Haha I understand the hatred, but I will defend the merits of Fahrenheit 'til I die.
0 and 100 correspond roughly to the limits of life (in fact 100 was based off the internal temp of cows if I am not mistaken- Fahrenheit was a cow farmer)
higher resolution (shush i still haven’t figured out how to use fractions and decimals)
Easier to divide into 3s and 2s. There are 180 degrees between 32⁰ (freezing) and 212⁰ (boiling). 180 is just a great number for dividing because it’s got basically all the prime factors I care about.
I realize none of these points will convince anyone that Fahrenheit is better, but I hope you will realize the choice isn’t terrible. And when I want a sane and scientifically relevant system, I will use Kelvin!
EDIT: Not sure where I heard the cow farmer thing. Fahrenheit apparently came from a family of merchants. I’m a bumpkin from a farm state haha. The real anchor point was supposed to be putting human body temp at (or near) 96. Interestingly, cows are awfully close to 100…
hahaha good one @benjito
I thought 100F was chosen for the temperature at which his wife had a fever, but I might be completely wrong about it, don’t quote me on that
I see your point for “higher resolution” but still…
When I lived in USA (for a brief period) I learnt to live in miles and ounces and whatnot… but I was unable to live in Fahrenheit…
No kidding. I live in two system household, and temperature conversions are miserable. I think it mostly has to do with the fact that there’s a shift (i.e b in mx+b is non-zero).
The I grew up with them and I still have to look up how many feet in a mile. Dont get me started on cups/quarts/gallons. But I’m willing to stick up for Fahr because there’s just one unit to worry about in that case.
Okay, so I would rank at 9k if all the games I had played had been on 19x19 boards. I understand,.
But that makes no sense! Almost all the games I have played here have been on 9x9 boards.
You cannot compute a ranking from 9x9 games and certify it as a plain “ranking”, implying that it is valid for 19x19 games. It is like calling an apple an orange.
It is very hard for me to shut up, because ignoring the extreme difference between 9x9 and 19x19 handicap stones makes no sense.
So, what do I believe is the correct way to change rankings? Obviously, to compute separate rankings for each board size, as I mentioned earlier. Seems an easy solution to me. Would give me a fair ranking for 9x9 and a question mark for 19x19, which is honest and accurate.
I don’t expect a ranking comparable to AGA or EGF ratings when I don’t play any 19x19 games!
No, that is not fully clear. Maybe you would not play that good on 19x19. But maybe you would. Anyway, the algorithm can still calculate a reliable rank for 9x9 for you, and it uses that to calculate your overall rank, combined with your less reliable 19x19 rank.
Well, the system needs to give you some rank to ensure that you as well als the people that play against you get a fair game. That’s the main point of ranks, you know? And assuming you are 9k when you play 19x19 is fairer than just assuming you’re a newbie on OGS, which you are not.
I can assure you, I would play very poorly on a 19x19 board, due to the unfamiliarity with the differences in strategy. You have failed to show that the OGS rank is “reliable” for predicting 19x19 play, as you claimed.
It is worse than that. Because OGS treats 9x9 games as 19x19 games, which is an error, the OGS rank is even more unreliable. This is the real problem. As I already said above, rankings should be separately listed for each major board size. So far, everyone here has ignored that important point, making the false assumption that 9x9 play predicts 19x19 play. Not at all: the handicap stones have dramatically different values relative to winning a game, and the strategies are different.
As for needing a rating (implied: for 19x19 play), that is untrue. If your software permits it for those having “?” ranking, I would simply set my own rank (as is possible on other systems), so I might set it at “17k?” and play unranked games for awhile, or leave my rank unknown and choose the rank of players I would like a game with. The discussion in this paragraph is irrelevant to me, since I probably will never have the time to play a significant number of 19x19 games. either here or elsewhere.
I’m unsure about getting involved at this point but I’m also not clear about why this is such an issue.
It seems that you are happy to play 9x9 on OGS - in which case matching with other 9ks here will give you even games more or less and everyone can be happy.
If anyone asks you how strong you are irl you can just say “well I’m 9k on ogs only playing 9x9” and again everyone will be happy.
if you ever want to find out how strong you actually are on 19x19, you can just play a few games on 19x19 and see what happens - you’ll probably find that your 9x9 experience serves you well. and if you don’t then that’s also fine.
And if you want to
then you can go ahead and do that on ogs anyway.
I’ll leave others to go over again the point that ogs doesn’t treat 9x9 as if it were 19x19 (handicap is scaled differently between the two, etc) but I hope you can see that if you want to find even 9x9 games then you are good to go as you and and with the ranking system as it is - all your new 9k opponents also used to be 14k last month so you are still likely playing the same even games against the same opponents as you were before.
Or turn on “hide ranks” and chill!
[as an aside, it’s not completely farfetched that the system can predict your 19x19 rank from only playing 9x9 games - your opponents likely play both board sizes and if you can beat then on one size chances are that you could given them a good game on the other board size - sure it might take a few games to get into the swing of things but it’s not like playing a completely different game. Your 9x9 skillz will still apply in 19x19 games.
OGS already tried separate ranks. (If I remember correctly for blitz/live/correspondence), it only leads to outdated rank for everyone. People often reached high sdk on live, then sandbagged on blitz/correspondence against their own will and so on …
In most cases their rank on settings which they use more rarely was VERY far from real.
When you reach high level on 9x9, your strength on 19x19 becomes higher too, its inevitable.
Difference in strength because board size is much smaller than difference between ranks that will be because outdated separate ranks.
So using overall ranks only leads to higher accuracy overall.
Your conclusion follows from erroneous premises. Think about it: in the middle of the game, a single mistake on 9x9 can easily lose the game: one mistake. In a 19x19 board, a single mistake is far less likely to have global implications, so you’d have to make a few mistakes to lose a game. Of course, I’m using intuition here, since we don’t as yet have an easy way to reason about go on different board sizes.
Computing a 19x19 rating using 9x9 statistics still makes no sense. I single handicap stone on a 9x9 board equals several stones on a 19x19 board.
OGS should compute separate ratings for separate board sizes. This is a far bigger issue than blitz vs. correspondence, which only deals with how quick we think, not with a fundamental difference in board complexity and strategy.
And to be clear, we all agree on this. You, me, OGS, ranking algorithm and system, etc, etc.
If you, as a 9k, play a 14k on a 9x9 handicap game, you will not give them 5 stones but rather some other amount of handicap according to whatever the scaling is (cleverer people than me will know)
[edit: about 1 stone and some Komi or 2 stones and some other Komi? At a guess]
Of course we take care of this. On 9x9 handicap is a mixture of handicap stones and Komi adjustments. The rule of 1 rank = 1 stone is only applied on 19x19 games. For smaller boards a similar (adjusted) rule is used.
Unintuitively, even for anoek, anoek found that when many users requested an overall rank instead of the split ranks we used to have, that the overall rank is actually a better predictor for specific board sizes than the rank generated just for that size. I guess the system is so clever that extra data is more useful and generates better accuracy than our attempts to segment it.
