, mainly because they start off where they belong, near the bottom, and in effect force better and more experienced players up the ladder by increasing the number of subleagues at every League level.  But, they delay proportional balance as it means more players have to percolate into the correct leagues.  Since we care more about achieving stratification, tell all your friends, and maybe all your enemies, to join us! 

Based on what I'm seeing, January or February was far too optimistic a projection.  Perhaps Jan or Feb 2020 is more likely?  It depends on several things.

• Will there only be ONE Dicktator I subleague, or multiple?  If it's capped at 1 stratification will be harder to achieve, but proportional balance will come faster.
• Might there ever be a League or Leagues above Dicktator I?  More Leagues: faster stratification, slower proportional balance.
• Will the promotion/demotion ratios stay the same forever?  If 15 grows, both stratification and proportion rates should increase too.
• Will the pool sizes always be 100 +/- 1?  Smaller pool size relative to the number of promotions/demotions gets to both faster.

The road to equilibrium might slow down, speed up or remain the same based on those answers.

Perhaps the ultimate question is "Do we want to wait that long or should we help it along?"

 

LeagueStability.xlsx

[Moe Overload]

We already know that there will be three Dicktator leagues, so nine leagues in total.

I just beat someone 34% higher level than me in the Sexpert 1 league, so it's completely possible to punch up.

[ZackMacKenzie]

17 minutes ago, Moe Overload said:

I just beat someone 34% higher level than me in the Sexpert 1 league, so it's completely possible to punch up.

In the post that few are reading I cover that.  I' m not saying it isn't possible to punch 30% or even 60% up.  I do it fairly regularly.  I'm saying that asking someone to punch 100%  - 200% up is asking too much and that has been the League I've been in for 2 straight seasons.

That, and removing the incentive to tank.

[Moe Overload]

Sorry I only skimmed your post. It was way too long. I agree on removing the incentive to tank. If the rewards were the same at the bottom of one league as the top of the previous one, tanking would be pointless.

[ZackMacKenzie]

21 minutes ago, Moe Overload said:

Sorry I only skimmed your post. It was way too long.

I know, I know.  That's why I tried to break it up into spoiler chunks with headings.  People could read whichever subtopic interested them and skip the others.  If only the close-spoiler tag wasn't bugged.

Next time, if there is a next time, instead of trying to put multiple spoilers in a single post, I'll put each one in its own post.  Each one should be short enough to read within a typical BM — that's the alleged attention span standard, right?

Sadly, I find that anyone can claim just about anything briefly, but they'll persuade close to no one unless they back it up with facts or at least details.  On the other hand, when adding details and facts people get put off by the wall of text, TL;DR, and again, close to no one is persuaded.  Until it can be shown that assertions without substance change more minds than detailed arguments I'm stuck being long-winded if I want to be persuasive.

I'm disappointed that few if any will get to the top-heavy joke, or will be too stressed-out by the length of the post to appreciate it.  Oh well.  It made me smile at least. :shrug:

[ZackMacKenzie]

Another Proposal (conscious attempt to be more brief)

Harder to describe, but may be easier to implement than other proposed changes, and increases the speed to equilibrium.

• Each League's players are ranked by veterancy, Mojo, level, whatever.
• Each pool/sub-league is built by selecting players, in order, by that rank (deterministic, effectively tiered sub-leagues).
• The 1st, and therefore toughest pool, gets the most promotions.
• The last, and therefore weakest pool, gets the fewest promotions, maybe as few as one!
• No change to the total number of promotions for the Season.
• Rewards can be changed or left as is, whatever the devs prefer.
• Greatly increases speed towards stratification-type equilibrium (by promoting more deserving players than undeserving players — see next point).
• Promotes fewer relative newbs into Leagues where they are completely out of their depth.
• Presents less opportunity for the highest ranked players to tank (since they're in the same pool, not distributed across all pools).
• Can be ended if/when stratification has been achieved.

Again, the main reason not to change rewards is to avoid 10,000+ complaints about Leagues "getting nerfed".  Changing the rules or League construction tends to viewed more as an acceptable tweak than changes to rewards which some see as "stealing what was rightfully theirs".  It's kinda like how sin tax sits better with most people than sales or income taxes do.  (They can be avoided, often affect 'other' people, and encourage us to avoid products we already know aren't good for us.)

Done.  (Shorter?)

[jelom]

5 hours ago, ZackMacKenzie said:

I'll try to remember to be more conversational

More????  😨😱

[Urist]

On 12/10/2018 at 10:45 PM, ZackMacKenzie said:

Static player base of 100,000, with all of them starting in Wanker 1

That was the number I assumed at first too, but upon further info, I found it to be WAY to high. The first W3 league launched on Oct. 4, with 1200 people. Which means that the initial player base put into leagues was ~53,000. There must have been some culling based on activity. I.E. People who hadn't been active in the last X days were not put into the first league group. And yes, I was talking about what you're calling "proportinal equilibrium." There's just no way to model player activity.

But yes, projections for Jan - Feb were way off, especially factoring in that people with 0 points in W1 do not promote, even if they are in "top 15" with 0 points. I have 6 accounts (only 1 of which I'm actually checking in every day) so I've been able to observe some of the lower tier leagues. And in the W1 leagues I've seen, at this point there are not 15 people per league promoting to W2. Probably around 10.

Anyway, the Wanker 1 league is what's going to cause problems with your ideas. In any attempt at making "tiered" subleagues, all the new players will be fighting each other for that 1 promotion spot, while the inactive, sleeping giants, all with 0 points will all be together, not getting promoted in the sub-league with the most promotion slots.

Another thing:

On 12/10/2018 at 6:14 PM, ZackMacKenzie said:

10x L250+,  14 more @L200+,  19 more @L150+,  34 more @ L100+

It sounds like the level spread across leagues is more balanced than you think. This week's S1:

1 player at level 140 (He must be having a great time. But he has 47 points so far, making him 77th place, so good on him for trying)

22 players at level 150-199

14 players level 200-224

15 players level 225-249

14 players level 250-274

30 players (including me) level 275-299

5 players level 300+

And there are 7 players with 0 points so far. Including players at level 292, 296 and 298. So look forward to seeing them in W3 next week. The level 298 guy is the only player other than myself who has more than 160k ego, so I'll finally be able to make 1st place in S1 and move on to S2. Because that's been my philosophy. First place or don't promote.

Anyway, I guess I'll re-iterate my previous advice. Endure. You never answered, but what level are you? It sounds as if you're low enough that you could be gaining a level every day. Or in less than 36 hours at least. That's the bright side in all this. As a super-active player at level 276 it takes a little over 48 hours for me to gain a level (5 levels in 11 days, based on being lvl 271 when I started this thread). It must be significantly more time for the semi-active players in lower tier leagues. So you'll catch up.

Edited December 13, 2018 by Urist

[ZackMacKenzie]

2 hours ago, Urist said:

1.  And in the W1 leagues I've seen, at this point there are not 15 people per league promoting to W2. Probably around 10.

2.  Anyway, the Wanker 1 league is what's going to cause problems with your ideas. In any attempt at making "tiered" subleagues, all the new players will be fighting each other for that 1 promotion spot, while the inactive, sleeping giants, all with 0 points will all be together, not getting promoted in the sub-league with the most promotion slots.

3.  Anyway, I guess I'll re-iterate my previous advice. Endure. You never answered, but what level are you? It sounds as if you're low enough that you could be gaining a level every day. Or in less than 36 hours at least.

1.  Cool.  Thanks for the info.  It's good to know that adjustments are being made, even local ones.

2. Wanker 1 wasn't really the focus of the proposal.  Since we already know that there are local adjustments being made, Wanker 1 could be exempt from the proposed changes.  It was already a special case since it couldn't demote anyone, and it's almost certainly still the most populated League.  Between those 2 factors it's the most volatile in terms of who has a mathematical chance of promoting.

3. Sorry, I either missed the question or forgot to answer it.  I'm L69, but I was only L66 yesterday.  Between high activity, the new XP rules coming into effect, and a new League cycle starting (i.e. the 1st pool of potential wins in over 5 days) I was somehow able to level 3 times today.  Up to L50 I was able to level at least 2x/day.  Since then it's been only slightly better than once a day.  So, three in a day was not something I expected to ever see again.

Yes, it is mathematically possible that I (or other low levels) can advance enough over the course of the week-long tournament to 'catch-up' to higher ups.  I'll have to see if the new XP tweaks make that easier or harder.  The most I've been able to overtake in a week, including booster use, has been 4 or 5, and the boosters seemed to be the biggest reason for that. 

Between the changes in XP and the changes in Pachinko, I think quality gear is even more important than before.  Before we at least had a chance at Legendary equipment in all 6 slots.  Now, it seems that number varies and may or may not reach 6.  I've only seen between 2 and 4 slots represented at any given time.  So leveling has accelerated and gear maintenance has been curtailed.  It'll be interesting to see how that affects things going forward.  Binge spending on Pachinko may not be as effective as it was before.  It's too soon to tell.

FWIW, I won't be posting any new suggestions here.  It's really just up to the devs to decide what to do, if anything, assuming they've seen or heard any of this.

 

[asdf10]

On 12/10/2018 at 11:23 PM, ZackMacKenzie said:

Again, the main reason not to change rewards is to avoid 10,000+ complaints about Leagues "getting nerfed".  Changing the rules or League construction tends to viewed more as an acceptable tweak than changes to rewards which some see as "stealing what was rightfully theirs".  It's kinda like how sin tax sits better with most people than sales or income taxes do.  (They can be avoided, often affect 'other' people, and encourage us to avoid products we already know aren't good for us.)

 

Honestly an adjustment of rewards will probably be a buff to the overall player base. Right now the rewards are concentrated in the top 4 slots of each league. The other 96+ slots have relatively crappy rewards. If the rewards were restructured such that 1st in the previous league was the same prize as 75+ of the next league the players would be getting a lot more rewards over all. So all 100+ players in a particular league would be happy with their reward. (Not counting the crappy experience gain for those finishing 75+, but at least they get the previous leagues 1st place reward).

[DHarry]

Also, promoting would be a good thing. Now its almost a punishment to end 5-15 b3cause you dont get a big reward but will be slaughtered the following week. 

[ZackMacKenzie]

On 12/13/2018 at 8:19 PM, ZackMacKenzie said:

Yes, it is mathematically possible that I (or other low levels) can advance enough over the course of the week-long tournament to 'catch-up' to higher ups.  I'll have to see if the new XP tweaks make that easier or harder.  The most I've been able to overtake in a week, including booster use, has been 4 or 5, and the boosters seemed to be the biggest reason for that. 

Level dinged first thing this morning and it looks like I won't be able to catch ANY of the just-out-of-reach players, boosters or not.  Very early returns indicate it might be harder now, or it was just an even weirder level spread than before.

[snowman22]

Just a data point.

The level spread between last week's league (S1) and this week's league (S2) is MUCH less than it was three weeks ago.

Three weeks ago 15th in S1 == 11th last in S2.

Now I estimate 15th in S1 == 50th in S2.

There is still the rewards being skewed problem but in my S2 at least the ego gap is not an issue.

[Urist]

Yeah, it's not much of a problem in higher leagues, because people have to be both powerful and active enough to get there. The trouble with the lower leagues is the high level inactives clogging up the spaces. My first venture into S2, my level spread is:

44 players 300-320

48 players 250-299

11 players 202-249

Looking at a lower level (35) account in W2, the level spread goes from 21 to 241. But a level 37 player is in 1st place so far.

[ZackMacKenzie]

19 hours ago, Urist said:

Looking at a lower level (35) account in W2, the level spread goes from 21 to 241. But a level 37 player is in 1st place so far.

Rank doesn't really mean anything for the first 48 hours...or more.  The last 4 weeks (maybe longer) I've been 1st to 5th going into day 3, and my typical finish has been 45th.  Most people knock off their immediately winnable matches ASAP.  By day 3 the lower levels have run out of them or are rapidly approaching that cliff and their rank sinks like a stone.

Case in point, I'm presently 1st in my W3 league, but only 69th in level, probably lower still in veterancy.  I figure 25th to 30th is my most likely finish.

[_shal_]

On 12/20/2018 at 6:22 PM, snowman22 said:

Just a data point.

The level spread between last week's league (S1) and this week's league (S2) is MUCH less than it was three weeks ago.

Three weeks ago 15th in S1 == 11th last in S2.

Now I estimate 15th in S1 == 50th in S2.

There is still the rewards being skewed problem but in my S2 at least the ego gap is not an issue.

I highly doubt that, to be honest. I got promoted in 2nd place from S1 just a week ago - a "real" second place, not a "tanked for a month beforehand and was way overpowered for S1 at the end" - and got nowhere near 50th in S2 the week after. I suspect you're severely underestimating how much the quality of those 15th-place-in-S1 players has degraded in three weeks as well.

Edited December 22, 2018 by _shal_

[snowman22]

6 minutes ago, _shal_ said:

I highly doubt that, to be honest. I got promoted in 2nd place from S1 just a week ago - a "real" second place, not a "tanked for a month beforehand and was way overpowered for S1 at the end" - and got nowhere near 50th in S2 the week after. I suspect you're severely underestimating how much the quality of those 15th-place-in-S1 players has degraded in three weeks as well.

There are 61 people < 160K in the new S2. There were lots of people in last week's S1 in the 140-160K range. I think the 14 best of these would easily finish halfway up the table in my current S2. Maybe the brackets vary a lot and mine is an easy one?

[Observer_X]

I would not be too simplistic, but as I see it, the main problems of the leagues resides one step back, i.e. in how the battle stats are calculated and, more importantly, obtained. As most of you know, it is out of the question that on the side of obtaining better equipment, paying players have an unquestionable advantage over free players. The same applies when obtaining the better girls, especially during legendary events, as paying players in general can put in enough resources in refills to obtain the best girls for their teams, and then upgrade them quickly. This means that in the same league there are players at the same level with very different amounts of EGO, and players with lower level with higher EGO than players with higher level. For example, in my actual league, there is a level 278 player with over 196k EGO: I am level 304 and have only 170k ego, after having only recently upgraded my stats and equipment. All this to say that there are only two ways to reequilibrate the situation: or some thresholds became included, making so that in a certain range of levels (as small as possible) the players, independently from the fact they are free or paying, can't go over or under certain ranges of stats, or the method of insertion inside a certain league, or, more in general, the method to assemble league turns, should get greatly improved compared to now. This obviously assuming that devs want balanced leagues, a thing that personally I am not prone to bet on.