PVE Guild Ladder

Hello,
I would like to suggest upgrading current PVE Ladder, as its wasting huge potential of this feature.

This is part of much larger issue that developers face (or ignore) with all games and servers. Satisfying all kinds of players, from most casual ones to most hardcore ones. In the end you have community that you can split into several layers of "tryharding", where the more casual layers consist of biggest % of population, therefore developers often focus on them. But the hierarchy doesn't work without the top layers - let me explain in wow terms. Sooner or later everyone in game wants to try all the end content of the game and then they need the good players to teach them / carry them / lead the guild to let them do that. Also most people will pretty likely have an urge to try being really good at the game, but if they aren't the type thats good at learning themselves, they need someone to show them / teach them / be their example - either guild member/leader/streamer, doesnt matter. Also many parts of this game are relying of existance of guilds (and people socializing in them) and guild dont work without good players leading them. The hierarchy needs all its parts, but for the top to exist, there needs to be some form of competition, because without any form of measurement and comparison to others, with all game content being defeated, there are no goals, tryharders have nothing to tryhard and so on.

Many people say that noone wants to tryhard 10 years old datadisc, yet once you go into any raid, people can go crazy and do anything to get any higher in their meters, because its the only kind of competition they get. This server is living proof that people wanna play this datadisc and playing means achieving new goals and competing - so start giving them a way!

__________________________________________________ _____________________

Curently the PVE ladder is only based on speed kill records - which causes multiple negatives:
- the ladder is in no way reflecting guild activity - most "top" guilds are long time inactive
- the ladder disregards any other aspects that make a good guild (no wipes, consistency, activity, multiple clears, etc.)
- over the years with everchanging class/raid functionality the top ladder is mostly just list of overgeared guilds that did special raids with some bug abuse raid setup to burst the bosses (quite possibly could use reset or at least removal of too old kills)

In the end the ladder at its current state is simply laughable and meaningless, therefore im gonna explain how to make it properly reflect guild quality and activity at the same time, with notes on how to implement it at its basic form and many possibilities of further development.

I hope I dont have to explain why having proper PVE guild ladder would help bring some competition and prestige to guilds on server, which is hugely lacking here. The current speed kill records can still exist as speed kill records, but separate from ladder. (in fact the speed records will be useful for determining time bonuses for my suggestion, but the speed records first need reset as there is big amount of speed kills that must have been bug abusing (rs25 with 1 healer - how? :D))

I will skip discussing what is a "guild raid" and what isnt - I guess it already is programmed in current ladder and its based on how many ppl in raid are from the given guild (idk the value but 80%+ makes sense)

Obviously everything liste below is open for debate, especially some specific number - however I will describe everything exactly to give you some image of what it could look like

Now on how to configure the ladder properly:
In order for the ladder to have meaning and value, it needs to not only reflect skill and top performance of guild, but also other qualities that all players want from a guild.

__________________________________________________ _____________________

1) Activity tracking
This part basically defines the main functionality of the point calculations and implementation - its copied from EPGP addon if you know, with slight modifications.
Each guild will have its so far gained score - lets call it X.
Each week (from ID reset to ID reset) the guild will be able to obtain some points (described in next parts) lets call them Y.

So on each ID reset the new guild score will be recalculated as newX = (X + Y) * d
- the guild score will NOT update during week, because it should be irrelevant which day of the week guilds raid

Where d is decay coefficient, which makes obtained points loose value over time - this is good not only for keeping the ladder fresh and reflecting on guild activity, but also allows for ladder functionality changes, where you can continuosly add features to the ladder or change points gained configuration and if you do it at the time of ID reset, it wont break the system or disregard previously obtained points and can continue normally being fair to everyone.

I suggest d = 0.8 = 80%, meaning 20% current point reduction, which corresponds to 90% of value reduction over 10 weeks, meaning that 90% of total points is gained during last 10 weeks, also it will only take 10 weeks to get guilds to 90% of normal point values. 10 weeks is 2 and quarter months, which is ideal period to say that guild is consistently raiding for long enough time while still being mostly reflection of their current performance.

__________________________________________________ _____________________

2) Week points calculation
Now the tricky part is what do we call guild quality - as we want overall score to best correspond to it. For example if 100 raiders joined same guild and they cleared hc ICC 15x a week, just coz they would have 2-3 simultaneous raids each day even thou the guild and raids would be total mess, I wouldnt say the guild is 15x better than guild that kills it once :D

Now lets say we have scores of guild from raids during a week:

We split it based on raid types: type - points from raids (guild can have many runs during a week with different points from those raids)
...
ToC10n - a1,a2,.. -> calculate A (total points from ToC10n runs - function described below)
ToC25n - b1,b2,.. -> calculate B
...

ICC25d - i1,i2,... -> calculate I
RS10d - j1,j2,... -> calculate J
RS25d - k1,k2,... -> calculate K

The points from each raid can obviously differ and will be very different for each raid type. First we are gonna calculate points gained from each raid type (A,B,..) and then the week points will be simply sum of those points A + B + ...
- its good sign if guild clears even lower raid tiers, so nax and uld can also generate some points and even though the lower raids are much less important, they will simply each give much less points in base, so here we can just add them

The real question here is how to calculate how many points to give guild for given raid type. If we just summed points from all runs of the given type - it would purely reflect quantity, so follow my formula:

here we choose quantity multiplier X and calculate total points I = i1 + i2*q + i3*(q^2) + i4*(q^3) + ...
where i1 > i2 > i3 > ... (sort from most points to least)
and q = (X - 1)/X
- the coefficients are geometric sequence with finite sum quickly converging to X.

this means that first the guild gets points from their best run of given type, then *q of second best run of that type and so on - so the best run of the week is most important, and the sum of points goes up to a X*single run
Examples:
with quantity multiplier 1.5 the best run gives 100% value, second 33.3% value, third 11.1%, fourth 3.7%, ...
with quantity multiplier 2 the best run gives 100% value, second 50% value, third 25%, fourth 12.5%, ...
with quantity multiplier 2.5 the best run gives 100% value, second 60% value, third 36%, fourth 21.6%, ...
with quantity multiplier 3 the best run gives 100% value, second 66% value, third 44%, fourth 30%, ...
- any real number >1 can be chosen

The best balance to me seems to be multiplier 2, where the points from 3rd or 4th raids during a week start telling the guild that more quantity really stops being as important and should also focus on quality - its very hard to get double points than average raid by having good raid quality (would have to be very clear 0% aura run in my example table), so quantity point gain should be similar.
- the quality comparison isnt so easy because those points won't be so just one factor multiplication, quality will have some base but with wipefests will be able to go to 0 and with some good performance and special bonuses go to multiples of base and will be very raid/boss specific
- also quality will mostly reflect what people do to kill it all the time, so some solid level of quality points will be easy to gain

__________________________________________________ _____________________

3) Raid points

Total points from raids will be mostly composed of points from bosses, which we can just sum again, because the differences in boss importances will already be encalculated into points from specific bosses. But performance on bosses isn't the only indicator on raid quality.

The overall raid quality has a simple indicator - clear speed. Therefore each raid should recieve bonus points based on how fast (if it was fast) it was cleared. This part requires some research on common clear times and the max speed bonus should be very hard to reach, I would imagine the value of maximum bonus be between 20-30% of points aquired from solid run.

The measuring of clear time should be from pull of first boss to kill of last boss (as it makes sense to clear the trash pre-first boss while creating the raid). Also for longer raids its a good motivator to also put sub-bonuses for speed of parts of raids, so that guilds that are partially progressing can already "compete" at least for bonuses in parts of raids - especially ICC wings, idk if you wanna to make this for nax/uld.

ICC wings however need to be measured from trash pull (except first quarter - there normally start from marrowgar pull) to last boss kill as clear speed of that trash is big component to the smoothness of the raid. (Technically the Crimson Halls and Frost Wing quarters already have scripted events that detect their pull, wont be hard to add one to plague aswell)
- the exact values

__________________________________________________ _____________________

4) Boss points

General formula will be first to take base value of boss points, then calculate all multipliers and than multiply - all factors are taken only from the last succesful kill except wipe multiplier (no kill - no points ofc)
- values in calculations will have to be very boss specific because different factors have very different importance for different bosses

a) Base points (boss and difficulty specific)
- assign basic point amount awarded for kill on specific difficulty (will be table for each boss and each difficulty/mode/aura)
- the values need to express relative value of the boss kills and the minimum value needs to be above 40 to give reasonable results after applying multipliers as we will most likely wanna display ceiled values on web for better appearance.
- example: https://ctrlv.cz/shots/2019/04/23/cwcJ.png

b) boss specific goals (multiplier)(boss specific) - biggest window for development
- can be multiple different bonus or minus % for all kinds of boss mechanics handling
examples: computing average time beasts survived on deathbringer and call it X, add 25-5*X % of points (can go to negative if average beast lived more than 5s)
or -5 % points for each Kinetic bomb that drops on BPC
etc., here its basically up to developers how much they wanna play with it
- first sum all the plus or minus % to gain total multiplier for given boss handling
- and you might wanna cap the max possible negative bonus with the amount of points lost for wipe - so that killing it tragically is still better than wiping, however I feel like wiping and going again would be better than for example killing princes for 10 minutes while flying 20x over the room
- cannot put boss points into negative

c) Wipe and death reduction (multiplier)(boss specific)
- percentage value of total boss points lost for each wipe (25% would mean that after 4 wipes you gain nothing for killing that boss) - again you can't go below 0 boss points
- plus at end of each boss kill, the % of people that died will also be apply for partial wipe reduction (if wipe reduction is at 25% and 2/10 people are dead when boss is killed, it will add extra 0.2*0.25 = 0.05 to wipe reduction)
- easy bosses should have wipe reduction high (can be more than 100%) because it proportionally also sets the death penalty
- you can still let some people purposely die and then ressurect them, only count alive people at moment of boss kill (and only alive people in same instance :D)
- wipe would trigger on boss reset, so sindra would have to be killd right after killing the 2 small dragons
- example: https://ctrlv.cz/shots/2019/04/23/tHVz.png

d) speed kill bonus (multiplier)(boss, difficulty, aura and mode specific)
- this will be table of values, where for each boss mode and difficulty (or with/without icc aura) will be base time and bonus per second below that time
example: Lord Marrowgar 25hc 30% - 60 - 0.01
meaning - on 25hc Lord with 30% aura, anytime guild kills it below 60seconds, they get 1% extra bonus for each second below minute - on 40s would be +20%

Example calculation:
Guild went on to do Professor Putricide 25hc with ICC buff that grants 600 basic points. During the second succesfull pull Goo hit someone 13x, so they got -6.5% for that and 4 people got to 3+ stacks of Plague Sickness earning them next -4%, however 0 people got affected by Choking Gas earning them +5%, adding all mechanic modifiers to total -5.5% points. With wipe reduction being 40%, because they wiped once they automatically lost 40%, and because 4 more people were dead at the end of second pull which was kill, they lost next 4/25 * 40% = 6.4% adding to wipe points loss to 46.4% point loss for wipes and deaths. Their kill was too slow to get any speed bonus (no points loss).
Totalling 600 * (1-0.055)*(1-0.464)*1 = 303.912 => 304 points awarded
base * mechanic multiplier * wipe/death multiplier * speed bonus

__________________________________________________ _____________________

5) Windows for further development
- the ladder basic info about the guild could have some fields fillable by guild master, working as official guild description (with shivtr/forum links etc),
- would be also nice to implement undergeared bonus points - to get good starting guilds a chance (always detect equiped gear randomly from 10-30s into fight, so people dont try to fake switch gear)
- later it can include extra motivators for people to do stuff that would be fun, yet noone does - such as ulduar runs in ulduar gear (its a big waste of having one of best wow raids and only facerolling them with endgear), or start making new custom raids, that will bring new real challenge
- giving the guild leader option on webpage to select a holiday week up to 4 times per year (meaning if the guild does no 25man raid, the guilds points will freeze over the duration) - and so would last week log and current week would show holiday on web




__________________________________________________ _____________________

Implementation notes:
The ingame implementation will be pretty easy to add to instance scripts and boss scripts just building on what you already have for the speed records,only adding more data to log.

Only real question here is what data to log and how to handle crashes. You can be nice and want to display list of all attributes from the fight including boss specific fails/bonuses - in which case you would log all measured constants in some form, or be lazy and just log the multipliers and than omit some details on web and just add some general formulas for people next to it. To avoid problems with crashes you would ideally want to update db after each kill/wipe, which I dont know if you do with icc wipe count for example, you must know how is your db handling the peak hours :D

Next big part is what data to log - unlike with current ladder, where you only keep one kill log for each boss for each difficulty for each guild. (always the fastest of them) Here you will want all raids from current and last week for each guild, but in the raid logs each boss will not need all players/specs etc (but if you want you can, at this point its gonna be moved to web server, which I guess doesnt have problems) only few numbers from each boss.

The data from current week obviously needs to be kept with all details - it will be needed to do the ID reset recalculation of total points, but I would also keep previous week - 1) for possible debugging rechecking 2) because people will wanna compare what raids are what guilds managing how and if only current week was displayed - people would mostly see just parts of week :D

Long term I would just log history of guild total points, to allow possible mass recalculations if some bigger problem occurs, because you could backtrack which guild gained how many points which weak. (up to like 20 weeks, older points have just 1% value after 20 decays)

Also for the success of this the important part is web appearance - the page pve ladder would be same, just guild details could also include some point highling if guild is in top 20 (and rank), some note from guild master would be also nice ...and:
1) Current Week (number of total points gained this week)
- list of raid types (also time and points from them)+ opens sublist of raids done of that type (and total points from that raid) - clicking on raid opens new page with a)
a) Raid info (total points)
- list of bosses and points from them + openable list of all factors how the points were gained/lost
- also raid speed bonuses

2) Same for previous week

Thats some basic layout that I imagine would be pretty good and fits perfectly with your web design aswell :D ...the key part is clear display or explanation of how the points are gained/lost in each step - if for example the boss mechanics wont be specifically each displayed in calculation and subresults, than there should be link to special page explaining the point system and listing all boss mechanics and points gained/lost with them.


----------------------------------------------------------------------------------
Thanks for reading, if you want more ideas, chat/talk or help coding, im here :D

Quote:

---

Originally Posted by fatlinek

...

---

And no response? ...

Damn one month and still noone red it :D

Good job, man. Idea is good, but u know it reqire testing and cheking scalling coeffecients. I am not sure they wanna collect so many data (especially boss specific goals in "Boss points", that also ruins idea unique runs.
I mean for example kitting adds on vdw or killing it, both ways works but your system will benefit only one of it. So basically i wanna ppl look for best way for fast run it, not just best using guide. I know that content is pretty old "u can't figure something new in 10 years old game" and so on, but i believe that when competition will start smb will find smth for sure. So theese points will be meaningless cause "best way" will ignore it or mc/kite.
And i am really missing theese ingame messages about the fastest boss kills ^^, that were up when i came on molten 1st time.

U dont need to test or check anything, the point distribution will be same for everyone and can be adjusted anytime later..

And the tactic reward points should obviously be for things that are necesarry to do no matter what tactic you have... obviously on VDW u can kite mobs, but on DB if your tactic includes letting adds live and kite them.... your tactic is just bad :D ...there are many measureable goals in fights that are clear indicator or good kill... and no matter what, this part of this ladder is most irelevant and least important...

Can we ask for a debug on the website at least, either you can't reach 25 heroic record, either you can't reach 10 heroic record, either you can't reach 25 normal record , either you can't reach 10 normal record

I get that not many people check those record but for those who do it's pretty annoying to not know what comp people have used, what record is it, obviously some record can't be beaten anymore (blood prince hp bug, some old buff who were stacking and so on)

Shouldn't be that hard maybe try to modifie the system so that we don't have to check multiple criteria (because it seem that it's the problem) but just put (25,25hc,10,10hc) would be 4 categorie wich is 1 less than right now with this "multiple criteria" problem

https://armory.warmane.com/pveladder

"The PvE ladder is flawed since a lot changed since 2017.
Biggest difference is the DPS where everyone used to be over 20k, that's not the case anymore and it makes it really hard for even the best guilds to beat a mediocre guild from 2017, this makes the PvE ladder obsolete and therefore invalid in terms of "facts".
I suggest a complete wipe of the current PvE ladder since it displays old guild that dosen't even exists anymore and the timer to beat is impossible to beat since a lot of spells used to be bugged to their advantage back in 2017. "

I can se they removed the kills from 2017 since I posted this in "suggestions and change" but some guilds, like example Rank 1 Guild on the PvE ladder on icecrown is "Ubda" which has 6 memebers left in the guild. Number 4 on the rank is "Royal Bloodline" which I think is inactive too.
Suggetion - Remove dead guilds from the ladder, even if they reform and get new people, it's not the same people that got them the points aka, it's not the same guild even though it has the same name.

Add more dungeons to the charts, Like Naxxramas, Eye of eternity and Obsidian Sanctum.
Also some of the kills, like Ulduar 25 Flame levithian in 16 seconds...
Lord marrowgar 25 kill in 31 seconds, Unreal, from some a Asian guild "Anti fuming". Now we all know the latency has a huge impact on your "Dps" and to have a such huge time gap between them and everyone else, I just find it unreal. When I look at "Lady deathwisper" and "deathbringer" they are not even in the top 50. some kills on this "ladder" looks really scetchy if you ask me.
If the rank 1 kill is 30 points and the rest is 5 points. that's a HUGE point advantage on 1 boss that goes toward a guild that's exploiting or simply dosen't exist anymore.

If we are going to have a ladder like this who Ranks the Guilds on the realm we need it to be true!
Ranking guilds on a realm is kinda big, and if you going to do it, it needs to be done properly.

It was suggested so many times that the ladder should be reset for multiple reasons, but still nothing... idk what could they be afraid of, literally only negative is people being afraid of it happening again so it would reduce their will to compete in it.... but that argument currently doesnt work because almost all those records are unbeatable because of some past bugs and the fact that we have 300ms server diff in peak hours now, reducing caster burst damage by +-20% so reset would be improvement in all regards right now
+ it wouldnt be so scary if it would happen repeatedly, at least it would reflect on current guilds, if you do the reset arround start of year, people will automaticaly understand that it might be yearly thing like the squish and take it as PVE seasons... which is a good thing..


btw I guess still noone red my topic hm? :D

AT LEAST RESET THE STATISTICS ON NEW YEAR IF YOU CANT READ MORE THAN 1 ROW OF TEXT thx :D

No need to be toxic. It might sound basic but it would need TONS of time and testing to make something stable and useful.

Well thats the point - not really, they have first step with these statistics and basic version of what I described isnt many steps from it ... just apart from tracking boss kills in boss scripts track the whole raids in instance scripts and than on web change point evaluation and display based on numbers I wrote.. the many additional stuff can be added later step by step..

In any case if they dont want help coding it that I offered and dont wanna do it themselves... I will just be the toxic caps locker and write AT LEAST RESET THE LADDER :D