Split from this thread: https://greaterthangames.com/forum/topic/the-best-heroes-a-statistical-community-effort-2798
Wouldn't changing numbers to fit pre-concieved notions be the exact antithesis to the purpose of statistics in the first place?
If the numbers are contradicting a common-sense assumption, it suggests that something is wrong with the method of number-gathering.
"Common sense isn't so common."
That doesn't mean it's inaccurate, only that too few people possess it.
Common sense is only useful if it actually reflects reality. Preferring common sense over actual data is foolish.
Data is only useful if it actually reflects reality. Preferring data over actual common sense is foolish.
You can collect data that supports a point, or collect data that refutes a point, and in either case just not be collecting other data which might throw off the results you're trying to "prove". Studies and statistics are extremely easy to bias.
When your expectations do not match your results, then either your expectations are wrong, or your model is. Assuming that either one is correct is foolish. Find why there is a descrepancy (or at least why there might be a descrepancy), and make your decision based on that.
For example: dclietz expected AZ to improve in H=5 games more than he did because he benfits from pumping. One possible explaination is that there are few enough pumping heroes that 5H doesn't significantly improve AZ's chances of being pumped. So now you can do math to investigate whether this is likely, or if the source of the discrepancy is elsewhere.
Insanity: doing the same thing over and over again and expecting different results.
Insanity: doing the same thing over and over again and expecting different results.
Insanity: doing the same thing over and over again and expecting different results. ~ Albert Einstein
Ooh, that's weird. :D
Wow... That was awesome, I seen each step of the splitting topic process. First I saw the new topic in the "Rules and Gameplay" forum, it was called "Envisioner Wrote:", then I saw the name change and the edit to the original post to add in where it is from. Then I saw it appear in "Off Topic". That was awesome!
I am glad that you got to be a behind-the-scenes witness to the thread-splitting problem in all its glory!
I think Paul's gone mad with his new power!
I have never thought that was an especially good definition of insanity. It doesn't really fit, say, a paranoid schizophrenic, for instance.
Depends on the cause and the effect, but when you're discussing statistics, it is more than apt.
It is a poor definition as there is plenty of insanity that does not involve doing the same thing over and over again, even in statistics.
It is not comprehensive and should be reversed to say:
Doing the same thing over and over again while expecting different results is a type of insanity.
That isn't catchy though, so we don't say it.
In reality I'd go with it being obsessive or stupid, those make more sense with the definition than insanity does.
Yeah, but Einstein is however long dead now?
It's not even a particularly apt description of insanity by itself. There are many situations where doing the exact same thing over and over will result in a variety of outcomes. A slot machine, for example, or a shooting range. (And there's always quantum mechanics.)
It's an appeal to authority, but Einstein wasn't actually an authority on mental illness. But he was a really smart dude, so everything he ever said must be true, right?
The entire field of statistics is based around the premise that repeating an action over and over will give you different results - but those results are predictable different.
Make of that what you will.
But everything falls within the mean. If it doesn't it's an indication that something has changed.
Jimmy, a slot machine is, in and of itself, an X factor in the equation. As for using a shooting range for an analogy, the person shooting should be improving if they do it often enough, and the gun is not in the same exact position with each shot. Statistics tries to minimize variables, or at least track them.
Oh no, statistics is great. I was mostly commenting on how Einstein's insanity definition is flawed for a variety of reasons. In the shooting range example, for instance, when guns are tested for accuracy they're bolted to the floor and fired mechanically, and they rarely ever put bullets in the same location even when the human factor is completely removed. Ballistics is a lot like stats, actually, in that the goal is precision but it's not a goal that can truly be achieved, only approximated.
I would imagine that about the same level of technology (measuirng the position and speed of individual air molecules and processing all that data fast enough to be relevant) would enable both firing every bullet exactly on-target and flawlessly predicting the outcome of a dice throw. In either case, I consider the chaos-theory mathematicians who call such measurements impossible to be quitters.
True randomness is impossible in a clockwork cosmos. That's part of why I find the idea so seductive in fiction.
I'm not familiar with the term "clockwork cosmos". Would you be so kind as to expound upon your meaning?
It means that all reality operates with the precision of an incredibly vast and cosmic machine, with every atom a "gear" in this vast "clock", so that nothing is ever truly random, but rather the deterministic effect of a hugely long and complex chain of cause and effect.
Ah.
I've often wondered about that, actually. Most randomness is like that of course, but I've wondered if there is true randomness at a quantum level. I can't really see how we would find out, though, besides taking two identical universes and running them to see if they diverged (which seems rather impractical to me). Or figuring out how all the quantum stuff actually works.
I think we just diverged. Again.
I'm not sure. Statistics defined by the nature of randomness, so discussing the nature of randomness isn't necessarily out of place in a statistics thread.
That being said, if we devolve into discussing whether our topic is on topic then we will definately be off topic, and then we will have to be moved to another thread which is solely focused on discussing whether this thread is on topic.
Quantum Mechanics might be interpreted as suggesting that true randomness is indeed possible. Then again, quantum mechanics can be interpreted as saying all sorts of nonsense if you look at pop-science reports of it from a layman's perspective. I have no idea what it actually says, and enjoy playing with the obviously-not-actually-right implications in fiction, but seeing it used that way but the otherwise-amazing David Brin has definitely made me wince a bit, so I should be careful how I employ these tropes in my own work.