Guild Wars 2

Statistics by Team (since using GW2PVP):
My Account—
Red Team-
37 games played.
Victory
15 (40.5%)
Defeat
22 (59.5%)

Blue Team:
55 Games played
Victory
34 (61.8%)
Defeat
21 (38.2%)

Stats from gw2pvp.de overall:
Victory Statistics by Team
32050 unique matches played on red
Victory
16546 (51.6%)
Defeat
15504 (48.4%)

34302 unique matches played on blue
Victory
20225 (59.0%)
Defeat
14077 (41.0%)

Seems to me it’s close – but more likely to win if on a ‘blue’ team?

There looks like a big enough sample size here of matches played over numerous months (from players registered on gw2pvp.de).

I’m starting to not like playing on ‘red is dead team’

This is purely through matchmaking statistics right? No hotjoins included?

Some of the maps have imbalanced elements, for example on Foefire the blue home point (waterfall) is easier to hold / harder to decap than the red home point (Quarry).

To a double-sided coin, the Strong Law of Large Numbers states that the probability of Heads and Tails will be 50% if it were to be flipped infinity.
Source

For the sake of argument, lets ignore team setups and pretend that Winning is 50% no matter who or what player is on either team. If Red and Blue were to be played indefinitely, there wouldn’t be a 50% win ratio between the two colors. Like Zinks said, it’s because the Blue and Red sides aren’t mirrored exactly. The probability would be above or below 50% (for or against) for the Blue or Red side.

The amount of games played since GW2 launch is still an incredibly low number to define the true probability of color’s wins/loss ratio because Matchmaking, as well as the often unpredictable skill factor of players, is NOT a true 50% algorithm. Also, the middle on Foefire map is definitely more of a reason why either Blue or Red will lose the map. The Middle on almost any map is often the deciding factor of which team snowballs first, rather than the win of either sides.

Yes. For example blue gets waterfall as homepoint on legacy which is a huge advantage.

What I think I’m trying to say is..
You’re more likely to win on blue comapred to red teams.
Its as if blue – your team has a higher MMR than the red team (on average).

There is nearly a 10% difference here – between red/blue matches.
I believe gw2pvp.de includes ranked/unranked games in that total counter.

These are also likely to be people who are actually interested in their overall performance.

Itsa a cute pet theory(nice Source), but Saiyan’s comment does not accord with axiomatic probability or formal statistical methods.

Saiyan.1704:

…the Strong Law of Large Numbers…

Almost, but actually totally irrelevant. Better luck next guess. This is a (kinda) rare case where statistical inference can be performed by EXACT testing…no approximation required, so behavior at the limit (dealt with by the LLN) is irrelevant.

The FORMAL result supports the OP’s claim; it is significant at a p value < 0.00001… in other words, there is less than a 1 in 100,000 chance the test has yielded a false positive
(wanna verify? H-naught: categorical variables are independent).

But this takes the terms of data descriptions to be appropriate…hint: they are not and there’s nothing surprising in this data……… garbage in garbage out, OP is abusing statistics

mikey.5380:

Itsa a cute pet theory(nice Source), but Saiyan’s comment does not accord with axiomatic probability or formal statistical methods.

Saiyan.1704:

…the Strong Law of Large Numbers…

Almost, but actually totally irrelevant. Better luck next guess. This is a (kinda) rare case where statistical inference can be performed by EXACT testing…no approximation required, so behavior at the limit (dealt with by the LLN) is irrelevant.

The FORMAL result supports the OP’s claim; it is significant at a p value < 0.00001… in other words, there is less than a 1 in 100,000 chance the test has yielded a false positive
(wanna verify? H-naught: categorical variables are independent).

But this takes the terms of data descriptions to be appropriate…hint: they are not and there’s nothing surprising in this data……… garbage in garbage out, OP is abusing statistics

Besides statistic propaganda, what’s your point? The data OP posted may support HIS claims but it’s not absolute, unquestionable data. Everything here is speculation :p

mikey.5380:

Itsa a cute pet theory(nice Source), but Saiyan’s comment does not accord with axiomatic probability or formal statistical methods.

Saiyan.1704:

…the Strong Law of Large Numbers…

Almost, but actually totally irrelevant. Better luck next guess. This is a (kinda) rare case where statistical inference can be performed by EXACT testing…no approximation required, so behavior at the limit (dealt with by the LLN) is irrelevant.

The FORMAL result supports the OP’s claim; it is significant at a p value < 0.00001… in other words, there is less than a 1 in 100,000 chance the test has yielded a false positive
(wanna verify? H-naught: categorical variables are independent).

But this takes the terms of data descriptions to be appropriate…hint: they are not and there’s nothing surprising in this data……… garbage in garbage out, OP is abusing statistics

Rants at OP that he is wrong. Looks Below.

Evan Lesh.3295:

It’s cool the community has the tools to find things like this. We saw the same thing and updated the matchmaking to remove the color bias so metrics were prettier. There will always be a best and worst player in the match and we can never make perfectly even teams, but we can at least make sure the best player doesn’t always end up on the same team color

Dev comes in, follows up with investigation, fixes Issue. Thanks Evan for looking into this!

Brief Anecdote, I activated gw2pvp for myself when it popped up and then immediately proceeded to go on a ten game losing streak. Most recently I’ve gotten back into my normal trend of games, but my Blue team win ratio is definitely better than my Red teams, by a small margin of 5%. Still working on bringing my gw2pvp win ratio to match my overall winratio.

Flip a coin 100 times and see if you get 50%.

All solo Q

When I queue with my engie, I end up on blue team more than I do red. The screenies are just a small sample but in general, I end up on blue 65-70% on engie.

Well, I dunno about sPVP, but I can say without a doubt that team color matters a lot in Keg Brawl. Red team fills first, which is a big advantage in a game that usually fills late or not at all. Additionally, the top players on the winning side of the previous match tend to get placed on the red team in the next match. And whatever method they use to assign teammates tends to get stuck and place the same good/bad players together for extended streaks. Also, volunteer reward is/has always been broken.

Not that ANet really cares about KB, but I do, so I thought I’d mention it.

And I though I was paranoid for thinking red won more often.

You should also check the code that determines who gets the NPC kill in Forest, or make it so the NPC gets downed and you actually have to finish it, because I swear I see red taking the kill most of the time when players of both teams are fighting by them.

If I was right about the first thing, I could also be right about the second.

Look at greatswords, A-net always had blue bias.

But the concept is interesting when you take into account that not all sides of the map are even. It’s easier to knock blue off home on skyhammer for example, and various blink spots are natively different to their respective side. Even if you think of the difference between Henge and Mine and how you would zone the point.

Oh and for fun lol~:

MithranArkanere.8957:

And I though I was paranoid for thinking red won more often.

You should also check the code that determines who gets the NPC kill in Forest, or make it so the NPC gets downed and you actually have to finish it, because I swear I see red taking the kill most of the time when players of both teams are fighting by them.

If I was right about the first thing, I could also be right about the second.

Actually you were incorrect. It was blue team that has the higher chance of winning. And that was because of how mmr distributed the teams.

MithranArkanere.8957:

And I though I was paranoid for thinking red won more often.

You should also check the code that determines who gets the NPC kill in Forest, or make it so the NPC gets downed and you actually have to finish it, because I swear I see red taking the kill most of the time when players of both teams are fighting by them.

If I was right about the first thing, I could also be right about the second.

There’s another thread about that. its the killing blow that gets the kill. If there’s conditions from both teams on him as in burning from both teams it’s basically a coin flip.

ReaperJr.5967:

No hotjoins included?

I highly doubt that anyone considers Hot Join to be reliable information for anything PvP related….

1) No restrictions on who’s on what team
2) People can freely switch teams
3) People really don’t take it seriously, it’s a sand box to test builds / get your feet wet

Wolfey.3407:

ReaperJr.5967:

No hotjoins included?

I highly doubt that anyone considers Hot Join to be reliable information for anything PvP related….

1) No restrictions on who’s on what team
2) People can freely switch teams
3) People really don’t take it seriously, it’s a sand box to test builds / get your feet wet

Which is also exactly why i asked.