Red alert. Soviets vs Allies balance (or lack thereof)

[Murto the Ray]

What name do you usually play under?

[Doomsday]

I think the biggest difference between allies and soviets is defensive structures.

Having played multiple RTS games in the past, it feels like in OpenRA economy is less important and build capacity / production is the most important thing. What I mean it is harder to turn massive economical advantage into production. It's really common to see players floating even in masters division. In Age of Empires 2 player can build 10 barracks and have 10x production compared to player with one barrack. In OpenRA production bonus for multiple barracks caps at 1,5x. For this reason easiest way to boost production is using each different production tabs at once. Allied players can utilize their defensive structure tab much easier than soviets. Allies can keep spamming pillboxes in one flank and be more or less safe there. Soviet factions can't really do the same because tesla coils are harder to spam because of their insane power cost. If they want to do the same, they have to invest more into power plants. Cheap and easier allied defensive structures allow allies to focus all military power in one area of the map without getting punished as hard as soviets.

tl;dr: Soviet defenses are harder to spam for gaining more assets on the field because of high power cost of Tesla Coils.

[klaas]

I'd love to see such statistics !

I've attached a small python script that can extract faction, name and win/lost from a replay file. All you need to do use the extracted info somehow, dont have time for this now.

To use it, you ust put all the replays in a folder, put this script in a folder above it, and it will generate an output file. Note: you will need python to run this.

[FRenzy]

Thanks Klaas ! That's really great

With my very poor programming skills, I managed to take on, and get out some results from it.

Tried out with the Masters Division of Season 1, here's what it gives :

COUNTRIES : ('germany', 'england', 'france', 'russia', 'ukraine')
Wins : 4, 16, 5, 17, 21
Losses : 3, 9, 5, 25, 23
W/L Ratio : 1.33, 1.78, 1.0, 0.68, 0.91
CAMPS : ('allies', 'soviets')
Wins : 25, 38
Losses : 17, 48
W/L Ratio : 1.47, 0.79

Although this is not a scientifically well-conducted study , this already shows some interesting tendencies : Allies win

[klaas]

Awesome, seems like your scripting skills are at least as good as mine. As for the result, I'm not surprised but it's nice to have some real data.

I'm a bit puzzled about the large number of Soviet players however. This is ulikely, since there are 3 players that always play Allies and only 1 that picked soviets.

Edit: this was for season 1, I'm stupid...

I just ran your script for all my replays (I almost always play random):

COUNTRIES : 'germany', 'england', 'france', 'russia', 'ukraine'
Wins : 24, 25, 16, 22, 17
Losses : 33, 44, 21, 33, 49
W/L Ratio 0.73, 0.57, 0.76, 0.67, 0.35
CAMPS : ('allies', 'soviets')
Wins : 65, 39
Losses : 98, 82
W/L Ratio 0.66, 0.48

Obviously, your script counts games that do not officially end as loss, so the number of wins is lower then it should be. Still, it gives the same general trend. I've made some small changes to your script to fix this and make it a bit more efficient, and only count games when the matchup is actually allies vs. soviets.

[Chimpo]

What name do you usually play under?

Me? I've had loads & rarely stick with any for very long, perhaps I should choose one. Recently its been Tits McGee, John El Chimpo and some others. My rl initials spell APE so I often choose monkey themed names

UK IP ending 121

[FRenzy]

I've made some small changes to your script to fix this and make it a bit more efficient, and only count games when the matchup is actually allies vs. soviets.

Yes awesome, that's what we need, thanks !

Still, it gives the same general trend.

Yes, here's what it gave me for RAGL S1 Masters, Allies vs Soviets matches, if I'm not mistaken :

COUNTRIES : ('germany', 'england', 'france', 'russia', 'ukraine')
Wins : 3, 14, 3, 5, 7
Losses : 3, 5, 4, 9, 11
W/L Ratio : 1.0, 2.8, 0.75, 0.56, 0.64
CAMPS : ('allies', 'soviets')
Wins : 20, 12
Losses : 12, 20
W/L Ratio : 1.67, 0.6

[zinc]

That could just be in part detecting certain better players having selected to play allies, which isn't the same as allies being better.

Also, I'm thinking the artillery has already been changed since those games took place?

[scorp]

That could just be in part detecting certain better players having selected to play allies, which isn't the same as allies being better.

in that case, it may be useful to have a look at some AI games (bots)

My Impression is that allied bots are generally more effective against other bots than soviet. Quite simply put, they throw out the same cheap infantry/arty trashmobs against their opponents as many human Players seem to do because it's cheap, effective, and little micromanagement is needed. And that's bots we're talking about here, the entire basewalking discussion isn't even touched by this.

However, i think ever since there are 5 factions instead of 2, balancing them fairly has become extremely hard. Whenerver a powergamer finds a new exploit tactic to use effectively in competetive gaming, a new discussion will arise.

[klaas]

Nice FRenzy, these stats are even worse then what I thought they would be. I've updated the script a little bit more, now it will automatically ignore multiplayer games , so you can easily run it on your own replay folder without multiplayer games skewing the results.

I'll try this tonight for the previous releases too.

If anyone wants his own replays tested but has no idea how to run Python scripts, you can 7z them and I can do it for you. Or you can install python and learn how to script, its not so hard, and can help with picking up girls in the bar (I wish).

My stats for 1vs1 games, this release:

COUNTRIES ('germany' 'england' 'france' 'russia' 'ukraine')
Wins ............20........14........12........12........18
Losses ........12........17..........3.........22........22
W/L Ratio ....1.67....0.82.......4........0.55.......0.82
CAMPS ('allies' 'soviets')
Wins ............46........30
Losses ........32........44
W/L Ratio ....1.44.....0.68

Interestingly, for the 0508 release the ratio is not as dramatic:

CAMPS ('allies' 'soviets')
Wins : 111, 93
Losses : 96, 108
W/L Ratio : 1.16, 0.86

note not all replays are processed correctly, not sure if there is still a bug in the code, or some replay files are somehow truncated at the wrong spot.

[Chimpo]

Great work Klass, knew it wasn't my imagination (or lack of skill hehe)

[anjew]

CAMPS ('allies' 'soviets')
Wins ............46........30
Losses ........32........44
W/L Ratio ....1.44.....0.68

CAMPS ('allies' 'soviets')
Wins : 111, 93
Losses : 96, 108
W/L Ratio : 1.16, 0.86

Shouldn't the Allies wins equal the Soviet loses and vice versa or am I a bit slow?

[OMnom]

CAMPS ('allies' 'soviets')
Wins ............46........30
Losses ........32........44
W/L Ratio ....1.44.....0.68

CAMPS ('allies' 'soviets')
Wins : 111, 93
Losses : 96, 108
W/L Ratio : 1.16, 0.86

Shouldn't the Allies wins equal the Soviet loses and vice versa or am I a bit slow?

Maybe they're including wins/losses from mirror matchups? In which case, trying to find quantitative proof is going to be really hard, since most of the top tier only use Allies in serious matches.

[klaas]

CAMPS ('allies' 'soviets')
Wins ............46........30
Losses ........32........44
W/L Ratio ....1.44.....0.68

CAMPS ('allies' 'soviets')
Wins : 111, 93
Losses : 96, 108
W/L Ratio : 1.16, 0.86

Shouldn't the Allies wins equal the Soviet loses and vice versa or am I a bit slow?

Maybe they're including wins/losses from mirror matchups? In which case, trying to find quantitative proof is going to be really hard, since most of the top tier only use Allies in serious matches.

There is still a problem somewhere, it could be caused by custom maps or corrupt replays. It has a pretty minor effect though.

The script ignores mirror matchups and team games.

You could adapt the script to work for TD too btw.

[FiveAces]

Chimpo, it seems to me that most of your matches tend to happen on maps with many chokepoints, which is where the Allies' strengths shine the most.
Try playing more open maps and have a mobile army, find flanks on your opponents instead of attacking into their line of base defenses
and make good use of the amazing utility tools the Soviet factions have (spy plane, paradrop, Iron Curtain).

On the topic of balance, I'd argue that Soviets are simply harder to play
(base defences notwithstanding, as there is definitely something awry with pillbox/turret spam).
Really good job with those scripts Klaas, but I think they might carry an inherent bias,
as most good players tend to gravitate towards Allies nowadays _because_ of the strength of pillboxes.

On the offensive, all factions are on even ground and I'd go as far as to argue that V2s are way more useful that artilleries in the open field.
Granted, they take a lot more micro (as everything Soviet does, see also Hinds vs Yaks),
but they just have more utility with their 100% accuracy and burst damage on demand compared to the immobile, frail and RNG-dependent artillery.

Additionally, I briefly skimmed my RAGL matches across season 1 and 2 and came to the following stats
(I play random factions; some matches did't happen due to no-shows or forfeits):
Played as Allies:16
Won as Allies: 13
Allied winrate: 81,25%
Played as Soviets: 18
Won as Soviets: 15
Soviet winrate: 83,33%