Let's use statistics to determine how many items there are

[James Raine]

gbonehead wrote:

Eric Morton wrote:

This morning, I ran a quick computer simulation that selected multiple samples of 1,000 random items from populations of various size.

Using this method, I found that a population of 1,100 +/- 50 pre-cull items fits with the results of both my first 1,000 views and Sean McGowan's first 1,000 views. (We're the only two voter's I'm aware of so far that have tracked and posted data for our first 1,000 views.)

There's a third person doing so. That's how I can do things like create this post-narrow graph of unique vs. total items seen. :)

Granted, infinite series do funny things, but the post-narrow curve appears to be converging on a number below 800.

As a colorblind person, your graph color selection sucks.

[Thomas LeBlanc]

gbonehead wrote:

Okay, ready? I've voted 606 times and seen 573 unique items (post-25%). My current estimate of the number of items out there is 694-695.

So rounding up to 700 gives a possible 489,300 pair combinations. I wonder how many votes have been cast since the cull? I would guess more than enough for each pair comabination to be seen at least 2 times. Maybe even 3 times.

[Mark Moreland]

Thomas LeBlanc wrote:

gbonehead wrote:

Okay, ready? I've voted 606 times and seen 573 unique items (post-25%). My current estimate of the number of items out there is 694-695.

So rounding up to 700 gives a possible 489,300 pair combinations. I wonder how many votes have been cast since the cull? I would guess more than enough for each pair comabination to be seen at least 2 times. Maybe even 3 times.

Without revealing how many total combinations there were this year, I will say that as of last time we were presented this particular fact by the numb-o-crunch-o-matic (Tuesday), approximately 14% off all possible combinations had not yet been voted on.

[Thomas LeBlanc]

Mark Moreland wrote:

Thomas LeBlanc wrote:

gbonehead wrote:

Okay, ready? I've voted 606 times and seen 573 unique items (post-25%). My current estimate of the number of items out there is 694-695.

So rounding up to 700 gives a possible 489,300 pair combinations. I wonder how many votes have been cast since the cull? I would guess more than enough for each pair comabination to be seen at least 2 times. Maybe even 3 times.

Without revealing how many total combinations there were this year, I will say that as of last time we were presented this particular fact by the numb-o-crunch-o-matic (Tuesday), approximately 14% off all possible combinations had not yet been voted on.

Hmm, I remembered the voting update number incorrectly. ~200k, not the 500k i had in my noggin.

So for the first 8 days, the average votes per day were about 25k. Add another 11 days (Tuesday) multiplied by the aforementioned average and increase that another 25% (to account for the reduction of time spent on holiday matters for the first week of voting, addition of message board carrots, increased contest awareness, and first/update days not full days of voting) brings the number to ~600k. Which is about 14% less than 700k. Then working backward leads to about 840 unique items post cull and disqualifications. Pretty close to what Epic Meepo has been guessing...

[gbonehead]

James Raine wrote:

As a colorblind person, your graph color selection sucks.

Ouch.

[James Raine]

gbonehead wrote:

James Raine wrote:

As a colorblind person, your graph color selection sucks.

Ouch.

No offense intended, it's just illegible.

[Curaigh]

I can finally add to this number.
15 items repeated giving me an average of 667.

There was one item I saw twice which had almost the same name as one on my list. Not quite, but close enough that a couple of typing errors might account for the difference. If so the number would be 17 and 558.

[Eric Morton]

Hm. Not quite sure how to incorporate that last data point, given that we don't know if it's 667 or 558. My gut tells me to go with what was actually recorded (for an estimate of 667), because 558 is more than one standard deviation from the mean, and because I know for a fact there are more than 558 items.

[Curaigh]

I would use the 667. Some items are just that close in name me thinks. :)

[Eric Morton]

At least one other poster has mentioned a pair of items with near-identical names, so we'll go with 667.

Average Estimate After 17 Samples: 819 items (1092 items before the cull)

Average Estimate, Dropping Outliers: 745 items (993 items before the cull)

[Curaigh]

16 repeats for 625.
Doubt I'll make another fifty votes tomorrow though and probably mute per your other thread. How do the two compare?

[Eric Morton]

The data are converging on gbonehead's estimate...

Estimate After 18 Samples: 808 items (1,077 items before the cull)

Average Estimate, Dropping Outliers: 726 items (968 items before the cull)

[Monica Marlowe]

I love this thread, my inner biologist brings me over whenever I see new posts. You guys are awesome.

[Feros]

11 repeats for 909.

Possibly an outlier?

EDIT: Holy crap! I just made Champion!

[Eric Morton]

An estimate of 909 is well within one standard deviation of the mean.

Average Estimate After 19 Samples: 814 items (1,085 items before the cull)

Average Estimate, Dropping Outliers: 738 items (984 items before the cull)

[Thomas LeBlanc]

Feros wrote:

EDIT: Holy crap! I just made Champion!

Congrats!

[Feros]

Thomas LeBlanc wrote:

Congrats!

Thanks! How many of us are there now?

[Jon Haire]

I've seen three champion voters.