Treaty > Supremacy

[JulianK]

9 times out of 10 treaty devolves into players making decks to jack up whatever unit is best for them and bruteforcing it cause its most optimal in team treaty, and proceeding to be pissed off cause of american gatling guns with protection that cant be beaten by anything

#NerfCoffeeMill

Your profile has to be:

1100 supremacy elo
0 treaty games
11/10 keyboard warrior

[harcha]

genuinely getting into aoe3 arguments is keyboard warrior behavior

[ShinkuroYukinari]

9 times out of 10 treaty devolves into players making decks to jack up whatever unit is best for them and bruteforcing it cause its most optimal in team treaty, and proceeding to be pissed off cause of american gatling guns with protection that cant be beaten by anything

#NerfCoffeeMill

Your profile has to be:

1100 supremacy elo
0 treaty games
11/10 keyboard warrior

Top 500 in Team Supremacy and playing team treaty primarily as of recent. Try harder.

[Mitoe]

Treaty is objectively harder than supremacy because I will never be good at it.

[Le Hussard sur le toit]

Treaty is objectively harder than supremacy because I will never be good at it.

You have it backward. Supremacy is objectively harder than treaty because people need to beat you in supremacy.

[Hazza54321]

Mitoes easy in both gamemodes so its hard to make a comparison

[Mitoe]

Mitoes easy in both gamemodes so its hard to make a comparison

Yikes, we play BO21 now?

[JulianK]

true aoe kings would bo21 7 sup, 7 tr, 7 dm games

[iron_turtle]

forgetting Empire Wars and Tycoon Mode.

[Aizamk]

don't forget about treaty 10, 20, 30, 60 and 90...
and king of the hill. there's that too.

[W_S_A]

9 times out of 10 treaty devolves into players making decks to jack up whatever unit is best for them and bruteforcing it cause its most optimal in team treaty, and proceeding to be pissed off cause of american gatling guns with protection that cant be beaten by anything

#NerfCoffeeMill

Your profile has to be:

1100 supremacy elo
0 treaty games
11/10 keyboard warrior

Top 500 in Team Supremacy and playing team treaty primarily as of recent. Try harder.

Gatling guns in treaty lmao. Try harder.

[JulianK]

don't forget about treaty 10, 20, 30, 60 and 90...
and king of the hill. there's that too.

we speak of relevant game modes only ... i understand a 1-game-mode nub wouldn't know which is which tho

[Hazza54321]

And you put deathmatch as a relevant gamemode

[LionHeart]

Virign treaty player vs Chad Supremacy enjoyer

[Zutazuta]

NR 40 HM NWS

[Goodspeed]

Since the greater than (>) symbol is ambiguous, we can't say if treaty > supremacy. By what measurement? Player base? Supremacy clearly wins. Game design? Supremacy clearly wins, since there aren't 40 minutes of no decisions and no player interaction. But let's focus on the claim made in the OP:

Treaty is the obviously more challenging and more difficult game mode to master which both paper and reality showcase.

If we take "challenging" and "difficult" to mean the same thing (if that was not the intent I invite the OP to let me know the difference) there are 2 claims here:
1. Treaty is the more challenging game to master on paper
2. Treaty has proven to be the more challenging game to master in practice

First, we need to talk about the word "master". What does it mean to "master" a game? Is it to play the game completely flawlessly? That would be impossible either way, so we can conclude that it's not what the OP meant. Is it, then, to be the best player in the world? That seems like a bit of a high bar as well, plus no one ever managed to be the best at both games so our analysis would soon reach a dead end. Is it to be within 100 ELO of the best in the world? We could potentially work with that, though it is still problematic. The best player in Go has an ELO of 3850. The second best, 3700. So there would still be only one player who mastered it, which sounds wrong intuitively. But let's ignore that issue and say to master a game is to be within 100 ELO of the best in the world.

So let's start with claim (1).
How do we measure how challenging something is to master on paper? Is it about how much time it takes a player with sufficient talent to get there? Or how smart one has to be to be able to do it at all? Both seem reasonable. Since the OP didn't specify, and since "the amount of time it takes" is more workable here, let's go with that.

To be within 100 ELO of the best player is to win at least 33% of your games against them. How much time would it take, on paper, to get good enough to beat the best sup player 33% of the time?
First we need to figure out which player we're talking about. Obviously not the current top players, who are trash. We would be talking about iamgrunt, who was the best player during the height of the game's competitiveness. How much time would it take, on paper, to get within 100 ELO of him at his peak?
I have no idea. No one has any idea. This is an impossible question. The same applies to treaty, of course. No one has any idea how much time it takes to get within 100 ELO of whoever the best player is/was there. It follows that the OP's claim cannot be proven or disproven, and is therefore to be considered frivolous, which in itself is some admittedly anecdotal evidence towards treaty players having low IQs but I digress.

Claim (2), then.
This one is problematic for another reason, which is that some skills needed for both games transfer between them, meaning no player in history has ever learned both of them starting with a clean slate. Either a player with treaty experience starts playing supremacy, or the other way around. For example, a treaty player comes to play sup and due to the transferable skills takes much less time to become good at it than they otherwise would have. Can we ever know how much time they would have taken had they started with a clean slate? No. Can there ever be a player who starts playing both games with a clean slate? No. Can claim (2) then ever be proven or disproven? No. Ergo, this claim is as if not more frivolous than the other.

TLDR: 0/10 post, go back to playing dead games LOL

[harcha]

I don't think your argument against claim 2 is particularly good. I think you can find other ways to measure if thing one is harder to master than thing two other than just time.

[Goodspeed]

I don't think your argument against claim 2 is particularly good. I think you can find other ways to measure if thing one is harder to master than thing two other than just time.

Well...?

[harcha]

I said what I said, that's my openis.

[Goodspeed]

Fine, I'll answer it for you.

A game's difficulty can be somewhat reliably estimated by the range of elo ratings. See: viewtopic.php?f=315&t=17041

... the difficulty of a game can be estimated by taking the amount of "skill levels" there are, where a skill level is a rating point where one wins about 2/3 of the games against the previous level, and loses 2/3 to the next level.
Intuitively, this logic seems sound. The more skill levels there are, the more there is to learn.

This is convenient for games that use ELO ratings, because with ELO there are ~100 points between the levels. That is to say, a 2000 rated player will win 2/3 of their games against a 1900 rated player.

According to the linked post, there are about 16 levels in supremacy, meaning there is 1600 elo between the lowest rated player and highest rated player. This would give AoE3 supremacy a difficulty rating of 16. (For comparison: Checkers at 13, AoE2 at 18, SC2 at 20, Chess at 20, Go at 36).
This criterium might not work here however, because treaty has such a small player base that the ladder may never have reached a "settled" state. It may have pre-DE, on the simple elo ladders, but I don't know what the range would have been back then. Does anyone remember?

Side note: Is there a DE ladder where we can see the full range of elos?

[Le Hussard sur le toit]

Fine, I'll answer it for you.

A game's difficulty can be somewhat reliably estimated by the range of elo ratings. See: viewtopic.php?f=315&t=17041

... the difficulty of a game can be estimated by taking the amount of "skill levels" there are, where a skill level is a rating point where one wins about 2/3 of the games against the previous level, and loses 2/3 to the next level.
Intuitively, this logic seems sound. The more skill levels there are, the more there is to learn.

This is convenient for games that use ELO ratings, because with ELO there are ~100 points between the levels. That is to say, a 2000 rated player will win 2/3 of their games against a 1900 rated player.

According to the linked post, there are about 16 levels in supremacy, meaning there is 1600 elo between the lowest rated player and highest rated player. This would give AoE3 supremacy a difficulty rating of 16. (For comparison: Checkers at 13, AoE2 at 18, SC2 at 20, Chess at 20, Go at 36).
This criterium might not work here however, because treaty has such a small player base that the ladder may never have reached a "settled" state. It may have pre-DE, on the simple elo ladders, but I don't know what the range would have been back then. Does anyone remember?

Side note: Is there a DE ladder where we can see the full range of elos?

Where did you get your stats for your difficulty rating ? I'd like to do some comparison. In particular I wonder at your 20 for chess.
I've just had a look at lichess's rating stats and found a 24 difficulty rating for blitz (lowest player 600 elo, top player currently 3000). I've a hard time agreeing lichess blitz is actually more difficult than standard over-the-board classical chess (if only because top online blitz players tend not to win otb classical tournaments, while top classical players tend to win online tournaments too).

(Besides I think you need to look at the rating distribution and the number of standard deviation if you really want a statistical assessment of the competitiveness of a game.)

[Goodspeed]

Fine, I'll answer it for you.

A game's difficulty can be somewhat reliably estimated by the range of elo ratings. See: viewtopic.php?f=315&t=17041

... the difficulty of a game can be estimated by taking the amount of "skill levels" there are, where a skill level is a rating point where one wins about 2/3 of the games against the previous level, and loses 2/3 to the next level.
Intuitively, this logic seems sound. The more skill levels there are, the more there is to learn.

This is convenient for games that use ELO ratings, because with ELO there are ~100 points between the levels. That is to say, a 2000 rated player will win 2/3 of their games against a 1900 rated player.

According to the linked post, there are about 16 levels in supremacy, meaning there is 1600 elo between the lowest rated player and highest rated player. This would give AoE3 supremacy a difficulty rating of 16. (For comparison: Checkers at 13, AoE2 at 18, SC2 at 20, Chess at 20, Go at 36).
This criterium might not work here however, because treaty has such a small player base that the ladder may never have reached a "settled" state. It may have pre-DE, on the simple elo ladders, but I don't know what the range would have been back then. Does anyone remember?

Side note: Is there a DE ladder where we can see the full range of elos?

Where did you get your stats for your difficulty rating ? I'd like to do some comparison. In particular I wonder at your 20 for chess.

I went with Carlsen's classical rating as the top rating and 900 as the bottom, but I could be off with the bottom. Not sure what I based that on at the time.
For the rest, see the linked post. Unless you already checked that and found it poorly sourced (it is), in which case lmk I'll source it properly

I've just had a look at lichess's rating stats and found a 24 difficulty rating for blitz (lowest player 600 elo, top player currently 3000). I've a hard time agreeing lichess blitz is actually more difficult than standard over-the-board classical chess (if only because top online blitz players tend not to win otb classical tournaments, while top classical players tend to win online tournaments too).

It could be that Lichess blitz is a more accurate measurement of difficulty than OTB classical because the official OTB classical ladder only contains rated players. Lichess has many players who only play online and don't take the game very seriously. If all Chess players were on the official ladder, Magnus' rating probably would be much higher. So 24 levels seems closer to the truth. The same applies to the checkers ladder I used, which probably also contains only club players.

(Besides I think you need to look at the rating distribution and the number of standard deviation if you really want a statistical assessment of the competitiveness of a game.)

Elaborate on this?

[Goodspeed]

So I found the DE ladders https://www.ageofempires.com/stats/ageiiide/. The sup ladder ranges from 2200 to about 400, meaning there are 19 levels. This is close enough to the 16 from before, but it is interesting that the range is wider. Maybe they use a modified version of the elo formula (which makes it no longer elo but I wouldn't put it past them). Treaty ratings range from 1700 to 600, meaning there are 11 levels. Of course with player bases as small as these, this is probably not a meaningful measurement when it comes to the objective difficulty of a game. BUT it is meaningful when we consider how hard it would be to get within 100 elo of the top. Because when there is much less competition, it's significantly easier.

Difficulty, it could be argued, depends on the level of competition. If we go with that definition, we can safely say sup is harder. If we assume similar competition levels, we can't say which is harder because the treaty player base has never been large enough for a properly settled elo ladder to emerge.

By the way, the treaty ladder includes team games. The theory doesn't apply to team ladders because the formula is different and they tend to have wider ranges because of this. A ladder for solely 1v1 treaty can be assumed to have an even narrower range. If someone remembers a pre-DE elo range for treaty 1v1 that would be great.

[dicktator_]

We don’t play rated games in treaty so the current elo range doesn’t mean much. I don’t remember elo ranges of the past but I do remember pr ranges, tr top players would usually hit around pr40-41 (Julian might’ve gotten higher in 2020 when he started winning every damn game), while I remember ep top players hitting pr45, and of course legit pr50 in the past. Getting within 100 elo of the best player would be insanely difficult either way, since I would not be within 100 elo of Julian if we had legit elo, so someone would have to get better than me to accomplish that in treaty.

[Goodspeed]

I guess PR ranges, while it won't allow us to deduce the amount of skill levels and compare to other games, will at least serve to compare the two AoE3 game modes. It's still not an ideal comparison though because team games muddy the waters, especially considering top treaty players played relatively more team games than top sup players. Also, still, the small player base in treaty may have prevented it from ever developing its "true" range. Hard to say how many active players you'd need for that to happen.

I suppose in theory, as long as you have active players at each skill level, it should slowly converge to the true range. But with a tiny player base you'd be dealing with players in the bottom few skill levels quitting more often because their win rates will be relatively lower. Why? Well, with a player base of millions, even on the lowest skill levels your win rate would be near 50% as you would only get matched up with players as bad as you. If instead there's only a couple of other players at your level, you would more often be playing against players above your level resulting in a poor win rate and this could be discouraging.