Reducing the luck in multiplayer.
Moderator: Forum Moderators
But the game isn't being developed for the masses; it isNiels wrote:
Cause of the bunch of stupid idead in this thread (and others) obout that thematic, i understand that none of the more involved people read this.
But im sure the majority never realy considered "Randomness without Replacement"
being developed for the developers and based on what
I have seen, they are smart enough to know of several ways
to reduce variance if they wanted to do that.
Have you thought through the implications of using a
system where you track how a person's luck is has
been in order to modify the chances to hit? This is
going to make playing the game well harder as you
will need to know what you current luck modifier
is and order your attacks to use up you good and bad
modifiers in the right attacks. This is not in keeping
with the KISS design principal espoused by the
developers.
If you care enough and have the time, you can maintain
your own fork with your change and try to convince
enough developers (not necessarily the current ones)
that your version is worth maintaining as a separate
fork to keep the project going. At least initially you
should be able to track updates to Dave's version
to get bug fixes. Since Wesnoth is already in the
leadup to 1.0, it is already the case that most of the
changes should be bug fixes which will make your
job easier.
Saying that you lost ONLY due to bad luck is a bit ofEllestar wrote:
I read that article before, but i'm more worried about a huge unluck that happens sometimes. For example, i just surrendered in a game where i dealt 70 damage (120 evaluated) and recieved 165 damage (113 evaluated). So i dealt 120/70=1.714 less damage and recieved 165/113=1.46 more damage. So i was at a disadvantage 1.714*1.46=2.503 to 1 ONLY because of unlucky rolls. That's worse than fighting while standing on a plain with 30% defence against someone on a mountain with 70% defence. In other words, such disadvantage can't be countered at all. What's the fun in that?
an extreme position. If you were in such a good position
that you were going to win with NORMAL luck, what
steps to you take to reduce the variance of combats?
This is an important part of the strategy for playing games
where luck is involved.
For example if you have a big lead it might be best to try
to force the battle to occur on terrain with poor defense
for both players to reduce variance even if your other
option is to fight in defensive terrain that is slightly
better for you than your opponent.
Similarly, once you get to that point you might start
recruiting more units with magical attacks or more
blows rather than the best on average units.
For Risk the normally visible luck usually isn't thatTippsey wrote:I duno sure luck is annoying some of the time but it;s awesome at other times. Every game has luck involved yet we complain less about those. Risk is luck based, monopoly is luck based hell anywhere dice are involved so is luck, and oh layoff the everyone.
important when a string of good or bad luck occurs.
The really important luck occurs at the end of the
game (at least when not using mission cards) when
you wipe someone out and you need to make a set
with their cards in order to wipe out another player.
Probably in Monopoly the effect is reduced by important
hits and misses being spread out in time rather than
being rolled back to back. So it strings of good or bad
luck don't stand out as much.
If you want an example of a game that gets people to
whine about bad luck, Titan is a good example. In battles
some things are similar to Wesnoth in that units fight
at full effectiveness right up until they are killed. So
that the difference between killing and almost killing a
unit can in some cases swing a whole battle. Also a lot
of rolls of groups of dice are made during the game, so
that there will often be some notably good or bad rolls
that will stick in people's minds (while lots of "normal"
rolls will be forgotten). So you can get a lot of whining,
but the game offers interesting strategic and tactical
choices that are there because of the randomness.
Exactly that is the Problem !!!!!Also a lot
of rolls of groups of dice are made during the game, so
that there will often be some notably good or bad rolls
that will stick in people's minds (while lots of "normal"
rolls will be forgotten)
Exactly, this should not be possible. If you understand the concept, you would see, that the only asumption a player can make in the alternativ system is: "If I missed, maybee i hit next time". What he will assume already. He still cant be sure, on an 80% atack it will possible be more likly to happen, but that is what he wants to see.Have you thought through the implications of using a
system where you track how a person's luck is has
been in order to modify the chances to hit? This is
going to make playing the game well harder as you
will need to know what you current luck modifier
is and order your attacks to use up you good and bad
modifiers in the right attacks.
Code: Select all
If you care enough and have the time, you can maintain
your own fork with your change and try to convince
enough developers (not necessarily the current ones)
that your version is worth maintaining as a separate
fork to keep the project going. At least initially you
should be able to track updates to Dave's version
to get bug fixes. Since Wesnoth is already in the
leadup to 1.0, it is already the case that most of the
changes should be bug fixes which will make your
job easier.
Except that isn't a problem. It doesn't matter if people complain about bad luck, really, so long as the system actually does work.Niels wrote:Exactly that is the Problem !!!!!Also a lot of rolls of groups of dice are made during the game, so that there will often be some notably good or bad rolls that will stick in people's minds (while lots of "normal" rolls will be forgotten)
PS: I don't know who you play monopoly and risk with, but when I play, people complain about bad luck all the time. Every single game.
What you don't seem to understand is that we, the developers, have rejected this idea. Even though it is rather easy to implement, it is something we do not WANT to implement. It is not even a question of it not being worth the effort; it is something that would not be wanted even if it took no new code at all. So, your only hope is to fork Wesnoth and try to convince people that your fork, with its randomness-with-replacement, is superior. I wish you luck.Niels wrote:If you read my first post, you would have seen, that my changes would take only about 10 lines of code. I wouldn't need a bunch of developers to implement it.If you care enough and have the time, you can maintain your own fork with your change and try to convince enough developers (not necessarily the current ones) that your version is worth maintaining as a separate fork to keep the project going. At least initially you should be able to track updates to Dave's version to get bug fixes. Since Wesnoth is already in the leadup to 1.0, it is already the case that most of the changes should be bug fixes which will make your job easier.
For I am Turin Turambar - Master of Doom, by doom mastered. On permanent Wesbreak. Will not respond to private messages. Sorry!
And I hate stupid people.
The World of Orbivm
And I hate stupid people.
The World of Orbivm
That isn't true at all. Players will be able to calculateNiels wrote:
Exactly, this should not be possible. If you understand the concept, you would see, that the only asumption a player can make in the alternativ system is: "If I missed, maybee i hit next time". What he will assume already. He still cant be sure, on an 80% atack it will possible be more likly to happen, but that is what he wants to see.
exactly what the new odds of hitting are. They get to
see the code which makes it relatively easy. Even if they
didn't it would still be possible to figure it out by keeping
statistics.
So you have added a new dimension to the game which
is figuring out what your luck modifier is and
manipulating your attacks to take advantage of good
luck on key attacks and burning of bad luck on
unimportant attacks.
Judging by the number of threads started about this topic, the feeling of randomness and unfairness is one of the things that annoy Wesnoth players the most. Rather than the way that the random generator works, I think this is due to the famous "chance to hit determined by terrain only" principle. Why? The player doesn't have enough of a chance to affect it.
Let me make a comparison: I play a boardgame called Advanced Squad Leader. It's a complex WWII squad combat simulation. In there you have an ever-present luck element: you roll the dice to determine the effect of your fire. However you don't get the same feeling of randomness as in Wesnoth. Why?
- Like Wesnoth, there are many, many factors that can affect your fire. Even though randomness is a part of the game, you can "stack up" elements that favor you.
- Unlike Wesnoth, these factors affect your chance to hit: attack an enemy in motion in open ground with 12 squads firing together under the direction of your best leader, and at the end of the round, if they are not dead or at least routed it will be a total surprise. And this kind of "surprise" is not commonplace like in Wesnoth.
- As the power of your fire (or the negative circumstances for the enemy) increases, it eventually becomes impossible to miss. Then bad luck means "really bad results", but it stays realistic: so much firepower can't have zero effect.
So basically, I think that instead of having all in-game modifiers (night/day, leadership, weapon type, etc) affect only the power of the attack, we must have some of them affect the to-hit chance percentage (I know, we have magical and marksman, but they are not enough). This way the player will be able through his strategy to alleviate the effects of the randomness, and will feel less frustrated.
- night/day is an ideal candidate to affect to-hit, since it deals about vision.
- it would also make sense that leadership affect to hit probabilities, since the fire/melee is better directed.
It would also be interesting to have gradations in the effect of the attack, instead of a yes/no system. So, the 60% would be a target number, but you would still get a weak result if you score 55 on the "virtual dice". 75 would give a stronger result than hitting the mark exactly. I think that with sufficient odds, a miss should actually be impossible: for example if the player's normal attack is 10 and he has very high to hit odds, the less he could deal is 2 damage.
Even if we absolutely want to leave the to-hit system as-is, I think it is worth discussing these propositions. At least we can see if they are a solution to the problem, and if they are, we can still decide to keep the problem.
Let me make a comparison: I play a boardgame called Advanced Squad Leader. It's a complex WWII squad combat simulation. In there you have an ever-present luck element: you roll the dice to determine the effect of your fire. However you don't get the same feeling of randomness as in Wesnoth. Why?
- Like Wesnoth, there are many, many factors that can affect your fire. Even though randomness is a part of the game, you can "stack up" elements that favor you.
- Unlike Wesnoth, these factors affect your chance to hit: attack an enemy in motion in open ground with 12 squads firing together under the direction of your best leader, and at the end of the round, if they are not dead or at least routed it will be a total surprise. And this kind of "surprise" is not commonplace like in Wesnoth.
- As the power of your fire (or the negative circumstances for the enemy) increases, it eventually becomes impossible to miss. Then bad luck means "really bad results", but it stays realistic: so much firepower can't have zero effect.
So basically, I think that instead of having all in-game modifiers (night/day, leadership, weapon type, etc) affect only the power of the attack, we must have some of them affect the to-hit chance percentage (I know, we have magical and marksman, but they are not enough). This way the player will be able through his strategy to alleviate the effects of the randomness, and will feel less frustrated.
- night/day is an ideal candidate to affect to-hit, since it deals about vision.
- it would also make sense that leadership affect to hit probabilities, since the fire/melee is better directed.
It would make a lot of sense to get a better chance of hitting surrounded units. Duh. However, to avoid favoring hordes of inexpensive units, or giving too much of an edge to the player that has the most units (even though he's likely to win if he's a good tactician) surrounding should take into account the quality of units surrounding the "victim". Taking either the levels or the cost into consideration seem the best solution to me.Dave wrote:As far as I can tell, these ideas both advantage a side with more units against a side with less.Cuyo Quiz wrote:We have multiple propositions for that, including the fatigue idea and my idea about sorrounding.
I think this could increase the importance of luck in many situations, since as soon as you get 1-2 units down, your situation will become hopeless.
It would also be interesting to have gradations in the effect of the attack, instead of a yes/no system. So, the 60% would be a target number, but you would still get a weak result if you score 55 on the "virtual dice". 75 would give a stronger result than hitting the mark exactly. I think that with sufficient odds, a miss should actually be impossible: for example if the player's normal attack is 10 and he has very high to hit odds, the less he could deal is 2 damage.
Even if we absolutely want to leave the to-hit system as-is, I think it is worth discussing these propositions. At least we can see if they are a solution to the problem, and if they are, we can still decide to keep the problem.
A basic fatigue system could help reduce very bad luck.
One example could be that the base chance to dodge/defend against hits would sink if a unit
is considered "fatigued".
I know of other games where fatigue comes into effect too, though some do it by reducing accuracy, and some others by reducing the speed.
To keep it simple, the fatigue could be that for every third hit on a unit, that unit's dodge/parry chance is decreased by 10%.
This may seem as almost non-important, but may reduce the frustration by a bit, as you can attempt to swarm enemy units to "get them killed".
Also, personally to make it clear, I dont have a very strong opinion (pro/contra) about it (luck factor in Multiplayer).
One example could be that the base chance to dodge/defend against hits would sink if a unit
is considered "fatigued".
I know of other games where fatigue comes into effect too, though some do it by reducing accuracy, and some others by reducing the speed.
To keep it simple, the fatigue could be that for every third hit on a unit, that unit's dodge/parry chance is decreased by 10%.
This may seem as almost non-important, but may reduce the frustration by a bit, as you can attempt to swarm enemy units to "get them killed".
Also, personally to make it clear, I dont have a very strong opinion (pro/contra) about it (luck factor in Multiplayer).
I suggest an option called "Accuracy Variance" which will be an server option to be set at the beginning of the game. Setting it will mean that if a unit has 70% defence, exactly 70% of the attacks thrown at it will connect. Also, "Glancing Blows" possibility would be added. So if you attack an elf archer in the forest with a troll welp, you will connect exactly .6 times, doing 6 damage if the original attack was 10-2. If the result is a decimal, it will still add to total damage assuming the attacked unit has a weakness or resistance to the given damage type. I'm no expert in programming but I'd assume this wouldn't be overly hard to do.
As to the issue of whether luck should be a factor in games, It is my personal opinion that large amounts of luck are detrimental to a game, but small amounts are not. Percentages from 30-70% are generally not good to see, as they mean that good and bad luck will have much more an affect on the game. I personally would rather see odds that are 1-20% and 80-100%, as these mean that you will only get that rediculously frustrating string of misses or hits less than one percent of the time. Nothing is more frustrating than in bay of pearls getting your whole merman army ripped up by ai trolls because they get extremely lucky and hit with 100% acc when they should be hitting at 40% acc, but at the same time having no luck in a game means that you'll end up like chess, with a bunch of "standard" ways to play the game and almost no variation or vareity over time, because the way the game unfolds can be predicted. Luck can be a tool of games that are on the less-complicated side to attempt to make the game more unpredictable.
As to the issue of whether luck should be a factor in games, It is my personal opinion that large amounts of luck are detrimental to a game, but small amounts are not. Percentages from 30-70% are generally not good to see, as they mean that good and bad luck will have much more an affect on the game. I personally would rather see odds that are 1-20% and 80-100%, as these mean that you will only get that rediculously frustrating string of misses or hits less than one percent of the time. Nothing is more frustrating than in bay of pearls getting your whole merman army ripped up by ai trolls because they get extremely lucky and hit with 100% acc when they should be hitting at 40% acc, but at the same time having no luck in a game means that you'll end up like chess, with a bunch of "standard" ways to play the game and almost no variation or vareity over time, because the way the game unfolds can be predicted. Luck can be a tool of games that are on the less-complicated side to attempt to make the game more unpredictable.
[edit] I just realized that my idea of fatigue wasn't entirely accurate. I think what I said still applies pretty well though. Correct me please if I'm wrong. [/edit[
"Accuracy Variance" sounds a lot like the card example given earlier. Either that or rounding the # of attacks would lead to nasty balancing problems.
Fatigue misses the problem. There are already chance issues where knights can repeatedly charge a unit with only a 30% chance of hitting, yet kill the unit in one blow. Repeatedly. Fatigue would only make that problem worse, as lucky blows from heavy units onto lighter ones would allow more units to join the party and demolish them.
Meanwhile, it doesn't really solve the problem where you have one or two heavy units, attack a unit with only 20% defense, and miss all the time. Of course if you hit at the very beginning it might help, but if you're unlucky and hit later you're sunk. Fatigue merely changes what it means to be unlucky.
My main frustration from the game is how quickly my units can die. I'm never able to figure out when my units have a good chance of dying because of the huge amount of chance involved... I just don't know whether I'm going to be lucky or not, and there simply are not enough units on some maps for me to prevent my guys from getting swarmed by hit-crazy scouts. Unless you're on a map which is relatively small or you play with more than usual money it's really hard to form formations that will assure your unit won't die out of the blue.
Moving the defense values a bit towards opposite ends of the spectrum seems like a better idea to me. Since BFW is modable it wouldn't be very hard to maintain such a thing, and it'd be nice option at the very least. Moreover, with unit/gfx changes you could make it much different from the original scenario, which would split the community less sharply, giving people multiple reasons to like unmodded BFW over the mod, rather than just simple percentage values.
"Accuracy Variance" sounds a lot like the card example given earlier. Either that or rounding the # of attacks would lead to nasty balancing problems.
Fatigue misses the problem. There are already chance issues where knights can repeatedly charge a unit with only a 30% chance of hitting, yet kill the unit in one blow. Repeatedly. Fatigue would only make that problem worse, as lucky blows from heavy units onto lighter ones would allow more units to join the party and demolish them.
Meanwhile, it doesn't really solve the problem where you have one or two heavy units, attack a unit with only 20% defense, and miss all the time. Of course if you hit at the very beginning it might help, but if you're unlucky and hit later you're sunk. Fatigue merely changes what it means to be unlucky.
My main frustration from the game is how quickly my units can die. I'm never able to figure out when my units have a good chance of dying because of the huge amount of chance involved... I just don't know whether I'm going to be lucky or not, and there simply are not enough units on some maps for me to prevent my guys from getting swarmed by hit-crazy scouts. Unless you're on a map which is relatively small or you play with more than usual money it's really hard to form formations that will assure your unit won't die out of the blue.
Moving the defense values a bit towards opposite ends of the spectrum seems like a better idea to me. Since BFW is modable it wouldn't be very hard to maintain such a thing, and it'd be nice option at the very least. Moreover, with unit/gfx changes you could make it much different from the original scenario, which would split the community less sharply, giving people multiple reasons to like unmodded BFW over the mod, rather than just simple percentage values.
Then do it!Phyvo wrote: Moving the defense values a bit towards opposite ends of the spectrum seems like a better idea to me. Since BFW is modable it wouldn't be very hard to maintain such a thing, and it'd be nice option at the very least. Moreover, with unit/gfx changes you could make it much different from the original scenario, which would split the community less sharply, giving people multiple reasons to like unmodded BFW over the mod, rather than just simple percentage values.
“At Gambling, the deadly sin is to mistake bad play for bad luck.” -- Ian Fleming
I agree with dave.
if you don't like the fact that it involves luck, start save loading, o learn to program and do something about it.
there's no point in complaining.
if you don't like the fact that it involves luck, start save loading, o learn to program and do something about it.
there's no point in complaining.
Oh no look out its a ray gun.
You should move to avoid the rays
the rays are coming out of the gun
if you are hit by the rays
you will be shot by the rays
the rays are fast so you should be fast to
can you win against the fast rays from the gun?
You should move to avoid the rays
the rays are coming out of the gun
if you are hit by the rays
you will be shot by the rays
the rays are fast so you should be fast to
can you win against the fast rays from the gun?
I discussed at length the distribution of damage previously. Instead of rehashing the same, lame objections to the current damage system, it would be good if people actually responded to the points in my essay. So far, no-one has rebutted any of my points, so I remain unconvinced that the distribution needs revision.
This quote is not attributable to Antoine de Saint-Exupéry.
Unsung/Dave:
Abusing saving by hitting ALT+S over and over again is not fun. If you're doing that the game has already failed to be fun to you and is thus pointless, because games have no other purpose. If the game forces you to do this to have a smidgen of fun then you are either playing the wrong game or the game itself is terrible. Whichever the player percieves is his choice... and most go for the latter.
As for doing something about it, I have. Considering the simplicity of the modification, the results turned out pretty good: I subtracted 10% from all defense values below 50% and added 10% to those above 50%. The only two issues with this I have noticed is that dwarves are a bit too strong on mountains and that drakes become too weak everywhere. Skirmishers, though tougher than before, can be dealt with using mages or other units on good ground. I probably should add 10% to magical and sharpshooter attacks too.
I don't even know WML. I just opened the text files and followed the tags. Tinkering with stock values is hardily difficult enough to be called "programming"... I actually spent more time last night failing to write a Python script that did the job for me than doing it by hand myself.
Ott:
I suppose that if you play on larger maps with a whole lot of units it isn't so bad. However, I like playing small maps because even with them it can take me an hour to finish a game... with the other three players being fast-thinking accelerated comps. I can't even imagine what it'd be like trying to do that with people... I don't have the time. Thus, I guess the problem is more noticable to me and others who play like me. Some of the time I simply do not have enough units to make any formation, and so when I place my units on the most favorable terrain possible and they die... It's like those large chance swings of yours are made even bigger than usual.
This also explains why some feel simply changing the gold to 150 is not enough, I think, since on small maps the number you have later in the game is not as affected as on larger ones by this change, unless you're crazy and are playing at 1 gold/village when there are only around 40 for four players.
Abusing saving by hitting ALT+S over and over again is not fun. If you're doing that the game has already failed to be fun to you and is thus pointless, because games have no other purpose. If the game forces you to do this to have a smidgen of fun then you are either playing the wrong game or the game itself is terrible. Whichever the player percieves is his choice... and most go for the latter.
As for doing something about it, I have. Considering the simplicity of the modification, the results turned out pretty good: I subtracted 10% from all defense values below 50% and added 10% to those above 50%. The only two issues with this I have noticed is that dwarves are a bit too strong on mountains and that drakes become too weak everywhere. Skirmishers, though tougher than before, can be dealt with using mages or other units on good ground. I probably should add 10% to magical and sharpshooter attacks too.
I don't even know WML. I just opened the text files and followed the tags. Tinkering with stock values is hardily difficult enough to be called "programming"... I actually spent more time last night failing to write a Python script that did the job for me than doing it by hand myself.
Ott:
I suppose that if you play on larger maps with a whole lot of units it isn't so bad. However, I like playing small maps because even with them it can take me an hour to finish a game... with the other three players being fast-thinking accelerated comps. I can't even imagine what it'd be like trying to do that with people... I don't have the time. Thus, I guess the problem is more noticable to me and others who play like me. Some of the time I simply do not have enough units to make any formation, and so when I place my units on the most favorable terrain possible and they die... It's like those large chance swings of yours are made even bigger than usual.
This also explains why some feel simply changing the gold to 150 is not enough, I think, since on small maps the number you have later in the game is not as affected as on larger ones by this change, unless you're crazy and are playing at 1 gold/village when there are only around 40 for four players.