AOE2DE without inca feudal vil rush

[IYIyTh]

So what?

Why do birds fly, Umdeuter? What is the meaning of life and why is it 42?

 

[Umdeuter]

They try to be Superman, I think, but I am not a scientist

 

[IYIyTh]

They try to be Superman, I think, but I am not a scientist

Neither is anyone in this thread.

Welcome to the internet.

 

[Umdeuter]

there must be plenty scientists on aoezone

 

[The Bloodless]

I think the discussion is that Incas are/aren't boring because they got feudal villager upgrades removed. It seems to be settled that they aren't underpowered and they weren't underpowered, and they aren't overpowered. A few pros picking civ for 3-4 games out of 2000 is not gonna shift the win rate significantly (common sense).

I like the way Incas are now myself because people have not been "forced" to go Inca rush due to it being a pretty strong strategy, so they're finding that Incas have a lot of other great options.

 

[nimanoe]

Ironically Professor HongeyKong introduced some weird data that said there was elevated utilization of Inca's over other civs. There was a lengthy discussion about that data set provided, where four instances showing 4 pro's significantly picking inca's over civs w/ a significance of over 2 std deviations.

This never happened, @HongeyKong tried to calculate the Standard Deviation of different pro's by how often a civ was played. In that image he posted the only person for which the civs lined up with how often they were played is for Hera, all the other pro's didn't have Inca's as their most played civ. Then there were a couple of pages of trying to let you understand, but for some reason you still don't.

After folks laughed that off, turns out here were at least 4 separate data sets someone came up with that indicated regardless of reason, that this was indeed happening post nerf.

Out of all games that were analyzed, Hera had 2 or 3 games where he picked Inca's. This would've boosted the winrate of Inca's by 0.11-0.16%
If you think that validates your "theory" that pro's are picking Inca's on purpose to prove they're fine then go ahead, but the winrate is still way higher compared to the other elo groups.

Btw: It's pretty ironic that you call HongeyKong a professor, when you're acting like a student that's unwilling to accept facts or learn anything.

 

[IYIyTh]

This never happened, @HongeyKong tried to calculate the Standard Deviation of different pro's by how often a civ was played. In that image he posted the only person for which the civs lined up with how often they were played is for Hera, all the other pro's didn't have Inca's as their most played civ. Then there were a couple of pages of trying to let you understand, but for some reason you still don't.


Out of all games that were analyzed, Hera had 2 or 3 games where he picked Inca's. This would've boosted the winrate of Inca's by 0.11-0.16%
If you think that validates your "theory" that pro's are picking Inca's on purpose to prove they're fine then go ahead, but the winrate is still way higher compared to the other elo groups.

Btw: It's pretty ironic that you call HongeyKong a professor, when you're acting like a student that's unwilling to accept facts or learn anything.

We must be looking at different data sets.

I realize this refutes your point earlier that this wasn't happening at all, but there are four examples where there is a 2 std dev significance or greater of picking Inca's for these four players, unless the data is wrong. It's not mine, so entirely possible.

Stop conflating your argument that the impact is minimal vs. it isn't happening. They are different arguments entirely.

We can't determine the true impact without validating every single pro and not even Professor HongeyKong has the time for that, though I would very much enjoy to see what he would find if he analyzed the entire playrange.

 

[nimanoe]

View attachment 194304

We must be looking at different data sets.

I realize this refutes your point earlier that this wasn't happening at all, but there are four examples where there is a 2 std dev significance or greater of picking Inca's for these four players, unless the data is wrong. It's not mine, so entirely possible.

Stop conflating your argument that the impact is minimal vs. it isn't happening. They are different arguments entirely.

So you looked for this post from Hongeykong but didn't read this? :

EDIT2: The numbers dont match the civs but that doesnt matter.

 

[nimanoe]

Just look at the sources and you'll understand you're wrong.

Hera played Inca's 22 times in his last 445 games: https://aoe2.club/playerprofile/199325
Viper played Inca's 10 times in his last 396 games: https://aoe2.club/playerprofile/196240
JorDan played Inca's 15 times in his last 501 games: https://aoe2.club/playerprofile/2858362
TaToH played Inca's 3 times in his last 158 games: https://aoe2.club/playerprofile/197388

 

[Umdeuter]

tbf here, it's not very smart to post this table like that

(it's also not very smart though to believe that 5 players all have the same ranking of civs played)

edit: but this is funny for sure

In that image he posted the only person for which the civs lined up with how often they were played is for Hera, all the other pro's didn't have Inca's as their most played civ. Then there were a couple of pages of trying to let you understand, but for some reason you still don't.

but there are four examples where there is a 2 std dev significance or greater of picking Inca's for these four players

he explained to you how the data is wrong dude

what are you doing

 

[IYIyTh]

I probably missed the second edit.

So Professor HK's data's borked. Great?

 

[nimanoe]

I probably missed the second edit.

So the data's borked. Great?

Not for your theory, as the data suggests that Inca's was right around the middle of the pack for JorDan, TheViper and Tatoh and we've already showed that Hera only picked Inca's 2 or 3 times, making the impact on the Inca's winrate minimal.

 

[IYIyTh]

Not for your theory, as the data suggests that Inca's was right around the middle of the pack for JorDan, TheViper and Tatoh and we've already showed that Hera only picked Inca's 2 or 3 times, making the impact on the Inca's winrate minimal.

Not necessarily disproven then either without analyzing a significant portion of that upper echelon of ELO. Great!

 

[The_Philos]

Isn't it pointless aswell because top 10 players with >60% winrate overall inflate every civ's winrate, simply as they are so much better than their regular opponent? (Given that they mostly don't pick civs)

 

[IYIyTh]

Isn't it pointless aswell because top 10 players with >60% winrate overall inflate every civ's winrate, simply as they are so much better than their regular opponent? (Given that they mostly don't pick civs)

Only if they aren't shown to have been picking Inca's disproportionately. Given that Inca's have such a small utilization rate (3rd lowest in their range,) it's not unreasonable to see an impact, given those same folks are full time aoe2 players while the rest of the elo range are not.

 

[nimanoe]

Not necessarily disproven then either without analyzing a significant portion of that upper echelon of ELO. Great!

True, but it's quite hard to do, as we can't check if someone picked a civ or if they randomed it. However, given that we've already seen the stats for four of the biggest pro players/streamers, I'd be curious which player you think has picked Inca's a lot. If you name some, we could analyze their stats and see if they've played Inca's more than one would expect, at least giving some validity to your claim.

 

[HongeyKong]

Just look at the sources and you'll understand you're wrong.

Hera played Inca's 22 times in his last 445 games: https://aoe2.club/playerprofile/199325
Viper played Inca's 10 times in his last 396 games: https://aoe2.club/playerprofile/196240
JorDan played Inca's 15 times in his last 501 games: https://aoe2.club/playerprofile/2858362
TaToH played Inca's 3 times in his last 158 games: https://aoe2.club/playerprofile/197388

This
At the time, while adding other pros numbers I coundlt be bothered to export csv from aoenexus or input the numbers from aoe2club accordingly, so I just spammed the numbers from top to bottom, as the purpose was just to show the pattern anyway.

Out of the four pros, Tatoh was the one who most probably have picked his top played civ (Mongols - 3.8xSD). The other threes top played civs are only 2.5-ish from their means, which is plausible to be obtained by randoming, but I wont be surprised if they did pick a civ every now and then.

EDIT: P-value for two-tailed 2.5 z-score is like 0.012

EDIT2: Just noticed the two top Argentinians picked Chinese A LOT 11

 

[HongeyKong]

True, but it's quite hard to do, as we can't check if someone picked a civ or if they randomed it. However, given that we've already seen the stats for four of the biggest pro players/streamers, I'd be curious which player you think has picked Inca's a lot. If you name some, we could analyze their stats and see if they've played Inca's more than one would expect, at least giving some validity to your claim.

Burden of proof 11

 

[IYIyTh]

Or -- it could remain a comment suggesting that the anomaly of the 1650 range win rate could very well be explained by such a factor that is at least evident in one pro player's civ utilization. In context with an overarching thread direction about the Incan nerfs great success in both dumbing down the game, and making it the third least utilized civ in the game in the 1650+ range.

 

[im the cookie]

Honest question:

Could that low pick rate be due to higher elos preferring random civ?

 

[IYIyTh]

Honest question:

Could that low pick rate be due to higher elos preferring random civ?

Your question is based on a supposition that the data would indicate isn't true, ie: it's why Chinese/Mongols/Franks/Huns/Mayans are so significantly picked above the others.

It's also more likely that the utilization rate for Inca's will actually drop further over time as the "try them after the nerf," phase passes.

 

[Elvaenor]

Or -- it could remain a comment suggesting that the anomaly of the 1650 range win rate could very well be explained by such a factor that is at least evident in one pro player's civ utilization. In context with an overarching thread direction about the Incan nerfs great success in both dumbing down the game, and making it the third least utilized civ in the game in the 1650+ range.

You keep going back to the fact that they're very low in utilization, but some civ has got to take that spot right? Why do you keep focussing on this so much? It's obvious that not all civs can be middle of the pack and some civs will just naturally be more disliked than others. Why are you so extremely insistent that the inca's shouldn't have this place but another civ instead? Just because you like to play their trush doesn't mean everyone should like the civ like you do (which people obviously don't or didn't)?

Edit: Also, their utilization on the 1650+ level is 1.7%. For a full random pick the chance should be around 2.7%, so inca's while obviously picked less than would be expected for full random picks it's still not as bad as you portray it to be. This together paired with their good win rates (even if pro's try to boost their win% I won't ever be conviced their real win rate would be below 48%) This does make it seem like there's not that much of a problem with the nerf, and the civ in its current state.

 

[HongeyKong]

Honest question:

Could that low pick rate be due to higher elos preferring random civ?

 

[IYIyTh]

Eh, not seeing how that confirms that's based on random civ.

They are the third least picked/played civ, which reflects a lack of confidence in utilizing the civ post nerf, like it or not.

 

[MrIster24]

there must be plenty scientists on aoezone

There are a lot of smart people in the community for sure.

Just look at the sources and you'll understand you're wrong.

Hera played Inca's 22 times in his last 445 games: https://aoe2.club/playerprofile/199325
Viper played Inca's 10 times in his last 396 games: https://aoe2.club/playerprofile/196240
JorDan played Inca's 15 times in his last 501 games: https://aoe2.club/playerprofile/2858362
TaToH played Inca's 3 times in his last 158 games: https://aoe2.club/playerprofile/197388

I'm happy to take a more formal look at some data if given a matching data set + question to answer, maybe I will when I get back to my comp. I'm much more fluent on the analysis side than using APIs to pull the data so I'll try to check it out.