LeapFish versus Estibot ? Wich is the best?

[Teddy7]

LeapFish versus Estibot ? Which is the best?

I want to know your opinion about LeapFish.com and Estibot.com

1. Which of these do you think is most accurate in DN valuation business? Why?

2. What ist your favorite Appraisal page? Why?

3. What are your best instruments to evaluate a good domain name?

 

[VURG]

1.estibot. The owner asks for errors and takes a lot of feedback, research and advice in order to get as close to accurate as possible. The site doesn't oversell the quality of the estimate as all estimates have sources of error. The estimates are better than anything else I've seen.
2.as above
3.instinct. I have a budget that I stick to and get a good idea of the bargains within that budget as they rarely come around.

Personally I don't use online appraisals.

 

[Markith]

Both of those systems have so many inaccuracies and it has become a popular practice to list a domain stating "APPRAISED AT $5k" when in reality and to a trained domainer it is worth under $100.

I use my noggin. Just being on the boards and observing the market/sales can give you a really good sense of a domains value.

good luck

 

[gazzip]

Personally I don't use online appraisals.

I don't either, they are pretty pointless IMO and can never be accurate as they can never determine the two most important things, what would a seller be willing to accept and how much an end user would be willing to pay for it !

If any automated appraisal system could ever come close to being accurate on 500 previous known domain sales figures then I might consider being slightly influenced by them, I very much doubt that will ever happen and IMO they are flogging a dead horse trying !

All they are doing is throwing numbers in the air, accurate ones...It does'nt appear that way to me - Sorry but thats just the way I see it

Estibot

Tandberg.com - automated = USD 24,436 > actual sales price = $1,500,000

SheMale.com - automated = USD 4,328 > actual sales price = $520,000

Iran.com - automated = USD 2,960,738 > actual sales price = $400,000

RealEstate.net - automated = USD 1,903,692 > actual sales price = $300,000

Campgrounds.com - automated = USD 2,235 as CampGrounds.com or
USD 968,884 as Campgrounds.com

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

LeapFish

Tandberg.com - automated = $281,274.00 > actual sales price = $1,500,000

SheMale.com - automated = $186,065.00 > actual sales price = $520,000

Iran.com - automated = $134,168.00 > actual sales price = $400,000

RealEstate.net - automated = $65,424.00 > actual sales price = $300,000

Campgrounds.com - automated = $14,541.00 > actual sales price = $175,000

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

Both of those systems have so many inaccuracies and it has become a popular practice to list a domain stating "APPRAISED AT $5k"

Yeah, It looks that way and before too long end buyers will be saying......LeapFish.com says its only worth $96 so thats all I'm willing to pay for it.....Great :tu:

If they are not accurate then all they are is a toy but will end buyers take it that way ??


.

 

[.X.]

EstiBot is still in "beta" so the work is on going right now to tweak it in, It is an excellent automated valuator imo, leapfish is so far off on the over all valuation of names, it is virtually useless imo.

 

[Mark]

1) Estibot - hands down

2) I've never put much faith in them really - But do like comparing recent sales on Estibot lately - Some are pretty close :tu:

3) I use Wordtracker as my main "tool" now - along with Google and Many years of Good and Bad Experience.
Your best Tool will always be Reading The Industry's resources out there already : Domain Forums / DNJournal.com / NameBio.com / Blogs (ricksblog.com , frankschilling.typepad.com).

 

[movingconcierge]

Machines will never be able to do certain things like read a human mind. The human mind is, after all, the ultimate determinate of what something is worth. That is why getting appraisals in a thread by other humans will always be better than an automated system.

The only exception to that would be if the maker of an automated system was able to solicit appraisals from a wide array of people in the industry and then just recall them later. It would still be inaccurate as it wouldn't be able to read the subtle nuances as other sales of similar names or other events that may impact a names value.

I do applaud the NPr who built estibot, but I look at it as a novelty and not a serious valuation tool. That's my opinion at least.


BTW: I am a professional estimator of services.

 

[accent]

The domain market is itself variable - more than any other I have seen. It is not at all liquid, with most sellers willing to wait for years, if not decades, for the right price. Other sellers try to flip in a few weeks. There are bottom feeders that buy almost anything if the price is low enough and end users who pay whatever it takes to get a particular domain, often a domain that few others have any interest in. All those get mixed into the sales records, with a large percentage never reported.

The type of analysis used in estibot, for the little I know of it, can handle such wide ranging variables. Not exactly, the random factor is too strong. Not perfectly, there is lot of noise in the input. But these same issues confront human appraisers - it sometimes happens that names appraised by humans at reg fee sell within months for $10,000. This market is that unpredictable.

 

[Brujah]

I like estibot. Keep in mind if it hasn't already been mentioned since I skimmed the thread, that estibot only accounts for the domain value in my understanding.. and not the additional values it may have due to development or related.

 

[webmasterinfo]

Both sucks soooo much.. I will never reply in a appraisal thread if someone mention LeapFish or Estibot and i suggest you as well to just ignore threads mention the links above.. They suck soooo much and not accurate at all.

 

[movingconcierge]

Please do form an opinion :bah: that's sooo funny.

Both sucks soooo much.. I will never reply in a appraisal thread if someone mention LeapFish or Estibot and i suggest you as well to just ignore threads mention the links above.. They suck soooo much and not accurate at all.

 

[estibot.com]

Thanks for pointing out the variation in the higher end domains. EstiBot was never meant to even try to estimate accurately the high end domains, rather just give a rough idea that a domain may be very valuable. In the Guide on the site it says that any valuation over 100,000 or so is just an indicator that the domain may or may not be worth hundreds of thousands if not millions.

It's doing pretty good in the lower end, which is where it's optimized more. It has flaws, for sure, but it's wrong to say it's just "throwing numbers in the air". This is a false statement, as EstiBot has been tested and there is a statistically significant correlation of EstiBot valuations and known sale prices. P<0.0001, R squared = 0.76 in a population of about 1,000 domains. So it is proven not to just throw numbers in the air, although it does make quite a bit of mistakes, and my job is to reduce the number of those obvious errors.

If you take 5 domains and don't like what you see, and at the same time demand that it is accurate with 500 consecutive domains, from a statistical viewpoint that's a self-contradictory argument and also neglect the fact that humans will not be able to accurately appraise 500 consecutive domains either.

Experienced domainers are better than EstiBot, but it may serve the less experienced ones in giving a starting point for human valuation (which is its intention). EstiBot does not claim to be a all-encompassing solution to appraising, rather just one tool you may like to use in your research. EstiBot also gives you keyword research data for your names at a glance: search popularity, search results, title popularity, anchor popularity, and backlink data. It may be useful to some.

By the way, inspired by this thread, I did some heavy testing and came up with Beta v2.01 Now many flaws again have been fixed, and it has gained in accuracy. EstiBot is still in its early phases and I will work to make it better. Plus I will try and add more keyword tools for the convenience of the user.

e.g. the domain names mentioned in this thread are now much closer:

tandberg.com sale price $1,500,000 estibot 29,143 <--- this is because backlink routine is broken at the moment
iran.com $400,000 2,984,711 <-- domain for a nation-state, surely worth more than 400k
shemale.com $520,000 510,682 (correct spelling is shemale, not she male as in above post)
realestate.net $300,000 282,331
campgrounds.com 175,000 968,884 (correct spelling is campgrounds not camp grounds)

These are nicely in the ballpark, with the exception of Tandberg, that's because my backlink checking routine is broken, and these types of domain are very hard to appraise automatically anyway.

As for iran.com - well, it's a nation-state, so the domain may well be worth 3 million next time around. Previous sale prices are not necessarily the absolute indication of current let alone potential value.

Thanks for discussing my system, this all helps me greatly during the beta testing phase, and my aim is really to make this a useful research tool, I would not even want anyone to use this tool as the sole source of information when they are buying or selling domains.

I did a little test myself to wrap up today's session. Here are about a 100 consecutive domains - the latest NameBio known sales in the order in which they appeared. They are mostly nicely in the ballpark, with some exceptions of course. Bazaar sales are hard because the sale prices are very low and obviously the buyer thinks the domains are potentially worth more. EstiBot thinks like the buyer. I'm always looking for the anomalies, and trying to find out why they happen. Many times they just happen because the domain market can be damned unpredictable

An appraisal of 500 for a known sale price of 3000 is good. An appraisal of 10,000 for a known sale price of 60,000 is good, and vice versa. These are indicators of ballpark value. Whether any given name is sold at 1,000 or 10,000 is, ultimately, anybody's guess, depends on the motivation of buyer and seller.

I challenge you to cover with your finger the Estimate and Known Sale Price and try to guess the sale price yourself before peeking. Can you beat EstiBot on 100 consecutive domains? 1 point for each domain that you get a closer estimate than EstiBot. It's not that easy....

Namebio.com latest sales 13/jun/07:
domain- estibot valuation - sale price

lebenswert.com $2,945 $3,388
vstat.com $1,269 $1,288
topexotic.com $531 $300
sharewise.com $338 $3,020
mestream.com $425 $500
iran.com $2,984,711 $400,000
connected.co.uk $73,332 $157,931
businessplan.com $741,070 $70,000
galabingo.com $64,074 $65,750
trackandfield.com $193,112 $57,000
luaus.com $6,953 $50,050
rightfit.com $3,601 $50,000
triple.com $81,355 $48,000
territory.com $53,574 $30,000
golla.com $8,749 $30,000
udbank.com $362 $25,250
pachildsupport.com $82,167 $25,250
specific.com $85,259 $25,000
answerman.com $5,163 $23,750
e-money.com $30,370 $23,000
careerexpo.com $3,833 $21,250
premios.com $22,683 $19,662
ecampaign.com $7,139 $17,750
lgd.mobi $1,398 $10,000
forexplatform.net $405 $7,877
atl.mobi $9,702 $5,500
webawards.org $527 $4,000
synchronicity.net $835 $4,000
nuclearsecurity.org $268 $3,271
saintlouis.net $5,301 $3,234
mature.info $8,454 $3,100
barker.org $904 $2,788
schoolloan.org $1,842 $2,500
edesign.net $4,634 $2,500
bro.org $3,411 $2,399
fays.net $103 $2,237
magnus.net $805 $2,000
farnell.net $241 $2,000
peoplegroup.net $287 $1,888
newliving.net $155 $1,888
educatedconsumer.net $42 $1,888
resultados.mobi $70 $1,792
sge.eu $103 $3,858
poker-games.eu Reg Fee $2,661
thecradle.co.uk $1,269 $2,000
saphosting.eu Reg Fee $1,596
kfje.com $76 $15 <---these are bazaar sales...bottom dollar
seasidelooks.com $117 $2
allculturedpearls.com $110 $1
dreamsflourish.com $127 $1
ironparkbench.com $269 $1 <---bazaar sales end
nzr.mobi $297 $250
seoa.com $475 $798
typoerror.com $948 $70
webpagedesign.info $1,273 $80
array.mobi $8,636 $250
sdv.mobi $3,024 $295
centralperk.net $78 $450
preciouslittlebabythings.com $20 $3 <----bazaar bottom dollar sales
veae.com $151 $20
indianahomes.info $55 $40
vdde.com $191 $15
malibucondos.info Reg Fee $21
pokerhome.us $64 $31
basketballdot.com $52 $5
jewishdot.com $83 $5
rewardsemail.com $184 $20 <---bazaar ends
sqe.mobi $1,119 $210
12n.com $1,203 $760
9ia.com $1,348 $1,350
exchanger.com $10,999 $10,000
horsetrainer.org $483 $388
blogposts.com $1,974 $580
neworleanspizza.com $4,732 $550
easybiller.com Reg Fee $480
icierge.com $36 $1,000
getemup.com $1,126 $10 <----bazaar bottom dollar
sublimeglow.com $591 $4
iherethat.com $548 $21
coolreport.net $32 $23
zeie.com $202 $15
zeoe.com $484 $15
zeoi.com $123 $15
smallplastics.com $532 $4
tlvz.com $448 $15 <---bazaar ends
2008.info $2,709 $608
carfinancing.info $106 $95
wxa.com $6,496 $5,719

Cheers!
Josh

 

[VURG]

Just a question. If Estibot is not designed for high end domains, than why give an estimate?

Why not give an automated response along the lines of:

Estibot estimates this domain to be worth more than $50000. Estibot specialises in domains worth less than $50,000. We recommend seeking a non-automated appraisal for an accurate valuation on this domain.

 

[user-7256]

Wow, looking good Josh!

How many domains has EstiBot appraised??

 

[.X.]

Thanks for pointing out the variation in the higher end domains. EstiBot was never meant to even try to estimate accurately the high end domains, rather just give a rough idea that a domain may be very valuable. In the Guide on the site it says that any valuation over 100,000 or so is just an indicator that the domain may or may not be worth hundreds of thousands if not millions.

It's doing pretty good in the lower end, which is where it's optimized more. It has flaws, for sure, but it's wrong to say it's just "throwing numbers in the air". This is a false statement, as EstiBot has been tested and there is a statistically significant correlation of EstiBot valuations and known sale prices. P<0.0001, R squared = 0.76 in a population of about 1,000 domains. So it is proven not to just throw numbers in the air, although it does make quite a bit of mistakes, and my job is to reduce the number of those obvious errors.

If you take 5 domains and don't like what you see, and at the same time demand that it is accurate with 500 consecutive domains, from a statistical viewpoint that's a self-contradictory argument and also neglect the fact that humans will not be able to accurately appraise 500 consecutive domains either.

Experienced domainers are better than EstiBot, but it may serve the less experienced ones in giving a starting point for human valuation (which is its intention). EstiBot does not claim to be a all-encompassing solution to appraising, rather just one tool you may like to use in your research. EstiBot also gives you keyword research data for your names at a glance: search popularity, search results, title popularity, anchor popularity, and backlink data. It may be useful to some.

By the way, inspired by this thread, I did some heavy testing and came up with Beta v2.01 Now many flaws again have been fixed, and it has gained in accuracy. EstiBot is still in its early phases and I will work to make it better. Plus I will try and add more keyword tools for the convenience of the user.

e.g. the domain names mentioned in this thread are now much closer:

tandberg.com sale price $1,500,000 estibot 29,143
iran.com $400,000 2,984,711
shemale.com $520,000 510,682 (correct spelling is shemale, not she male as in above post)
realestate.net $300,000 282,331
campgrounds.com 175,000 968,884 (correct spelling is campgrounds not camp grounds)

These are nicely in the ballpark, with the exception of Tandberg, that's just one of the types of domains that are very, very hard to appraise automatically.

As for iran.com - well, it's a nation-state, so the domain may well be worth 3 million next time around. Previous sale prices are not necessarily the absolute indication of current let alone potential value.

Thanks for discussing my system, this all helps me greatly during the beta testing phase, and my aim is really to make this a useful research tool, I would not even want anyone to use this tool as the sole source of information when they are buying or selling domains.

I did a little test myself to wrap up today's session. Here are about a 100 consecutive domains - the latest NameBio known sales in the order in which they appeared. They are mostly nicely in the ballpark, with some exceptions of course. Bazaar sales are hard because the sale prices are very low and obviously the buyer thinks the domains are potentially worth more. EstiBot thinks like the buyer. I'm always looking for the anomalies, and trying to find out why they happen. Many times they just happen because the domain market can be damned unpredictable

An appraisal of 500 for a known sale price of 3000 is good. An appraisal of 10,000 for a known sale price of 60,000 is good, and vice versa. These are indicators of ballpark value. Whether any given name is sold at 1,000 or 10,000 is, ultimately, anybody's guess, depends on the motivation of buyer and seller.

I challenge you to cover with your finger the Estimate and Known Sale Price and try to guess the sale price yourself before peeking. Can you beat EstiBot on 100 consecutive domains? 1 point for each domain that you get a closer estimate than EstiBot. It's not that easy....

Namebio.com latest sales 13/jun/07:
domain- estibot valuation - sale price

lebenswert.com $2,945 $3,388
vstat.com $1,269 $1,288
topexotic.com $531 $300
sharewise.com $338 $3,020
mestream.com $425 $500
iran.com $2,984,711 $400,000
connected.co.uk $73,332 $157,931
businessplan.com $741,070 $70,000
galabingo.com $64,074 $65,750
trackandfield.com $193,112 $57,000
luaus.com $6,953 $50,050
rightfit.com $3,601 $50,000
triple.com $81,355 $48,000
territory.com $53,574 $30,000
golla.com $8,749 $30,000
udbank.com $362 $25,250
pachildsupport.com $82,167 $25,250
specific.com $85,259 $25,000
answerman.com $5,163 $23,750
e-money.com $30,370 $23,000
careerexpo.com $3,833 $21,250
premios.com $22,683 $19,662
ecampaign.com $7,139 $17,750
lgd.mobi $1,398 $10,000
forexplatform.net $405 $7,877
atl.mobi $9,702 $5,500
webawards.org $527 $4,000
synchronicity.net $835 $4,000
nuclearsecurity.org $268 $3,271
saintlouis.net $5,301 $3,234
mature.info $8,454 $3,100
barker.org $904 $2,788
schoolloan.org $1,842 $2,500
edesign.net $4,634 $2,500
bro.org $3,411 $2,399
fays.net $103 $2,237
magnus.net $805 $2,000
farnell.net $241 $2,000
peoplegroup.net $287 $1,888
newliving.net $155 $1,888
educatedconsumer.net $42 $1,888
resultados.mobi $70 $1,792
sge.eu $103 $3,858
poker-games.eu Reg Fee $2,661
thecradle.co.uk $1,269 $2,000
saphosting.eu Reg Fee $1,596
kfje.com $76 $15 <---these are bazaar sales...bottom dollar
seasidelooks.com $117 $2
allculturedpearls.com $110 $1
dreamsflourish.com $127 $1
ironparkbench.com $269 $1 <---bazaar sales end
nzr.mobi $297 $250
seoa.com $475 $798
typoerror.com $948 $70
webpagedesign.info $1,273 $80
array.mobi $8,636 $250
sdv.mobi $3,024 $295
centralperk.net $78 $450
preciouslittlebabythings.com $20 $3 <----bazaar bottom dollar sales
veae.com $151 $20
indianahomes.info $55 $40
vdde.com $191 $15
malibucondos.info Reg Fee $21
pokerhome.us $64 $31
basketballdot.com $52 $5
jewishdot.com $83 $5
rewardsemail.com $184 $20 <---bazaar ends
sqe.mobi $1,119 $210
12n.com $1,203 $760
9ia.com $1,348 $1,350
exchanger.com $10,999 $10,000
horsetrainer.org $483 $388
blogposts.com $1,974 $580
neworleanspizza.com $4,732 $550
easybiller.com Reg Fee $480
icierge.com $36 $1,000
getemup.com $1,126 $10 <----bazaar bottom dollar
sublimeglow.com $591 $4
iherethat.com $548 $21
coolreport.net $32 $23
zeie.com $202 $15
zeoe.com $484 $15
zeoi.com $123 $15
smallplastics.com $532 $4
tlvz.com $448 $15 <---bazaar ends
2008.info $2,709 $608
carfinancing.info $106 $95
wxa.com $6,496 $5,719

Cheers!
Josh

Amen...Rock on Josh...Estibot is coming along great, I have compared recent sales for 3 days straight, Estibot is very close to the sell price VS the valuation price, I don't see it as a novelty at all, It's the real deal, And is still in beta, There is no doubt in my mind, Josh will work hard to get it as accurate as possible, He has put his heart and soul into the Estibot project, And it shows with the accuracy of the valuations.

 

[estibot.com]

Just a question. If Estibot is not designed for high end domains, than why give an estimate?

Why not give an automated response....

Valid thinking - thanks for the suggestion. I think the reason for this is twofold. 1. The results I'm getting with high end domains are still generally good enough to justify showing the appraisal value for the user's reference. 2. While sensible, the strategy you suggested would in my mind take away some of the fun. After all, while I am serious about this project and do not intend this to be just another novelty item, I cannot deny that it has some entertainment value, and I want it to be enjoyable, too.

Cheers
Josh

 

[gazzip]

Josh, I realize you have put in a lot of hard work into your project and my comments/opinions are not intended as any sort of personal attack on you (or the owner of leapfish). They are just my “opinion” and after all, opinions are like As*H***s – everyone has them ! Maybe It would have been better to have kept my opinions to myself but here they are.


I did not and am not “demanding” anything Josh – what I said was >

“If any automated appraisal system could ever come close to being accurate on 500 previous known domain sales figures then I might consider being slightly influenced by them”

Meaning > Sarcasm, its not going to happen, even if you picked a random selection of 100 names never mind 500 the results would never be “accurate” (but I guess that would depend on your definition of what you think “accurate” means.) The results would need to be consistent at any number.

If you are going to call something “accurate” then your results need to be based on something that is proven or of a standard, in the case of domain appraisals all you have to go on is previous sale prices.

If your/or any automated system cannot correctly & consistently come up with a very close match of known previous sales then it is very misleading (or just plain bullshit) to call it “accurate”.

Your site says things like >

“No other domain appraisal system, whether automated or commercial human appraisals, can match EstiBot in overall predictive value”

“Besides, EstiBot is surprisingly “accurate” for values up to $100,000.”

You signature says > Free, Instant, “Accurate” Domain Appraisal

LeapFish.com says > Our famous domain name value estimator provides an “accurate” domain appraisal!

FACT - So we have two automated appraisal systems both claiming to be “accurate” but coming up with totally different results from each other and the actual sold price of the domain which is the yardstick or standard they both use to compare its accuracy…….. hmmm.

I totally agree you would have the same problem with human appraisals, so at the end of the day an appraisal is just an "opinion" based on a number of factors that are known about or taken in to account by that individual at that moment in time.

You say "An appraisal of 500 for a known sale price of 3000 is good. An appraisal of 10,000 for a known sale price of 60,000 is good, and vice versa."

No it is not good and should not be called “accurate” not even on a statistical level.

10,000 for a known sale price of 60,000 is not accurate, not close and could even be argued that it’s not even in the ballpark.


When I said “all they are doing is throwing numbers in the air” IMO what you are trying to do by creating an “accurate” automated domain appraisal is like asking a Real Estate agent to appraise a house based on its size, number of rooms, age, condition etc but not knowing crucial factors like which country, city, town, or street it is located in = impossible.

Number crunching does not alway accurately apply to any field.

In the case of domain appraisals many of the most important factors from an automated process are all missing and always will be,you’re taking out the “human” and emotional value of both buyer and seller and also missing other important elements like mindshare / brandability, value due to income from PPC revenue, market potential / commercial possibilities and the intended use/plan for the domain by the buyer.
(and there’s probably a whole bunch of other very important factors I don’t know of)

Which is why I see it as “flogging a dead horse” – How can it ever be accurate if you leave out even one – never mind many of the important factors of why a particular domain can sell at the price it does.

I just wish people would’nt present these things as accurate when they are most definetly not. They are a novelty, a toy, a laugh and IMO should not be presented as anything else.


Call it Guestibot or Ballpark Appraisals but please don’t call it accurate.

Does it pass any of the following criteria ?

Accurate

1) Conforming exactly to fact; errorless.
2) Deviating only slightly or within acceptable limits from a standard.
3) Capable of providing a correct reading or measurement: an accurate scale.
4) Acting or performing with care and precision; meticulous: an accurate proofreader.

No
Can any automated domain appraisal system ever be accurate ? - No
Are human domain appraisals always accurate ? - The vast majority of time - No


JMO

.

 

[estibot.com]

Maybe It would have been better to have kept my opinions to myself but here they are.

Not at all, I welcome and greatly appreciate your criticism and suggestions.

First of all I'd like to say that you are correct about the term "accurate". I used this term because I wanted to convey that I have statistically tested my system against other automated systems and some commercial appraisals as well, and the fact is that EstiBot beats them when it comes to hard, statistical analysis of hundreds of domains in various price ranges.

I think statistically. If you have EstiBot valuate 500 consecutive sales and have the other systems do the same, and also 10 human domainers, in the statistical analysis of the results, I am confident that EstiBot easily beats the other automated systems and may beat many of the humans, depending of course how experienced they are in domain appraising.

But as I said, you are absolutely right that "accurate" is the wrong word for it. Your real estate example was brilliant. I have now removed all such claims from my website and my signature.

Inspired by your criticism, I have also added more guidance on the site, emphasizing that an EstiBot valuation is just one tool one may use when appraising domain names, and the EstiBot appraisal can be used as a starting point for human appraisal. This may indeed be useful to many, especially beginning domainers who are not sure where to start with their appraisals. Punch in a few names and see how they compare to each other, and also look at the keyword research data that EstiBot provides. It may be helpful.

You say "An appraisal of 500 for a known sale price of 3000 is good. An appraisal of 10,000 for a known sale price of 60,000 is good, and vice versa."

No it is not good and should not be called “accurate” not even on a statistical level.

I agree that accurate is not the word. But it can be good; it is anyone's guess whether a single domain name will go for 500 or 3000. What I'm saying is that it is an acceptable ballpark valuation, keeping in mind that in the end, valuation is never an exact science, machine or human.

Nothing can be said "on a statistical level" when discussing individual appraisals.

10,000 for a known sale price of 60,000 is not accurate, not close and could even be argued that it’s not even in the ballpark.

Accurate - no, but I think it is in the ballpark, given that domain prices range from $1 to $14,000,000 and the ultimate sale price can be surprising.

Which is why I see it as “flogging a dead horse” – How can it ever be accurate if you leave out even one – never mind many of the important factors of why a particular domain can sell at the price it does.

I see where you are coming from, but while I agree with you that it can never be consistently "accurate", I think a service like this can be useful, if it is good enough as a ballpark valuation system and keyword tool. That's why I don't see it as flogging a dead horse.

They are a novelty, a toy, a laugh and IMO should not be presented as anything else.

With the same criteria, that could be said for most human appraisals as well. While they are interesting, and some humans are very very good appraisers, for the most part they present no proven validity,and often wildly subjective. At least EstiBot has been tested to some extent and the statistical correlation is there.

I strongly disagree that EstiBot is a toy. It can be fun, and sometimes makes wild mistakes that make you laugh, but it's more than that. It is very much a work in progress, in BETA testing, and it is better in some domain categories and tlds than others, but given some time, I can make it much better in all categories.

Take a look at my most recent test. From NameBio today, 3character .coms in the descending order of sale date.

ykj.com 5,617 USD 5,100
9sl.com 360 USD 607
ahl.com 207,500 USD 95,711
12n.com 760 USD 1,205
9ia.com 1,350 USD 660
wxa.com 5,719 USD 6,819
bu0.com 260 USD 401
xy4.com 412 USD 507
1uj.com 150 USD 598
6jr.com 140 USD 997
isj.com 7,049 USD 348,795
lqk.com 6,000 USD 5,782
6aw.com 300 USD 648
iji.com 9,600 USD 8,053
cy1.com 532 USD 872
jnd.com 6,700 USD 6,802
yhh.com 5,400 USD 5,353
yg0.com 200 USD 383
pat.com 63,000 USD 144,998
1vz.com 485 USD 332
c5n.com 10,000 USD 209
w0r.com 173 USD 910
u08.com 600 USD 684
56e.com 300 USD 681
gqu.com 4,123 USD 5,340
ens.com 19,000 USD 32,366
v3x.com 2,100 USD 2,566
gxh.com 6,782 USD 7,341
9mb.com 213 USD 1,186
67d.com 186 USD 398
5lq.com 146 USD 558
4c0.com 146 USD 353
1j9.com 106 USD 559
0bv.com 146 USD 641
c08.com 160 USD 1,014
6md.com 210 USD 399
9ue.com 210 USD 400
nee.com 15,000 USD 47,199
l9d.com 142 USD 160
clk.com 17,500 USD 35,916
06f.com 175 USD 1,970
s3l.com 133 USD 287
0lj.com 100 USD 525
0lm.com 100 USD 305
qir.com 6,650 USD 14,926
ot9.com 279 USD 308
bcf.com 71,200 USD 18,795

I challenge other systems and even humans to repeat that...um...accur....ahem...predictive value

Also, you get these appraisals instantly, and for free.

It's not a toy, not just a novelty thing. It's more than that.

Gazzip, thanks for taking an interest and bothering to write solid criticism. It is only with this kind of feedback from smart domainers in this forum that I am able to constantly improve EstiBot.

It may yet surprise you

 

[Teddy7]

I opened this thread and I am positivley surprised to see the interesting comments about this apraissal system. Pesrsonaly I "discovered" estibot about two weeks ago and I am satisfied with this program. I also use Leapfish and I have found there other apraissal tools that I like. My Question, Josh, is why didn't you implement some elements from Leapfish to complete this appraisal program? It would be an excellent apraissal tool when it would include:
Archive.org
Google Search Results:
Yahoo Search Results:
MSN Search Results:
Search Engine Score:
Maybe you have already implemented some of these things. I appreciate your hard work on this program. Keep up the good work!
Oh I forget something:
It would be useful to check also the free TLD (I use eurodns.com) It helps alot in making decisions.

 

[Teddy7]

An other idea is to look if all other reg. Domain names are active or are in parking modus. If all other domain names are active, is my domain more worth. Can you think to implement this tool to?