Here are some of the ones near the top:

Bog Imp (9th), 1000.00, 0.00, 1000.00, 1000.00, 1000.00, 0.00, 1
Spell Blast (6th), 250.49, 249.51, 250.49, 500.00, 0.99, 0.00, 2
Lumbering Satyr, 250.00, 0.00, 250.00, 250.00, 250.00, 0.00, 1
Framed!, 98.97, 0.02, 98.97, 98.97, 98.97, 0.00, 1
Horror of Horrors (9th), 75.00, 0.00, 75.00, 75.00, 75.00, 0.00, 1
Island (10th), 50.00, 0.00, 50.00, 50.00, 50.00, 0.00, 1
Island (RAV), 22.71, 0.00, 22.71, 22.71, 22.71, 0.00, 1
Firebreathing, 50.00, 0.00, 50.00, 50.00, 50.00, 0.00, 3 (note no version, this may be some promo?)
Wall of Shadows (CH), 50.00, 0.00, 50.00, 50.00, 50.00, 0.00, 1
Smash (RAV), 35.00, 0.00, 35.00, 35.00, 35.00, 0.00, 1

Then, if you go to the cards worth 9.99, there are a whole slew of 0.01-0.25 commons set to this price. I can only surmise "9.99" is a default of the code behind the pricelist generator if it doesn't find/never found any sales; or it could be an overflow error.

I'd like to bring this to the attention of whomever manages the programming of the pricelist generator, to find a way to catch these inaccuracies that are obvious to a human eye with some good code that will keep a lookout for these kinds of way-over-the-curve prices and replace them with a correct archived price or a "baseline" price (if the high price is the only one found in the dataset).

I keep an archive of the price files that I've downloaded for MWS, and I had to go back a year or more, but I did find old data for the examples above that was correct.

Speaking of archives, is a collection kept (say, once monthly forever and 4 once-a-week archives of the weekly prices for the last 2 months) of the historical pricelists on the server? If so, a way to search the archive by cardname to show graphical trends of pricing over time would be AMAZING.
As a seller, knowing the price trend for cards from this well-made pricelist would be quite useful!

Thank you and I hope the badly-priced cards can be fixed soon.

__________________
There's no need to fear, UNDERDOG is here!

All your Gruul Nodorogs are belong to me. Trade them to me, please!

Report rules violations.

I understand if a human reader comes across these kinds of obviously wrong prices that common sense would jump in and say Hey, however I deal with software that does not have the human element and imports raw price data from your list for every card. I'd guess many hundreds of people use some magic catalog/inventory program that has functionality to download the pricelists straight into the software, and there is no automated filtering; only when the end user happens across a bad price would he/she realize their deck's price figure is incorrect at that moment, or their collection's value is overstated. Each user finding such errors would have to edit their data.

Due to the computerized nature of the pricelists I think its reasonable to ask the list script programmer to add more code to catch these out-of-norm prices better at the source, so the data is more correct for all who download it. At least I'm bringing it to your attention

A quick take on a fix is to establish reasonable bounds for prices found. We know no card outside of the power 9, alpha, beta, and some unique promos should ever have a price over, say, $500. No cards in sets younger than 2000, unless promo/foil, are worth more than $200 (roughly, AFAIK). Something like this could be used to screen failing card prices to the list maintainer in a report, and once in a few days this list could be gone through and plugged in to a file the list generation script sees such that disapproved card prices are replaced with better ones (as long as the failing price is still present).

Another approach is as I hinted at, keep a normal baseline pricelist that has had a human sort through to fix known repeat offenders and other obviously wrong prices. The generation script would then compare its new list to the baseline, and any final prices far beyond the curve would be set to the baseline instead. Prices set this way should be sent in a report to the programmer.

In either solution, after research on why particular cards are coming up in the reports, the programmer can ascertain the pattern for the source sales and engineer code edits to better catch the outliers- like you stated, possible art prints or somesuch. The end result is much improved source data that people can better rely upon and feel good about this website, rather than question the data as a whole due to errors somewhere in 32000+ entries.

At the very least, I hope someone can pass along the particular cards I posted to the person who can edit the pricelist script, figure out the pattern, and close the filtering hole in the code that let those bad prices get through. If there are no further qualifying sales for a card, causing it to drop off the list, no price is better than a wrong price.

Thank you!

you sound like a ridic lazy seller who wants to reap and leech benefits, and do didly himself.
if you're a programmer, and thinks you know what's wrong.. why dont you deal with it yourself? I've been using the guide long enough to know what's wrong with it too.. cept I know what data I can cull, or what I NEED to cull. you seem like you really dont have a clue of anything.

©

^ translation of the above in case you missed it. Go F your lazy ass self.

__________________
Since it is obviously inconceivable that all religions can be right, the most reasonable conclusion is that they are all wrong.
-Christopher Hitchens

Above is pointing out some errors in the raw data that magictraders produces in hopes that everyone can benefit from a more accurate list, and suggesting ways to improve it with some boundary checks.

How do you read anything more than above as being helpful to the owners of this site and the price list?

After some leg work myself, Planeswlk also supplies scripts that help folks needing similar fixes. On his site with updates is a QBasic script. After some brief downloading of QBasic (free from Microsoft) and some path changes, you can generate the fixed TXT files used by MWS for importing prices.

Unfortunately it's hard to know whether the prices reported are bogus (like all the 9.99s and too-high numbers), as the script still needs to trust the data coming from MOTL. This is probably why the request is here to help fix this at the source.

quote:

---

Originally posted by digitek:
Coolio... if you did any "leg work" yourself, you'd realize Planeswlk has been helping magic trader users (and Magic Workstation users, like myself) for over 4 years now, running his own site to post fixed price information for everyone. There's another MWS specific site that has a long history of updates every couple of months since 2005 when MWS stopped working with automatic updates.

Above is pointing out some errors in the raw data that magictraders produces in hopes that everyone can benefit from a more accurate list, and suggesting ways to improve it with some boundary checks.

How do you read anything more than above as being helpful to the owners of this site and the price list?

Regards

---

i've done a lot of the legwork myself for more than that period of time. I was here when the pricelist was implemented, i remember the staff discussions with regards to it, and at the time, we knew there were gonna be issues.. so I completely disregarded any and all data outta it for the first 2 years is was available. as the loopholes and leaks were plugged slowly, i eventually looked over my analysis and compared it with the pricelist and found it was closer and closer to being a more accurate picture of the prices, i started using it more, but that doesnt mean i take it at face 100%. legwork also includes common sense.. the 9.99 raw n of 1's I'd figure someone else would eventually figure out that unsold store listings were included in the count, and as thus, should be blanket disregarded. seriously ppl.. you have a brain, use it

__________________
Since it is obviously inconceivable that all religions can be right, the most reasonable conclusion is that they are all wrong.
-Christopher Hitchens

quote:

---

Originally posted by digitek:
That post helps clear things up a lot, thanks coolio. Can we agree that some further processing could be used to reduce some of these numbers, so the custom fixes that Planeswlk is talking about wouldn't be needed as much?

---

there's the pipe delimited.. import that with excel.. sort by raw n descending, and delete all the rows with raw n 1? how hard is that? takes no more than 1-2 mins.. TOPS

__________________
Since it is obviously inconceivable that all religions can be right, the most reasonable conclusion is that they are all wrong.
-Christopher Hitchens

quote:

---

Originally posted by coolio:
there's the pipe delimited.. import that with excel.. sort by raw n descending, and delete all the rows with raw n 1? how hard is that? takes no more than 1-2 mins.. TOPS

---

you can write scripts to do this as well. I'm just not sure how...

__________________

-Caquaa
caquaa@yahoo.com

Buying MM counterspells for my counterspell collection. Click me!

That approach seems dangerous - there are plenty of valid rows which have raw-n of 1. Wiping out all those rows is over 75% of the file. In most cases the estimate is still valid, so I wouldn't want to lose that info.

Some heuristics would be great to apply here, so month after month if a price varies more than X percent, it would stand out. Maybe that combined with a Raw-N of 1 would do the trick.

If you know of a place where I could download previous price files, I'd be happy to take a look at this.

Regards

quote:

---

Originally posted by digitek:
In most cases the estimate is still valid,

---

no.... they really aren't

©

__________________
Since it is obviously inconceivable that all religions can be right, the most reasonable conclusion is that they are all wrong.
-Christopher Hitchens

Based on my findings and comparisons with online retailers and ebay auctions, I'm finding in most cases that is not true... they are fairly close to perceived value.

Furthermore, many stores use that price guide, regardless of Raw N to set their in store prices (I know 3 of our local stores do).

So there certainly thousands of stores and users that think otherwise... perhaps they aren't aware of the mystical "Raw N" rule. =)

quote:

---

Originally posted by digitek:
Even though you are saying over 7700 cards prices in the MOTL price guide are invalid...

Based on my findings and comparisons with online retailers and ebay auctions, I'm finding in most cases that is not true... they are fairly close to perceived value.

Furthermore, many stores use that price guide, regardless of Raw N to set their in store prices (I know 3 of our local stores do).

So there certainly thousands of stores and users that think otherwise... perhaps they aren't aware of the mystical "Raw N" rule. =)

---

Anyone using a single data point for anything is simply dead wrong. If that value isn't compared to anything else, then it's reliability is nil. It's also obvious that such a thing is NOT happening, or the cards pointed out in the example would all be ridiculously priced at these "many stores" that use the price guide. Obviously, they're only using such pricing when it meshes with other data. If there are in fact people stupid enough to always rely on every single price in the MOTL price guide or any other single data source, then they really have no business being in business.

__________________
There's no need to fear, UNDERDOG is here!

All your Gruul Nodorogs are belong to me. Trade them to me, please!

Report rules violations.

quote:

---

Originally posted by digitek:
Even though you are saying over 7700 cards prices in the MOTL price guide are invalid...

Based on my findings and comparisons with online retailers and ebay auctions, I'm finding in most cases that is not true... they are fairly close to perceived value.

Furthermore, many stores use that price guide, regardless of Raw N to set their in store prices (I know 3 of our local stores do).

So there certainly thousands of stores and users that think otherwise... perhaps they aren't aware of the mystical "Raw N" rule. =)

---

i grow weary of debating with someone who is utterly clueless on how certain things in the secondary market when it pertains to magic, works. believe what you wish.

©

__________________
Since it is obviously inconceivable that all religions can be right, the most reasonable conclusion is that they are all wrong.
-Christopher Hitchens

I'm not sure if the inner-workings of the MOTL price guide are explained in detail anywhere (I don't see your points in the FAQ or elsewhere on the site), could you point us to where Raw N is explained?

For instance if a card in the past has been worth $1.00, and for whatever reason no one sold that card recently, perhaps in absence of recent activity, it's value is still $1.00. Is that perhaps what the guide uses for it's pricing when there is not enough data to move the price otherwise?

I'm also finding it humorous that somehow I'm now defending a portion of the MOTL price guide that someone who's been involved with it for years is saying is useless. One could argue that if such experts in this area don't think half of the guide is useful, perhaps those entries shouldn't be reported in the first place... =P

quote:

---

Originally posted by digitek:
Here I thought we were having a healthy discussion...

I'm not sure if the inner-workings of the MOTL price guide are explained in detail anywhere (I don't see your points in the FAQ or elsewhere on the site), could you point us to where Raw N is explained?

For instance if a card in the past has been worth $1.00, and for whatever reason no one sold that card recently, perhaps in absence of recent activity, it's value is still $1.00. Is that perhaps what the guide uses for it's pricing when there is not enough data to move the price otherwise?

I'm also finding it humorous that somehow I'm now defending a portion of the MOTL price guide that someone who's been involved with it for years is saying is useless. One could argue that if such experts in this area don't think half of the guide is useful, perhaps those entries shouldn't be reported in the first place... =P

Regards

---

Maybe that explains why you are mistaken in thinking that these values are valid. If you don't understand how something works, it's generally a good idea to not get into extended discussions about how to use it properly.

The price guide is based purely off of recently completed Ebay auctions. If there hasn't been a completed auction of a given card, then it won't show up in the list. Raw N shows how many auctions are considered in a given value.

As for cards with a Raw N of 1 not being useful, no one is saying that. It's better to see a price that I can see is possibly off-base that I can then compare to other sources than to not see any price at all.

__________________
There's no need to fear, UNDERDOG is here!

All your Gruul Nodorogs are belong to me. Trade them to me, please!

Report rules violations.

"nil" as coolio was pointing out, is commonly known in scripting languages as the absence of having a value period. It's different than "null" which is most commonly equated to zero. The equivalent of nil here would be "no recorded sale or authorized evaluation ... ever"

Many items in the world are priced on a single sale, such as public art auctions (which is essentially what eBay is). That is different than nil - but I certainly agree one should be more skeptical on it's value than higher N.

I've been using store.tcgplayer.com, which has a nice store front that basically compares retail prices from many stores (with a small inflation they tack on for their service). For those cases with a single sale, comparing with store values might be a great way to see if a value is too high (for instance, Bog Imp 9th edition available for 8 cents at ChannelFireball). The site appears to be pretty scriptable, so I'm sure that could be used as a good "upper limit" cross reference.

And to be fair... any Raw N still needs investigation for accuracy. Never underestimate the power of gullible buyers in large quantities. =)

quote:

---

Originally posted by digitek:
"nil" as coolio was pointing out, is commonly known in scripting languages as the absence of having a value period. It's different than "null" which is most commonly equated to zero. The equivalent of nil here would be "no recorded sale or authorized evaluation ... ever"

---

Actually, considering I'm a computer programmer, I happen to know damn good and well that "null" NEVER equates to zero. That's a very common rookie programmer mistake that is learned from very quickly. Null very specifically means an absence of any value. Nil is a replacement for null in a few specific languages. I think it's a fair statement to say that a single data point's reliability is nil, as in assigned no value, even though I specifically used nil instead of null because it's much less commonly associated with programming.

You do realize that assigning prices based on public art auctions is far and away different than doing so for Magic cards, right? When only 1 item in the world exists, yes, you can generally only price it based on single sales, since people don't tend to buy art in order to sell it right away, thus an obvious lack of data points to actually draw an actual conclusion by nature. The prices of a specific art piece still can vary quite widely from sale to sale, a perfect example of why that single data point was unreliable.

__________________
There's no need to fear, UNDERDOG is here!

All your Gruul Nodorogs are belong to me. Trade them to me, please!

Report rules violations.

Obviously MTG is different that unique art (unless you look at the one of a kind altered-art cards out there). But even with MTG, with a base value of any sort (even if it's based on previous sales when the set came out), a single value plus historical data is useful - more useful than no data at all.

I actually started a separate thread on asking for historical data which could be used to add further input for low Raw-N numbers, showing historical trends, etc. There's actually a pretty nice site being developed by someone else called the Black Lotus Project, which shows really nice graphs based on historical data as well. If MOTL could make previous guides available, I'm sure we could push the history back even further. Very interesting to see the impacts of Core and Expansion sets on prices...

quote:

---

Originally posted by digitek:
I work for an embedded engineering company (one of the largest embedded chip providers), and the reliance on NULL being equal to zero is absolute. If that were not the case would have millions of devices behaving improperly. If you are aware of an embedded language where NULL is not equal to zero (i.e. the expression (NULL == 0) is FALSE) I would love to hear about it.

---

I honestly am not knowledgeable about that field, so I couldn't tell you the truth of that. I was going by your comments regarding scripting, where you are 100% beyond a shadow of a doubt wrong. If you wish to compare apples to oranges, I am unable to have an intelligent discussion with you.

__________________
There's no need to fear, UNDERDOG is here!

All your Gruul Nodorogs are belong to me. Trade them to me, please!

Report rules violations.

Either way, neither discussion precludes my original response, which is that a single piece of information combined with historical data, is more valuable than no information at all. Therefore equating that single value to nil is invalid, which was done above. nil is the absence of any information (or NULL in a your scripting language of choice).

quote:

---

Originally posted by digitek:
Either way, neither discussion precludes my original response, which is that a single piece of information combined with historical data, is more valuable than no information at all.

---

Thank you for completely and utterly ignoring my point. Of course when you combine it with historical data, it's value increases. That's the exact reason why I pointed out that taking a single data point WHEN IT ISN'T COMPARED TO ANYTHING ELSE has a reliability that is nil. As soon as you start bringing in historical data, you have additional information and the reliability naturally increases. Seriously, it's not like we're talking quantum physics here. I have no clue what your agenda is or why you're trying so hard to twist things around to try and make some incomprehensible point, but I have no desire to continue trying to correct you. I'm done.

__________________
There's no need to fear, UNDERDOG is here!

All your Gruul Nodorogs are belong to me. Trade them to me, please!

Report rules violations.

But both of these points aside, this thread started with a poster that was trying to improve the accuracy of the price guide and inform users and creators of possible ways to do that.

It seems that the only responses have been criticizing the poster and basically hinting that users should do this work themselves. Are you speaking on behalf of the price guide creators and saying you are not interested in constructive forum posts with suggestions of this nature?

quote:

---

Originally posted by digitek:
I don't think I was every trying to correct you.

---

Then you don't pay any attention at all. You posted about how many stores use the MOTL price guide to set values regardless of the Raw N. I pointed out how anyone using a single data point was making a major mistake because the reliability was nil, and that it obviously wasn't a true statement to begin with or else the data problems pointed out in the first post would be reflected by these supposed stores. You then went on about a point about the values of nil and null were supposedly being used wrong and how they relate to reliability. I was the only one to ever use the term nil until then, so it's pretty clear you were still responding to me.

quote:

---

Originally posted by digitek:
But both of these points aside, this thread started with a poster that was trying to improve the accuracy of the price guide and inform users and creators of possible ways to do that.

It seems that the only responses have been criticizing the poster and basically hinting that users should do this work themselves. Are you speaking on behalf of the price guide creators and saying you are not interested in constructive forum posts with suggestions of this nature?

---

I've never said anything remotely like that, but thanks for yet again trying to put words in my mouth. Taking issue with specific statements that you personally have made in no way reflects any opinion regarding the initial poster or their ideas.

__________________
There's no need to fear, UNDERDOG is here!

All your Gruul Nodorogs are belong to me. Trade them to me, please!

Report rules violations.