ESOC elo ladder

[momuuu]

If you look at Elo in chess, top players switch position after a tourney by playing against each other. It's simply the norm.

Which could be problematic for a game as volatile as aoe3. Play bad for a few games or get a few bad maps/match ups and you'll drop down the rankings rapidly. If one chess player beats another, that says much more than when one aoe3 player beats another.

[Garja]

Well depends on how close their respective rating is. Elo is conceived so that the lower player always gains more so losing the position is as easy as getting it back.
In any case, associated risks are always related to what use is made of the ratings. So far it is only used for seedings.

[Veni_Vidi_Vici_W]

why are there four frenchmen in the top 10? are you sure it's working correctly?

The fact that Tit and Kynesie are nr1 and 2 at the moment, clearly shows that the current rating is not representative, galaxy's away from it in fact. Sure they are not bad, but far from top.

Here are some calculations, including the case about India (JP elo info used since Esoc info not showing these days, though it wouldnt make much of a difference):
- Tit total 1v1 games = 3446. Games played with India = 2618. So he plays 76% of his 1v1 games India (2618 / 3446)
- Total 1v1 games won = 2396 (70%). Total 1v1 games won with India = 1931
- Winrate with India = 1931 / 2618 = 74%

Games played without India = (3446 - 2618) = 828
Games won without India = (2396 - 1931) = 465
So, 1/1 winrate WITHOUT India is = 465 / 828 = 56% = X

X = what? Take into account some free Iro all-in rushes from some time ago, known lagg issues and you are left with what exactly?
Plus, which solid topplayer (to belong in the top rating) plays actually 76% one civ?

Sure some of the Indian winrate comes from cherry picking opponents, some of it from people not countering a sepoy rush properly, but even taking that into account, the above argument still stands.

As mentioned before in this topic, without QS you can play the same people over and over, or choose who you play (and who you dodge), making the current ELO far less acurate thani it otherwise would be.

Btw EP team, you could also make India houses free next time, similar to Sioux...

[Kaiserklein]

Savage veni

[Garja]

If India is OP (I believe that's the point) then I'd argue that the EP Elo clearly represents it.

[[Armag] diarouga]

why are there four frenchmen in the top 10? are you sure it's working correctly?

The fact that Tit and Kynesie are nr1 and 2 at the moment, clearly shows that the current rating is not representative, galaxy's away from it in fact. Sure they are not bad, but far from top.

Let me elaborate, including the case about India (JP elo info used since Esoc info not showing these days, though it wouldnt make much of a difference):
- Tit total 1v1 games = 3446. Games played with India = 2618. So he plays 76% of his 1v1 games India (2618 / 3446)
- Total 1v1 games won = 2396 (70%). Total 1v1 games won with India = 1931
- Winrate with India = 1931 / 2618 = 74%

Games played without India = (3446 - 2618) = 828
Games won without India = (2396 - 1931) = 465
So, 1/1 winrate WITHOUT India is = 465 / 828 = 56% = X

X = what? Take into account some free Iro all-in rushes from some time ago, known lagg issues and you are left with what exactly?
Plus, which topplayer plays actually 76% one civ?

Sure some of the Indian winrate comes from cherry picking opponents, some of it from people not countering a sepoy rush properly, but even taking that into account, the above argument still stands.

As mentioned before in this topic, without QS you can play the same people over and over, or choose who you play (and who you dodge), making the current ELO far less acurate thani it otherwise would be.

Btw EP team, you could also make India houses free next time, similar to Sioux...

It doesn't prove that India is op, but rather that Tit can only play India, which is a fact.
Likewise, irishfaithful could only play France (he played all his rated games with France), Darwin could only play Russia (72.5% with Russia, and about 50% with the other civs), mikawanosika could only play China (70% with China, less than 50% with the other civs), and so on.
Does it make these civs op? Not really, it just shows that some players can only play one civ.

What would be relevant however, is if one guy could play every civ at top level, and had a crazy win rate with one civ.
I don't know what's your issue with India, it's probably a bit too easy to play atm, but it's really not op.

[Veni_Vidi_Vici_W]

If India is OP (I believe that's the point) then I'd argue that the EP Elo clearly represents it.

Thats part of it, though it would not be the complete case for whether India is too op (even though I think its too good ye). But you missed the other half (idk how). If you are (without taking India into account), only having such a relatively low winrate when not even playing a lot vs the 'better players', then I dont think you belong in the top spots of the ranking.

[Kaiserklein]

India is definitely too strong. And it can be played on literally any kind of map

[Mitoe]

why are there four frenchmen in the top 10? are you sure it's working correctly?

The fact that Tit and Kynesie are nr1 and 2 at the moment, clearly shows that the current rating is not representative, galaxy's away from it in fact. Sure they are not bad, but far from top.

Let me elaborate, including the case about India (JP elo info used since Esoc info not showing these days, though it wouldnt make much of a difference):
- Tit total 1v1 games = 3446. Games played with India = 2618. So he plays 76% of his 1v1 games India (2618 / 3446)
- Total 1v1 games won = 2396 (70%). Total 1v1 games won with India = 1931
- Winrate with India = 1931 / 2618 = 74%

Games played without India = (3446 - 2618) = 828
Games won without India = (2396 - 1931) = 465
So, 1/1 winrate WITHOUT India is = 465 / 828 = 56% = X

X = what? Take into account some free Iro all-in rushes from some time ago, known lagg issues and you are left with what exactly?
Plus, which topplayer plays actually 76% one civ?

Sure some of the Indian winrate comes from cherry picking opponents, some of it from people not countering a sepoy rush properly, but even taking that into account, the above argument still stands.

As mentioned before in this topic, without QS you can play the same people over and over, or choose who you play (and who you dodge), making the current ELO far less acurate thani it otherwise would be.

Btw EP team, you could also make India houses free next time, similar to Sioux...

I don't know that India is really as strong as everyone makes them out to be. Part of the reason they're successful and have a high pick rate is because they are consistently good into many different situations. They're solid on TP, non-TP, livestock, water; they have a decent rush, they defend rushes well, they have a solid boom, they have diverse unit compositions, etc.

But, they don't really excel more than many of the other civs in a lot of these areas. They're just not bad at it. Overall they're slower than other civs and don't scale that well past the mid-game, and a lot of civs have timings that are hard for India to hold.

I think a large part of the reason they're successful is because so many other civs are just unplayable in certain situations: German/Otto/Iro are practically unplayable on non-TP, other civs are unplayable vs water, Brit suffers on low-resource maps, etc.

I think in general making civs more consistent across different types of maps would go a long way to proving that India is not as strong as it is often perceived.

That being said, they are definitely a good civ, I just don't think they're overpowered.

[Kaiserklein]

If India is OP (I believe that's the point) then I'd argue that the EP Elo clearly represents it.

Thats part of it, though it would not be the complete case for whether India is too op (even though I think its too good ye). But you missed the other half (idk how). If you are (without taking India into account), only having such a relatively low winrate when not even playing a lot vs the 'better players', then I dont think you belong in first spot.

Well it's a new Elo, which means that there isn't a big rating difference between players. So playing vs the "better players" or vs lower players doesn't make that big of a difference

[momuuu]

why are there four frenchmen in the top 10? are you sure it's working correctly?

The fact that Tit and Kynesie are nr1 and 2 at the moment, clearly shows that the current rating is not representative, galaxy's away from it in fact. Sure they are not bad, but far from top.

Let me elaborate, including the case about India (JP elo info used since Esoc info not showing these days, though it wouldnt make much of a difference):
- Tit total 1v1 games = 3446. Games played with India = 2618. So he plays 76% of his 1v1 games India (2618 / 3446)
- Total 1v1 games won = 2396 (70%). Total 1v1 games won with India = 1931
- Winrate with India = 1931 / 2618 = 74%

Games played without India = (3446 - 2618) = 828
Games won without India = (2396 - 1931) = 465
So, 1/1 winrate WITHOUT India is = 465 / 828 = 56% = X

X = what? Take into account some free Iro all-in rushes from some time ago, known lagg issues and you are left with what exactly?
Plus, which topplayer plays actually 76% one civ?

Sure some of the Indian winrate comes from cherry picking opponents, some of it from people not countering a sepoy rush properly, but even taking that into account, the above argument still stands.

As mentioned before in this topic, without QS you can play the same people over and over, or choose who you play (and who you dodge), making the current ELO far less acurate thani it otherwise would be.

Btw EP team, you could also make India houses free next time, similar to Sioux...

I don't know that India is really as strong as everyone makes them out to be. Part of the reason they're successful and have a high pick rate is because they are consistently good into many different situations. They're solid on TP, non-TP, livestock, water; they have a decent rush, they defend rushes well, they have a solid boom, they have diverse unit compositions, etc.

But, they don't really excel more than many of the other civs in a lot of these areas. They're just not bad at it. Overall they're slower than other civs and don't scale that well past the mid-game, and a lot of civs have timings that are hard for India to hold.

I think a large part of the reason they're successful is because so many other civs are just unplayable in certain situations: German/Otto/Iro are practically unplayable on non-TP, other civs are unplayable vs water, Brit suffers on low-resource maps, etc.

I think in general making civs more consistent across different types of maps would go a long way to proving that India is not as strong as it is often perceived.

That being said, they are definitely a good civ, I just don't think they're overpowered.

orrrr
Accept that you can't actually balance well with such a diverse map pool?

[Mitoe]

I never said you could, but we could do better than this.

[BrookG]

Any mathematician interested?

[Aykin Haraka]

I am not even playing 1v1 for 3 months something like that, you can try to be #1
including da 2000's games, even 2010's, veni nice stats, I mean very nice stats solly
tag me next time, you are cute here

"""Take into account some free Iro all-in rushes""""
xd what ?

[Hazza54321]

ep india

[[Armag] diarouga]

If India is OP (I believe that's the point) then I'd argue that the EP Elo clearly represents it.

Thats part of it, though it would not be the complete case for whether India is too op (even though I think its too good ye). But you missed the other half (idk how). If you are (without taking India into account), only having such a relatively low winrate when not even playing a lot vs the 'better players', then I dont think you belong in the top spots of the ranking.

He's on the top spots of the ranking because he only plays India, thus him being bad with other civs is irrelevant.

The reason why he's on the top spot tbh is that he doesn't play rated games anymore. If he played against me, kaiser, raphael, erik or Mitoe, his elo would surely go down. All things considered, the real issues with the EP ladder is that you don't have to be active, and that you can dodge some players.

[Garja]

You have to be active. But you can point sit because there is no idle status or decay.
And ye you can cherry pick players which is why EP ladder should use quicksearch.

[[Armag] diarouga]

And ye you can cherry pick players which is why EP ladder should use quicksearch.

Yes

[Papist]

The real question is why people care so much about ratings in a 13 year old video game

[Cometk]

well they're used as the measure for seeding tournaments, which is designed to produce the highest quality matches possible at the latest stages of the tournament. a faulty seeding system might induce, for example, something like _h2o vs blackstar in the ro16, which would be far too early of a round for two legendary players to meet in. that also has a monetary effect on those individual players

[Kaiserklein]

The real question is why people care so much about ratings in a 13 year old video game

As an esoc staff member I'd have expected you to know tourneys are seeded by ep elo

[Hazza54321]

Even without regarding tournaments, reaching milestones can be satisfying

[deleted_user]

The real question is why people care so much about ratings in a 13 year old video game

Why do people care about anything? Life is broth (boiled)! Hoohoo!