On 1/3/2024 10:17 PM, MK wrote:

I bet none of you wouldn't dare find that out. ;)

This sounds like a favorable bet to me. So if I understand
correctly, if at least one person reading this doesn't dare
to find out, then I win the bet?

---
Tim Chow

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

Tim Chow <tchow12000@yahoo.com> writes:

On Wednesday, January 3, 2024 at 10:17:25 PM UTC-5, MK wrote:

Do you think your winning and losing PR's will be the same or at
least very close to your overall PR?

Even if your "winning PR" is higher than your "losing PR," it doesn't necessarily mean that when you play with a higher PR, you're more
likely to win.

There is a nice article with some statistics done in R:

http://freerangestats.info/blog/2016/03/19/elo-pr-luck

Bottom line: I you are asked to predict the outcome of a game and may
choose whether you are give the luck rating or the PR to do so, choose
the luck rating.

Best regards

Axel

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 1/4/2024 7:59 PM, MK wrote:

On January 4, 2024 at 6:27:38 AM UTC-7, Timothy Chow wrote:

On 1/3/2024 10:17 PM, MK wrote:

I bet none of you wouldn't dare find that out. ;)

So if I understand correctly, if at least one person
reading this doesn't dare to find out, then I win the bet?

Hmm, well, I'm not sure teacher. If two negatives
make a positive, wouldn't it mean "all of you would
dare"..? ;)

Right! So you're betting that all of us would dare. If at least
one of us doesn't dare, then I win the bet.

---
Tim Chow

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 1/4/2024 8:19 PM, MK wrote:

The point here is to compare two average PR's, one for
games won, one for games lost

That's what I did. In my example, PR for games won was 8.1;
PR for games lost was 7.8.

---
Tim Chow

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 1/5/2024 4:02 PM, MK wrote:

On January 4, 2024 at 8:55:39 PM UTC-7, Timothy Chow wrote:

On 1/4/2024 8:19 PM, MK wrote:

The point here is to compare two average PR's,
one for games won, one for games lost

That's what I did. In my example, PR for games
won was 8.1; PR for games lost was 7.8.

Okay, but you also added a bunch of other stuff to
it and claimed that it proved something that it didn't.

Specifically, which sentence that I wrote do you say is wrong?

---
Tim Chow

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 1/7/2024 5:08 AM, MK wrote:

On January 6, 2024 at 2:23:25 PM UTC-7, Timothy Chow wrote:

But does that mean I'd rather play with a PR of 9
than a PR of 3? No.

Here you switched from average PR's to a specific
pair of invidual PR's and tried to generalize but you
can't generalize based on two arbitrarily picked PR's
the same way you can do with two averages which
are themselves generalized PR's for wins and losses.

I agree that one can't generalize. But in my original post,
I did not attempt to generalize. I simply asked the question,
does that mean I'd rather play with a PR of 9 than a PR of 3?
And the correct answer to that question is no, as I said.

If I play with a PR of 9, then my winning percentage
is 8/17, which is less than 50%, but if I play with a
PR of 3, then I win 100% of the time.

If you had lost the game you played with a PR of 3,
then your average winning PR would be 8.666 and
your average winning PR would be 7.5 and so your
"winning PR would be higher than your losing PR"
by even a wider margin, but then you would say that
when you play with a PR of 3, you lose 100% of the
time. There is no end to such illogical deduction...

But in my original post, I did not make any illogical deductions.
The illogical deductions were introduced by you.

Of course this example is contrived, but it illustrates
the difference between your expected PR conditional
on winning, and your expected win rate conditional
on your PR.

Again, this is not true. PR is a result calculated based
on the luck+skill=1 fallacy.

What I said was true. The example was indeed contrived. It
also illustrated the difference that I stated.

So, if you tell me that you played 25 games and that
in N1 of them your average PR was 8.1 and in N2 of
them your average PR was 7.8, then ask me in which
set I would "expect" you to have won more, I would
say the set with the higher average PR.

Similarly, if you tell me that you played 25 games and
that you won 10 of them and that you lost 15 of them,
then ask me in which set I would "expect" your average
PR to be higher, I would say the set of 10 that you won.

If you don't confuse apples and oranges, there is no
difference between "PR conditional" and "win rate
conditional", if you can ask the questions correctly.

I won't quibble with what you say here, but none of it
contradicts what I said in my original post. It only
contradicts what you imagined I said, not what I actually
said.

---
Tim Chow

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 1/8/2024 2:12 PM, MK wrote:

On January 8, 2024 at 7:02:15 AM UTC-7, Timothy Chow wrote:

On 1/7/2024 5:08 AM, MK wrote:

On January 6, 2024 at 2:23:25 PM UTC-7, Timothy Chow wrote:

But does that mean I'd rather play with a PR of 9
than a PR of 3? No.

Here you switched from average PR's to a specific
pair of invidual PR's and tried to generalize but you
can't generalize based on two arbitrarily picked PR's
the same way you can do with two averages which
are themselves generalized PR's for wins and losses.

I agree that one can't generalize. But in my original
post, I did not attempt to generalize.

Actually, by misgeneralizing using arbitrary PR's, you
were trying to show generalizing using average PR's
was also wrong but you failed at it.

That is not what I was trying to show.

I simply asked the question, does that mean I'd
rather play with a PR of 9 than a PR of 3? And the
correct answer to that question is no, as I said.

The word "THAT" in your question referred to your
observation immediately above your question. It
was a nonsensical, stupid question with no logical
transition from what preceeded it.

The question makes sense and its answer is no.

Instead of using PR 9 and PR 3 numbers, if you had
used PR 9 and PR 6 numbers, your winning percents
would be 8/17 and 1/7, thus the answer to whether
you should better play with a PR of 9 than a PR of 6
would be "Yes".

But that is not the question I posed.

You asked a question and answered it. What would
you call that if not a "deduction", (i.e. drawing of a
conclusion by reasoning)?

I call it asking a question and answering it.

What I said was true. The example was indeed
contrived. It also illustrated the difference that
I stated.

There is no problem with it being contrived. It failed
despite being intentionally contrived to illustrate a
difference because your trying to contrast a pair of
arbitrary PR's against a pair of average PR's failed.
It's just as simple as that but you either don't make
an effort to understand, incapable of understanding
or too conceited to accept even if you understand.

It did not fail to do what I intended to do.

You said: "Of course this example is contrived, but it
illustrates the difference between your expected PR
conditional on winning, and your expected win rate
conditional on your PR."

Maybe what you don't understand is that I wasn't
arguing against what you were stating/claiming,
which may wery well be true but your example failed
to illustrate it.

I know you weren't arguing against what I was stating/claiming.
Nor was I arguing against what you were stating/claiming.
It's only in your own mind that there is such an argument.

I could have just left it at saying this much but I went
one step further to use your own example to ask the
same question, in both diresctions (i.e. PR conditional
and win rate conditional), to show you that your example
illustrated the opposite of what you had intended to.

It illustrated what I intended it to illustrate. If you misunderstood
what I intended to illustrate, and found that my example failed to
illustrate what *you* thought it intended to illustrate, then that's
your problem and not mine.

---
Tim Chow

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 1/9/2024 2:03 AM, MK wrote:

Would you care to clarify what exactly your were trying
to show? Otherwise, this is just an empty assertion.

I see that you've finally realized that you can't find any errors
in my original post, so my work is done.

---
Tim Chow

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 1/11/2024 3:46 AM, MK wrote:

On January 9, 2024 at 6:49:36 AM UTC-7, Timothy Chow wrote:

On 1/9/2024 2:03 AM, MK wrote:

Would you care to clarify what exactly your were trying
to show? Otherwise, this is just an empty assertion.

I see that you've finally realized that you can't find
any errors in my original post, so my work is done.

Whoever told you that must have fooled with you.

Wouldn't be the first time that you've fooled with me!

---
Tim Chow

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 1/11/2024 3:45 PM, Timothy Chow wrote:

On 1/11/2024 3:46 AM, MK wrote:

Whoever told you that must have fooled with you.

Wouldn't be the first time that you've fooled with me!

This time I swear that it wasn't me but maybe
you are beginning to think that you are me..?

MK

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)