Life In 19x19

Kirby wrote:

Toge wrote:Why is the question relevant? It's the same game no matter what rank you are.

I am starting to agree more and more with Toge's sentiment here. It is the same game no matter what rank you are.

I highly disagree with this.

One of the things I've found when straddling the line between two ranks is that the higher rank is more likely punish small mistakes than the lower rank. In other words, you have work out these kinks to make it in the higher rank, whereas complacency is forgiven in the lower rank.

If your goal is to improve, or to play the best you can, then given an option between the two ranks, there's really no question as to which would be more desirable.

LocoRon wrote:...
I highly disagree with this.

One of the things I've found when straddling the line between two ranks is that the higher rank is more likely punish small mistakes than the lower rank. In other words, you have work out these kinks to make it in the higher rank, whereas complacency is forgiven in the lower rank.

If your goal is to improve, or to play the best you can, then given an option between the two ranks, there's really no question as to which would be more desirable.

I agree that it takes longer to improve if you are playing a weaker player. These days, though, I am trying to play every day. So if I play a weaker player, sure, they might not punish me for small mistakes. But I should definitely win, then.

Like I said, I'm on a bit of a winning streak right now, and I expect it to continue if I truly am stronger than the rank that I have right now. So eventually, I will get to the higher rank, and those that are stronger can punish my mistakes.

In the meantime, I can refine my current skills by ensuring wins against those that don't punish my mistakes well.

Yeah, it takes longer than if I was simply bumped up a rank, and could play against stronger players more easily. But I don't think it's a waste to practice winning against those that don't punish my mistakes.

Kirby wrote:I've won 12 of my last 13 games on KGS on a relatively new account, but I still haven't ranked up.

There's just not enough data here -- were these games all even games against the same ranked players, or or properly handicapped games against stronger/weaker players? Were they the same player, or all different?

If all even or properly handicapped, then your probability of winning hasn't changed. If you flip a coin 13 time and get 12 heads, the probability of getting a head on the next flip is still only 50%.

If you want rank up, then you need a number of unexpected wins against stronger players (or improperly handicapped opponents). Even loses against the same settings (when combined with these wins) can still cause your rank to increase.

Recall getting an initial solid rank, you need at least 2 wins and 2 loses, because the system needs to establish both an upper and lower bounds.

Sure, there are other ways to "cheat" the system for a short time, but it will eventually level out again anyway.

When playing even games against stronger players, or adjusting the handicap against stronger/even/weaker players, you are essentially telling the system that you want to "self-promote". So now you have to prove to the system that you deserve the promotion by establishing both an upper and lower bound at that new rank.

xed_over wrote:If all even or properly handicapped, then your probability of winning hasn't changed. If you flip a coin 13 time and get 12 heads, the probability of getting a head on the next flip is still only 50%.

But that is not at all the situation. The system does not know this probability, no one knows. After 12/13 heads it is reasonable to assume that the coin is skewed, and this is what KGS does. KGS tries to understand what skew is the most likely given all the previous flips, i.e. determines which relative strengths would have given the actual game results with the maximum probability.

In particular you don't need to decrease the handicap to get promoted. If you are winning enough KGS will promote you, even if you always play with the suggested handicap.

Some reasons for not promoting are that those people were overranked, or on the lower limit of their rank, or you have played a lot recently and those 12 wins are not so much if you have other 100 games of which you lost 50.

xed_over wrote:...
There's just not enough data here -- were these games all even games against the same ranked players, or or properly handicapped games against stronger/weaker players? Were they the same player, or all different?

Yep, they were all at the appropriate handicap, and just against various (different) random players. The loss I had was also at an appropriate handicap, where I was playing a stronger player.

xed_over wrote:If all even or properly handicapped, then your probability of winning hasn't changed. If you flip a coin 13 time and get 12 heads, the probability of getting a head on the next flip is still only 50%.

Well, hopefully the probability of my winning has changed because of my studying.

xed_over wrote:Sure, there are other ways to "cheat" the system for a short time, but it will eventually level out again anyway.

Yes, it will level out eventually. If you read the other posts that I've made in this thread (even my last post), I think that you'll see that I'm agreeing with this.

xed_over wrote:When playing even games against stronger players, or adjusting the handicap against stronger/even/weaker players, you are essentially telling the system that you want to "self-promote". So now you have to prove to the system that you deserve the promotion by establishing both an upper and lower bound at that new rank.

I'm not sure why you are talking about this, because this is not what I'm doing. I have been playing properly handicapped games.