Login

russian armor

Factions and tournament

10 Dec 2019, 09:11 AM
#21
avatar of Maret

Posts: 610


why not, t34/85 are gonna act like panthers. don't really see a problem with that, maybe the t3 t34/76 could be locked behind an upgrade or tier 4 like mobilized reserves. And on the docs that got t34/85 we could change that with IS2 for gmca/aw and kv1 on armored assault.
The better option would be nondoc kv1 at t4 and t34/76 at t3 but who knows.

No need broke non-broken things.
10 Dec 2019, 09:15 AM
#22
avatar of Sander93

Posts: 2066 | Subs: 3

I don't think it's a good idea to use tournaments for allies vs axis data. Unlike automatch, allies always start on spawnpoint 1, and axis start on spawnpoint 2. Any stats that point to allies being stronger than axis can just as easily be attributed to the tourney map pool overall favoring spawn point 1.


There are many more issues with tournament "stats". Low sample size (causing unreliability because RNG/mistakes/spawns/etc rather than faction balance can heavily influence the results of the low amount of games), potentially corrupt data (if a good player wins versus a worse player 3-0, and they happened to pick Allies first, Allies get a 66% win ratio while they would've had 50% with 4 games played, or Axis would've had 66% if they happened to pick Axis first, and with the low sample size this does not get cancelled out on the average), or the fact that there's a huge price pool and players will generally only play safe with tried and tested strategies (OKW Grand Offensive) instead of taking risks with off-meta strats that could potentially work better (Noggano snatching G2 of the finals with Defensive Tactics, and almost getting G4 with OST Infantry Doctrine). Or the fact that most 1v1 players simply dislike playing UKF (as proven by low pick rates even at a time they were very competitive) regardless of how the faction is.

In conclusion, tournament data is indicative at best, but never conclusive.
10 Dec 2019, 10:21 AM
#23
avatar of Vipper

Posts: 6661 | Subs: 1

..
I don't find extended range permanently through vet to be a good way to balance power level. If you make the extended range to have some sort of limitation as reducing fire angle akin spearhear or for limited time duration as a munition exchange, i would find it less afoul.
You can swap around the power level of the unit between vet as you wish, that doesn't make extended range a good option IMO.
...


The problem with that is that range advantage is not the same.

Tiger starts with 45 range and it is out-ranged by 15 going down to 10 with veterancy.

IS-2 starts with 40 and it is out-ranged by 10 going down to 0 with veterancy.

By removing the range bonuses Tiger will always be out-ranged by 15 and IS-2 by 10, while the allied TDs will be cheaper and with better chance to penetrate at max range.
10 Dec 2019, 11:37 AM
#24
avatar of SunTzu

Posts: 26

For OST, if the problem lies in T3 and T4, is it enough that the prices of the tanks are reduced and is the more "scarce" ones?
10 Dec 2019, 11:48 AM
#25
avatar of elchino7
Senior Moderator Badge

Posts: 6469 | Subs: 1

jump backJump back to quoted post10 Dec 2019, 11:37 AMSunTzu
For OST, if the problem lies in T3 and T4, is it enough that the prices of the tanks are reduced and is the more "scarce" ones?


In the same way UKF is not as useless in teamgames, that kind of change has bigger repercussions in other modes.
10 Dec 2019, 11:50 AM
#26
avatar of SunTzu

Posts: 26



In the same way UKF is not as useless in teamgames, that kind of change has bigger repercussions in other modes.


so how can you enhance OST without making it OP in the other modes? Adding the puma? Giving a type of light tank?
Decreasing the costs of some things?
10 Dec 2019, 11:57 AM
#27
avatar of Katitof

Posts: 14015 | Subs: 7

jump backJump back to quoted post10 Dec 2019, 11:50 AMSunTzu


so how can you enhance OST without making it OP in the other modes? Adding the puma? Giving a type of light tank?
Decreasing the costs of some things?

You don't.

Ost and UKF will never be tournament friendly, because they are designed and balanced with specific playstyle in mind and that playstyle isn't going to win you anything at tournament level, because its not aggressive enough, ost and UKF are methodical snowballing factions that respond to threats, let the game run long enough and you'll be overwhelmed, players know that and pick factions that peak earlier to end the game faster.

That doesn't mean the factions themselves aren't competitive or good.

Ost is probably THE strongest faction in team games and it always held that position, it only briefly went to 2nd place when UKF was released and still imbalanced.

You can't really balance 5 different factions with completely different playstyles for the level of competitiveness, where losing a single squad can snowball into losing the game.
10 Dec 2019, 12:03 PM
#28
avatar of Stug life

Posts: 4113



There are many more issues with tournament "stats". Low sample size (causing unreliability because RNG/mistakes/spawns/etc rather than faction balance can heavily influence the results of the low amount of games), potentially corrupt data (if a good player wins versus a worse player 3-0, and they happened to pick Allies first, Allies get a 66% win ratio while they would've had 50% with 4 games played, or Axis would've had 66% if they happened to pick Axis first, and with the low sample size this does not get cancelled out on the average), or the fact that there's a huge price pool and players will generally only play safe with tried and tested strategies (OKW Grand Offensive) instead of taking risks with off-meta strats that could potentially work better (Noggano snatching G2 of the finals with Defensive Tactics, and almost getting G4 with OST Infantry Doctrine). Or the fact that most 1v1 players simply dislike playing UKF (as proven by low pick rates even at a time they were very competitive) regardless of how the faction is.

In conclusion, tournament data is indicative at best, but never conclusive.
then what's indicative ?

auto macth stats ? no cause they are random and u get put with lower skill lvl

stats? nope cause they are tailored to the factions

roster ? nope cause each faction is unique

how do we do it ?
10 Dec 2019, 12:11 PM
#29
avatar of Katitof

Posts: 14015 | Subs: 7


auto macth stats ? no cause they are random and u get put with lower skill lvl

Actually, that's very indicative due to sample size alone.
Not all games are imbalanced due to skill difference and you can sort them out anyway(if relic got basic API tracking for it).

Add a sprinkle of feedback from top level players and you have a bigger picture.
10 Dec 2019, 12:32 PM
#30
avatar of SunTzu

Posts: 26

jump backJump back to quoted post10 Dec 2019, 11:57 AMKatitof

You don't.

Ost and UKF will never be tournament friendly, because they are designed and balanced with specific playstyle in mind and that playstyle isn't going to win you anything at tournament level, because its not aggressive enough, ost and UKF are methodical snowballing factions that respond to threats, let the game run long enough and you'll be overwhelmed, players know that and pick factions that peak earlier to end the game faster.

That doesn't mean the factions themselves aren't competitive or good.

Ost is probably THE strongest faction in team games and it always held that position, it only briefly went to 2nd place when UKF was released and still imbalanced.

You can't really balance 5 different factions with completely different playstyles for the level of competitiveness, where losing a single squad can snowball into losing the game.

I don't find that OST is balanced, an overly defensive faction is not good.
10 Dec 2019, 12:34 PM
#31
avatar of Bananenheld

Posts: 1593 | Subs: 2


why not, t34/85 are gonna act like panthers. don't really see a problem with that, maybe the t3 t34/76 could be locked behind an upgrade or tier 4 like mobilized reserves. And on the docs that got t34/85 we could change that with IS2 for gmca/aw and kv1 on armored assault.
The better option would be nondoc kv1 at t4 and t34/76 at t3 but who knows.

That sounds great. Would give soviets the extra punch which they really need right now
10 Dec 2019, 12:36 PM
#32
avatar of Sander93

Posts: 2066 | Subs: 3

auto macth stats ? no cause they are random and u get put with lower skill lvl

That wouldn't matter too much as long as the sample size is large enough, because then on average every faction would have a roughly equal amount of games won due to skill gap. Sadly though we do not have access to up to date and complete automatch stats either.

Ideally we'd have access to big data (the game collecting information on literally everything) like most modern games, but CoH2 does not have that.

In the end, the best we can do is use (somewhat outdated) automatch and tournament stats as indicators (but as an indication only), while using our gut feeling along with talking to top players (because ultimately only at top level can we eliminate as much external influence, i.e. l2p issues, as possible) and keeping an eye on community sentiment to get a rough estimate of what the faction balance is currently like. Obviously not much of a scientific method, but I'd say it's been working well enough for the last 1.5 years.
10 Dec 2019, 13:11 PM
#33
avatar of blvckdream

Posts: 1682


That wouldn't matter too much as long as the sample size is large enough, because then on average every faction would have a roughly equal amount of games won due to skill gap. Sadly though we do not have access to up to date and complete automatch stats either.

Ideally we'd have access to big data (the game collecting information on literally everything) like most modern games, but CoH2 does not have that.

In the end, the best we can do is use (somewhat outdated) automatch and tournament stats as indicators (but as an indication only), while using our gut feeling along with talking to top players (because ultimately only at top level can we eliminate as much external influence, i.e. l2p issues, as possible) and keeping an eye on community sentiment to get a rough estimate of what the faction balance is currently like. Obviously not much of a scientific method, but I'd say it's been working well enough for the last 1.5 years.


Automatch always tries to match players with equal elo. That can mean rank 100 faction X vs rank 500 faction Y because they have the same elo rating due to either faction imbalance or lower player numbers in a ladder. That's why automatch stats are pretty much meaningless because they will always tend to be very close to 50% apart from maybe the top 1% of each ladder.
10 Dec 2019, 14:07 PM
#34
avatar of Sander93

Posts: 2066 | Subs: 3

Automatch always tries to match players with equal elo
That's why automatch stats are pretty much meaningless because they will always tend to be very close to 50% apart from maybe the top 1% of each ladder

I'd say that only really applies to the middle bracket, while good players will win the majority of their games regardless of elo matchups, so the top ~10-25% will still tell us a decent enough story. This also happens to be the part where player skill is consistent, factions are used to their full potential, and learn to play issues are mostly absent. Again nothing of that is even close to anything conclusive, but it does helps forming the bigger picture. I wouldn't call them flat out meaningless.

I'd say the top 25% wr stats of Siphon's automatch stats collection were pretty similar to what I thought faction balance looked like across multiple gamemodes (1v1 in the case of the linked example) at that time. Again though, it only somewhat reinforced what I/we already thought, we never used any of those stats as a baseline for adjustments.
10 Dec 2019, 16:47 PM
#35
avatar of EffenNewbie

Posts: 1325

then what's indicative ?

auto macth stats ? no cause they are random and u get put with lower skill lvl

stats? nope cause they are tailored to the factions

roster ? nope cause each faction is unique

how do we do it ?


Automatch stats are probably better because automatch tries to get equally ranked opponents. In tournaments, the highest seeds play the lowest so the skill level difference is as large as it can be within the pool of players. Most of the matches were 3-0 or 3-1. You can't combine a bunch of mismatches and then make conclusions on balance.

Also, I don't think any amount of tuning would greatly increase people playing UKF in these tournaments. It seems almost fashionable for players to dislike playing UKF.
10 Dec 2019, 19:42 PM
#36
avatar of Siphon X.
Senior Editor Badge

Posts: 1128 | Subs: 2


Automatch stats are probably better because automatch tries to get equally ranked opponents. In tournaments, the highest seeds play the lowest so the skill level difference is as large as it can be within the pool of players. Most of the matches were 3-0 or 3-1. You can't combine a bunch of mismatches and then make conclusions on balance.


Well, as blvckdream correctly points out, automatch stats don't really work that way. It will not try match players of equal skill, but instead players of similar ELO, which in turn is a function of player skill AND faction performance. Basically, better players playing weaker factions will ideally be matched with less skilled players playing a stronger faction.

As Sander correctly mentions this breaks down at both ends of the ladder, because eventually there will be nobody strong enough available to perfectly match a player with a strong faction. However, as I also point out in the referenced article is that win rates on the ends are also strongly influenced by how many players in a certain skill range play these factions. E.g. if more games are played on axis side by high level players, they inevitably be matched with worse player, inflating their win percentage and so on.

Edit: Moved the other stuff to the appropriate thread...
10 Dec 2019, 20:32 PM
#37
avatar of Doomlord52

Posts: 636

So, who wants to setup a "balance" tournament: an obnoxiously long series where top-ranking players need to play a TON of games against each-other as every faction (and combination), solely for the purpose of getting win:loss statistics? I'd suggest a similar format WCS, except each player matchup needs to consist of a "best of 5" of every faction combination (5x USF vs. Ost, 5x USF vs OKW, 5x etc....).
:clap:

Really though, considering the player base of CoH2, getting a complete statistical breakdown of balance that's as accurate as it needs to be is very unlikely. If the game had daily concurrent player peaks over 100k it would be possible, since top-level play would actually consist of only top players. Unfortunately, even with the free give-away a little while ago, its only reached 22k peak concurrent players, and that average is now down to around 7k (based on steamcharts).

Even so, I still think the WCS is a better source of balance data than Automatch. Players in the WCS were all in the top 150 for 1v1, whereas in automatch it's not uncommon for top 100 players to be matched against top 300-500 players, since that's the closest available.



11 Dec 2019, 00:37 AM
#38
avatar of EffenNewbie

Posts: 1325



Well, as blvckdream correctly points out, automatch stats don't really work that way. It will not try match players of equal skill, but instead players of similar ELO, which in turn is a function of player skill AND faction performance. Basically, better players playing weaker factions will ideally be matched with less skilled players playing a stronger faction.

As Sander correctly mentions this breaks down at both ends of the ladder, because eventually there will be nobody strong enough available to perfectly match a player with a strong faction. However, as I also point out in the referenced article is that win rates on the ends are also strongly influenced by how many players in a certain skill range play these factions. E.g. if more games are played on axis side by high level players, they inevitably be matched with worse player, inflating their win percentage and so on.

Edit: Moved the other stuff to the appropriate thread...


I said *probably* because there are a lot of pitfalls, like you've pointed out. You're still more likely to get something usable out of the automatch results. You would have to throw out all of the games between mismatched rankings, take out matches where it took place on a map that produces biased results, etc, etc. You'd also have to put a limit on how high of a ranking you would want to use. Some things like T70 abuse don't really happen at 4 digit rankings.

Even then, one of the problems with ELO is that you can get to a ranking by being good or by being mediocre a lot.
11 Dec 2019, 00:45 AM
#39
avatar of EffenNewbie

Posts: 1325

So, who wants to setup a "balance" tournament: an obnoxiously long series where top-ranking players need to play a TON of games against each-other as every faction (and combination), solely for the purpose of getting win:loss statistics? I'd suggest a similar format WCS, except each player matchup needs to consist of a "best of 5" of every faction combination (5x USF vs. Ost, 5x USF vs OKW, 5x etc....).
:clap:

Really though, considering the player base of CoH2, getting a complete statistical breakdown of balance that's as accurate as it needs to be is very unlikely. If the game had daily concurrent player peaks over 100k it would be possible, since top-level play would actually consist of only top players. Unfortunately, even with the free give-away a little while ago, its only reached 22k peak concurrent players, and that average is now down to around 7k (based on steamcharts).

Even so, I still think the WCS is a better source of balance data than Automatch. Players in the WCS were all in the top 150 for 1v1, whereas in automatch it's not uncommon for top 100 players to be matched against top 300-500 players, since that's the closest available.





Sure, but thow out UKF because nobody is going to play it anyway. Then instead of having the #1 seed play the #32 seed in the first round, have the #1 seed play #2, #3 play #4, etc.

The early rounds in the WCS were usually mismatches because they were supposed to be mismatches (and the selection team did a good job). The finals were 3/2 splits because they really were the few elite players. They were the #1, #2, #3, and #5 seeds.
1 user is browsing this thread: 1 guest

Livestreams

Korea, Republic Of 39
unknown 31
Georgia 3

Ladders Top 10

  • #
    Steam Alias
    W
    L
    %
    Streak
Data provided by Relic Relic Entertainment

Replay highlight

VS
  • Ostheer flag tightrope
  • The British Forces flag Kobal
Faymonville Approach
Faymonville Approach Honor
Honor it
3
Download
Download
19
uploaded by Kobal

Board Info

235 users are online: 12 members and 223 guests
Katitof, MMX, Serrith, Aerohank, Sander93, Stormjäger, Vipper, |GB| The Lnt.599, Tomy0, Musti, SgtJonson, Crecer13
144 posts in the last 24h
1400 posts in the last week
4159 posts in the last month
Registered members: 74872
Welcome our newest member, WedeQQ
Most online: 918 users on 27 Oct 2019, 01:03 AM