Round 1 was disgraceful...

[FML|WorpeX]

Why do you keep saying FFA is largely luck? I completely and totally disagree. If FFA was all luck we would not see the same winners every season. We would not have accurate predicons in FML and furthermore we would be seeing weaker players winning matches they shouldn't win.

I think you're just using "luck" as an excuse for your losses. Their is luck in FFA, yes, but not nearly to the degree you imply.

[rygorych]

FFA is luck when it's among players of equal skill.

FML is not really in indicator of that.. you have 4-5 absolute nuts of rice's caliber, about 10 "quite goods" and the rest are "decent". (and 1 bad, being me... and 1 horrible, being rabbit).

Could you honestly predict the outcome of S10 final? No freaking way. When you have 4 ppl with as much as skill as those guys, it comes down to LUCK.

Luck like Das being sandwiched, then running around like a headless chicken being cut off because human* expoed the wrong way. Luck like items drop, luck like a player deciding to go one way to creep and not the other, luck like a bugged creep not bugging out for once, lucky crit, etc.

Now if you put up a game like Dase, Darkness, Htrt and lets say Blooder... well, I have a fucking bridge to sell to anyone who bets on Blooder (who's quite good, but obv not the others' caliber).

[rygorych]

And I actually don't cite lack as an excuse for my losses... I call it what it is in general.

In my games, I try to minimize the impact of luck. My entire macro strat is based on as little luck as possible. And it's not entirely accurate of course for me to say that, because luck isn't really the factors out of my control (e.g. wreck losing 1v1 to duck).

So there is:

1. Your skill
2. Factors outside your control/skill
3. Luck

[KiLLyoSeLF]

It's been 3 days so I finally have to ask...wtf does purple sky mean? Is it some nerd reference I don't recognize? Or is it some like deeply deeply nerdy Warcraft reference I'd have to have played WoW or something to have learned?

Q-veta help me. Right he could be saying "pink polar bear!" a trillion times for all it means to me, because that's about as much as "purple sky" makes sense to me.

Or is purple sky some form of saying, "I suicide you when I know who you are because I'm afraid of you actually beating me in a game where I can't blame a noob for not 2v1ing you well enough to make the 2v1 work!"

[DarKNeSSCaLLs]

Luck will always be present in FFA games, but if you're saying a game is 100% based on luck, sorry I disagree with you. I've come out of really bad situations before due to clever thinking + strategy. Outsmarting/outplaying your opponents can be a key factor to victory.

[FML|Renaud]

I think hes referring to his mass chim in the sky

[rsm]

If we aren't taking spawn location into account there isn't a single factor of luck in FFA. I'm baffled at the amount of people who think there is besides the usual spawn position, or item drops. It is all skill, anybody who thinks it is luck simply is bad at ffa. I wouldn't be able to go 90% with about 50 ffa games on ladder if it was luck. OFC nowadays i don't even find at 10-0 ffa....

[rygorych]

Of course it's not a 100%

In a game with top 4 skilled players, e.g. HTRT, Dase, Maga and choose 1 more who's head+shoulders above everyone else (let's say Target in his prime), it will be very high. I'd say 60-70%. On a low-creep map, even higher due to item drops (e.g. pipe on Monsoon or Gold Rush).

I mean, think about it.

We can all pretty much agree, the players have about same micro, same macro, same manip.

So as they come into the game, their chance to win is:
HTRT 25%
Dase 25%
Maga 25%
Target 25%

Now, let's say Dase is position fucked near HTRT and is fucked on every creep he tries. His chance to win goes down to let's say 5%, while others are up to 30%, HTRT being fed  to 35%. Had dase spawned in another direction, or at least didnt feed Htrt, it could have been different, right?

Of course you can argue, HTRT is actually just an awesome player and is even better than Dase... Sure, that's possible, but then we're not in equal game after all.


So when I say luck is a #1 FFA winning factor, I'm talking about the games where everyone is of equal skill.

I honestly am not sure how you argue against this...


Now Duck what you're talking about is a long run streak, and you're absolutely right, in the long run, the effect of luck diminishes (you get to play various maps, various positions, various creep drops), and overall effect of skill increases. However when I talk about luck, I talk about a single game with even skill distribution. This is simple stats people.


And a little more for those who are math-inclined/nerrrds:

(click to show/hide)

In statistics, it is agreed that the golden number is "30" when it comes to proving consistent effects of factor X on outcome Y.

So, for example if Ebo claims to win 80% of non-noob games, he will need to play 30 non-noob games and indicate an overall winning rate of 80%.

If, for example, Ebo plays only 10 games, and wins 8 of them (also 80% winning rate) this is not yet an indicator of his being right... it could be assumed as "luck" until he plays the full 30 games, after which it could be concluded that his particular skill indeed contributes to the winning rate of 80%, rather than luck.

Furthermore, the each experiment needs to exactly the same (e.g. vs exact same players).

What all of the above basically means is that it is very hard to judge skill statistically, even based on ALL of FML seasons. We can certainly see "obvious" trends, but because almost every game is against different players, there is a huge soup of factors and maps, and thus true level of skill/luck effects can not be calculated, not even in another few years of running the league.

Fundamentally, talking about a single game rather than an array, and assuming difference in skill is not a factor, it comes down to luck (60-80% depending on map) and the variation of external factors.

 

[FML|red7z7]

we can test that hypothesis actually. put them in the same game 10-20 times and get a statistical result and check if its significant ^^


srsly though, can we? it would be great to see

[rygorych]

Absolutely, and for example if out of 40 games,

Htrt wins 12
Dase wins 10
Maga wins 10
Target wins 8

We can say that the players are approx. same skill, but Htrt is the slightly better than the rest while Dase = Maga.

Of course here is where the long-term race imba kicks in and we can either conclude that Target is the worst out of 4, or simply NE is least powerful in the long run while Hu is the most powerful.

TLDR: we don't have enough data to really say anything besides the obvious trends over the years. Let's play the game and enjoy it instead of arguing semantics.

[y.zenchenko]

I wanna show my opinoin on this topic. Out of my ffa-experience I can claim the following (talikng about b.net ot FNL games):
80% or may be 90% of games I've lost are lost because of my own mistakes and not because of someone else's luck or something else.

that's it

[FML|Renaud]

Well, I see your point, but it's questionable.
Because often (at med-high level game) position luck is unlucky, because it gets you teamed. 

but i see what you mean

[rygorych]

You're right Renaud, and then there are also external factors like:

-interpersonal conflicts among players (e.g. I'm more inclined to hit Rabbit)
-lag
-how tired/distracted a player is
-cat on the keyboard
-race hate (e.g bias vs orc/hu)

...which are not really luck, but for any given particular match could be assigned same statistical value as luck.

[FML|Renaud]

-cat on the keyboard

In the king of FFA finals, at some point a cat jump on me (not the keyboard ty god) was pretty lucky it was during some idle times!

[shiKaka]

I had the opportunity to sit down with HTRT the french free for all master. Please take a look at what he had to say about being FML champion.


me - teach me to be FFA CHAMPION!@#%
htrt - $$$$$$
me - i will teach you how to get laid!
htrt - See that's your first problem, the key to getting being FML champion is NOT getting laid.