Conquer Club

Is there a way to figure out the average score for all CC players with scores above 1000 points (Since the average score of all CC players will have to be 1000)?

Add em up and divide by the amount of numbers you counted (:

write a program that will read the page, remember each score, and take the average.

thats the easiest way i can think of.

Go to the scoreboard, and find the page where 1000 turns to 999. Then take that page number, divide it by two, and then go to THAT page number. There's your average score.

Why above 1000? Any particular reason?

A slightly inaccurate but quicker way is to take samples from the user base, and find the standard deviation, and use that to generate a curve, and integrate that, and find the average value... which takes statistics and calculus, so it's probably not worth it.

Itrade wrote:Is there a way to figure out the average score for all CC players with scores above 1000 points (Since the average score of all CC players will have to be 1000)?

The average score for all CC players is not 1000

There have been several resets, so it will be about 999.

hecter wrote:Go to the scoreboard, and find the page where 1000 turns to 999. Then take that page number, divide it by two, and then go to THAT page number. There's your average score.

thats not average. thats the middle number.


Lets say, for example, we have 5 members (yeah, thats really small, but it works for the example).
each of those 5 people has their own page (again, very overly simplified)

but the scores are arranged like this:

1. 10000000000000
2. 1500
3. 1300
4. 1200
5. 999

by your explanation, you would go to page 4, because that is the first to be over 1000

by the way you explained it, you would divide 4 by 2, leaving 2. page 2 would be at 1500, but that is clearly not the average, because the insanely huge score first place has would raise the average a lot. The same thing would happen here, but with less of a huge effect.

counting all would be the only accurate way to do it.

My way would give the median average, your way would give the mean.

Didn't you take grade 6 math?

hecter wrote:My way would give the median average, your way would give the mean.

Didn't you take grade 6 math?

yes i did. I was simply responding to your answer saying that it would give the median, not the average.

The guy in the first post asked for average, which is why i left the comment i did after yours.

Itrade wrote:Is there a way to figure out the average score for all CC players with scores above 1000 points (Since the average score of all CC players will have to be 1000)?

If he wanted median, wouldn't he have said median instead of average?

hecter wrote:Go to the scoreboard, and find the page where 1000 turns to 999. Then take that page number, divide it by two, and then go to THAT page number. There's your average score.

and you said average, hecter, not median.

Itrade wrote:Is there a way to figure out the average score for all CC players with scores above 1000 points (Since the average score of all CC players will have to be 1000)?

pm sully!

Hotdoggie wrote:

Itrade wrote:Is there a way to figure out the average score for all CC players with scores above 1000 points (Since the average score of all CC players will have to be 1000)?

The average score for all CC players is not 1000

There have been several resets, so it will be about 999.

I don't recall any resets, when were they?

wcaclimbing wrote:

hecter wrote:My way would give the median average, your way would give the mean.

Didn't you take grade 6 math?

yes i did. I was simply responding to your answer saying that it would give the median, not the average.

The guy in the first post asked for average, which is why i left the comment i did after yours.

Itrade wrote:Is there a way to figure out the average score for all CC players with scores above 1000 points (Since the average score of all CC players will have to be 1000)?

If he wanted median, wouldn't he have said median instead of average?

hecter wrote:Go to the scoreboard, and find the page where 1000 turns to 999. Then take that page number, divide it by two, and then go to THAT page number. There's your average score.

and you said average, hecter, not median.

It's all averages. Median is an average, mode is an average, mean is an average. I gave him the easiest one to figure out. If he wants to go and add up all the scores than he can go right ahead. Or he could just go my way.

Hecter's way is nice and easy, and it gives us 1248, which puts me below average.

I'm still interested in the count-'em-all-and-divide-by-how-many-there-are average, though.

Itrade wrote:Hecter's way is nice and easy, and it gives us 1248, which puts me below average.

I'm still interested in the count-'em-all-and-divide-by-how-many-there-are average, though.

You could try going through page by page, and only take say, 5 values from each page. You'd still get a good 2500 values though...

Yeah, I say a few per page would be good, but take many from the top page.

By the way, congrats on winning 600 points in less than 2 minutes by sheer luck, Hecter. You owe wcaclimbing big time...

hecter wrote:It's all averages. Median is an average, mode is an average, mean is an average. I gave him the easiest one to figure out. If he wants to go and add up all the scores than he can go right ahead. Or he could just go my way.

I know this is just semantics, but I've never heard median mod or range described as an "average". They are are methods of describing statistical distributions, just like the mean is but I don't think that is the definition of average. In any case, average is usually synonymous with mean.

I can't think of a very accurate way of determining the average score other than doing a decent amount of hand calculation (or more likely, excel calculation). Since scoreboard data isn't available in spreadsheet form you have to copy and paste individual values which is going to be a big pain if you want to look at a huge data set.

But as an estimate, take the first score on each page and average them. It will be similar though certainly not exact.

I did the first and last scores of the first six pages before getting bored. The average is 2068.25, poo-maker's score is obviously having a big effect there.

If I were to do the same thing on the last six pages before 1000, I should come up with a close enough estimate of the average, right? Something that would maybe be close to the the median of 1248, perhaps?

Itrade wrote:I did the first and last scores of the first six pages before getting bored. The average is 2068.25, poo-maker's score is obviously having a big effect there.

If I were to do the same thing on the last six pages before 1000, I should come up with a close enough estimate of the average, right? Something that would maybe be close to the the median of 1248, perhaps?

Why are you doing the first and last scores? The last score on page 1 will be directly next to the first score on page 2.

Just take the first or last score from each page.

Haha, I only noticed that a few pages from the end. Oh well.

Anyway, the average I got from that is 1547. This is all terribly inconsistent. I think I might've done something wrong.

Perhaps someone can program that score-counting program and save a whole bunch of time?

I would think Sully's method would be horribly skewed for the first page, but probably reasonable for all other pages, due to outliers. The implied average score on the first page is 3255 taking an average of the top score on the first and second two pages. This is higher than the actual average of the top 25 scores on the board.

Perhaps a better estimate might come from taking an average of the median of each page? But that would probably result in too low a figure for the first page. I think the first page is the most problematic, and most likely to skew results - after that I would expect the distribution would be flatter and flatter down to 1000.

Well if we can just find the average for the first page and use it in all other calculations it would make all other results a whole bunch less skewed. There is the problem of recording down 250 scores, though...

Itrade wrote:Well if we can just find the average for the first page and use it in all other calculations it would make all other results a whole bunch less skewed. There is the problem of recording down 250 scores, though...

The average of the first page is 2485.952

Itrade wrote:

Hotdoggie wrote:

Itrade wrote:Is there a way to figure out the average score for all CC players with scores above 1000 points (Since the average score of all CC players will have to be 1000)?

The average score for all CC players is not 1000

There have been several resets, so it will be about 999.

I don't recall any resets, when were they?

I think he's talking about examples like krusher and other multies.

agreed, pm sully

sully800 wrote:

hecter wrote:It's all averages. Median is an average, mode is an average, mean is an average. I gave him the easiest one to figure out. If he wants to go and add up all the scores than he can go right ahead. Or he could just go my way.

I know this is just semantics, but I've never heard median mod or range described as an "average". They are are methods of describing statistical distributions, just like the mean is but I don't think that is the definition of average. In any case, average is usually synonymous with mean.

I can't think of a very accurate way of determining the average score other than doing a decent amount of hand calculation (or more likely, excel calculation). Since scoreboard data isn't available in spreadsheet form you have to copy and paste individual values which is going to be a big pain if you want to look at a huge data set.

But as an estimate, take the first score on each page and average them. It will be similar though certainly not exact.

there ya go