Why am I better as white?

[negapesuo]

Based on my recent KGS games (past month), I'm displaying a bias towards winning as white.

Overall win ratio is a bit higher due to climbing, but I think the win ratio as white is still something worth noting. Could this give me any insight into my play style?

EDIT: Nevermind... it just hit me that I play my games as white when playing against people weaker than me, while I play my games as black when I play people stronger or even. It makes sense that there is a win bias for white. Feeling dumb....

[Knotwilg]

An untested hypothesis of mine is that the lower our rank, the more White will win, because of komi. Komi compensates for nearly perfect play. At the other extreme, given random play, komi will on average tilt the result into White's favor. We don't play randomly, but we don't play perfectly either, so I think you will find a deviation in White's favor in general.

[jlt]

On the other hand, weaker players make larger mistakes, and a small komi of 7.5 becomes insignificant...

I looked at my games on KGS. From March 1st to June 30, I played 68 even games at about 5k level.

As white: 22 wins, 17 losses
As black: 14 wins, 15 losses.

It seems I am doing better as white but this is not significant. If I toss a fair coin 39 times, the probability to get heads at least 22 times is about 26%.

P.S. I just realized that IGS allows to make searches on past games.

Click on your nickname > User Info > My page > Search match results and game records > Change search conditions.

I haven't played recently on IGS, but here are my stats anyway. I had 155 even games between Jan 1st 2018 and now, at levels 13k+ -- 8k (I play so sporadically on this server that I was sandbagging in the beginning of 2018).

As white: 44 wins, 28 losses
As black: 48 wins, 35 losses

[kyulearner]

As for KGS, it is a common situation.
Even for bots, they get approx 40% winrate playing stronger players, and approx 60% winrate playing weaker players.
This probably means that ranking difference in KGS is wider than is justified by the number of handicap stones.

[gowan]

An untested hypothesis of mine is that the lower our rank, the more White will win, because of komi. Komi compensates for nearly perfect play. At the other extreme, given random play, komi will on average tilt the result into White's favor. We don't play randomly, but we don't play perfectly either, so I think you will find a deviation in White's favor in general.

It is often said that weaker amateur players make so many mistakes early in the game that komi becomes meaningless until the small endgame phase. On line most games have short time controls so it is hard to count carefully; we probably have a feeling that the game is close, but I think that is likely mostly based on "eye balling" assessments of the position. If this is how things go then the komi is kind of a gift to white at the end of the game, creating a bias towards white.

[Knotwilg]

I took a random sample of 5 players on KGS and looked at their winning percentage with White and Black. The difference is W% - B%

19%
21%
20%
7%
10%

Based on "the rule of 5", I am 93.75% confident that the median of the difference in winning with White and Black lies between 7% and 21%.

Larger samples will give a better idea on the distribution but we can already say that we can't say yours (13%) is exceptional.

You are likely to be just like most of us.

[Bill Spight]

I took a random sample of 5 players on KGS and looked at their winning percentage with White and Black. The difference is W% - B%

19%
21%
20%
7%
10%

Based on "the rule of 5", I am 93.75% confident that the median of the difference in winning with White and Black lies between 7% and 21%.

Larger samples will give a better idea on the distribution but we can already say that we can't say yours (13%) is exceptional.

You are likely to be just like most of us.

Does KGS follow the traditional handicap method, which favors White? E.g., with a 1 rank difference, does the weaker player take Black, with a komi of 0.5? It is more equitable for White to give 6.5 komi or for Black to take two stones and give 6.5 komi.

[Uberdude]

Remember that White has a half stone advantage in handicap games assuming the player ratings are an integer apart, so is expected to win more than half. Also if a 4.7 kyu plays a 4.3 kyu will KGS make the stronger 4.3 kyu white with Komi or are colours random?