Computing ratings, spreadsheet?

Originally posted by Grampy Bobby
Interesting thread, Sonhouse. Here's some grist for your mill... five and past year opponent weighted, simple and to the point:

1) Wins + Draws = Non-Loss % (Draws are often both prudent and valuable. Of the three chess outcomes, only one hurts).
2) Loss%.
3) Non-Loss% divided by Loss% = Non-Loss/Loss Ratio.

* Example: Your 65% wins + 6% draws = iating realistic match-ups.



..........................................................

Thanks for adding the "Draws are often both prudent and valuable" comment.

I think ("my opinion is" ) sometimes people get very righteous about draws, and I think it is because they fail to appreciate the "big picture" aspect of chess competition, especially when the game is part of a larger competition where other games will determine to outcome.

One time Greenpawn34 and I were in a tournament, and we drew both games.

From my perspective, I walked away with a pair of draws against the highest-rated player in the tournament, and allowed myself more time to work on the other games.

From his perspective, he neutralized the #2 guy, and (from his perspective) his most dangerous threat in the competition. He basically gambled that he could rack up a better score than me against everyone else, and win.

He was right, and won it. A few righteous souls gave us a boatload of grief, but we both had valid, logical, "game-theory" reasons for what we did.

Of course, we also dogged each other about who should and should NOT have taken the draw, but the result indicates to me that GP was more correct than I was.

Sorry about the length of this, but my point is that I support the logic behind the first point- a draw is the Caissic equivalent of "live to fight another day".

Paul

Hi Paul.

To be honest there was no game plan.

I just played the postion in front of me.

I was losing in my White after missing or ignoring what was a good move,
I saw it, Ithought I had better, It was not so, so offered a draw.

And in my Black it was heading for for a real boring position
(though again you were better and stats show white does better from that thing).

As for mopping up the others. I was lost in 3 of those games but
managed to spin them.

Then suddenly I was winning and really looking forward to the final v SG.
(for a while it looked like he would not make it as he too toiling in a few games.)

Cannot recall the righteous souls having a go. (shrug),

Who were they. I'll publish their losses on the blog. 😉

You can get a free grading calculator from the Scottish website.

http://www.chessscotland.com/grading.htm

Thought I'd wait till you had worked one out for yourselves before telling you. 🙂

Originally posted by greenpawn34
You can get a free grading calculator from the Scottish website.

http://www.chessscotland.com/grading.htm

Thought I'd wait till you had worked one out for yourselves before telling you. 🙂

The ones here are the official (unofficial? ) RHP versions. 😏

(But I'll check out the Scottish one also.)

Originally posted by greenpawn34
Hi Paul.

To be honest there was no game plan.

I just played the postion in front of me.

I was losing in my White after missing or ignoring what was a good move,
I saw it, Ithought I had better, It was not so, so offered a draw.

And in my Black it was heading for for a real boring position
(though again you were better and stats show wh ding.htm

Thought I'd wait till you had worked one out for yourselves before telling you. 🙂

I remember our conversations after the fact a little differently than this, but it appears the PMs drop off after a bit. As I recall you told me you entered the tournament on a lark because it was one person short, and were looking to get through it quickly.

Perhaps you were just BSing me in the PMs, but you had rationales for your actions, and they made perfect sense to me.

I also remember you telling me you would have resigned the one game if I had declined the draw offer! 😞

The other part was that bit over in the general forum where you and Very Rusty were going back and forth, and Kewpie intervened. I got involved because VR was in full "Pharisee Righteous mode", and I got on him for busting me in public when I was in his Clan (which I left after I realized that he thinks Matthew 7:1,2 doesn't apply to him).

Edit: http://www.playtheimmortalgame.com/board/showthread.php?threadid=132445&page=1

Long story, but bottom line is that draws have a useful purpose (cf. Karpov-Kasparov 1st match marathon!)

Can't remember all the PM's but it sound like me.

I hate having a game load on but if I thought I had a
good chance in those games I would have played on.

Maybe a bit of speech play to lighten my game load and hope you
would not look too deep. Blinded by the grade. 😉

A lot of my draws are endings I'm winning but really cannot
be bothered be playing on. Sometimes you are forced to play on in the
100% wins because they won't resign. No fret.

And I resigned in the lost endings before they really became endings
So if we had played on and I saw nothing but misery then I would have jacked in.

I hate endings, what can I say.

So there was no game plan about drawing with the No.2 and I'm sure
you did not think that either. Anyway who cares?

I do recall entering because they needed one player.
I was just there for the ride.
But when I saw a chance to clash with SG. I rolled along to see if it would happen.

A good clean strong player SG. He is going through a magnificent slump.
Always thought he had too many games on his plate.
But what the heck. he's enjoying himself.
(I see he is down to 9 in an effort to halt the slide.)

Me and VR always have a tangle - no harm done.
Best to stay out it.

Edit 1:

Never knew that thread contiued after I left.
I rarely go back to rake over coals.

If you I would have just shrugged my shoulders and moved on.
Pulling out of his clan made it all seem worse than what it was..

I've never been nor expect to be in an OTB situation where I agree
a draw with a strong player to concentrate on the rest.

Here I am capable of dropping ppints all over the place and do not trust
RHP grades. So it is a risky tactic if you want to win a section.

Don't hold grudges, life's too short.

Originally posted by Paul Leggett
Thanks for adding the "Draws are often both prudent and valuable" comment.

I think ("my opinion is" ) sometimes people get very righteous about draws, and I think it is because they fail to appreciate the "big picture" aspect of chess competition, especially when the game is part of a larger competition where other games will determine to outcome.

first point- a draw is the Caissic equivalent of "live to fight another day".

Paul

27 July '10 applied this 'RHP Strength Rating', just out of curiosity, to the expanded stats of the top seven highest rated players on this site (Weyerstrass, Capa Preta, Fantal, David Tebb, Altern8, Ilywrin, Alzheimer). Weyerstrass' Non-Loss/Loss Ratio with 100% Wins and Draws skews the average well beyond 3-sigma meaningless. Midpoint 'RHP Strength Rating of 1750, however, does seem useful for the moment as a high end benchmark. Here's a current look at GP, PL and GB:


* Greenpawn34: 76% wins + 17% draws = 93% Non-Losses divided by 7% losses = 13.3 x 100 (eliminating the decimal)
= 1,329 Non-Loss/Loss Ratio.

Non-Loss/Loss Ratio = 1329 x 1560/1893 (5-year opponent/personal average rating)= 1,095 x 1447/1955 (1-year averages)
= 810 RHP Strength Rating.


* Paul Leggett: Non-Loss/Loss Ratio = 614...... RHP Strength Rating = 407.


* Grampy Bobby: Non-Loss/Loss Ratio = 400...... RHP Strength Rating = 365.



...............................................................................

Originally posted by greenpawn34
You can get a free grading calculator from the Scottish website.

http://www.chessscotland.com/grading.htm

Thought I'd wait till you had worked one out for yourselves before telling you. 🙂

For others' benefit, I looked at the Scottish grading calculator. It calculates ratings differently than RHP, so although it may be just dandy for Scottish players, I don't think it''s going to do us any good here. 🙂

GP RHP Strength Rating = 810
PL RHP Strength Rating = 407

😏

Don't look at the average grades of opponents, that means nothing.

Originally posted by greenpawn34
GP RHP Strength Rating = 810
PL RHP Strength Rating = 407

😏

Don't look at the average grades of opponents, that means nothing.

My old college roommate (Bralen) and I have played (I'm guessing) over 100 games. He's in the 1500 range.

We hooked up after a long period out of touch, and he's the guy who led me to this site.

I also have a friend on the site, Niculae, who is in the 1400 range, and we've played over 100 games I bet.

Sometimes somone on the site with an axe to grind will give me grief about my average opponent's rating, but I think they forget we play for fun. I like my old and my new friends, and we enjoy playing. And whatever points I do get, I give them right back with interest when they take me out at the knees.

Whatever the ratings, you still have to play the games!

Originally posted by greenpawn34

GP RHP Strength Rating = 810
PL RHP Strength Rating = 407

😏

Don't look at the average grades of opponents, that means nothing.

What does?

Originally posted by Paul Leggett

My old college roommate (Bralen) and I have played (I'm guessing) over 100 games. He's in the 1500 range.

We hooked up after a long period out of touch, and he's the guy who led me to this site.

I also have a friend on the site, Niculae, who is in the 1400 range, and we've played over 100 games I bet.

Sometimes somone on the site with an axe ...[text shortened]... n they take me out at the knees.

Whatever the ratings, you still have to play the games!

My experience and sentiments as well, Paul.


...........................................

"...with an axe to grind will give me grief about my average opponent's rating."

My oponents avarage are lower than yours. 🙂

Stats and figures can all be twisted, especially grades.

Originally posted by greenpawn34

"...with an axe to grind will give me grief about my average opponent's rating."

My oponents avarage are lower than yours. 🙂

Stats and figures can all be twisted, especially grades.

Conversational focus: possible improvements to the rhp site's rating system.

No other subjective issues or out of context comparisons are on the table.



.........................................

Originally posted by Mad Rook
[b]Here's my spreadsheet version. I hope it works with Excel. I guess the big difference is that I used semicolons instead of commas for separators. I probably should have started in column A instead of col B, but that's water under the bridge. I don't think it matters much.

My version is longer than Uries' version, but mine also factors in provisional oppo ...[text shortened]...

In cells E24,25,26and E29,30,31 It says #Value in all those cells. I input my rating and another rating close but nothing happened. I assume the numbers will pop into the output cells as soon as I type in the two ratings, right? Not provisional so I put nothing there. Do you know what #Value means in those cells?

Originally posted by sonhouse
In cells E24,25,26and E29,30,31 It says #Value in all those cells. I input my rating and another rating close but nothing happened. I assume the numbers will pop into the output cells as soon as I type in the two ratings, right? Not provisional so I put nothing there. Do you know what #Value means in those cells?

Yeah, the numbers should pop into the output cells as soon as you hit the return key after entering the second rating number.

#Value means there's something wrong with the formula or the formula in that cell is trying to reference a cell that contains text instead of a number.

First, check to see if you have good output numbers in these three cells:

Cell E20 - If the two ratings are close, the number in this cell should be near 0.5

Cell F24 - This cell should contain either 0.5 or 1 (Actually, if you didn't enter provisional information, then cell F24 should be 1 )

Cell G24 - This cell contains the K-factor, so the number in this cell should be either 32, 24, or 16, depending on the first rating number.


If all three of the above cells seem OK, then the problem has to be with the formula in the cells you mentioned. Note that cells E29-31 reference cells E24-26, so if there's a problem with cells E24-26, then there will probably also be a problem with cells E29-31 (which seems to be the case). Double check the formulas in these cells.