Arete.7019

Just to be clear, the numbers don't really support that very strongly. I added another month to the sheet to make this more obvious. 2024 is pretty average across 4 years and 2023 is higher than average.

The werdes sites numbers don't really seem correct, they don't line up with what I recorded from the API or with the current matchup, but idk whats going on there. Anyway here's the last four years for roughly the same week using their numbers https://docs.google.com/spreadsheets/d/1JwPQEX9Q-KHcCOazF3FgGeOy7oubm5kMWOZL4ymyOVU/edit?usp=sharing. Lowest numbers being right during peak COVID seems kinda off, but I guess you could make some fun conclusions from that data like "peak boonball balancing increased numbers but WR decreased them". I'm sure many in these forums will love that take.

Sure, idk why it needs more than the "I wrote it down" that your numbers had but: https://docs.google.com/spreadsheets/d/10ZUqXnC8iYgDgfFnm0Wr2_FSkeHHxQgWw9sEiO_JCvI/edit?usp=sharing https://docs.google.com/spreadsheets/d/1OAW9gUpdOGgSvvplBbqMkoYyCgSf30Fn1r-7ht1ikk0/edit?usp=sharing

I recorded them from the API on those weeks

Sure, here's some numbers: 24-6-7 -> 24-6-14 matchup (Final matchup pre WR) Total K+D EU (all tiers) - 807801 Total K+D NA (all tiers) - 738752 24-8-2 -> 24-8-9 (Final matchup pre Janthir Wilds API shutdown) Total K+D EU (all tiers) - 803780 (-0.5%) Total K+D NA (all tiers) - 770414 (+4.2%) If you still want to bet that numbers are down 30% I'm willing to take that bet, let me know 🙂

So if you want the % of total warscore in a skirmish to gain an equal % of the VP for that skirmish. The VP is based on activity yeah, but my whole point is that warscore isn't a perfect measure of team strength. This is my only real disagreement, if warscore perfectly assessed team strength then weighting VP from it exactly would be fine, but with warscore being a fuzzy measure the current split is better. I brought up that example of EU T3 this week because I really doubt that blue is 90% of the strength in those offhours skirmishes, but they'd be getting 90% of the score. Same thing for the primetime skirmishes.

Fixed that, thanks I wasn't trying to say that the higher % is better, just that it meant 1st/3rd were closer. I think the new distribution of 50%/33%/17% for 1st/2nd/3rd is fine, having a bigger gap encourages players to go for the win harder because it matters more. I definitely agree that there could be much better in-game incentives to focus the winner. T3 structures being terrible to fight in and the winning server being the most likely to have T3 structures is the main thing that leads to the winner not being focused imo. Yeah, I think I get what you're saying, but my problem is that what if a B player during the night is doing 60 avg score and a B player during the day is doing 50? You could have exactly equal populations at any skirmish and it still wouldn't end up fair if most of the spread comes from offhours.

I think you pretty much get it, its very difficult to get a large advantage in prime time because dodging fights and especially because hiding in objectives is so strong. T3 EU this week is a good example https://imgur.com/a/kBP6DKr. Check out the prime time skirmishs and skirmish 5, 17, 29 where blue side is dominating. If VP was proportional to warscore the spread would be 11.4 VP in 5, 10.3 in 17. Compared to 13 and 25 where the spread would be 17.7 and 2.5. Green doesn't even win skirmish 13 despite being a stronger team than red. Warscore isn't doing a great job of measuring team strength.

Yeah ok, we're on the same page

I'm a bit confused if we even disagree here. Say you have 0 healing power and 10% outgoing healing. You use a skill with a base healing of 1000. It will heal for 1100 (1000 * 1.1) (on allies only, since outgoing healing doesn't increase self heal).

.5 is a 50% decrease, 1.5 is a 50% increase. 1.45 is correct if you have an outgoing healing increase of 45%. Also base healing on a skill can be affected by outgoing healing modifiers fwiw, this is easily testable in game, just go use rice balls in the raid training area

Yeah, but I think where we're at odds here is that you're overrating how viable warscore is for measuring the difference between teams in prime time. I believe a team could easily be much stronger in prime time but only win warscore by a relatively small margin, while a team that is moderately stronger in offhours could easily gain a massive margin on warscore. I don't think the current VP are significantly out of line with the real populations, you can take a look at this spreadsheet I made with data from two weeks ago https://docs.google.com/spreadsheets/d/1Y65Sh8-PuB5nEGAV1sGlFgYVEjfCqb7t93tpZxiR150/edit?usp=sharing. % activity on the summary pages is equal to the % of total K+D in the region. K+D as an activity measure probably somewhat overvalues prime time and undervalues offhours, so I wouldn't be surprised if the % of total VP is equal to the % of population per skirmish.

If 20% of your population plays in one skirmish and 8.58% of the population plays across 3 skirmishes, you can't give each skirmish the same amount of points without that 20% of the playerbase getting screwed on their ability to determine the outcome of matchups. 8.58% of players in those three skirmishes are gaining far more points per player than the 20% of players in the single skirmish. Now imagine that the skirmish with 20% of the playerbase always wins, while the 3 skirmishes with 8.58% of the playerbase always lose. In this alternate reality numbers are exactly equal in each skirmish for every team and VP gained is the old 5/4/3. Can you see why the players in the 20% population skirmish might be upset that the VP given is equal regardless of population?

You're getting the wrong answer because you're applying what is presumably supposed to be an increase for your outgoing healing modifier as a 50% decrease. The equation E = (ACD) + (AB) gives you (.5 x 1000 x 1.2) + (.5 x 1500) (1.5 x 1000 x 1.2) + (1.5 x 1500) 600 + 750 1800 + 2250 E = 1350 E = 4050 Your calculator gets this right, so I'm really confused how you managed to get it wrong? https://imgur.com/GPPWATQ

Its not that I think anet's numbers are that far off, its that: 1) Players in those skirmishes are highly disorganized and easy to dominate for the small amount of competent groups playing at those times 2) There aren't enough competent groups to possibly spread across all worlds in all skirmishes, so there will be giant disparities in many of them They wouldn't have a proportional amount of impact with your changes though. The spread is the entire mechanism that determines who is winning or losing, not the total amount of points gained in a skirmish. Primetime could give 500/499/498 points and it would have less impact than an offhours skirmish that gave 5/3/1. This is fair, making each player have an equal impact is also an important part of the VP changes. By lowering the spread in prime time matchups you'd be giving a larger percentage (and therefore larger influence on matchup position) to other skirmishes. This directly reduces the impact players in that timezone have on a match. You can even see this in your earlier post where the offhours heavy red server moves to 1st instead of the more primetime balanced green server. The majority of players in that region will have better matchups if the green server advances and red doesn't because a lot of red's win would be coming from a very low population TZ. I agree that there is potential for play hour manipulation with the current system, but I don't think in the real world this is an actual problem. There are few groups that will log out when good content is available and few groups that will keep playing when no content is online. I'd also be very surprised in anet doesn't take multiple weeks or months of data for those calculations to make it harder to game the system.