Luck in multiplayer

General feedback and discussion of the game.

Moderators: Forum Moderators, Developers

Daravel
Posts: 309
Joined: August 30th, 2008, 10:39 pm
Location: England

Luck in multiplayer

Post by Daravel » April 26th, 2014, 10:13 pm

I know what you're thinking, not *another* luck thread, but I'm going to try and do this without ranting, ask some proper questions and hopefully foster some form of discussion.

I'll start off by saying that I am not talking about singleplayer at all, this is purely about multiplayer and the experience a) it's meant to generate and b) the experience it does generate. So, me and my friend have recently begun playing some 2v2 and we've met with variable success, neither of us are the best players, but I don't think we make overly stupid moves and try to use ToD, terrain and tactics etc. However we keep experiencing some pretty crippling defeats and I am going to lay a large proportion of this down to luck (not all, but a lot).

In many of these games (unfortunately I have no screenshots) I have gone into very significant negative luck in the early turns of the game which results in the loss of 2 or even 3 units in unlucky circumstances which ultimately leads to the loss of the game. Now, I know the two common arguments against this:

1) Plan for luck - Which I always do, is often very difficult and sometimes impossible. There is a difference between thinking "my archer has 50% chance to hit, he has 4 attacks and I only need 2 hits to kill, therefore, statistically, I'll kill their unit" and "their unit is in open ground, I can get three units on him, statistically doing 35 damage" and then going on to do 10. Sometimes the luck swing is that bad, that you can, instead of kill a unit and wounding another, end up doing minimal damage to one unit. Now here's the kicker, that your enemy now has two near full health units left which means you're now vulnerable. Then their next turn happens and they go on to get very lucky (+75% or more) and instead of wounding two of your units, manages to kill 2 and ZOC a third. None of this through bad planning on your part, all of it, every single bit, on lucks part. And that turn of luck can lose a game.

2) Luck balances out - This is tied to point 1. On average, luck evens out, it's a rule of maths, regression to the mean and so forth. But this is not a valid counter-argument. Take the above example, where instead of me dealing 35 damage, I deal 10 and instead of them dealing 20 damage they deal 35 (very common), this is a 35 damage swing and possibly a 4 unit swing (2 of theirs alive, 2 of mine dead). That compounds on the next turn when I have less units to hit with and they have more, then I have to spend gold to get myself back on par - and in the retreat lose towns. The fact that, in the last few turns when you have 3 units left, that you overall luck is finally sitting around 0% doesn't make a bit of difference because your opponent has dealt 150 more damage than you in real terms.

The point is, at very early points in a multiplayer game, significant luck swings can ruin a game, and I say ruin because I enjoy Wesnoth and I play other people for the challenge of tactically out thinking them, not for rolling a die. The moment luck (or any arbitrary game mechanic) makes a sound move by a player pointless, then the game has failed (my opinion).

I promised to ask a question, and here it is

Most successful multiplayer games (League of Legends, Call of Duty, Starcraft) which have large 'professional' tournaments are not luck based (based on my knowledge), this is because luck can deny player skill. So, what exactly is the aim for multiplayer in Wesnoth? I'm aware that the intention isn't to become an international phenomenon and that the game is made by volunteers and reflects their wishes, but is it fair to say that the multiplayer was a direct port of the singleplayer with the capacity for multiple human players added in? Or was maintaining the heavy luck element a conscious choice?

With that last question, I'm not 'getting at' anyone, I am just interested whether the role of luck in a multiplayer game was ever discussed and a deliberate decision made? Because I'll be honest, I don't think I'll play many more games because I feel like I'm losing games where I'm in an excellent position and make perfectly sound moves but get overly penalised by a massive swing in luck.

User avatar
zookeeper
WML Wizard
Posts: 9740
Joined: September 11th, 2004, 10:40 pm
Location: Finland

Re: Luck in multiplayer

Post by zookeeper » April 26th, 2014, 11:12 pm

Daravel wrote:So, what exactly is the aim for multiplayer in Wesnoth? I'm aware that the intention isn't to become an international phenomenon and that the game is made by volunteers and reflects their wishes, but is it fair to say that the multiplayer was a direct port of the singleplayer with the capacity for multiple human players added in?
I think that's fair considering how MP was added later and was not in the original version(s).

User avatar
beetlenaut
Developer
Posts: 2391
Joined: December 8th, 2007, 3:21 am
Location: Washington State
Contact:

Re: Luck in multiplayer

Post by beetlenaut » April 26th, 2014, 11:18 pm

Daravel wrote:statistically doing 35 damage" and then going on to do 10. Sometimes the luck swing is that bad
Often the luck swing is that bad! "Planning for luck" means not to getting into a situation where luck like that would make you lose, because you are pretty much guaranteed to get luck like that at least once per game. So I would say that that was, indeed, poor planning. One of the hardest things about Wesnoth is knowing when it was poor planning, and when it was the RNG, but the truth is, it's almost never the RNG: I have seen posts by top ladder players acknowledging that luck decides the game about 1 out of 20 times. But that means that the other 19 out of 20 are decided by the strategy.

Here is the rationale for luck from the game's designer.
Campaigns: Dead Water,
The Founding of Borstep,
Secrets of the Ancients,
and WML Guide

User avatar
tekelili
Posts: 1038
Joined: August 19th, 2009, 9:28 pm

Re: Luck in multiplayer

Post by tekelili » April 26th, 2014, 11:23 pm

I will avoid the classic defense "luck is a good design in war games" (because it was repeated to insanity over years). But just a couple of questions about your points:

1- What outcome do you expect from a strategic game in wich players perfomance at same skill level? A tie?

2- You mention some professional games withouth luck, but there are also other professional games like lot of sports, that solves ties in a more random enveiroment like penalty series, because they consider better for the game solve issue so rather than repeat game. It is this a bad design?

EDIT: Btw, LoL has somthing called "critical chance"
Be aware English is not my first language and I could have explained bad myself using wrong or just invented words.
World Conquest II

TheCripple
Posts: 103
Joined: March 19th, 2011, 3:30 am

Re: Luck in multiplayer

Post by TheCripple » April 26th, 2014, 11:41 pm

Daravel wrote: 1) Plan for luck - Which I always do, is often very difficult and sometimes impossible. There is a difference between thinking "my archer has 50% chance to hit, he has 4 attacks and I only need 2 hits to kill, therefore, statistically, I'll kill their unit" and "their unit is in open ground, I can get three units on him, statistically doing 35 damage" and then going on to do 10. Sometimes the luck swing is that bad, that you can, instead of kill a unit and wounding another, end up doing minimal damage to one unit. Now here's the kicker, that your enemy now has two near full health units left which means you're now vulnerable. Then their next turn happens and they go on to get very lucky (+75% or more) and instead of wounding two of your units, manages to kill 2 and ZOC a third. None of this through bad planning on your part, all of it, every single bit, on lucks part. And that turn of luck can lose a game.
That doesn't seem like planning for luck at all. Take your archer example: There's a 25% chance that they won't kill the unit. That's a pretty big chance, and it's a matter of weighing the risk (which is quite possibly a dead archer and broken attack plan). The same thing applies to the 10 out of 35 damage - it's a lower chance, but that still involved a pretty big risk, and if that loses the game there's the question of why you were staking that much on it to begin with. Usually it's either a mistake, or having already essentially been outplayed so that there aren't better options.

As for luck in games and mainstream competitive games not having it, I have one word: Poker. It's way, way bigger than Starcraft, League of Legends, and the other video game contenders combined. It has professionals who can win consistently enough to make money. It has a level of randomness in it that makes Wesnoth look positively deterministic.

Daravel
Posts: 309
Joined: August 30th, 2008, 10:39 pm
Location: England

Re: Luck in multiplayer

Post by Daravel » April 26th, 2014, 11:45 pm

beetlenaut wrote:
Daravel wrote:statistically doing 35 damage" and then going on to do 10. Sometimes the luck swing is that bad
Often the luck swing is that bad! "Planning for luck" means not to getting into a situation where luck like that would make you lose, because you are pretty much guaranteed to get luck like that at least once per game. So I would say that that was, indeed, poor planning. One of the hardest things about Wesnoth is knowing when it was poor planning, and when it was the RNG, but the truth is, it's almost never the RNG: I have seen posts by top ladder players acknowledging that luck decides the game about 1 out of 20 times. But that means that the other 19 out of 20 are decided by the strategy.

Here is the rationale for luck from the game's designer.
I've bolded the main bit I'm going to refer to.

In the games I've played recently I've had several turns (6,7,8 or so) where, on looking at the statistics there has been a significant swing of luck, and one game in which the swing was so bad, it was literally 120 damage overall (that is what I lost in luck plus what they gained in luck). At what point do you look at the luck based damage change and think it's a bit ridiculous? There was no way I could win that game, 120 damage is 3-4 units worth and it's just not possible to win the game with that kind of deficit (when you factor in the loss of villages as well); which goes back to my point of, when an arbitrary game mechanic undoes a players positioning, tactics etc. then the game has failed (for it has not tested mine, or my opponents skill, it has simply tossed a coin and determined I lose). Now of course, the 120 damage deficit is going to be very rare, but there are still plenty of cases where there is significant swings.

How do you plan for it though? It's my ToD and not theirs, I have recruited a few more units and can move most of them into attack and the rest to support/block, I move up and can be pretty confident, that even with a bit of bad luck, I will kill 1 unit and wound another. Instead I get horrible luck and do minimal damage, but it's ok, because it's my ToD and they shouldn't do too much damage. Then my opponent goes and gets very lucky and takes out two of my units, this then repeats the next turn (which happens all too often). If I'm planning for that, is my best option to not attack? (In which case, I'd never attack!) By definition of the RNG, you can not plan for the luck, you have to make your best play with what you have, you have to attack at you ToD and use the terrain as best you can, you can't sit there thinking "what if I do 70% less damage and they go on to do 50% more damage?"

Also, don't forget the other type of luck. Luck when you don't need it and they do. When you get really lucky in dodging hits on a unit you expected to die, or scoring a kill when you wanted another unit to get the level. Then your opponent misses all their hits but scores 5 30% hits on your best unit. These can all make the luck look even, when in fact, their luck was 'luckier' - in that they achieved far more with their extra hits/dodges than you did. Think of those games where your damage dealt is that same as theirs, yet you've managed to kill half the units? Is that poor positioning or because the RNG placed your hits all over the place? Yes, it may be bad positioning sometimes, but it clearly isn't if you miss most of your hits with 4 units but then go on to get a bit lucky with odd units here and there on other turns.

I do appreciate that part of this is me denying my bad moves and blaming more on luck than perhaps I should. But I think we're talking about two slightly different situations. I'm referring to really edge case situations where they score 9/10 hits while you score 1/10 and the amount of damage is totally out of proportion, these situations, I don't think can be planned for (they are different from those in which you need to ensure you have 2-3 extra hits available to make sure you have the kill).

My final point (and opinion only) is, in multiplayer, should you need to plan for that? This is really my main question. When you take the point of multiplayer, is it not to test skill? Luck never, ever, has a place in skill. Grandmasters wouldn't play chess if there was a random chance one person started with two queens, footballers would play if one team sometimes got a smaller goal, etc. Please don't mistake me, I'm not demanding change or trying to insult/bring down the developers/Wesnoth, I'm just wondering if this had ever been considered/is this what everyone really wants for multiplayer?

Edit for other people's posts:
tekelili wrote:I will avoid the classic defense "luck is a good design in war games" (because it was repeated to insanity over years). But just a couple of questions about your points:

1- What outcome do you expect from a strategic game in wich players perfomance at same skill level? A tie?

2- You mention some professional games withouth luck, but there are also other professional games like lot of sports, that solves ties in a more random enveiroment like penalty series, because they consider better for the game solve issue so rather than repeat game. It is this a bad design?

EDIT: Btw, LoL has somthing called "critical chance"
1- Chess is the perfect example, a completely balanced game in every sense, not a single advantage can ever be conferred. Obviously Wesnoth is more complex - in which it's about unit choice, terrain usage, when you attack, who you use to attack - no chance of a tie in Wesnoth, ever.

2 - Penalties are not random, they are skill based entirely. The only aspect of a sport I can think that is luck based is a) weather and b) coin tosses.

It does, and Riot hate it and wish they never used it. Which is why Ashe's passive got changed. Crit chance is horrible, horrible design and why I believe a lot of pro-gaming leagues play with no crit in tournaments. Which is why TF2 is only good for fun.
TheCripple wrote:
Daravel wrote: 1) Plan for luck - Which I always do, is often very difficult and sometimes impossible. There is a difference between thinking "my archer has 50% chance to hit, he has 4 attacks and I only need 2 hits to kill, therefore, statistically, I'll kill their unit" and "their unit is in open ground, I can get three units on him, statistically doing 35 damage" and then going on to do 10. Sometimes the luck swing is that bad, that you can, instead of kill a unit and wounding another, end up doing minimal damage to one unit. Now here's the kicker, that your enemy now has two near full health units left which means you're now vulnerable. Then their next turn happens and they go on to get very lucky (+75% or more) and instead of wounding two of your units, manages to kill 2 and ZOC a third. None of this through bad planning on your part, all of it, every single bit, on lucks part. And that turn of luck can lose a game.
That doesn't seem like planning for luck at all. Take your archer example: There's a 25% chance that they won't kill the unit. That's a pretty big chance, and it's a matter of weighing the risk (which is quite possibly a dead archer and broken attack plan). The same thing applies to the 10 out of 35 damage - it's a lower chance, but that still involved a pretty big risk, and if that loses the game there's the question of why you were staking that much on it to begin with. Usually it's either a mistake, or having already essentially been outplayed so that there aren't better options.

As for luck in games and mainstream competitive games not having it, I have one word: Poker. It's way, way bigger than Starcraft, League of Legends, and the other video game contenders combined. It has professionals who can win consistently enough to make money. It has a level of randomness in it that makes Wesnoth look positively deterministic.
I think you misunderstood, it wasn't 10/35 damage, it was that statistically I would do 35 damage (the max may have been 50 for example) - example: a 5-10 attack on 70% chance to hit, you'd expect 7 hits (35), but you only score 2 (10). I'm not talking about expecting to score 35 damage and whenever that doesn't happen I feel hard done by. I'm talking about when, you do virtually no damage and then your opponent goes on to do a lot more than they should.

A lot of people say "if that move goes wrong and that would mean you lose, then you shouldn't make that move". There must be loads of times in everyone's games where if 2-3 of their units died and you failed to kill 1,2,3 of their units that it would spell doom for your game.

Poker is almost entirely luck based though, so it isn't the same.
Last edited by Daravel on April 26th, 2014, 11:56 pm, edited 2 times in total.

User avatar
iceiceice
Developer
Posts: 1056
Joined: August 23rd, 2013, 2:10 am

Re: Luck in multiplayer

Post by iceiceice » April 26th, 2014, 11:50 pm

1. Briefly, you are not properly planning your moves if you are saying to yourself "hmm 4 strikes on 40% defense, that means the most likely outcome is 2 hits." Wesnoth forces you to plan for *all* possible outcomes. In many ways the game is very defensive by design -- whenever you decide to attack, you have to deal with the possibility that all or most of your units may run up and do basically nothing and be left standing next to the enemy on bad terrain.

2. As tekelili points out, it isn't even true that most other games, even professional games, have less luck than wesnoth. Dota 2 probably has more money in its professional scene than any game right now by a significant margin. Dota has, since inception, had things like critical strike, whereby every attack randomly has a small chance to do something like 3x or 4x damage, usually vaporizing your unit. Some units have a small chance to "stun" or "entangle" a unit they attack, making it unable to move or fight back. If you get "first hit bash", that usually means you are dead. These are not obscure abilities, in some cases it will be a professional team's entire strategy to combine these abilities with others so that they have chances to permanently stun an enemy, or so that they can try to win a fight by critical striking your entire team at once.

I regard it as a very good thing that wesnoth does not have critical strike.

There are some things that real-time games may do to try to mitigate the importance of luck swings, but they aren't appropriate in a turn-based game. http://dota2.gamepedia.com/Pseudo-random_distribution
(Although I think there is an "experimental" version of wesnoth that uses something like this.)

Edit: I read your post again, I have the following thought. Luck is a significant enough factor in wesnoth that, yes, against a skilled opponent, there will be some fraction games, say 5-10%, that you just can't win because you had bad luck at key moments. It depends on the map and the matchup. It is always very demoralizing when you have a lot of misses, and many players feel a strong urge to give up when this happens, even experienced players. But it's part of the game, you learn to deal with it and to make plans less susceptible to it, and we ultimately feel that it's an essential part of the game.

Daravel
Posts: 309
Joined: August 30th, 2008, 10:39 pm
Location: England

Re: Luck in multiplayer

Post by Daravel » April 27th, 2014, 12:05 am

iceiceice wrote:1. Briefly, you are not properly planning your moves if you are saying to yourself "hmm 4 strikes on 40% defense, that means the most likely outcome is 2 hits." Wesnoth forces you to plan for *all* possible outcomes. In many ways the game is very defensive by design -- whenever you decide to attack, you have to deal with the possibility that all or most of your units may run up and do basically nothing and be left standing next to the enemy on bad terrain.

2. As tekelili points out, it isn't even true that most other games, even professional games, have less luck than wesnoth. Dota 2 probably has more money in its professional scene than any game right now by a significant margin. Dota has, since inception, had things like critical strike, whereby every attack randomly has a small chance to do something like 3x or 4x damage, usually vaporizing your unit. Some units have a small chance to "stun" or "entangle" a unit they attack, making it unable to move or fight back. If you get "first hit bash", that usually means you are dead. These are not obscure abilities, in some cases it will be a professional team's entire strategy to combine these abilities with others so that they have chances to permanently stun an enemy, or so that they can try to win a fight by critical striking your entire team at once.

I regard it as a very good thing that wesnoth does not have critical strike.

There are some things that real-time games may do to try to mitigate the importance of luck swings, but they aren't appropriate in a turn-based game. http://dota2.gamepedia.com/Pseudo-random_distribution
(Although I think there is an "experimental" version of wesnoth that uses something like this.)
I think I've referenced most of these points in my above post, but also briefly:
-If you plan for a) getting lucky and killing more, then you attack in full, b) getting normal luck, then you attack as planned, c) negative luck, then you attack more cautiously and d) horrible luck, then you don't attack at all (to prevent your guys getting slaughtered in the open). So, if you plan for all those, what action do you take? Considering this:
- You know what your units can do and what is the 'most likely'
- you know it's your ToD, you have decent terrain, good unit numbers/composition, tactics
- you do not know if you're going to be crippled by bad luck

Provided you planned your attack well, you can only ever plump for option b, anything else is foolish. As I said above, you make the best plan and you go with it. By definition you do not know when the bad luck will happen. Remember, I'm not saying, I have 4-4 attack and 50%, therefore I get 10 damage, I always try and over-hit if I can, I'm talking about when even if you have 3 extra units, you still don't kill a unit.

Remember, this also applies to defending, when your units, in good terrain, just get pummeled and they instantly wipe out two defenders in 70% terrain and manage to ZOC out a third (when this is extreme luck on their part, not, just a little bit of luck).

2. Sure, I exaggerated. Some games will have heavy luck elements. I'm not really sure if those players think they're any good, by relying on random chance to win games. I know for a fact, whenever I get luck in Wesnoth, I feel bad, I feel like I've cheated.
iceiceice wrote: Edit: I read your post again, I have the following thought. Luck is a significant enough factor in wesnoth that, yes, against a skilled opponent, there will be some fraction games, say 5-10%, that you just can't win because you had bad luck at key moments. It depends on the map and the matchup. It is always very demoralizing when you have a lot of misses, and many players feel a strong urge to give up when this happens, even experienced players. But it's part of the game, you learn to deal with it and to make plans less susceptible to it, and we ultimately feel that it's an essential part of the game.
Not to be facetious, but, have you decided it's an essential part of multiplayer? That's actually my real question (under all the semi-rant), is has this been discussed and is it worth having that discussion? I'm not even against a small bit of luck, I can cope with that, my real problem is when there is the huge swing in luck that just leaves you feeling like you've wasted X amount of time on that game. When it comes to multiplayer, do you want any % of games to be 'unwinnable' due to a random mechanic?

User avatar
tekelili
Posts: 1038
Joined: August 19th, 2009, 9:28 pm

Re: Luck in multiplayer

Post by tekelili » April 27th, 2014, 12:16 am

2 - Penalties are not random, they are skill based entirely.
I are you really telling me that (basically having 3 options for penaty be throwed: right, left and middle), there is no randomess in wich side each player (penalty thrower and goal keeper) will decide to perform? That means that if the same player throws 100 penalties to the same goal keeper each outcome will be decided only for skill without any randomess involved? Sorry, but I rather see this issue like bassed on players skills: they have some %chance to score.
Be aware English is not my first language and I could have explained bad myself using wrong or just invented words.
World Conquest II

User avatar
iceiceice
Developer
Posts: 1056
Joined: August 23rd, 2013, 2:10 am

Re: Luck in multiplayer

Post by iceiceice » April 27th, 2014, 12:16 am

The fact is, we like the game, the game is very successful, and with time and experience we have all generally gotten used to the amount of luck involved. It was pretty much conceived this way, so that it would always be very difficult to properly plan things out in wesnoth. See WesnothPhilosophy. Edit: Actually, what I meant to link to was this: http://forums.wesnoth.org/viewtopic.php?t=21317 Read it carefully! This is straight from the horse's mouth. (Dave is the creator.)

Do you have a concrete proposal to reduce the amount of luck? There are certainly people who have been interested enough to make experimental, alternate versions where the role of luck is somehow different. But I don't see that there is any simple way that the game could retain it's character while reducing the impact of luck swings.

chris123
Posts: 4
Joined: April 14th, 2014, 12:09 am

Re: Luck in multiplayer

Post by chris123 » April 27th, 2014, 12:45 am

I believe that the problem with the RNG is that the chances of hitting doesn't build upon it self. What I mean is, if I had 50% chance to hit that is going to be 50% to hit weather I hit or miss my other shots. You brought up league of legends as an example of a game that doesn't depend on luck, but I seem to remember they have critical strikes ;) , though their chance of hitting a critical strike builds upon itself.

User avatar
iceiceice
Developer
Posts: 1056
Joined: August 23rd, 2013, 2:10 am

Re: Luck in multiplayer

Post by iceiceice » April 27th, 2014, 1:02 am

Yeah, that's what I meant when I referred to this page. http://dota2.gamepedia.com/Pseudo-random_distribution
That kind of thing is probably okay in a real time game, because people don't have time to react to these kind of chance-to-hit changes. But in a turn-based game it's not a good idea. IMO.

abhijit
Posts: 96
Joined: October 4th, 2011, 1:27 am
Contact:

Re: Luck in multiplayer

Post by abhijit » April 27th, 2014, 3:12 am

Daravel wrote: example: a 5-10 attack on 70% chance to hit, you'd expect 7 hits (35)...
No you don't :) You have 10 hits, each of which can fail 30 times out of 100.
Winner of The Alternate Frontier

TheCripple
Posts: 103
Joined: March 19th, 2011, 3:30 am

Re: Luck in multiplayer

Post by TheCripple » April 27th, 2014, 4:18 am

Daravel wrote:Poker is almost entirely luck based though, so it isn't the same.
A handful of professionals are able to consistently win major tournaments. Below that level, a fair number of people are consistently able to make some money on the side of whatever else they do competing in the sort of events that the first group doesn't bother showing up for (casino dealers are a similar case). Below that level, a number of people manage to lose money consistently. It's clearly enough of a skill based game that skill frequently determines who wins, despite the heavy randomness. Somehow, I can't help but think Wesnoth is the same way, as it is significantly less random and also has a great many random instances in the typical game.

User avatar
8680
Moderator Emeritus
Posts: 742
Joined: March 20th, 2011, 11:45 pm
Location: The past

Re: Luck in multiplayer

Post by 8680 » April 27th, 2014, 5:21 am

Daravel wrote:Because I'll be honest, I don't think I'll play many more games because I feel like I'm losing games where I'm in an excellent position and make perfectly sound moves but get overly penalised by a massive swing in luck.
Have you tried any of the randomness-reducing add-ons that are floating around? Perhaps you would prefer playing with one of those.

Post Reply