Guess the grade #68

I got it - now!

 

Log in or Sign up to hide this ad.

An old rattler!

Finally, I'm not off by 1! LOL

 

eh? doom and I had the same pics this round... a change in ranking should be mathematically impossible unless he dropped a score this round. In fact, given his 23 vs. my 30 overall picks,,, this weeks results should have impacted his score more negatively than mine..... you using one of those voting calculators from FL ?:devil:

 

I found a bug in my program. I am not sure last week's score were counted last week. They are this week. Sorry for any confusion.

 

I've been thinking about how to reward more guesses rather than the abitrary dropping every 10 grades. I'd like to see a "number of guesses correction factor". We can take your calculated score, subtract it from 1, then multiply by the number of guesses to get the final score. That would reward people for guessing. The more guesses you have, the higher your correction factor and the higher your score. That would also even out the abstained votes better. You could not be rewarded for not guessing, but if you have the guts, you're rewarded. The formula would be:

(1-calculated score) x number of guesses = corrected score

Using the new method the standings would be:

bahabul 0.429 30 17.13
Leadfoo 0.462 28 15.06
BadThad 0.5 30 15.00
ddoomm1 0.381 23 14.24
Duke Ka 0.538 28 12.94
mark_h 0.556 29 12.88
swhuck 0.714 23 6.58
ldhair 0.455 12 6.54
Anarcho 0.765 18 4.23
tmoneye 0.8 16 3.20

 

Does not handle numbers >1 (there are lots in fact most), does not relate to scores and IMHO, over corrects for the number of guesses.

I am not at home so I don't have access to all the numbers et al. How about simply dividing the score by [cube root](# guesses / max # guesses). I would have to look at the numbers, but I think cube root would work better than square root.

 

Here is what that would give you;

Correction cort'd score

1. 0.915 0.416 ddoomm1 0.381 23
2. 1.000 0.429 bahabully 0.429 30
3. 0.977 0.473 Leadfoot 0.462 28
4. 1.000 0.500 BadThad 0.500 30
5. 0.977 0.551 Duke Kavanaugh 0.538 28
6. 0.989 0.562 mark_h 0.556 29
7. 0.737 0.618 ldhair 0.455 12
8. 0.915 0.780 swhuck 0.714 23
9. 0.843 0.907 Anarchos 0.765 18
10. 0.811 0.986 tmoneyeagles 0.800 16

 

There are so many ways to look at this.
There are those that just can't be here to vote and those that just don't know, so they don't vote.
Some are just playing a strategic game to be on top. We also have folks that join in late.
I don't have a clue if there really is a formula that is fair to all. We just need to keep it fun for everyone.
One thing is for sure, I don't belong in the top 10 with as few times as I was able to vote.
I'm good with anything rlm. It's all fun.

 

Like you say, "There are so many ways to look at this." There is not right or wrong, but infinite shades of gray.

Anywho, as it stands now, you must have 1/4 of the maximum available guesses to be counted. That actually DQ's about 80% of the people voting. Do you really think more should be DQed?

BTW, in my mind, perhaps the most important reason for running these is not who "wins", but to help those learning to start grading. They might not even partidipate, but by watching and reading, they can learn a lot.

 

64 ok i said it

 

lol it's your Guess the grade just score it how you want. It is about learning and that's it. Not who wins.

I have been looking at the Teletrade site I click the picture to maximize it and cover the top of the screen with a piece of paper. Then guess what the grade will be. I've done hundreds of these now and am getting better and better. You can do lots in one setting and pick your coin type if you want to be better at one or are worse at it and need more practice.

 

LOL....I forgot about the "also rans". OK, no problem....continue on with however you want to score it. I was just trying to dream up something different.

 

63 this time around!!