Updated Statistics (#4)

Potentially, will take a look. And let you know

1 Like

SOTL made a video using your site.

Conclusion - Win rate diverts more at higher elo. In other words civs are more unbalanced at higher level.

I think it’s just that eco bonusses have more influence at the highest level.

Right. And maybe better use of timing bonuses.

1 Like

And also very low skill players not using bonuses correctly…

At lower elos, civs that are “autoplay” carry the player, but even so it doesnt matter ar very low elos when players dont have a clue about competitive gaming

I would stress though that whilst the numbers do show that civ choice matters more at higher levels it doesn’t show that civ choice doesn’t matter at lower levels. In particular the difference in the effect size is actually quite small implying that whilst there is a difference the difference really isn’t that big.

3 Likes

There speaks the true Statistician :smiley: (you will I think prove most useful)

1 Like

Yeah. Civ pick will matter all the time regardless of elo level anyway.

Difference of difference is very small. 11

1 Like

Do we have enough sample size for water maps? I’m not implying they are very necessary. Water meta is very strongly established.

Sorry forgot to reply to this, currently have about 11k matches for >2000 Elo I think so yer just about enough to start making outputs for it. Not made up my mind yet whether to add it to the site or whether to just upload it to gdrive and provide a link here. Will let you know once I’ve created it and uploaded the content somewhere

Not even close unfortunately. Maybe if I combined all matches across all patches I might have something. I guess the water hasn’t been significantly touched in a very long time so it might be representative still. Will have a play and let you know

1 Like

Maybe just upload the link to drive.

Whoa. People just don’t play water maps.

As an aside I am currently seeing the Dravidians dominate on team nomad. Admittedly it could just be limited data or bad code but yer wasn’t really sure what to make of this. Were people expecting Dravidians to be a top tier team nomad civ ?

EDIT:

4 Likes

Heres the headline plot for the >2000 Elo open maps group:

The full collection of plots can be found here:

3 Likes

Yes indeed, Dravidian is really too strong in nomad

1 Like

To be clear though it’s only team nomad they appear to just be average in solo nomad.

Edit: Solo Nomad >1200 Elo

Probably the Thirisadai? I can imagine it has huge impact in TGs.

I did. Their TB is huge for nomad. Persians is considered a good nomad civ due to having +50 wood, food. But that’s their civ bonus. For Dravidians it is basically +25 wood for all of your team mates.

Thanks. Bengalis is better than Italians. Now I can sleep in peace knowing that Italian AoE fans can not sleep at night. 11

2 Likes