Characters are being updated now for Dragonflight Expansion!
- Item Level Character Rankings
- SimDPS
Also check:
- Guild Item Level Rankings
- Item Level Character Rankings
- SimDPS
Also check:
- Guild Item Level Rankings
Comments: 21
Improved Scoring for Tier15, the Long Version
11 years ago
This topic is a complicated one that needs further explanation. Many players are posting comments without understanding the subject.
WoWProgress is the oldest and the most advanced WoW Rankings website. It was launched 5 years ago and Sunwell was the first raiding tier for it.
Our scoring system was designed for Sunwell gating. Now it seems many people don't remember what gating is.
Gating is not related to skipping bosses
When you defeated a boss, there was a gate that didn't allow you to see next one. Top guilds had to wait for new content to open every time they killed a boss.
Game developers decided when it's time to open the next boss. Once the gate is open, all the guilds that have defeated the previous one, started trying the new boss simultaneously (I'm not counting region time difference here).
Gating means that all competitive guilds start working on the next boss simultaneously
Even if you killed the previous boss faster, you weren't have any advantage on the next one.
Instead of one big race, the raiding tier contained several small races.
Point decay and counting total amount of points from several bosses was a perfect solution for gated content.
Now I'll explain why the system isn't working so well when there is no gating.
With point decay, every day you spend sitting on a boss costs you some amount of points.
There are different scores attached to different bosses.
Lets say there is a raid with 3 bosses: boss1, boss2, boss3.
They have daily scores attached: 10 for boss1, 10 for boss2, 50 for boss3.
It means every day you spend working on boss1 costs you 10 points.
boss1 and boss2 are intermediate bosses and boss3 is the final and the most complex one, which is why there are more points for it.
Looks absolutely logical, right?
Harder bosses cost more points. If you kill the harder boss faster, you get more points comparing to the easier boss kills.
Now there is a trap in this logic.
Example:
Some guild lost a day on boss1 because they made a mistake in tactics, or their key players had a birthday party, etc (the reason doesn't matter here).
This mistake costs them 10 points (1 day).
Quiz question: what is the flaw in this logic?
Part 2 is here
Kernel how can i contact you? I want to ask you something.
kind regards
Esmaticx
They also lose 10 and 50 points from 2 & 3 bosses. Exactly what we experienced in t14 content - had ~2 weeks break at xmas/new year, couldn't catch up with other guilds, even though we progressed the bosses faster.
Callie#11693
The only difference is that now spectators can have a better idea of just how well various guilds are doing in their progress towards the final result, as before it was misleading.
Take Vigil for example:
http://www.wowprogress.com/guild/us/mal-ganis/Vigil
They skipped Council and then made it to 4/13 by killing the next two bosses, including Megaera, which appears to be an even harder boss than Council in 25, yet they are still on less points than a guild that has killed council and still sits at 3/13.
Second...no one has finished the race. As of right now the furthest progressed guild is 11/13 H again meaning the rules are being applied evenly.
Third, you point out one of the inherent problems with comparing the US and EU but didn't bring it to a logical conclusion. Measuring it this way gives both the US and EU guilds a more even playing field. Both groups have a lockout of 7 days meaning they have the exact same amount of time to down the content as any other guild. TBH the US even still has a slight advantage because they will have received their full 7 day try a full day ahead of EU guilds.
Finally this scoring addresses the previous ordering issue. For example, back when the leading guilds were stuck on H Dark Animus Blood Legion was showing as 1st even though they were third to kill H Primordius. This was because of the early advantage they received on the previous 7 bosses that the EU teams didn't even have the opportunity of matching.
P.S. This is a game. Specifically this is a game that other people (for the most part) are playing while a majority sits on the sidelines cheering. Don't try to bring out the holier than thou ehtics card to a completely trivial issue.
Alchav#1643
wowprogress is not the World of Warcraft raid organizer. They aren't affiliated in any way.
We are changing our progress measurement system just before weekly reset.
If you skipped a boss and your guild is strong, go kill it now after the reset.
I think the value of each boss should be dynamic and based off difficulty. The boss that has been killed the most in 10 man is worth 100 points. The boss with the second most kills in 10 man is worth 200 points, etc. The same can be done in 25 mans based on 25 man kills. So if for instance Twin Consorts is absolutely brutal in 25 and not that bad in 10 it would be worth more for the 25 man progression than the 10 man because less guilds would have killed it relative to some other 25 man boss while 10 man guilds would skip some bosses to do Twin Consorts. The sum of all the bosses leading up to the end boss would remain the same for 10s and 25s. Thanks to guilds skipping, easier bosses get adjusted to be worth less points than the more difficult bosses. Update boss kill values at the end of the week or autmagically as the ordering for number of times a boss has been defeated changes. Also, despite it being terrible trying to compare 10s and 25s it at least does better than the skipped bosses are worth less because 25s may be back loaded in difficulty on an instance whereas 10 mans might have a harder time with the 3rd boss and an easier time with the 10th or 11th... Anyway, /my2cents
Lol apparently you've forgotten what it is like to try and discuss something with WoW players. No matter the overwhelming amount of evidence you will never get a consensus on "what is more difficult." Draztal weighed in on the forums earlier today (I guess it's yesterday by this point) in a discussion of this very thing.
Just an example to demonstrate why that sort of point scaling would be ridiculously complex: bosses that require tight dps, healing or tanking requirements. My guild personally has been successful with healer intensive fights, decent at tank intensive fights and fairly poor at dps checks. Other guilds I've spoken to ran into a road block at stone guard and feng because of the tanking requirements but simply obliterated the next three bosses (generally considered harder) because their tanks were their weak links. The essence of what I'm getting as is difficulty is relative and when you try to ascribe points in a system with relative difficulties it's going to be messy to say the least.
Most killed boss: worth 100 pts
Second most killed boss: worth 200 pts
Third most killed boss: worth 300 pts
etc,.
or by some weighted function of the most killed boss. Below is just some arbitrary example values not following a function but just to give an idea where using the number of guilds that are killing something you can dynamically scale boss value and just make the total number of points allocated to bosses a constant and put a min and a max points to get some distribution.
Most killed boss (1000 guilds killing it): 100 pts
Second most killed boss (500 guilds killing it): 200 pts
Third most killed boss (499 guilds killing it): 201 pts
Fourth most killed boss (200 guilds killing it): 500 pts
Dethkrik#1795
This exactly. Taking this current progression as an example, 10 man guilds should be getting vastly more points for killing Council of Elders or Megaera than for killing Tortos or Ji-kun. On the flip side, based on its obvious difficulty, Dark Animus should be worth more than the rest of the pre-end bosses. While everyone may not be able to agree on what is difficult and what isn't, it fairly easily becomes apparent from seeing the number of guilds that are able to down certain fights which are road blocks and which aren't.
I think it is also time to do away with any sort of combined 10 and 25 man rankings. While some encounters may turn out equally difficult in 10 and 25, I think it is grossly unfair to both 10 and 25 man guilds that all of the encounters give the same points over both difficulties. Some encounters are always much simpler on 25 man, and some much simpler on 10 man. While it may simply be that Paragon are just that supreme, I think Twin Consorts going down after 4 hours on 10 man vs the day and a half Method and BL have spent on it without a kill is a prime example of the complexity and tuning being vastly different across the 2 raid sizes on some encounters.
In theory it looks good, but in practice it's not applicable. Chaning boss points will make things hectic in the beginning of the race. Guilds will be very confused if boss points are not settled.
So in the end, your progress will be rated correctly and you will all (us included) have a fair ranking.
Stingray
re1gn1te#1228
re1gn1te#1228
Мгва
So now (as far as I understand) WoWProgress only counts timestamp of the boss kills, and this is what matters.
Also those still during the progress, what matters for their ranking in the intermediate ranks is also the date of their latest kill. So 2 guilds a 6/13 HC would be ranked with the "who did the latest kill faster".
I am all fine with that, but in that case you dont even need points.
I estimate that you keep them handy in case you see "gated" content again, like in Tier 14, where the tier was split in 3 mini raids and a "time decay" point system would be meaningfull once again for the final boss of each mini raid.
In the mean time having a page with 20 guilds all having the same amount of points is a bit missleading, or confusing, at least in my opinion. But i am good no matter what you chose.
Point system always carries an amount of subjectivity, especially when the point rewards are not flat numbers per boss.
Still remember lootship having more points than the bosses before it in heroic for examble. Also it is meaningless to get point decay (that is carried all the way though progress) for bosses not available to you yet. Like Asia's 2 days delay to access the same content vs US for examble.
There is not a perfect system, but i apreciate all the info i can find in this site, it is invaluable.
A deletion of the 10/25 fake race from the default home page though, would be the greatest change this site could bring.
If you didn't have the reduced points for "skipping" (or probably following progression difficulty order), the system will still work fine. Eventually this guild will need to kill the harder boss. The only problem is if the tier ends while they picked up some "easier" kills. Same basic issue happens with the reduced point system as well though. Someone ends up in an awkward place if they don't clear at the end of the tier and "skipped" bosses.
Please let guilds follow the natural difficulty progression on heroic and don't create an artificial scoring system to encourage us to kill bosses in a certain order.
Strife#1977
Straife
Since when has the linear order of Heroic bosses been in any way related to the difficulty of the encounter? Elegon was vastly easier than Spirit Kings. Lei Shi was often done weeks ahead of anything else in Heroic ToES. And perhaps the best example of all: back in the day, Lady Deathwhisper was far harder than all but 3 bosses in ICC (Putricide, Sindragosa, Lich King) despite being boss #2.
Heck this isn't even going to work as intended, because regardless of this bizarre change, a "naturalistic" order is going to emerge. Guilds will skip around and "suck up the crap rank for a couple weeks" to get gear to make harder bosses die sooner than they would if they were tackled under geared. All this does is artificially obscure the ranking system.
If a boss would take two weeks to do under-geared but can easily be downed with some strategically placed gearing, people are going to talk about this on MMO Champion, everyone is going to do it in that order, and thus the entire system will collapse because the point of the change will have been collectively ignored. If the community at large decides Boss #3 is enough a pain of the ass you want to do 4,5,6,7 before it, everyone will do it in that order to save the total amount of time it takes to do those 5 bosses. The point of the change becomes completely ignored because people will figure "you such up the point loss for a few weeks, ignore your rank, and then rubber band back later, having saved weeks instead of doing it in the "linear order".
You are doing false assumption that we are going to force someone to kill bosses in linear order.
Again, there is nothing wrong with skipping bosses.
It is okay if no one will follow the linear path. Then all guilds will get the same amount of points.
But if a guild was strong enough to kill some bosses without skipping, they get more points.
Guilds that didn't skip Heroic Council are stronger than ones that skipped the encounter.
If you feel that skipping this boss was a mistake, come back next week and catch up.
Now lets compare it with point decay system.
Council would give much more points than Tortos and Ji-Kun.
Skipping Council would reduce your points _forever_, and that's actually very strong pressure to not skip the bosses.
So following your logic, the point decay system is worse than current scoring.
Stingray
It is also worth noting that only unless a guild fails to defeat the final boss will their penalty from the earlier encounters be permanent. If they fail to defeat the final encounter, the decay can again further differentiate them from amongst the other 11/13 (H) guilds instead of them all sitting at 8250 points.
People can't kill an end boss this tier (so maybe would need to be re-addressed if they could) without clearing the rest of the bosses, so for the top tier guilds, the rankings will return with top kills. One guild isn't necessarily weaker they just choose to spend their time on different bosses.
Who is your target audience? People following the initial race for first few weeks or the rest of the community after that (I'm sure you have that data).
Removing decay is fine, but also remove the reduced points for "skipping" bosses. Rankings will work mostly fine for the initial race and be greatly improved (imo), without influencing guilds to kill certain bosses, for the rest of the world.
But right now the table is a bit misleading. Some guilds have the same points as others but are not on the same place. I know it looks this way because it is secondary sorted by Killdate, but only by reading around on the site.
If you would add your second sorting parameter to the table, it would make things clear on the first look.
10man and 25man will never be perfectly tuned, and some bosses will always be harder on one of the difficulties. So the only thing you get from keeping them in the same column is 10 vs. 25 drama, just put them into two different columns?
I know that people will say, "just press the 25man tab", but the fact is that the majority of the people who visits wowprogress is not doing that. What's on the frontpage is the only thing people will see.
Surely you can just award 1 point for a kill so guild son 10/13Hc have 10 points and those on 6/13 Hc have 6 points.....why is ther a need for this 750 number??
Also if you wanted to award less points for people skipping bosses then they can get 0.5 points then.
But i agree with a poster above...yes it will work in the long run with being able to see who killed what first as it will be in a logical order, but many people want to see how much faster one guild was than another.
Say in 4 months time i look back and see Method and BL in position 1 and 2 with the same points...i would have to click on the guilds to actually see how much faster 1 guild killed it than the other....i mean with this way 1 guild may have killed it 30 mins before the other and be in position 1 and 2, or that 1 guild may have killed it 2 weeks after the other and still be in the same position and i wouldnt really know without investigating further.
So tbh each way has its pro's and cons.
Your new way of ranking guilds is definitely better, but it will take some time for people to undestand they have to look at ranking and not the score.
So you should probably either make the score reflect the new method (i.e. redefine it using the same method the ranking is calculated) or if you want to keep it for legacy compatibility with other thing, juste hide it from us, because in it's current state it's only confusing.
Ie $decay=true where ($bossnumber-1).status=undefeated (something along those lines)
So if boss 8 is killed, and boss 7 is still undefeated, then the decay will still be happening on boss 8. The decay stops once boss 7 is killed.
This will allow 2 guilds to rank properly where 1 guild skips a boss but kills it next reset vs a guild not skipping (ie stuck on boss 6) but manage to kill boss 7 and 8 next reset. Depending on the time stamps if Guild A kills boss 7 and then boss 8 before Guild B kills boss 7 in the next reset, they will rank higher. If it is after, then guild 2 will rank higher (as the points will be matched to register as if a guild killed 2 bosses on the same day) yet the time stamps will obviously show that boss 8 was killed the week before.
Personally I dont mind either system (as my guild barely ranks under 2k) but just thought if you prefered the decay system, it should still work with the above method in place (unless im missing something).
It doesn't realy seem logical to me to penalize guilds, that skipped boss, just to get more kills = better ranking with old system and then u change the rules in the middle of week :/
(For example, we are probably gonna lose realm first only because the second guild was stucked on the Horridon and did not skipped Council (not killed yet) after u announced the change. So when both guilds will get at the same progress, lets say 5/13HC, we are just gonna be second even if we killed all the bosses first?)
Also regarding the rank time adjustment by region, all I can say is that WF is WF no matter what and it should stay that way by default, however having a filter-like option to enable time adjustment can not hurt anyone.
For me this grading system doesnt mean anything in a Worlwide view due to the fact you don't take in considerating that some areas can play very earlier than other. nimber of hours/mns after server opening is more representative for me, and is the real grading.
Hope I wrote clearly, Engly is not my mother tongue.
Strife#1977
Straife
Jin'rok->Horridon->Tortos->Maegera->Ji'kun->Council->Drumuru-> The rest.
Many guilds are even doing Tortos before Horridon.
Council is being left, intentionally, until 3 heroics after them are killed. Truly, doing "Council" first is in no way indicative of a stronger guild. If anything it is indicative with a more reckless and poorly organized one. If Guild A spends 3 weeks doing Tortos, Maegera, Ji'kun before touching council, and I get three, two, and one kills of those bosses on heroic before I go to council, Guild A will have a substantial gear advantage over Guild B, that spent two of those three weeks on Council. That gear advantage will mean a faster, if delayed Council kill, but more significantly, will mean that Guild A, with more weeks of more Heroic boss kills under its belt than Guild B by the time they get to Drumuru, will have a substantially easier time meeting its tight enrage.
Thus the entire justification for the change has collapsed. No one is doing it, exactly as predicted. The currency in raiding is Time. You have 7 day a week guilds, 4 day a week guilds, 5 day a week guilds, 3 day a week guilds, etc. They all maximize their progression within the confines of the hours they raid per week. They will, entirely logically, sacrifice short term ranking for long term benefit. Killing more things faster sooner is always better than killing fewer things, slower. It keeps guilds motivated for one, but also it means that with more gear, they can make more of their limited time on harder fights in the future. After all, why engage Council with an Average guild iLevel of 515 when you can do it in 3 weeks with an Average guild ilevel of 523 because you killed three other heroic bosses in the interim? That's madness and any guild leader who does that is a fool.
And considering that no guild (in the US 25 top 50) is taking WoWProgress on it's de facto mandate that to get full points, you need to clear in order, it's hard to escape the conclusion that the change in ranking method has been a comprehensive failure. People are just ignoring getting 250 points, because nobody is actually thinking and acting on "we'll jump up if we kill Council first".
People in these threads called this from a light-year away.
During the first week of heroic progression there were guilds that didn't skip bosses and they were ranked higher, which is more relevant than ranking by amount of killed bosses.
After you kill hard bosses you skipped before, your score points increased significantly which is also good for relevancy.
Their rankings will depend on timestamps of their latest bosskills.
We cannot predict which path to the last boss is the most efficient for your guild.
If a guild didn't try to game the boss kill order and just killed them without skipping, and haven't lost too much time, we assume that the guild is stronger and our rankings reflect that.
On my server, 25man Jinrokh same day.
Horridon guild1 3 days before guild2.
Tortos Guild1 1 day after guild2..
With 3 HC guild 2 is ranked higher, decay system it would have been the other way around most likely, so what changed?
- score
- time of last boss kill
Zefrenus
Same issues with other guilds doing 10man and show progress as 25man
sorry i dont know where post this so i put it here :p
You could update progress of the other guilds too, please let me know if not fixed.
As for the site itself is still great but I feel it needs more of a community, a forum/blog/news page of some sorts. Like the videos consider allowing guilds to post screenshoots of kills and maybe allow users to comment on guild pages (if they allow) this will give more interaction between guilds and fans like wow raider .net
btw have you considered parsing logs on wowprogress? Would be great to rank players dps, heals, etc, this info is way better to have for guild recruitment.
I am finding the scoring system a bit meaningless (since it just represents the number bosses killed). I totally get the part where in the old system a delay in boss 1 costs you on boss 2. But why not make boss 2 decay depends on the difference in time between boss 1 and boss 2?
For example let's take Method and Envy as an example of 2 EU 25 man guilds.
Method beat Envy to Boss 11 boss by 3 days. However they beat them to Boss 12 boss by 20 days. That means that Envy spent 17 more days on Boss 12 (that includes 2 additional weeks of gearing from 2 additional resets on bosses 1-11). Still Envy and Method get the same amount of points on Boss 12? Why?
I think a better alternative is have Envy take a penalty of 17 days decay on boss 12 (instead of the actual 20days difference). Another alternative is have the boss points decay with every reset. So if you kill the boss on first reset (while you still have mostly gear from previous tier), you get full points, second reset you lose X points, at 10th reset(after you farmed previous bosses and have full Tier 15 gear) you lose 10X points, etc.
There could be some form of detection.
While a good system in theory, Why isn't it working when a 25man guild kills Lei Shen on 10N and then gets a 25 H. Jin'Rokh Kill without having killed 25N Lei Shen?