Guild Wars 2

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken

Ooh, interesting probability there for us to roll DH and Mag. I would love to see that matchup happen.

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken

Ark Bladesteele.2943:

Ooh, interesting probability there for us to roll DH and Mag. I would love to see that matchup happen.

The prophecy is real.

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken

32% HoD/DH/SBI matchup probability? Make it happen, Anet.

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken

FA 0% chance of T3
YB <1% chance of T3
SoS <10% chance of T3

T2 is almost locked.

Arcadio.6875:

FA 0% chance of T3
YB <1% chance of T3
SoS <10% chance of T3

T2 is almost locked.

T1 locked, T2 almost locked.

That’s a statement that the ratings system/match making algorithm is poor.

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken

FA has a tiny chance of swapping with TC lol.

SoS has a less than 2% chance of swapping with DB.

Blockhead Magee.3092:

Arcadio.6875:

FA 0% chance of T3
YB <1% chance of T3
SoS <10% chance of T3

T2 is almost locked.

T1 locked, T2 almost locked.

That’s a statement that the ratings system/match making algorithm is poor.

lol no that’s a statement about people stacking servers.

Hmm I thought you had to be within 100 Glicko to even have a chance. TC/fa > 100 SOS/db > 100…

Liston.9708:

Hmm I thought you had to be within 100 Glicko to even have a chance. TC/fa > 100 SOS/db > 100…

Nope. RNG is exactly that: random. We’ve been matched with servers 300 glicko higher than us and 200 glicko lower than us. It does happen.

Ark Bladesteele.2943:

Liston.9708:

Hmm I thought you had to be within 100 Glicko to even have a chance. TC/fa > 100 SOS/db > 100…

Nope. RNG is exactly that: random. We’ve been matched with servers 300 glicko higher than us and 200 glicko lower than us. It does happen.

If that is true, how was t1 100% locked for so long? Once the first domino falls maybe there is an unavoidable ripple effect though……

I’m going to guess that it has something to do with being exactly 100 rating points apart roughly translates into a 1% chance. So being 130 rating points apart starts bringing the chance closer to 1% and above 0% (or 0.001%)

When you get a match-up with very large glicko differences happens, it is because there are no other servers with similar glicko left to match-up with, which is when you get 300 glicko difference etc. Because they system MUST match you up against someone. If there was enough closer options it would have picked those instead, but servers are to different in size for the glicko to do its job as it is supposed to.

The Human Error.

Tier 1 been the same since 16 months ago. Pretty sad.

Liston.9708:

Ark Bladesteele.2943:

Liston.9708:

Hmm I thought you had to be within 100 Glicko to even have a chance. TC/fa > 100 SOS/db > 100…

Nope. RNG is exactly that: random. We’ve been matched with servers 300 glicko higher than us and 200 glicko lower than us. It does happen.

If that is true, how was t1 100% locked for so long? Once the first domino falls maybe there is an unavoidable ripple effect though……

The server has +/-100(or maybe a bit more) random roll. But it is always possible to be matched with servers with a much higher/lower glicko rating.

E.g. DB has a small chance of swapping position with SoS. If that happens, that means they are matched up with a server(FA) that is 200 glicko higher than them.

Exciton.8942:

Liston.9708:

Ark Bladesteele.2943:

Liston.9708:

Hmm I thought you had to be within 100 Glicko to even have a chance. TC/fa > 100 SOS/db > 100…

Nope. RNG is exactly that: random. We’ve been matched with servers 300 glicko higher than us and 200 glicko lower than us. It does happen.

If that is true, how was t1 100% locked for so long? Once the first domino falls maybe there is an unavoidable ripple effect though……

The server has +/-100(or maybe a bit more) random roll. But it is always possible to be matched with servers with a much higher/lower glicko rating.

E.g. DB has a small chance of swapping position with SoS. If that happens, that means they are matched up with a server(FA) that is 200 glicko higher than them.

But you need to compare the lowest of one tier and highest of the one below. Sure fa can be 200+ if SOS/db swap. The question was what is that breakpoint ? Lowest t1 and highest t2 is 130ish apart – that is where I thought the 100 came into play. It can’t be 200 or t1 would not have been 100% locked the last several months….. Chaba may have the right answer……

WTB tier roll magic find.

Liston.9708:

Exciton.8942:

Liston.9708:

Ark Bladesteele.2943:

Liston.9708:

Hmm I thought you had to be within 100 Glicko to even have a chance. TC/fa > 100 SOS/db > 100…

Nope. RNG is exactly that: random. We’ve been matched with servers 300 glicko higher than us and 200 glicko lower than us. It does happen.

If that is true, how was t1 100% locked for so long? Once the first domino falls maybe there is an unavoidable ripple effect though……

The server has +/-100(or maybe a bit more) random roll. But it is always possible to be matched with servers with a much higher/lower glicko rating.

E.g. DB has a small chance of swapping position with SoS. If that happens, that means they are matched up with a server(FA) that is 200 glicko higher than them.

But you need to compare the lowest of one tier and highest of the one below. Sure fa can be 200+ if SOS/db swap. The question was what is that breakpoint ? Lowest t1 and highest t2 is 130ish apart – that is where I thought the 100 came into play. It can’t be 200 or t1 would not have been 100% locked the last several months….. Chaba may have the right answer……

That is a good point. Either the random roll range is not 100 or there is some other restriction on the glicko roll.

the range is capped at 100 for each server. but when you’re talking about one server switching places with another, there are two servers involved. if the higher server rolls -100, and the lower one rolls +100 they can switch as long as they’re no more than 200 points apart.

since TC and FA are estimated to be 128 points apart, that is easily within the 200 point range. it’s not possible for one server to move 128 points just due to its own random roll, the other one needs to help. for example, if TC rolls -100 then FA only needs a roll of 29 or higher. but if TC rolls -50 then FA must roll 79 or higher to beat the 128 point gap.

for more info:
https://www.guildwars2.com/en/news/big-changes-coming-to-wvw-matchups/

(note that that particular article doesn’t mention the 100 point random limit. I think that was added later)

-ken

I think we’re looking for explanation of why your analysis always showed T1 locked with 100% chances of BG/JQ/TC when both FA and TC’s ratings were within 200 points of each other but now it shows less than 100% chance.

ah, now I understand the question.

ok. 100 is the hard cap, but that doesn’t mean that +100 or -100 will always be possible. the amount of randomness a server can get depends on its ‘deviation’ value. the exact upper value is 10 + (0.45 x Deviation).

TC has a predicted deviation of 127.7872, so 10 + (0.45 × 127.7872) = 67.5

so TC can’t possibly roll higher than +67.5, or lower than -67.5 because their deviation isn’t high enough to allow it. likewise, FA has a predicted deviation of 131.3507, which means their random roll will be between +69.1 and -69.1.

add 69.1 and 67.5 together and you get 136.6, which is high enough to get past the 128 point gap. in previous weeks this wasn’t the case.

note that the amount of variation is different for every server, because every server has a different Deviation value (and the values change from week to week). this complicates the math and makes it so that there’s no single number that’s the limit for everyone (or even a single number that’s correct for one server week after week). this is why everyone just uses 100 points as the limit, because at least that is the same for everyone.

-ken

Thank you for that explanation

Bet you $100 TC wins this week they got too close to FA

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken

I see both NA T1 and T2 are completely locked this week.

-ken

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken

.02 fa in t1 – go go go….

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken

TC/YB/Mag is possible lol.

0.000465%

Arcadio.6875:

TC/YB/Mag is possible lol.

0.000465%

About like a legendary drop – hate to waste that rng roll on a matchup…..

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken

Hallelujah

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken

I see NA T1 and T2 are completely locked again, and T3 is half-locked (Mag and SBI are guaranteed, but the third spot still has some variability left).

-ken

IMHO that’s to be expected after the extreme glicko cliff that was occurring between T2 and T3 matches for so long. The lack of a spring season meant that no shake up could occur and the entire system is allowed to stabilize. It will continue for awhile until players get tired of it and as long as T2 guilds don’t start engaging in the cross-server match-fixing that T1 does.

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken

Lol I love how EU RNG basically sticks its tongue out at any statistical predictions

God bless EU RNG.

sorry about the delay. I calculated these numbers this morning but for some reason I wasn’t able to access the forums.

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken

Hi IoJ! We meet again. Welcome to the living dead tier
-DR

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken

here are the estimated matchup probababilities for the upcoming matchups. “na1” and “eu1” files show the probabilities of getting any single opponent, “na2” and “eu2” files show the probabilities of getting a particular pair of opponents. the last two files are the estimated rating outcomes for the current matchups, assuming that final scores don’t meaningfully differ from today’s scores.

-ken