Luck in Wesnoth: killing competitive MPlaying

[Boy]

I am many years MP player. And not the worst one :)
And I want to say my words about "managing luck" in MP games.

What is real luck in MP duel?
Guide How to truly understand luck (and never ever use "s" key again).

Assuming that both players are equal in skill.

1. Map.
Combination of map & factions (matchup).
Ex.: Silverhead Crossing & Drakes, or Silverhead Crossing & Knalgan vs Elfs.

2. Matchup.
Probably this should be added to statement#1, but I want to distinguish it for easy explanation.
Ex.: Matchup not always bad for drakes vs undead, it very depends on map (small maps favor undead).

3. Position.
Combination of your position (as P1 or P2) & your faction.
Ex.: P1 orc rushing at Freelands.

4. Leader.
Ex.: mage or marksman could boost your attack with removing defending units from near villages, also would boost your defense by easy retaking your vills with fast enemy units killing.

5. Traits.
Unit traits overall and specially important traits in current matchup.
Ex.: Buying all game only high-HP elfs fighters or saurians.

6. Important battle moment.
Luck in important moments of battle.
Ex.: Killing unit in village - no heal for him next turn.
Ex.: Assassin miss and killed next turn - total disaster: your spent 17 gold for not even nothing, but only for feeding XP for your enemy.
Ex.: You very need only 1 last important blow and did not get it.

7. Overall combat luck.
* Overall continuous luck during one battle or all battles.
It simply means you would lose and no matter how skilled you are. The only possible way to win here - if you was lucky in st.#6.
* Unluck in faction's ToD and luck when it's not faction's ToD.
Ex.: orcs miss at night when they stronger, but hit at day. Math is OK, reality is NOT OK.

8. High-DMG hit.
Hit high-DMG attacks and miss low-DMG attack.
According to math - it's summary totally OK, no luck. But in real life it's matter!
Ex.: Your leader do tremendous amount of DMG by blowing all hits in sequence and next your unit miss. Math is almost OK, reality is NOT OK.

9. Killing.
Killing unit.
It's important to kill unit, not only harm (except if you intentionally don't want to kill badly wounded enemy unit to make him heal and for a long time eating upkeep).
Also here - killing 1 enemy unit with a lot of HP and good defense with only 1 your unit (1-by-1 kill).
Ex.: killing fencer at 70 with only 1 marksman. Or killing fencer at 60 with 1 clasher.

Some statements could have variants.
Ex.: Statement #5 "Unit traits" - there could be next variations:
* both players units have equal traits (both normal/better/bad);
* yours units have better traits & enemy units have normal traits;
* yours units have better traits & enemy units have bad traits;
* enemy units have better traits & yours units have normal traits;
* enemy units have better traits & yours units have bad traits.

[5dPZ]

Hi there,
Played this game for almost 20 years, regarding on RNG in Wesnoth, I think I have reached the final stage of any sort of relationship - acceptance.

If you search my earlier posts that are >10 yrs old, you might find a lot of post of me, complaining about RNG in this game.

But now, I just accept it. Accept the fact that for some games, they are meant to be your loss. It is OK, RNG makes you lose, then you lose, not your fault, not RNG's fault, it is the nature of RNG.

I think the the most important assumption in this whole discussion is:

Assuming that both players are equal in skill.

Ok, there are two possibilities:
When you see that both players are not equal, for example, you can clear see your skill is better than your opponents (this is often the case, because when you are the worse player, you can't even tell your opponent is better than you because you lack that understanding). CAN you make up some of the bad RNG with the skill difference? If you can, then this is good, that means you just proved you are the better player that could win a game despite of some bad RNG. If you can't, then, back to acceptance - this is just game that is meant to lose, even to a less-skilled player. He can make some stupid move, like take a trap village and get 6-sided attacks and survive with good RNG, and level up, etc; which makes you lose the game that you know you should have won, but it is all fine, this is Wesnoth, this sort of sh*t happens once a while with RNG - accepted.

Now, let's go back to your assumption. Let's say that both players are truly "equal in skill".
So now what? IF everything is perfectly balanced, and RNG is totally fair, with both players at completely equal skill - how does this game end? Is it gonna be a permanent stall with both players staring each other with an army? Do you want a game like this?
In such case, RNG is necessary - when you two are equal in skill, you need a tie breaker - that is RNG, a coin flip, to decide who will win this game - this is a good thing. And again you should definitely accept this, I mean, how would you think, would be better to decide a equal-skilled game, if players skills are truly equal? You need this coin, you need RNG, to end this torture of brain cell killing stall for both of you.

For a very real example - I just played a 1 vs 1 game yesterday. I played some turns of units positioning, etc for the first 6 turns or so. I can tell my opponent is a decent player, no play flaw detected. Then we engaged in battle - first turn, I got +60% damage taken, my ZOC line got broken and village taken. Then, for the entire 3 or 4 turns of fight, I got -30% dmg dealt, +10% dmg taken the whole game - my army got wiped, villlages taken etc. The game is decided. I said GG and left. Sure, it was not a pleasant experience, but I know in my heart that my opponent is a decent player, he would deserve this victory with a +40% overall RNG in favor - I would get this easy win ANY game, too, had I had that sort of RNG - The game is "fair", it is just decided that this game should be his win, instead of mine.

Acceptance, very important final stage for any sort of relationship, Wesnoth included.

[Yomar]

Well yes, skill counts, but nothing beats luck (or bad luck, depends on wich side you are) even if some players deny it, if you are a skilled player you often win, but once in a while (and maybe more), some games are so floaded with bad luck and in turn good luck for your enemy that even a clearly better player can lose against a worser one, and luck factor's importance increases with the skill of your enemy, better your adversary plays and less bad luck is needed to lose the game, also because you have less errors to exploit.
Then think of this, the good thing aboout RNG is that in this way, you never know if you will win, it would be boring to becthe best Wesnoth player and always win, at least in this way there is always the incognity of what will happen during the match.
Besides, at the end of an very unlucky game, you can say to yourself "Yah I played well, but with this kind of luck, there was no way of winning."

[Hejnewar]

Well yes, skill counts, but nothing beats luck.

No. Skill beats luck most of the time in wesnoth. There are games where the player who won took 25% more damage and dealt 50% less damage and still won without much trouble against a fairly average player. And skill can take you even further. Knowing how to play when you have bad luck is something that splits very good and amazing players and when two amazing players face each other games will usually take much longer just because luck loses its effectiveness, both of them know how to deal with it. Skill in wesnoth goes much much deeper than most players realise it does.

[5dPZ]

when two amazing players face each other games will usually take much longer just because luck loses its effectiveness, both of them know how to deal with it

I do not fully agree with this. Luck affect the game equally as long as two players have about equal skill levels - no matter both of them are newbs, average, or amazing players.

In your saying - two amazing players have longer games - only because RNG shifts side - it favors player A first, then B, then A, then B. Sure, you will have long games like this. But if RNG is consistently unbalanced, then the game would still be a short one no matter how amazing the two players are - don't forget amazing player can also take adventage of good RNG and snowball small advantage into game-winning factors.

Like I said, if, over 10 turns of game, if the RNG is >30% shift, then the game will be decided, no matter how amazing you are, given that your opponent is as good as you.

[Hejnewar]

That would be the case if comeback mechanics didn't exist in wesnoth, but the do and so, these players have the ability to abuse time and difference in units to regain what was lost at least in some part thus gaining more opportunities for possible favourable exchanges. 

There are also anti snowball mechanics in wesnoth, that prevent winning player from snowballing too much. That is why even if they can take advantage, game will take much longer, because the winning side will be crippled and losing will have advantage. You seem to completely ignore this fact and look only at luck and skill of a player. 

30% shift is a massive shift. It is really not seen that often and the longer game is the less likely it is to be the deciding factor. 

[Boy]

* I am not complaining at luck, also I don't call to change anything.

* I just want to show HOW IS FRAGILE COMPETITIVE GAMEPLAY.

* I want to show people, that PERCENTS are bad measure. Never use percents while describing luck because it would be in most times a LIE (or half-true). If want to describe duel game, describe all info as I pointed in my statements. And still if you want to use percents - use raw HitPoints (damage) instead - it would be a way MORE INFORMATIVE than percents!

[max_torch]

* I want to show people, that PERCENTS are bad measure. Never use percents while describing luck because it would be in most times a LIE (or half-true). If want to describe duel game, describe all info as I pointed in my statements. And still if you want to use percents - use raw HitPoints (damage) instead - it would be a way MORE INFORMATIVE than percents!

If you want to see a value that is not a percent to describe luck then maybe you would be satisfied with looking at Expected Value (EV) instead of probability.

[beetlenaut]

Back when competitive play was new, and balance was being evaluated, there were many posts about the effect of luck. (Though I'm not going to bother digging them up.) The consensus of ladder players in those threads was that luck made the worse player win in about 1 out of 20 games. So it happens occasionally, but if you think that bad luck is making you lose once or twice a day, then it's not the RNG at fault!

In my experience, the people who make this argument often take risks they shouldn't. For example, if you put an important unit in a key position with a 20% chance of getting killed, it's not the fault of the RNG if that happens and you lose. You should expect the 20% chance to happen four or five times per game. If you needed to take that risk, then you were not playing as well as you thought.

This is a surprisingly deep game, and it takes much longer to get good at it than most people think. Like 5dPZ said, it can be hard to notice the many tiny things your opponent did better than you. It may seem that you got unlucky when they just outplayed you. I agree with the OP that percents don't tell the whole story, but I don't agree that the game is fragile.

[Boy]

it's not the RNG at fault!

First of all, it seems that you are talking with your own thoughts )
Again, I am not complaining at luck.
I saw a lot of s***it, there is no need to explain for me anything )
And yes, you are talking right things in general.

>it's not the RNG at fault
Imagine a elf-orc game, where all elf fighters are 38/39HP. Now imagine that you are default orc with starting grunts. Next step - imagine that your grunts could not hit at 1st night.

>it's not the RNG at fault
Imagine another situation - "70 def curse". When almost all your orc army, including leader (6 units in total) attacking 1 footie in your vill for 2 turns and did not kill him. While another enemy units taking initiative at battlefield.

This examples for better understanding, ofc they are rare, but they are real.

[beetlenaut]

I said, "If." You left that word out of the quote, and it was important. If you are having bad luck in more than 1 out of 20 games, then it's not the RNG. That part was not my own thought either. The rest was of course.

If your first example actually happened to you, it would be one of the games you might lose by bad luck. I agreed that they exist. However, this should only happen in about 1 out of 3300 games*. It's too rare to prove that the game is fragile, and the fragility was the part I didn't agree with.

In your second example, attacking a footpad in a village with orcs is the kind of risk I was talking about. Failing to counter your opponent's other units will cause you to lose--not the RNG. (If the footpad can be killed with one hit, it might be worth trying with one or two units, but using most of your army for two turns is poor luck management.)

*I assumed the elves were on defenses of 40%, 50%, 60%, and 60%, which seemed typical.

[Boy]

It's too rare...

Both examples from tournament! games.

Rather pointless to battle at forum )