[vlad74]

WOW, Reece,

your post was #15,000 on page 600!

Great statistics, can not rep you however, done it too often in the past.

[Domainace]

Thanks, Reece. This is a keeper.

Now for the stupid question. I got the technical explanation for geographic mean. But if I had to choose one or the other, is the median or the geographic mean more reliable in this case? I mean, can anyone provide a practical example of where the difference might be vital?

The reason I ask is because I am making up a simple spreadsheet using this info by which I can quickly "score" names. If anyone wants a copy, please PM me. For the moment, I am going with median....cuz I haz hearda that 'n.

[Nsane Domain]

That's some serious info Reece.

Thank you very much!

Thanks to the others that offer useful insight and data on this thread as well.

Repped those I could.

[snoop]

Originally Posted by krx

But the higher quality names continue to come down, little by little. My view is this won't last forever - maybe another month, maybe another year, maybe another day. Then they *should* begin to rise. I would not be surprised to see the value of some of the quad prems and VCVCs being picked up now on the cheap (low xxx) double, triple, or even quadruple in value in the foreseeable future (ie, a few months to a few years). It's a gamble - so observe the standard caveat and don't invest more than you can afford (considering renewal fees, etc) and don't put all of your eggs in one basket. But I suspect it will be a good gamble for those who've done their homework here.

When things crash very rarely do they suddenly rise again. For the market to get back to where it was will take many years and many renewal payments in my view. It will take the US economy to turn around, the domain market to turn around and for speculative hot air to return to the market. Last time the domain market crashed it took two years to reach its lows and it was probably another 3 years before people thought the market was "healthy", ie a 5 year wait for anyone holding highly speculative names.
????: NamePros.com http://www.namepros.com/short-domain-discussion/358387-the-llll-com-sales-discussion-thread.html

Whatever you buy make sure you have a clear plan on how you'll get your money back. The people who are just buying in the hope the market will rise are probably the fall guys for the people still getting out, that is how it has been for the last 12 months and I can't see it changing any time soon. Don't read into the BS that the boom will return soon, the market won't bottom out until that kind of speculation is washed out of the market, and there is still a lot of speculative hot air in the LLLL.com market. Lastly don't make the mistake of buying just because something is "cheaper than a year ago", those prices of the past were bubble prices.

[tzsxc]

Originally Posted by -REECE-

Haven't quite gotten through all the letters yet but wanted to post what I've looked at so far. Will edit this post once complete with full details. Current data looks at January 2009 sales from A through U by first, second, third, and fourth letter position, looks at overall letter performance, and considers averages, medians, and geometric means. Sample sizes have been fairly consistent throughout at about 100-200 per letter. I haven't really had much of a chance to look at the data yet and do intend to spend more time analyzing it another day, however some things do jump out, such as the good performance of the letter U against weaker premiums such as F,G, and H.

What is the saying. There are lies, lies and then statistics. The biggest wrench in all of this is the small sample size and based on mostly dropped sales. Also if you notice AEIOU tend to have higher ratings. This means "U" is skewed by the whole pronounceablility premium. My 2c.

[-REECE-]

CVCVs do distort findigs a little bit (a lot when averages are taken, very little when medians are taken). One may want to discount values for vowels in the 2nd and 4th position. Nevertheless, U still does report very respectable numbers in the 1st and 3rd position where there is very little distortion (VCVC/VCCV, etc don't distort findings whatsoever when a median is taken). It would have been nice to have a larger sample size, however I took all the sales in January to make this report -- I didn't want to start taking sales from other months out of fear of further distorting findings because of different prices at those times.

While the letter position findings may be of limited use due to their sample sizes, the overall performance of individual letters across all four positions has very respectable sample sizes in the 100-200 range which should be more than sufficient to accurately report prices. Medians/Geometric means take out outliers -- CVCVs and other large pronounceable sales should have very little distortion on the findings.
????: NamePros.com http://www.namepros.com/showthread.php?t=358387

Originally Posted by tzsxc

What is the saying. There are lies, lies and then statistics. The biggest wrench in all of this is the small sample size and based on mostly dropped sales. Also if you notice AEIOU tend to have higher ratings. This means "U" is skewed by the whole pronounceablility premium. My 2c.

Not sure what you mean by that. I try to produce honest reports whenever I produce them. If you take the 900+ sales in January and do exactly what I did, I expect you'll end up with the same result. I have no hidden agenda here and have long believed (as have many) that some of the weaker premiums are no better than some of the stronger semi-premiums (the case being made here for "U" and perhaps for "W" tomorrow when I get around to crunching the rest of the data). Dropped sales are sales just like any other sale -- there's no reason they should be treated any differently. The auction environment they often occur in would suggest to me they should be given much more weight than private sales which are difficult to verify did in fact take place or sales at non-auction venues which might not accurately reflect true market demand (eg. buyer overpaid, seller sold too low). I have about 30 LLLL.coms at present -- no reason for me to have a biased opinion on whether a certain LLLL.com letter is worth more than another.
????: NamePros.com http://www.namepros.com/showthread.php?t=358387

I'm glad some people have found the info useful.

Originally Posted by Domainace

Thanks, Reece. This is a keeper. Now for the stupid question. I got the technical explanation for geographic mean. But if I had to choose one or the other, is the median or the geographic mean more reliable in this case? I mean, can anyone provide a practical example of where the difference might be vital? The reason I ask is because I am making up a simple spreadsheet using this info by which I can quickly "score" names. If anyone wants a copy, please PM me. For the moment, I am going with median....cuz I haz hearda that 'n.

Cheers Bill. That's not a stupid question -- it's a good one actually. Your best bet would probably be to try both and observe which corresponds better to real world data over the next few weeks. While they may report very different numbers in some cases for individual letter positions (due to sample size, exlained as per the above quote), they both report relatively similar findings overall (eg. You won't find either measure suggesting "Q" is better than "A" or even that "F/G/H" are better than "A"). The most accurate results in the information provided would definitely be the geometric mean of all reported sales per letter including letters in positions 1 through 4, whereas the individual letter positions are less reliable due to limited sample sizes resultant from an insufficient number of sales which have been observed. If the market stays relatively stable over the next few months, it may be possible to get a much better idea of where things are at with respect to values in certain letter positions -- the way the market has been bouncing up and down since the buyout has unfortunately made longer-term analysis difficult.

[tzsxc]

no distortion in the 3rd position? You counted youi for 10k.

anyways, you're analyzing based on dropped names. There is already a filter before you even parse the data.

I'm not saying you have agenda. Statistics are dangerous. I admire your effort to contribute to the forum. I, on the other hand, shouldn't just criticize. But i must point out that there is inherent fallacies.

[-REECE-]

10k gets weeded out in a median.

If there are say 10 numbers:

1. $10
2. $10
3. $20
4. $50
5. $100
6. $100
7. $1000
8. $1000
9. $2000
10. $10000

Median is found by taking position (N+1)/2, where N is the sample size (10 in this case). This would be position 5.5 in this case which would be calculated by taking the average of position 5 and position 6 (both $100) and a median of $100 would be obtained. So any large figures really don't distort findings so long as there are not so many large figures that it significantly distorts where the median is taken. If we replace the $10,000 by $10,000,000, it doesn't change the median (however it would seriously alter the average, hence why I recommend people don't use the average to base buying/selling decisions, however it is still an interesting piece of information to have, made slightly more realistic by capping it at $10,000)

Originally Posted by tzsxc

no distortion in the 3rd position? You counted youi for 10k. ????: NamePros.com http://www.namepros.com/showthread.php?t=358387 anyways, you're analyzing based on dropped names. There is already a filter before you even parse the data. I'm not saying you have agenda. Statistics are dangerous. I admire your effort to contribute to the forum. I, on the other hand, shouldn't just criticize. But i must point out that there is inherent fallacies.

[4LTR.org]

Thanks for posting all the very useful info Reece.

[SlimPickins]

Fantastic number crunching, reece! thanks! sp

[Domainace]

Thanks, Reece. I was really struck by the poor performance of B, and especially A, in the 4th position.

[bmugford]

Thanks for the info Reece. I would rep you if I was able to.

Brad

[DNK.it]

egmo.com | Bidding History
Winning Bid: 900 USD
Reserve met!
Winning Bidder: Bidder 11

kahy.com | Bidding History
Winning Bid: 130 USD
Reserve met!
Winning Bidder: Bidder 6

ajqa.com | Bidding History
Winning Bid: 70 USD
Reserve met!
Winning Bidder: Bidder 2

mkcf.com | Bidding History
Winning Bid: 70 USD
Reserve met!
Winning Bidder: Bidder 2

uloh.com | Bidding History
Winning Bid: 160 USD
Reserve met!
Winning Bidder: Bidder 3

qorv.com | Bidding History
Winning Bid: 70 USD
Reserve met!
Winning Bidder: Bidder 2

jsev.com | Bidding History
Winning Bid: 410 USD
Reserve met!
Winning Bidder: Bidder 3

mkcf.com | Bidding History
Winning Bid: 70 USD
Reserve met!
Winning Bidder: Bidder 2
????: NamePros.com http://www.namepros.com/showthread.php?t=358387

faru.com | Bidding History
Winning Bid: 1,500 USD
Reserve met!
Winning Bidder: Bidder 1

ftbh.com | Bidding History
Winning Bid: 191 USD
Reserve met!
Winning Bidder: Bidder 3

iyoz.com | Bidding History
Winning Bid: 120 USD
Reserve met!
Winning Bidder: Bidder 3

wgfb.com | Bidding History
Winning Bid: 151 USD
Reserve met!
Winning Bidder: Bidder 4

uoia.com | Bidding History
Winning Bid: 165 USD
Reserve met!
Winning Bidder: Bidder 3

heyg.com | Bidding History
Winning Bid: 800 USD
Reserve met!
Winning Bidder: Bidder 1

SEDO weekly sales

[krx]

Originally Posted by bmugford

Thanks for the info Reece. I would rep you if I was able to. Brad

Ditto. Too fond already...

Really fantastic piece of work, Reece. Haven't had a chance to pour over it yet but looking forward to it.

[YAYnames.com]

Can someone give me an idea on what the current/average minimum for a quad-premium LLLL.com is? Any insights appreciated.

Thanks.

[FusionSpy]

I'd say around $120-ish is min. for lower quad premium names.

[aric]

jqte.com

appraise?

[Nem0]

Originally Posted by krx

????: NamePros.com http://www.namepros.com/showthread.php?t=358387 Really fantastic piece of work, Reece. Haven't had a chance to pour over it yet but looking forward to it.

Yeah nice job, Reece. Definitely something to thoroughly go over when can find the time though.

Originally Posted by aric

jqte.com appraise?

D-P with Q that expires in 8 months, $12 max. Probably closer to $8-$10 if you sell right now. Although it does have "QT" in it and some people really seem to like those types alot.

[EG.domains]

ceky.com $303 (NJ)
cdfo.com $210 (NJ)

[WORDSWORTH]

Originally Posted by Nem0

Yeah nice job, Reece. Definitely something to thoroughly go over when can find the time though. ????: NamePros.com http://www.namepros.com/showthread.php?t=358387 D-P with Q that expires in 8 months, $12 max. Probably closer to $8-$10 if you sell right now. Although it does have "QT" in it and some people really seem to like those types alot.

Agreed. I would say around the $10 mark. My advice would be to sell now before the renewal date has too much of an impact on the price, unless you plan to hold.

[Semtex]

Today on TDNX:

khvf.com $20
nynk.com $31
uvvw.com $30
hixv.com $20
sxxk.com $42
mukg.com $115
renewals have been added

Today on Ebay:

qyet.com $35
ugfh.com $20.51
isxq.com $20.51

Today on Namejet:

luut.com $281

[Semtex]

Today on TDNX:

qblf.com $20
yjbn.com $20
reuj.com $31
kbmz.com $50
dfku.com $36

renewals have been added

Today on Namejet:

ltse.com $350
ddap.com $340
lrmm.com $140
femp.com $410
rimc.com $610
phcg.com $201
isck.com $152

.

[Nem0]

All LLLL.com sales (eBay) today:

PFVQ $8.50
NQFV $10.00
MQHV $8.50
XVSC $8.50
LWKY $10.50
XESW $15.50
WSZA $10.50
ZOCW $10.50
QWZT $8.51
YVZO $20.50
XJIH $13.52
TJCV $15.09
IYJX $8.51
FJVE $30.00
DZFR $20.25
CLTQ $23.51
FBGQ $7.52
XUQH $14.49

Alot more than usual for a Saturday.




-

[senty4love]

Rjew.com currently at ebay at 152 :-)

[anuj291]

I think its time for most of us to give up... i dont see the market going up for domains for atleast another 2 years or so...

So for all the long time investors its ok.. and for all those who wanted to hold the names for a year or so and then sell them its bad luck ...