On Randomness and Game Longevity

[zookeeper]

There should simply be alternative game mechanics (Age of Heroes, anyone?) for those who want pure skill.

I don't have anything against less-luck mods, but a completely deterministic one doesn't sound like a good idea to me. Reasons? The fact that the randomness is just too deeply integrated to the gameplay mechanics. Changing that (in a mod) would change the gameplay so much that we'd have to change a heckload of other things too: rebalance all units, simplify all stats, change how the combat system works, etc. Again, I'd be curious to see whether it works (really), but it could just as well be done as a completely new game (could be based on Wesnoth of course and reuse most of the stuff, but basically I think it would need to be a standalone game).

I think the thing that you overestimate a bit is the chances of just changing the logic of how hits and damage are dealt and still having the game work and be interesting, if everything else is kept as it is. I suspect that a lot of other things (complete rebalancing and simplification of unit stats, rule changes) would need to be changed to make a deterministic version good. The idea of "a deterministic Wesnoth" is not a bad one as such (I like it), but I believe it would require it to be more like a spin-off game than a simple mod to work well enough.

[Higher Game]

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 think my example of the 100 meter dash fits this example perfectly. The current Wesnoth system simply doesn't allow one to gauge very subtle differences in skill between expert players, and with such simple gameplay mechanics, any differences in skill are simply going to be subtle. Wesnoth is a simple game for simple players (20-30 hours is simple, period), but it fails when it comes to experts, and in the campaigns, since long term strategy is overemphasized.

Wesnoth is more or less finished. Making a new game would be silly because so much effort and work has gone into Wesnoth. A mod would do the job perfectly.

[khamul]

$0.02:
The frustration in Wesnoth for me comes from the relatively high likelihood of dealing no damage, for pretty much any given unit. I can't think of any comparable games, offhand, where it is even possible to attack and do no damage. I suspect there's a reason for this.

I'm not saying that this removes strategy, or makes the whole game dependent on luck, or anything like that. I'm just saying that it is the single biggest factor in winding me up.

I'd be interested in seeing a mod of Wesnoth where all units have a base damage that's always applied, and THEN do the normal CTH approach. So a spearman, for example, would immediately do 2 pierce damage, and then calculate the normal 7-3 damage according to the current rules. Obviously, this would require a lot of rebalancing, but the guaranteed damage would help reduce frustration at being unable to achieve anything, and would help players see each hit that followed as lucky, rather than each miss as unlucky.

The automatic 'base' damage could be modified by CTH: so a unit doing 5 base damage against a unit on 60% defense would only do 40% of the base: 2 damage.

[appleide]

Completely deterministic wesnoth like chess: Lose all replayability in single player campaigns when you figure out how to beat them.

[Higher Game]

Chess is very replayable. You're also forgetting that the AI recruits randomly, so there would still be a need for dynamic strategies.

[Kestenvarn]

Oh, for cryin' out loud... it is a troll.

[Sombra]

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 think my example of the 100 meter dash fits this example perfectly. The current Wesnoth system simply doesn't allow one to gauge very subtle differences in skill between expert players, and with such simple gameplay mechanics, any differences in skill are simply going to be subtle. Wesnoth is a simple game for simple players (20-30 hours is simple, period), but it fails when it comes to experts, and in the campaigns, since long term strategy is overemphasized.

Wesnoth is more or less finished. Making a new game would be silly because so much effort and work has gone into Wesnoth. A mod would do the job perfectly.

Wesnoth is simple, Chess is very simple, Go even more... still somehow the same players win mostly again and again.

To reduce luck you could play bigger maps. One the one hand you claim that its a simple game on the other hand you claim that you think its difficult to handle the long term strategy. Do you try MP games and challenge the big names => tournament thread?

If you can beat them on a 30%-60% basis I would agree that the game is to luck dependent and needs more strategy elements.

[Xeron_]

well i haven't read the entire 4 pages of text you lads and ladies have produced by now, but i have to state, that i find the current level of randomness actually very good.

Someone else has already stated wesnoth is not chess. - i agree. - wesnoth is similar, because you have to think what possibilities you have got, how the enemy could react and what possibilities/threads would appear out of the possible reaction of you enemy in caseX etc.
-However wesnoth is also a game based on military strategies. A military strategist is always in some way dependent on what results his soldiers reach in the actual battle. - and still people have always differenced between gifted generals and those who were not.

If you think about the matter more openly, - don't you think, that you can avoid phases of bad luck by playing a bit more defencively and not relieing on small chances too much?
And on the other hand, don't you think the game would actually lose a lot of it's flair if you wouldn't smash your fist on the keyboard now and then after your unit has refused to land an important hit? Don't you think, that wesnoth would turn cold and plain if a small hope, an act of despair can get you out of a deadly situation now and then - even if it's just a 15% chance?

I myself think, that these emotional situations are more valueable than strict calculations and statistiks.

Of course, if you want to play Wesnoth with a million people watching your game and beting millions of dollars you will like predicatbility and continuity, but if you play the game for fun, as most users do, fun should be the aim rather than a maximum of 2 days of thinking time as chess players may take it.

[Higher Game]

Oh, for cryin' out loud... it is a troll.

The only "trolling" I have done is when I picked on JW a little (after he rebuted me harshly when I was still a real newbie), who, ironically, identified the RNG as a source of some of his frustration.

I don't even think the randomness is a serious problem for casual multiplayer. It's only a problem for very long campaigns and for hardcore players who want (and require) a very accurate balance to test their skills.

Think of the campaign as a long series of 1's and 0's, like 1000. 50% chance for each to occur. If a string of 8 0's pops up, it's such a prolonged, catastrophic loss that it sets back the entire campaign. It's extremely unlikely (1/256), but there are 1000 numbers to go through, so it's a strong possibility, except with some save/reloading. It's possible for extremely good luck (1/256), but good luck doesn't count for nearly as much. It takes ~4 inflicted kills for a unit to get a promo. It takes 1 receptive kill for a unit to, uh, die.

Would you watch the Superbowl if each team had a 50% chance of getting 10 extra points at the beginning? It's random, and balanced, but it is still unappealing.

[Darth Fool]

Would you watch the Superbowl if each team had a 50% chance of getting 10 extra points at the beginning? It's random, and balanced, but it is still unappealing.

No, but I might watch it if they had a random coin flip to determine who got to choose which side and who got to receive first. Randomness is even less of a factor in long campaigns than in short Multiplayer games. In a campaign, you should be able to survive 8 misses in a row, especially later in the campaign, unless you are playing on hard.

[Higher Game]

In a casual multiplayer match, a match could be lost from, say, 4 0's in a row. This isn't a big deal, since it's just an hour or so (the string of #'s is shorter), and it's easy to get over an unlucky match. It's a bigger problem in a long campaign, when you realize that losing those elvish riders 4 maps ago screws you when you need to escort a distanct hero back to your base, and scouts just don't have enough muscle (hypothetical example). So, you go back 4 maps and it takes 4+ hours, instead of just a 1 hour unlucky match.

And believe me, it does take about 4 maps to get a promo. That's about 1 kill per match, a reasonable expectation. Some kill 2, kill a single level 2, and some get killed. It evens out to about 1/match.

Also, the luck factor in multiplayer can lead to some funny situations that both players can enjoy, as long as they don't take it as a serious test of skills. It's a fun diversion. This just isn't there when it's against a heartless AI, or a heartless hardcore player.

Check out Rabin's comment here. Lots of games use rigged RNGs to eliminate extremely variable luck; it's nothing to be ashamed of.

http://www.aiwisdom.com/bytopic_randomness.html

[JW]

And believe me, it does take about 4 maps to get a promo. That's about 1 kill per match, a reasonable expectation. Some kill 2, kill a single level 2, and some get killed. It evens out to about 1/match.

And believe me, it's very possible to promote 3+ units every map.

Randomness is a bigger problem in MP than in SP specifically because MP games are shorter: randomness has a much bigger effect.

[Zhukov]

Oh, for cryin' out loud... it is a troll.

Not quite the term I would use, but yeah.

[Sauron]

I'd be interested in seeing a mod of Wesnoth where all units have a base damage that's always applied, and THEN do the normal CTH approach. So a spearman, for example, would immediately do 2 pierce damage, and then calculate the normal 7-3 damage according to the current rules. Obviously, this would require a lot of rebalancing, but the guaranteed damage would help reduce frustration at being unable to achieve anything, and would help players see each hit that followed as lucky, rather than each miss as unlucky.

The automatic 'base' damage could be modified by CTH: so a unit doing 5 base damage against a unit on 60% defense would only do 40% of the base: 2 damage.

How about reading my sig . The linked thread describes a mod that does what you expect AND MUCH MORE.

[Sauron]

(...)f you assume all your attacks miss it makes no sense at all to do any offensive moves!!! (...)

You don't assume that all your attacks will do 0 damage, you assume that all your attacks MIGHT do 0 damage. I will give an example later.

I see no reason for assuming I make ANY damage. IF assuming I make EV leads to disaster (and it almost always does if we take into account SINGLE attack!) the best planning is betting on below EV. Now what is the lower limit for damage in such a "worst case management"? It is 0. It is no MIGHT, it is MOST LIKELY. Otherwise you are gambling.

(...)
The point is, you have to take into account the fact that some times your attacks will do no damage and try to minimize the risk while maximizing the benefit. The best players are very good at exactly that. This is a different type of game from a game where all results are 100% predictable like chess, and it requires different skills. This doesn't make it better or worse, just different.

1. Every (non-counter)attack makes you expose your units. More or less, but always.
2. If attack goes wrong (I already and by default assume you made all needed calculations., placed the cannon-fodder and bruisers at their positions to reduce the exposure. Fine. Now comes the counter, that is now MORE likely to hit you, due to weak position. IF your enemy is not unlucky - he gains advantage. All the time using position, that is stronger than your. This makes him gain local advantage. Of course next time you might get lucky, but opponent has now more units to counter. No skill allows to manage bad luck that exceeds some value, that depends on the situation on the board.