On Randomness and Game Longevity

General feedback and discussion of the game.

Moderator: Forum Moderators

Post Reply
Founding Developer
Posts: 7071
Joined: August 17th, 2003, 5:07 am
Location: Seattle

On Randomness and Game Longevity

Post by Dave »

Partly as a result of this topic: http://www.wesnoth.org/forum/viewtopic. ... 21&start=0 I though I would write some thoughts on randomness in Wesnoth and its implications on keeping players entertained.

By far, the biggest gameplay-related complaint about Wesnoth is that it relies too heavily on chance. The best-laid plans can fall apart based on an unlucky streak, and a poor strategy can win if the player is lucky.

Now, my first observation is that compared to 'd20' type gaming systems, Wesnoth isn't all that random. In a d20 system, each character typically gets just one attack per round, and must roll a certain amount on the die to hit. Even if they do hit, the damage they inflict is random, so even a successful hit might inflict a very small amount of damage.

So, why do some players find Wesnoth's system so frustrating? There are two main reasons I can see.

One of them is that a d20 system would also be frustrating for these players. The most complaints come from people who play lots of multiplayer, and d20 systems are more often used in RPGs for PCs-vs-NPCs.

The other one is that d20 type RPGs tend to give the attacking player more scope for determining the chances. The chance to hit in such games is typically a combination of the attacker's ability (To-Hit-Armor-Class-0, or THAC0, for instance), and the defender's defense value (often called Armor Class). If the player really wants to eliminate a certain enemy, they can do so by sending their most accurate attackers who will almost certainly guarantee success.

In Wesnoth, there are only a few units that can guarantee a certain hit rate (mages, etc), but most of the time, the player has little control of their chances of hitting. This lack of control breeds frustration.

However, this is all part of the game design. It is a fundamentally simple system, and the entire design revolves around the way chance-to-hit is determined. I don't think a simple change can make Wesnoth more enjoyable for the frustrated; rather I think it will make it less enjoyable for everyone.

I will completely agree that Wesnoth's system wlil probably start to get quite tiresome for players who have put 100+ hours into the game (and less, for some). We must remember what Wesnoth's game rules were originally meant to achieve: the game was a single player only game and it had only one campaign. It was designed to keep a player entertained for 20-30 hours. It was never designed to scale to the point of 'expert' players dueling each other over and over on the same maps, perfecting their strategies.

I am of course, glad that what people do with Wesnoth today is far beyond its original goals. However, we must keep the original design constraints of Wesnoth in mind.

Personally, what I would love to see happen is for some people to take Wesnoth's lead, and develop some games that they find entertaining. For instance, an original turn-based strategy game that is designed for fun, balanced, player vs player combat. However, it seems that the number of people who have the skill, time, and patience for such an undertaking are limited.

An alternative would be to modify Wesnoth's design and game rules to make it more suitable for online play between skilled players, as well as less frustrating.

My suggested change would be something like this:

Make damage inflicted by a hit random. Make the randomness determined by how easy the target is to hit. If they are in poor terrain, the damage might be random, 100-200% of the 'base' damage. If they are in good terrain, it might only be 100-120%.

Give attacks an 'accuracy' rating, and units a 'parry' rating. The 'accuracy' rating is just a number which determines how accurate the attack is. The 'parry' rating is two numbers: parrying skill and number of parries.

When one unit attacks another, the unit gets to try to avoid taking any damage from each blow. Their chance of avoiding a blow is given by modified_parry_rating/(attack_accuracy+modified_parry_rating) where modified_parry_rating = parry_rating*terrain_modifier

The defender can only parry up to their number of parries.

The key is that most units will have a very low parry rating compared to typical attack accuracy. Most attacks will hit, though the damage will vary somewhat. Only occasional elite units will have very high parry ratings, and be able to parry off many attacks. Even then though, the attacker will still typically have a good chance to hit. (For instance, a 'typical' accuracy rating might be 10, while a typical parry rating might be 2. A very good parrier might have a parry rating of 10, and thus in ideal terrain be able to parry 50% of the attacks, and only up to their number of parries limit).

Magical attacks would be unable to be parried.

Now, this idea is certainly more complicated, and is nothing more than a rough idea of what I would do if I wanted to remove the randomness-related frustration from Wesnoth. I have no doubt that many people will heartily dislike the idea, and I don't really blame them: I rather like the simplicity of the current system.

“At Gambling, the deadly sin is to mistake bad play for bad luck.” -- Ian Fleming
Glowing Fish
Posts: 855
Joined: October 3rd, 2004, 4:52 am
Location: Portland, OR

But ego is critical...

Post by Glowing Fish »

I think the biggest problem with random numbers showing up in gameplay is that people are getting way too egotistical about the game. It used to be that multiplayer was just a diversion. I myself have only played a few multiplayer games (I just don't have the time to put into it).
I don't see the purpose of having these perfectly symmetrical maps with no randomness, so you can "prove" that you are the better player. What is wrong with playing the game, having some bad luck, and losing? It doesn't mean you aren't a real man! You don't have to win, and you don't have to have a totally fair game, to have fun.
Actually, the only part that the randomness really bothers me is in campaigns. In long campaigns, you have to think about what units you will need ten scenarios from now, and then having a Great Mage miss four shots against an Orcish Warlord than then kills him can be very frustrating. But in a multiplayer game...if you do lose, so what?
I like the game mechanics just fine as they are now.
Don't go to Glowing Fish for advice, he will say both yes and no.
User avatar
Inactive Developer
Posts: 4453
Joined: November 26th, 2005, 7:41 am

Post by Sapient »

The attack/parry-ratings branch sounds like a nice proposal. The key challenge would be making it easily understandable; although, of course, it would never be as easy to understand as the current %def model. ;)
http://www.wesnoth.org/wiki/User:Sapient... "Looks like your skills saved us again. Uh, well at least, they saved Soarin's apple pie."
Posts: 221
Joined: January 11th, 2006, 8:51 am
Location: Barad-Dur, Mordor

Re: On Randomness and Game Longevity

Post by Sauron »

Hmmm, I must write sth I would rather not.

In my opinion the amount of players in wesnoth is partially the result of an addiction, similar to the one of the players in casinos. Addiction of adrenaline/ risk. Of course you can laugh at this, but many ppl consider playing Wesnoth like this: "I lost, RNG was against me in crucial moment. Time to play again, mabye this time it will not spoil my plans." - and they play ... .

The huge influence of RNG makes people seek for some golden-playing-rule, that will allow them to manage risk better than opponents. There's none, since you can only stick to strong terrain hoping the statistics will solve the rest. The alternative is maneuver - and even having best plan you can end up exposed (my last game: 3 magical attacks, each of them killing one - all missed - chance - 2,7% - and no way to screen the mage).

If we strip the RNG influence off the game it will certainly get boring much faster - rules/resistances/defenses/specials are simple enough to get learned by heart pretty fast. Paradoxally - in my opinion - the fighting system makes people remain attracted to game for much more time.

Having written all of that - I suggest analysing the tweak-up proposed in 1.0.5 version of my mod - the averages-based amount of damage. I think it is less complex than what you propose - and requires no changes to rules/abilities and so on.
Last edited by Sauron on September 22nd, 2006, 5:50 pm, edited 1 time in total.
Customize yourself random factor in game:
GET my mod [available as C++ sourcecode and compiled Windows executable] for wesnoth 1.6.4
at http://saurons-mod.zor.org/
Mod thread
Posts: 267
Joined: July 18th, 2006, 11:52 am

Post by PingPangQui »

You are completly right about that. Its that simplicity + gambling that people make addicted to it.
Inactive Developer
Posts: 1272
Joined: August 27th, 2004, 8:05 am
Location: Berlin, Germany

Re: On Randomness and Game Longevity

Post by Rhuvaen »

Dave wrote:By far, the biggest gameplay-related complaint about Wesnoth is that it relies too heavily on chance. [...]
However, this is all part of the game design. [...]
It was never designed to scale to the point of 'expert' players dueling each other over and over on the same maps, perfecting their strategies.
I am probably a mid-level player when it comes to tournament-style playing, and can say that despite the tactical randomness I will consistently lose against expert players and consistently win versus low-intermediate players. There is a lot of skill involved, after all. It's when players are evenly matched and are able to exploit the other's mistakes that the tactical outcome of a small number of attacks actually decide the game.

And why shouldn't it? If you pit two equal forces with equal strategic positions and equally trained commanders against each other on the battlefield, what do you expect? A draw? Or a decision of fate? The alternative to having chance decide the game in those situations is a stalemate (which often is worse for a game).

Chance is a part in all strategic wargames - and often, to a much greater extent than in Wesnoth - simply because chance, or rather the "fog of war" or unknown factors play such a big role. Strategic games that don't feature chance rely sorely on the player projecting his strategy and the opponent making countering moves. They are usually far more simplistic than any given Wesnoth scenario for players to be able to do so. If anything, Wesnoth needs to become simpler rather than more complex if chance was to become less of a deciding factor - otherwise it would be nearly impossible for any but the most trained players to successfully bring "their" strategy to a conclusion.

Is it really chance that is the problem? I can see that if I were trying to build my skills in tournament play and contend with the top-level players, as I would get close to matching their skill, there would be more and more situations where in my perception the randomness would decide the game. But is that really true? Haven't I been hinging my strategy on too specific tactical outcomes? What were the mistakes leading up to the situation where everything was decided by one attack? I think that subjective perception plays a role here, too.

If it really is true that we have reached a stage where chance is becoming the single most determining factor between expert players in MP, then let's look back at what Wesnoth and the people contributing and playing have achieved:
  • - a basic system that lends itself to MP and attracts a community to it
    - a finely tuned balance of units and factions
    - a set of balanced maps that offers variety and consistency
    - a growing community of players of skill
Wesnoth multiplayer has actually evolved into a very mature environment to come to face this "problem"! Now we could say: okay, this is all getting boring, let's go back to stage one and change the basics (just nuke our little ecosystem and start from bacteria again :twisted:). Or we could ask "what is the next step?".
Posts: 267
Joined: July 18th, 2006, 11:52 am

Post by PingPangQui »

I totaly agree with that too. After all its that simplicity (i.e. simple rules, not simple game play) and that bit of luck that makes that game so great.

And for those who think simple rules = simple game play - just think of chess or Go. They are both quite simple according rules (especially Go) but very demanding in game play.

Wesnoth is just the greatest computer based strategy game ever.
Retired Developer
Posts: 1086
Joined: September 16th, 2005, 5:44 am
Location: Hamburg, Germany

Post by Yogibear »

I have been reading a couple of threads about randomness in wesnoth and after a while i thought: If two players of similar skill play each other, the RNG is mostly deciding who wins.

Then i read a sentence from Doc, where he stated the following theory (it's not a cite, i don't remember the exact words): "If two players of similar skill play 100 games, the outcome will very seldom be like 52:48. In fact, only a neglectable amount of games will be decided by luck."

Since then, i paid a little more attention to this. Not long ago, i played Soliton, who by no means can be regarded of being more skilled than me.

Game 1 found him to make a mistake right in the beginning, loosing a gryphon in one turn. It also found me having excessive luck (i think i was 200 or more below EV damage taken for a long time). But i made a couple of mistakes: And they decided the game, Soliton won.

In game 2, he (IMO) misjudged the situation and decided to overrun me. He had some luck on his side (not so much but still better than EV) but he failed, loosing enough units to make me win the game in the end.

Like almost every mp player, i have been cursing the RNG a lot, but after these experiences i start asking myself if it is rather a question of perception than of reality. I must admit though, that i didn't do a serious analysis of a big number of games to prove that.

I also remember to be quite astonished a couple of times, when after having my mage miss all three attacks i looked for EV and found i was still on the better side of it...

BTW, i am playing wesnoth for quite some time now (although not every day as many others) and i still can't get enough of it :D .
Smart persons learn out of their mistakes, wise persons learn out of others mistakes!
Retired Developer
Posts: 1086
Joined: September 16th, 2005, 5:44 am
Location: Hamburg, Germany

Post by Yogibear »

Yogi Bear wrote:Not long ago, i played Soliton, who by no means can be regarded of being more skilled than me.

A linguistic misclick, this is supposed to read "Soliton, who is definitely more skilled than me".

Sorry for that, Soliton :oops: .
Smart persons learn out of their mistakes, wise persons learn out of others mistakes!
User avatar
Posts: 1401
Joined: February 7th, 2005, 7:04 am
Location: Poland

Post by Tomsik »

Randomness is very important thing, without it wesnoth would be boring. Frustration is caused only by extreme cases, so why change whole system dramatically instead of just making these extremes happen less frequently?

Perhaps there could be small amount of % added/substracted to thing that RNG produced based on your damage dealt/damage dealt EV.
Dragon Master
Posts: 1012
Joined: February 11th, 2006, 1:04 am
Location: Somewhere

Post by Dragon Master »

Well, so far accuracy and randomness has been proposed to be made more tame by adding more variables (counterproductive, I know). People have suggested that instead of keeping just the chance to hit, we add defense variables, attack variables, chance to hit, and chance to parry. The current system is much better IMO. Somehow changes don't seem useful for a game that has been built around another system. Don't get me wrong, it's perfectly fine to make a game or branch with different rules, but the mainline Wesnoth doesn't seem to need this.
User avatar
Multiplayer Contributor
Posts: 499
Joined: February 17th, 2006, 6:26 pm
Location: Chicago

Post by jb »

Well, I see a few things going on here....

First of all, too much of anything is too much. Even addictions. This includes Wesnoth. People love cocaine, but after 3 days of watching the sunrise you have to walk away for awhile. It's easy to get burned out on something if you try to get too many miles out of it.

Secondly, I think Sauron is exactly right about the gambling type of addiction. "Here we go baby, all i need are 2 hits from my mage....big money, big money - no whammies!" And then when your mage only hits once you curse out loud and throw your empty beer cans at your girlfriend and call her a fat slut because she refused to clean up after you.....

Also, on making the game different....Traits & Random factions are right now the best game changer around. It's always the same map, same terrain types, same day/night cycle, etc. something like wheather might help a bit, more water in spring...snow in winter, that type of thing. But what is really needed is a random map generator that makes better random maps. Then you'll never have a repeat scenerio, and things will always be fresh, if somewhat unfair at times.

Dave, I'm not sure why you're using D&D referance to a battle system as most computer games do not use a system based on 5% intervals. And you sound half like you want to give up on wesnoth all together and half like you want to re-invent it. Wesnoth is a good game. The stats and % are all very good, terrain and movement and traits and factions and att types....it's all good. Some people are just getting burned out from playing too much. They will be back, and so will you.
My MP campaigns
The Altaz Mariners - with Bob the Mighty
Posts: 103
Joined: July 29th, 2004, 4:50 pm

Post by Ewan »

In a different part of my life, I play historical miniature tournament games: take a historical army pre-gunpowder, design a points-bought force from the troops available to that army, go forth to fight against other opponents anywhere from Sumerians to medieval knights.

Similarly to Wesnoth, there are a bunch of terrain effects, weapon-armour interactions, and so on; and then a random factor (in fact, a few different types of random factors). That random factor can - rarely - win a game for you or lose it, if you happen to get a extreme set of rolls at the wrong time; but that's only going to happen if you're playing someone of sufficient skill that the imbalance caused is enough to tilt the game. Otherwise, you were going to win or lose anyway - and indeed a big part of the skill is in setting up attacks that are sufficiently imbalanced in your favour that dice do not matter! All of this is true in Wesnoth also; and now we have the debate on desired level of randomness which I've seen before :).

My take: I whine about dice like anyone else, and try to remember to thank them when appropriate! But even in the depths of bad dice, (i) that gives you something to blame other than skill, and (ii) I know it balances overall. So why worry, really? Plus, both Wesnoth and miniatures allow you to influence randomness - e.g. if you want to control your chance to-hit, take mages! If you want to live and die by random swings, take units like trolls with huge damage but small numbers of blows. So not only is there no problem, there's an added player-choice which alllows for superior play to be rewarded. And that's a good thing.
Dragon Master
Posts: 1012
Joined: February 11th, 2006, 1:04 am
Location: Somewhere

Post by Dragon Master »

Well technically, there is nothing random about Wesnoth. There may be chances that certain attacks will hit or not, but you always know the exact chances that your attack will hit or miss.
User avatar
Doc Paterson
Drake Cartographer
Posts: 1973
Joined: February 21st, 2005, 9:37 pm
Location: Kazakh

Post by Doc Paterson »

jb wrote:Wesnoth is a good game. The stats and % are all very good, terrain and movement and traits and factions and att types....it's all good. Some people are just getting burned out from playing too much. They will be back, and so will you.


Of course players are going to lose the occasional game due to luck, but that's why people need to play multiple games - Trust me- patterns do emerge, and the better player will always end up winning more games. I think that a lot of uneccessary stress and complaining stem from people taking individual games far too seriously.

Is it difficult for the RNG-Is-Terrible-Whiners to understand why the same players continue to dominate multiplayer? Probably.

Is it any surprise that the Final Four of our tournament of 16 players are the very four that most guessed it would be? Funny that the RNG didn't bump at least one of them off, isn't it? :wink:
I will not tell you my corner / where threads don't get locked because of mostly no reason /
because I don't want your hostile disease / to spread all over the world.
I prefer that corner to remain hidden /
without your noses.
-Nosebane, Sorcerer Supreme
Post Reply