PDA

View Full Version : What Duel statistic using doesn't show



Capoupacap
04-11-2017, 12:09 PM
Hi there,


During last dev stream, you showed us the duel statistic for each character. You then argue that because of these stats, conqueror for example need a nerf to balance win ratio in duel.

Mostly your balancing, as Romano said, mean to just play on some numbers and see how it impact duel statistic. It not a bad way to go, it works for some character but in the case of conqueror it seems a real rework is needed.

Can we wait for such a rework for the conqueror ???

cragar212
04-11-2017, 12:27 PM
Although caring about the game they are an extremely inexperienced team.

Meta data is a big part of balance but they are clearly relying far to heavily upon it. A good example was when Roman aka the ****ing dude in charge of balance said raider is pk killer. He said it was a mistake because they didn't have enough data at the time... Wtf anyone who had a general knowledge of the game would of thought hmmmm that doesn't look right.

They are causing serious damage to the game with their balance practices. Balance should be based off data + community feedback + ACTUALLY TESTING STUFF. Balance should also be small tweaks done monthly not these big nerf and buff patchs every 2 - 3 months.

Draghmar
04-11-2017, 12:49 PM
I'd remove community from affecting too much updates and balance because it's really hard to get proper conclusion when people are mostly come here to vent them self or whine on something - this *is* reality in most games I'm afraid.

Charmzzz
04-11-2017, 12:53 PM
They are causing serious damage to the game with their balance practices. Balance should be based off data + community feedback + ACTUALLY TESTING STUFF. Balance should also be small tweaks done monthly not these big nerf and buff patchs every 2 - 3 months.

You forget that "ACTUALLY TESTING STUFF" does take time. Example: you release your game. After 1-2 Weeks you make some charts with the data you got. You read in forums that "XY" is OP, but from internal testing you KNOW there is a counterplay for it. Then you see in your charts that one class has an absurd high Winratio, AND people complain about it on the forum. So you take a look at it, test it, which takes at least 2-3 days. Then you think about possible fixes and test them, which takes at least 4-5 days (don't forget you have to code it!).

How on earth should a "fast tweak" be tested? Or a tested fix inserted into the code really fast? It is not possible, only people who have no idea of programming would call for fixes every 2-3 weeks... If there is a gamebreaking glitch or bug, sure, it must be hotfixed asap. But I did not experience nor hear from this level of bug in For Honor until now.

Concerning statistics: they are right. There is no argument against it, even if people keep crying on the forum that PK is too fast or how annoying a vortex spamming warden/valk is. If they only win 50% of their duels they seem to be in line. If there is a hero with 70%+ winrate, something is wrong on him. Even in unfavourable matchups he has the upper hand to get these winrates...

Manlorey
04-11-2017, 02:10 PM
Concerning statistics: they are right. There is no argument against it, even if people keep crying on the forum that PK is too fast or how annoying a vortex spamming warden/valk is. If they only win 50% of their duels they seem to be in line. If there is a hero with 70%+ winrate, something is wrong on him. Even in unfavourable matchups he has the upper hand to get these winrates...
Agree with you on that, problem is, that there may be a chance their statistics are unreliable because of quitting frequency. People quitting their matches (http://forums.ubi.com/showthread.php/1628051-People-like-this-are-an-example-why-quitting-needs-to-count-as-a-loss) are a problem, which can easily influence the statistics for a hero by some degree (though, how much we do not know).

But why do people quit? Some quit to save-cheat their statistics, but some also quit because of the matchmaking system which has the capability of putting lvl 10 chars together with fully equipped 108, and be sure that the whole team consists either of low level chars or whole of the team is on 108 gear. That, and also the pleasure of joining dominion games with 700 points or more of the enemy team which is quite questionable, make people quit too and I do not think one is too blame for that. So this leads to another problem, possibly directly affecting the statistic.

Charmzzz
04-11-2017, 02:39 PM
Agree with you on that, problem is, that there may be a chance their statistics are unreliable because of quitting frequency. People quitting their matches (http://forums.ubi.com/showthread.php/1628051-People-like-this-are-an-example-why-quitting-needs-to-count-as-a-loss) are a problem, which can easily influence the statistics for a hero by some degree (though, how much we do not know).

But why do people quit? Some quit to save-cheat their statistics, but some also quit because of the matchmaking system which has the capability of putting lvl 10 chars together with fully equipped 108, and be sure that the whole team consists either of low level chars or whole of the team is on 108 gear. That, and also the pleasure of joining dominion games with 700 points or more of the enemy team which is quite questionable, make people quit too and I do not think one is too blame for that. So this leads to another problem, possibly directly affecting the statistic.

I did not see the stream, which statistics they take? Hopefully they take Duel because it is a mode not influenced by gear and because it is 1on1 it gives the best comparison between the possible matchups. If people quit in Duel it is very likely because they would lose anyway, so these Matches should be counted as a Win for the remaining person.

I don't think they take statistics from Dominion / Elimination and compare a Prestige 1 Gear 24 LB with a Prestige 5 Gear 108 PK. I know people here on the forum like to call the people of Ubi dumb and stuff, but that is most likely not true. Game Design and Balancing is pretty tough, and to get a job in one of the leading Game Software Studios is not easy. If it would be easy, why did those complainers not assign for it? ;)

Capoupacap
04-11-2017, 04:04 PM
The big issue I see for the dev team is between their statistic on duel (for average and pro "what a joke" player) and how these heroes are used in dominion.

Let's take three example:

Shugoki: quite poor in duel but in 4vs4 it can really ruin the game....New thing is now to have 2 shugoki in the same team.....You can balance all the way you want, it won't change anything

Lawbringer: It's been a month this "I stay still maybe he won't see me then i rush and push him downhill or maybe on these spikes" hero is pissing everybody in 4vs4. Now they OP this hero like hell

Conqueror: 4vs4 or Duel, people complain about a hero that spam when their design didn't give any other capacity....And they make it so strong with revenge. Then the nerf come and a very strong roomer in dominion is ruined. Now it's a "stay there, block (try to), and pray we arrive on time to res you".


I hope that when the dev team say they will show us how they work on balancing is true. Using statistic is interesting but you don't need them when you see how Heroes is abusively used in the game after a month.

CandleInTheDark
04-11-2017, 05:24 PM
I know people here on the forum like to call the people of Ubi dumb and stuff, but that is most likely not true. Game Design and Balancing is pretty tough, and to get a job in one of the leading Game Software Studios is not easy. If it would be easy, why did those complainers not assign for it? ;)

I have a friend who took a game design course, other than coding it requires a lot of high level math. So pretty certain it isn't as easy as people believe.