Popularity of techs and other statistics

[ffm2]

over_70_perc.zip over_10_under_70_perc.zip

[ffm2]

under_10_perc.zip

0_perc.zip

[Deicide4u]

 

Does anyone research this tech? Is there a point to it, besides interesting historical trivia and an additional spearmen unit for the Spartans?

[Grautvornix]

Sometimes: if I don't have enough wood but plenty of metal and food. (happens rarely but it provides the same as standard hoplites I think)

[Atrik]

1 hour ago, Deicide4u said:

Does anyone research this tech? Is there a point to it, besides interesting historical trivia and an additional spearmen unit for the Spartans?

It's a powerful tech/unit. But it's hard to adapt eco to it as you are now forced to send any new recruit into battle. Also it's competing with other metal hungry options, even more powerful like champs...

[ffm2]

For those questions I made these charts. 9% of the sparta players do that research at about min. 16. Apparently I did it 2 times too, but it must have been at a time where I already won and just research techs that I find and can afford because I don't aim for this unit.

[real_tabasco_sauce]

lovely graphs @ffm2

I find the tech and unit to be most useful if I've lost a lot of spears and want to protect skirms (with brasidas) even when i am depleted of wood.

[mimi01]

the most popular are all essential things needed to unlock more. 

[ffm2]

Since the discussion came up I checked my replays for the popularity of heroes. The script is much generated by AI quickly to get a answer, but the question isn't that complex.

Every replay with a commands.txt and metadata.json is counted. No check for cheats or how good the players are.

If you want to run it you need to place it in .local/share/0ad/replays/heroes/A_heroes.py.

Edit: Now only the first hero per player in that game is counted. The first result was a bit less than I expected and this reflects preferences better.

 

A_heroes.py

Edited Tuesday at 12:02 by ffm2
Updated

[Dakara]

On 09/03/2025 at 11:43 AM, ffm2 said:

Marian Reforms (roman_reforms) with 11% of roman players researching this tech are a trap (most of the time).

Skrimisher cost 50 f 50 w
have a gather rate 0.5 berry 0.25 grain 1 meat 0.75 wood 0.5 stone 0.5 metal

after Marian Reforms the skirm costs 50f 50w 15m
and the gather rates are 0.24 berry 0.12 grain 0.49 meat 0.37 wood 0.24 stone 0.24 metal

A woman has gather rates
1 berry 0.5 grain 1 meat 0.7 wood 0.35 stone 0.35 metal

Marian reforms are only useful if
A) You are about to deal the finishing blow
Your economy can rely completely on trade

Because you men now costs metal but can gather it less efficient than woman (and woman gather it very bad in comparison to normal men already)
No other upgrade is considered here.

after the reform they still have one unit with normal cost and normal rate of resource (the spear i guess)

[Deicide4u]

29 minutes ago, Dakara said:

after the reform they still have one unit with normal cost and normal rate of resource (the spear i guess)

Correct. It's basic rank spearman, just a different model. Spears are slower than javs, though.

Marian Reforms is powerful, but it's mostly a flavor tech. You can build multiple Army camps near your opponent's territory and train swordsmen that are almost as good as Marian Legionaries.

[real_tabasco_sauce]

nice, thanks @ffm2! There are certainly meta heroes, but honestly there's a pretty good mix here.

[ffm2]

Some skewing factors: Hippocrates is available in P2 which works in his favor in the statistic.

Spartan have free heroes as team bonus. So any team with them could send a less favorable hero on a suicide mission (E.g. sending them straight for farming women). Although I think the major benefit is the supporting aura on a full army at the first big clash.

The single percentage heroes could see a buff at least.

A big reoccurring argument is dancing with heroes. I set it usually on passive and retreat and still get blamed. I don't dance full on "left right click". I lure around the army when I notice a focus. But the game should be in the way that the player uses all options available. Maybe a solution on side of the game design can be found.

[Deicide4u]

51 minutes ago, ffm2 said:

Maybe a solution on side of the game design can be found.

The solution is to implement a proper target priority system. Targeted unit gets far enough away? Stop. Enemy units nearby? Attack them instead.

The main question is: how much is "far enough away"? Alternative is to reduce the target priority on enemy heroes to a minimum. That way, you'd have to manually target a hero 99% of the time.

[Servo]

I used all the techs except probably the counterintelligence which I couldn’t find out how it works now, whether I play SP or MP. So don’t remove them.

The assessment is imo based on land maps to which most players play. Lack of metals and stones are just the hindrances to research. 

[Atrik]

48 minutes ago, Deicide4u said:

The solution is to implement a proper target priority system. Targeted unit gets far enough away? Stop. Enemy units nearby? Attack them instead.

That's more or less what's in com mod rn. The retargeting happens every landed attack.
You'll still have a lot of problems due to targeting with those passive* 'targeting systems': 
One big limitation of targeting is currently when controlling large army walking into enemy base, your men will likely split into groups, and large amounts of them will start walking into the opposite direction of where you made an attack-move in order to chase an isolated female for example.

A active targeting feature on the other hand will allow users to determine targets, and retain some form of micro.

1. You can mitigate hero dance with 1 click instead of being forced to make 200 - here the passive targeting might mitigate it for 0 clicks
2. You can choose which units your will chases with 1 click instead of being forced to make 200  - here the passive targeting cannot mitigate pathings
3. You can spread damage of your ranged units with 1 click instead of being forced to make 200 - some passive targeting by @real_tabasco_sauce aim to mitigate it for 0 clicks.

As well, depending on the refresh rate of your passive targeting, the active targeting can be much lighter on performance.

Active targeting feature thread with selection box

*Here I used passive as in "not user controlled" and active as in "user need to make an input".

[ffm2]

In a former post I made a statistic about the most popular civs which I wanted to move in this thread. So I revisited it. For the description of the progress score check some posts ago in this thread. Quickstart replays are not in and I hope I also got all market exploits out.

These are the civ pics of the top 2000 in my replays:

These are civs with the one best progress score at minute 13 of each civ:

 

Now I did a mean value for each top 5 of that civ:

 

 

And so on:

I checked some random samples for correctness and it seems plausible to me.

Ptol is quite behind. I don't dislike them so much, they have strong slingers, camels to harass. The score focuses more on eco but still their food trickle should help a bit. 

Edited yesterday at 12:02 by ffm2

[wowgetoffyourcellphone]

14 hours ago, ffm2 said:

Ptol is quite behind

Surely the last 3 charts illustrate that it's Kush who needs a boost, nay?

[chrstgtr]

3 hours ago, wowgetoffyourcellphone said:

Surely the last 3 charts illustrate that it's Kush who needs a boost, nay?

I don't think it necessarily says anything. It's interesting, sure. But there is noise here that limits it's usefulness.

Notably, this doesn't account for differences in player skill for each civ. You would need to control for player differences to make real conclusions. It also obviously lacks any data for military strength. 

For example, I personally find ptol to be one of the stronger civs. But, in my perception, it isn't chosen by highly skilled players that often while it is frequently chosen by lesser skilled players. Many players using ptol also build settlements before min 13, which naturally slows down their eco boom, but may give them a strategic advantage that isn’t captured in the statistics  

I would also consider iber to be a decent to good civ. But whenever it is chosen it automatically boosts other civs too, especially gauls, because of it's team skirm bonus. So it's relative value is depressed. 

Edited 9 hours ago by chrstgtr

[ffm2]

8 hours ago, chrstgtr said:

8 hours ago, wowgetoffyourcellphone said:

Surely the last 3 charts illustrate that it's Kush who needs a boost, nay?

I don't think it necessarily says anything. It's interesting, sure. But there is noise here that limits it's usefulness.

It's not the full picture, but still shows something. Kush could use a buff (always careful with overcompensation) which can be also seen in the first plot of the general popularity.

 

Ptol is really puzzling. In @MarcusAureliu#s crossword he describes them as strong economic civ. One thing in team boni of gaul and ptol is the ptol support gaul (food trickle) while gaul gives nothing back (cheap skirms, not available). Also it could be that they produce fewer but also need fewer to achieve the same, altering the score.

 

Here I limited to only count the top 3 scores of a individual player to counter the bias of good players.

[MarcusAureliu#s]

Just economy wise i dont see what big advantage Persians have over Mauryans just to give an example. I imagine that players who play Persians or Gauls regularily try to boom and get to champs strategy wise, while Kush players see no advantage in lategame and go rush instead. In my opinion ptol has their power spike in early game till early p3.  @wowgetoffyourcellphone I agree with boosting Kush though, at a similiar skill level Kush player loses about 80% vs non archer civs according to my subjective perception that i am too lazy to support with evidence.