#14 fits rather well for the OP's argument.
#6 for jacob
#6 for jacob
The OP is not guilty of a hasty generalization. He looked at the entire situation via the graph. My either/or was pretty either/or. I mean, what options do you have other than responding to me or agreeing with me, as far as the discussion is concerned?
And you are still evading my post, the one just before you started talking about vegans and BLTs.
That would be #7 on your list.
And you are still evading my post, the one just before you started talking about vegans and BLTs.
That would be #7 on your list.
What some fail to understand is there is not a dispute with the data collected by the OP. He accurately recorded the available data he choose to interpret and made his graphs and formula, supporting his premise.
But, and it's a big but, the data collected is by no means close to all the data and information available. Where are the findings of actual players and their provisional ratings after playing in their 1st OTB events?
The OP's "formula" is based on a players established (voluntarily given in profiles) FIDE rating. A graph is made. A line is drawn through the center. Median estimates are made to predict an event (A players estimated FIDE rating in an OTB tournament, based on on-line blitz ratings no less). 3/5 minute games as compared to a 90 minute game. 1st time ratings and an eventual established rating are hardly ever the same.
Crystal ball predictions. What data is given to support predicting the future?
But, and it's a big but, the data collected is by no means close to all the data and information available. Where are the findings of actual players and their provisional ratings after playing in their 1st OTB events?
The OP's "formula" is based on a players established (voluntarily given in profiles) FIDE rating. A graph is made. A line is drawn through the center. Median estimates are made to predict an event (A players estimated FIDE rating in an OTB tournament, based on on-line blitz ratings no less). 3/5 minute games as compared to a 90 minute game. 1st time ratings and an eventual established rating are hardly ever the same.
Crystal ball predictions. What data is given to support predicting the future?
Vegans and BLT's is a very simple example of faulty logic. May have gone well over your head. Try figuring out the faulty logic in the statement.
Obviously, the formula will slowly change over time as FIDE ratings and lichess ratings inflate and deflate. But that probably isn't what you are referring to, since that change would be so slow. The OP is using inductive logic to say that because there is a strong correlation today, and ratings deflate/inflate very sloewly, chances are the strong correlation will still exist tomorrow. Is it completely foolproof? No, of course not. Nothing is, at least as far as I know. But it there is evidence for his prediction, and there is currently no evidence to contradict it.
Need I mention that you are still evading my post before the BLTs? It's post #172 in this thread.
Need I mention that you are still evading my post before the BLTs? It's post #172 in this thread.
And while we're on the topic of logical fallacies...
#2-Ad hominem: "What type of fool are you jacob531?"
#7-Evasion: I've already been over this. Please respond to post #172.
#11-Red herring: your last post draws attention away from you not responding to post #172.
#15-Straw man:"You [referring to me] demand my [referring to yourself] age." Actually, I never demanded it. I merely noted that you evaded an earlier question for no apparent reason. I am (and would have been) perfectly satisfied with the explanation that you don't want to give out personal info online. But there was no explanation when I made the comment.
But....post 172 is still my main point.
#2-Ad hominem: "What type of fool are you jacob531?"
#7-Evasion: I've already been over this. Please respond to post #172.
#11-Red herring: your last post draws attention away from you not responding to post #172.
#15-Straw man:"You [referring to me] demand my [referring to yourself] age." Actually, I never demanded it. I merely noted that you evaded an earlier question for no apparent reason. I am (and would have been) perfectly satisfied with the explanation that you don't want to give out personal info online. But there was no explanation when I made the comment.
But....post 172 is still my main point.
Jacob531 wrote: "
Obviously, the formula will slowly change over time as FIDE ratings and lichess ratings inflate and deflate. But that probably isn't what you are referring to, since that change would be so slow. The OP is using inductive logic to say that because there is a strong correlation today, and ratings deflate/inflate very sloewly, chances are the strong correlation will still exist tomorrow."
The OP has said no such thing. (Of course, a correlation can be made for the existing data for those that already have established ratings). The intent of the exercise is to "predict" a FIDE OTB rating for players who do not have one, based on data available here at Lichess.
Simply put, all the available data from many other sources was not considered, such as in practical terms, no evidence is given of results from FIDE tournaments for 1st time players. The OP is trying to invent a crystal ball to predict the future. Any such attempt should at a minimum include actual data from FIDE tournament results. His "formula" is based entirely on voluntary information given in profiles here and blitz/classical ratings. Hardly scientific. Even still, establishing "a median" of existing ratings can never be interpreted as a methodology to predict the unknown. What is known is longer games, correspondence games with no engine help, are a far better indicator than 3/10 games.
FIDE ratings given in profiles were established how long ago? Ratings increase with experience and practice. The ratings given are very likely to be much higher than their initial rating which the OP is trying to predict.
Obviously, the formula will slowly change over time as FIDE ratings and lichess ratings inflate and deflate. But that probably isn't what you are referring to, since that change would be so slow. The OP is using inductive logic to say that because there is a strong correlation today, and ratings deflate/inflate very sloewly, chances are the strong correlation will still exist tomorrow."
The OP has said no such thing. (Of course, a correlation can be made for the existing data for those that already have established ratings). The intent of the exercise is to "predict" a FIDE OTB rating for players who do not have one, based on data available here at Lichess.
Simply put, all the available data from many other sources was not considered, such as in practical terms, no evidence is given of results from FIDE tournaments for 1st time players. The OP is trying to invent a crystal ball to predict the future. Any such attempt should at a minimum include actual data from FIDE tournament results. His "formula" is based entirely on voluntary information given in profiles here and blitz/classical ratings. Hardly scientific. Even still, establishing "a median" of existing ratings can never be interpreted as a methodology to predict the unknown. What is known is longer games, correspondence games with no engine help, are a far better indicator than 3/10 games.
FIDE ratings given in profiles were established how long ago? Ratings increase with experience and practice. The ratings given are very likely to be much higher than their initial rating which the OP is trying to predict.
Post #172.
Since you are referring for the umpteenth time now to your precious post #172:
I have read it (and some previous posts in this thread) and basically you asked two questions:
1) How old is mdinnerspace?
2) What is his mathematical education?
Is that a correct characterisation so far?
If so: the answer to both of these questions will be "rather anecdotal" (to paraphrase one of the scientifically strictests minds here) as:
1) as mdinnerspace has already pointed out his age will not contribute (or take away) anything to (or from) his argument;
2) and the same goes for his mathematical education: regardless of him being Terrence Tao in disguise or never having mastered addition his argument is either accurate or wrong.
So, before he is IMHO obligated to answer, how about stating what your reason is for asking seemingly irrelevant questions.
For the record, here is my own opinion: i think OP did put in some serious effort to arrive at a complete meaningless result: there is an observably weak correlation between blitz (i count everything up to 10min/G as blitz) strength and tournament play. As i already said an example for this are Caruana and Nakamura, both about the same strength in tournament but one significantly stronger in blitz than the other.
One can of course calculate some statistical middle but because of a too big deviation this number will be meaningless when applied to a real existing person and his chess skills. For the same reason one can calculate an average height of a population but that will be meaningless if you want to know if suit A fits person X.
krasnaya
I have read it (and some previous posts in this thread) and basically you asked two questions:
1) How old is mdinnerspace?
2) What is his mathematical education?
Is that a correct characterisation so far?
If so: the answer to both of these questions will be "rather anecdotal" (to paraphrase one of the scientifically strictests minds here) as:
1) as mdinnerspace has already pointed out his age will not contribute (or take away) anything to (or from) his argument;
2) and the same goes for his mathematical education: regardless of him being Terrence Tao in disguise or never having mastered addition his argument is either accurate or wrong.
So, before he is IMHO obligated to answer, how about stating what your reason is for asking seemingly irrelevant questions.
For the record, here is my own opinion: i think OP did put in some serious effort to arrive at a complete meaningless result: there is an observably weak correlation between blitz (i count everything up to 10min/G as blitz) strength and tournament play. As i already said an example for this are Caruana and Nakamura, both about the same strength in tournament but one significantly stronger in blitz than the other.
One can of course calculate some statistical middle but because of a too big deviation this number will be meaningless when applied to a real existing person and his chess skills. For the same reason one can calculate an average height of a population but that will be meaningless if you want to know if suit A fits person X.
krasnaya
@krasnaya
Post #172 has nothing todo with age or mathematical education. In that post, @Jacob531 asked the other person to look at the original scatter plot (posted at imgur.com in the original post).
In that plot, we find that -- based on actual, systematic evidence --
that the observed relationship between self-reported OTB skills and online ratings is super strong. If you've ever done data analysis related to human subjects, you would know that the kind of strong association we observe between FIDE and Lichess ratings hardly ever occurs by luck (or at all) in the social sciences.
You have presented no evidence to support your claim that blitz and tournament play skills only show a "weak correlation". In contrast, all the available evidence suggests that the relationship is in fact very strong (with the usual "these are self-reports" caveat).
Post #172 has nothing todo with age or mathematical education. In that post, @Jacob531 asked the other person to look at the original scatter plot (posted at imgur.com in the original post).
In that plot, we find that -- based on actual, systematic evidence --
that the observed relationship between self-reported OTB skills and online ratings is super strong. If you've ever done data analysis related to human subjects, you would know that the kind of strong association we observe between FIDE and Lichess ratings hardly ever occurs by luck (or at all) in the social sciences.
You have presented no evidence to support your claim that blitz and tournament play skills only show a "weak correlation". In contrast, all the available evidence suggests that the relationship is in fact very strong (with the usual "these are self-reports" caveat).
This topic has been archived and can no longer be replied to.