We scraped all the data from every auction.
We got every bid, every bidder, and organized them into a searchable database.
We then computed how many auctions each bidder was involved in, how many of those auctions the bidder had placed a bid other than the initial bid (of generally $60), and computed what percentage of the auctions they won compared to the auctions they had bid on.
Here are some numbers:
Amount of auctions we were involved in 11,973
Number of unique bidder Id’s: 8,890
Amount of auctions we won: 4,771
Amount generated from all the auctions we were involved in : $13,200,000
Now Snapnames.com said there were 1,000,000 auctions during this period which means we were involved in, and have detailed records from about 1.2% of all auctions.
Snapnames.com also said that the bidder Halvarez was in 5% or about 50,000 auctions.
Going through our history we found Halvarez was in 2,023 of the auctions we were involved in or around 17%, obviously a much higher percentage than the overall average.
Of course we like to think we were in many of the better, higher priced domain auctions which may account for the higher participation rate of Halvarez in our auctions.
Out of the 2,023 auctions Halvarez placed an initial $60 bid getting him into the auction, he placed another bid in 590 of those or 29.2%.
Out of the 590 auctions he placed a bid in other than the initial bid, he won just 34 or 5.8%.
Compared to the 2,023 auctions he was in, Halvarez only won 1.7%
For comparison purposes, we placed bids (other than an initial bid) in 7,300 of the 12,000 auctions we participated in or 61.7%. We won 38.9% of all auctions we had an initial $60 bid in and 64.5% of all the auctions we placed additional bids in.
We then computed this same information for each of the 8,890 bidders compiling a list with the number of auctions each bidder was in, number of bids each bidder placed, number of auctions each bidder placed above the initial bid, and the win rate for each bidder.
We looked at those bidders that had a high bid rate but low win percentage.
We then went in and actually looked at each of those questionable bidder’s accounts to see each of the bids placed and their bidding patterns.
Bottom line we could not find any of these high volume, high bid, and low win accounts that matched Halvarez’s bidding pattern.
If you look at Halvarez’s bidding pattern on other domains, except those he apparently wanted to win, it was mostly “one timers”.
Typically in an auction bidders will increase their bids as other bidders increase their bids.
So a typical bidder will bid $1K, then $1,500, then $2,500, then $3,500, then $5,000, and so on, until they win the auction or until the high bid is more than they are willing to spend and drop out.
What you see with Halvarez is an initial $60 bid and then a “one time” bid which pushes against the high bidders proxy.
Having access to the system, Brady knew what the proxy bids were and he used this information to push the sales price close to the max proxy bid.
He was careful not to push you right to your limit.
If you saw time after time the auction ended at your maximum proxy bid you, would know something was up.
So Brady seemed to push the bidding up to somewhere between 80%-90% of your high proxy bid, seemingly dependent on the price range of the bid.
For example, lets say you had a proxy bid of $15,000, like I did on GreatJob.com, he put a one time bid close to the ending time of the auction of $12,500 pushing my proxy to the winning bid of $12,750. The previous high bid from a real bidder was $8,888.
So other than the initial bid of $60, he placed no others except for a $12,500 bid pushing on the proxy.
Another example of an auction where I was not the high bidder is AcAdapters.com. Halvarez placed a one time bid of $12,500 when the last real bidder was at $7,900, pushing the winning bidder bid to $12,750 based on a proxy.
Yet don’t get the idea that all Halvarez did is push up 4 and 5 figure bids.
A lot of his bidding activity was at much lower levels.
I won auctions with a bid as low as $80 where it was just Brady and I in the auction, and my proxy was $111 – like in the case of thejammer.com.
In another auction, makeabetterlife.com, that only Halvarez and I were in, my bid was pushed to $100 while my proxy was $111.
Themotive.com was another auction just between the two of us. In this case he pushed my proxy to $105.
I know it’s small change.
Nickel and dime stuff.
Amazing that a guy who must have cashed out for substantial dollars when Snap sold to Oversee would screw with $100 auctions.
But he did.
Time and time again.
In all there were 211 auctions of mine that I won in which Halvarez was the second highest bidder, and of the 12K auctions, he was the second highest bidder in 339 of them.
In terms of comparing all bidders, he was the 5th most frequent second bidder, winding up in 2nd place a whopping 58% of the time, while only winning 5.7%.
No other bidder came close.
Yes there were other bidders that had a high percentage of second place bids, but none had the volume as Halvarez.
The other substantial second place high bidders were all substantial winning bidders, those domainers you all know and widely regarded as having the best portfolios.
The fact that bids were placed to push up my proxy, not only on the four and five figure domains but on domain auctions as low as $80, tells me one thing for sure: Brady used a script to automatically place these bids, pushing up people’s proxies.
The first auction I have record of him pushing my proxy bid up was on 3/27/05 and the last one took place just a couple of months ago on 9/1/09.
So 50,000 auctions over 53 months, lets call it 1590 days, comes to almost 32 auctions a day.
But wait, Oversee says that 75% of Brady’s bidding was done before they acquired Snapnames.com.
Oversee acquired Snapnames.com in May 2007, so if that figure is correct then 37,500 of the auctions took place in that 26 month period or about 48 auctions a day, every day, 365 days a year.
If we use that number and see that bidding took place on domain auctions with prices as low as $80 I would say there is not way Brady was sitting around bidding on these auctions, there had to be a script.
So now we return to the main issue.
Were there other bidder IDs that were used by Brady for shill bidding?
So we ask are there any other bidders who had the same bidding pattern of Halvarez?
We couldn’t find any.
We not only looked at the stats described above, but we looked at the percentage of time that each bidder wound up being the second highest bidder vs. their percentage of winning bids.
What did we find?
No one came close to Halvarez.
In 58% of the auctions Halvarez bid in he came in second, while only winning 5.7%.
Moreover no other bidder had the sheer number of auctions they bid on like Halvarez 590 with such a low winning percentage.
Now for some disclaimers.
Of course once again we were in only in 1.2% of all auctions. That means we have no records on 98.8% of the auctions. We believe our sample should be representative of the whole, but of course it may not be.
Is it possible that Brady had other bidding IDs that he used, but did not use them in auctions we were involved in?
Its possible, but doesn’t seem all that probable.
Next disclaimer, we only looked at high volume accounts.
Those that had a significant amount of auctions, bids, and second place finishes.
Could Brady have set up say 500 different bidder ID’s and shill bid on just one auction using that ID?
I guess he could, but we also recognize that if he did, we would have no way of figuring that out from the bidding history.
We need volume to determine patterns.
You cannot see a pattern on an account which only had 1 or 2 auctions in 5 years, nor can you make a determination based on one or two questionable bids.
Did we go through all 8,890 bidders to see every bid they placed?
We looked at every bidder with statistically troubling numbers that had volume of transactions
In those accounts we either were able to recognize the domainer associated with the bidder ID, or when we looked at the detailed bidding in that account, saw normal bidding patterns.
Nothing that looked like the proxy pushing bidding of Halvarez.
1. We do not think there was another meaningful bidder ID used for shill bidding.
2. All of the auctions in which Halvarez pushed our proxy bid up with his shill bidding, Oversee properly noted and gave us credit for in the settlement agreement they sent.
3. Halvarez seemed to be especially interested in adult domains. This may also account for why he was in a higher percentage of my domain auctions than overall.
4. I hate Frank, Kevin, Buydomains, and Bonkers for outbidding me on a ton of great domains as well as costing me a fortune on the ones I won.
Now for the outstanding issues.
Although we cannot pinpoint any meaningful shill bidder other than Halvarez, there are still issues.
What to do when Halvarez was the 3rd or 4th bidder?
Before I ran the numbers I was concerned that I would find Brady started the high bidding off in most cases, triggering other high bids, winding up being the 3rd or 4th highest bidder, which Oversee is not offering compensation for.
Frankly this happened, but not as much as you would think.
However where it did happen its troubling.
Take the auction memoria.com for example, where Halvarez put in a one time bid of $10K , when at that point the high bid was just $6,100. After the $10K bid of Halvarez was placed, two other domainers bid on the domain back and forth until it sold for $10,805.
These are the most troubling for me as we all know if one guy doesn’t bid $10K maybe no one ever does.
Maybe the high bid winds up at $8K in this example, but then we will never know.
In another auction for the domain Clash.com, Halvarez placed bids for $22,500, $25,000 and $27,500. The auction found two other bidders before closing at $30,500.
Since there were two bidders higher than Halvarez, no compensation is offered by Oversee.
What would have happened if bids by Halvarez had not been placed?
Impossible to say.
We have seen many instances where a domain gets re-auctioned on NameJet.com for non-payment weeks after the first auction. The typical result is the domain almost never sells for close to the second bid of the first auction, but usually in the range of 50% of the first auction.
It has to do with the dynamics of the auction, psychology of bidders, certainly a topic worthy of a book more than a simple blog post.
I know one things for sure, you can’t stuff the genie back into the bottle.
You can’t predict what the outcome of an auction would have been if a bid had not been placed, no more than you could predict what would happen if the auction was rebid at a different time or place.
I do think some compensation is in order for these situations, but it’s well above my pay grade to figure out the right solution.
Does this mean we think you should accept the settlement?
That is a question which requires analysis of many other issues which go beyond this post.
One thing for sure, before you decide to settle read and then re-read the release that Oversee wants you to sign.
Its incredibly broad and overreaching.
It seeks to free them from any and all liability whether it is related to this exact situation or not.
If you have a meaningful settlement you should get the advice of counsel.
Are there other issues beyond these?
There are some bidding anomalies we have not been able to resolve.
There are some domains that appear as they were won by one bidder but went to another domainer.
These issues are for another day, another post.