Bidding Strategy For Great Collection Auctions
Walkerlover
Posts: 934 ✭✭✭✭
My question is if you are outbid during a GC auction. Do you throw in a test bid to see if you are outbid by a larger amount just to see where you are? Or do you try to get back to being high bidder?
Most importantly do you just wait and throw in your best bid in the last 15-20 seconds. Or do some of you throw in your best bid somewhere in between to discourage others from bidding. My questions pertain to inexpensive coins under $500. Thanks for your insights
Best Answer
-
PerryHall
Posts: 46,339 ✭✭✭✭✭
Just like eBay. Snipe it with your maximum bid just before it closes.
Worry is the interest you pay on a debt you may not owe.
"Paper money eventually returns to its intrinsic value---zero."----Voltaire
"Everything you say should be true, but not everything true should be said."----Voltaire4
Answers
@Walkerlover
All of the above.
Sorry to be vague, but it really depends on the coin, and even the time of year sometimes.
I often bid in the last minute, and so do most others from my experience. Test bids are used to at least see if you can assess a secret max high bid from others, and you can take it from there.
Shills are there, so just be smart & patient. A war can develop very quickly too.
Bargains are to be had, and timing matters. The previous auction results are helpful, but not written in stone.
I have seen a few coins that I have bought show up on past auctions. There is some turnaround with popular series and attractive toners. I once checked on my coin that's in my profile/avatar above, it was still in someone's Registry with our hosts lol..
If I really want a certain coin, good chances someone else does too, but not every time. Once in a while, the price guides get completely shattered for a nice coin.
Good luck!
Fixed closing time which means you have to be highest bid at close to win. I place my MAX bid as close to closing time as I can. I place a lowball bid to put on my bid list as soon as I see a coin I have interest in that is currently below my max bid (some start higher than I want to pay)
Better to wait - if you bid aggressively early in the auction, you give others a chance to mentally overcome what may have been their maximum comfort level. They may do more research, realize they missed an important attribute of the coin, etc.
Better to bid with just a few minutes left. If you wait until the last 60 seconds though, you may have trouble bidding on all the lots you are interested in if they close in rapid succession or you may miss out if you have a computer or phone connection problem at the last minute.
I place my best bid and then ignore it. If someone wants to bid more, 99.99% of the time, I don't care. If I do care in a live auction, I'll bid in the live auction. For a timed auction, if you really care you are basically in a lottery if you try to snipe. Does your bid get in in time? Is it the high bid? Is my internet stable?
Auctions really should extend bidding until bidding stops. I don't know why GC went with this format.
Pretty sure it's because @ianrussell knows exactly what he is doing. His objective is to achieve the highest sale price possible, both for himself and for his consignor.
I'll bet that means having a fixed closing time, where people know they have to get their highest proxy bids in before a fixed stop. The alternative is people slow playing it, knowing the auction won't close so long as anyone is bidding.
Live auctions don't lend themselves to this, but online auctions do. Do you honestly think the 800 pound gorilla in the space, eBay, doesn't know what it is doing, and has been leaving money on the table all along?
You are looking at it from a bidder perspective, and "don't know why GC went with this format." The simple answer is their objective is the exact opposite of yours as a bidder.
Finding yourself "basically in a lottery if you try to snipe" encourages nuclear bids. Extending an auction indefinitely does not.
Nuclear bids result in higher hammer prices when you have two people competing. You as a bidder want lower hammer prices. Not complicated.
I like the fact that there is no live auction! In fact, I have all but stopped using Heritage since I would win the internet auction, only to lose the "Live" auction. Drove me mad, so I just quit bidding! I might go back to their internet only auctions but why bother.
For $500> coins I will bid what I am willing to pay, might be early in the auction or it might be late. The timing is of little concern to me, if the bids are already over what I will pay then I move on. Sub $500 coins are not scarce, there will be another one, maybe the very next week. No reason to get into a bidding war or overpay for generic coins like this.
My Collection of Old Holders
Never a slave to one plastic brand will I ever be.
I don't care if it's live or internet, but a new bid should extend the time. Most real auctions work that way including sites like Proxi-bid. Timed auctions aren't good for sellers or buyers.
.
First I do not have any special or inside knowledge as to why. This is only an observation.
Also I am not promoting any particular auction method (although I do of course have a preference).
When I check, GC in recent times is auctioning between 5000 to 6000 lots per week (Sunday). The majority of these appear to hammer below $200 and the vast majority below $1000 (all a best guesstimate but generalized for that). The lots 'go off' about every 3 seconds (again estimate). Therefore GC can auction about 1200 lots per hour and can complete a Sunday auction of 5000 to 6000 lots within 5 hours. Checking GC it looks like 5163 lot for this week (Sunday). 5 hours is still a fairly lengthy auction but reasonable amount of time I guess.
If a different format from a 'hard close' was to be used then it would appear to either extend the auction time and possibly easily double it. Or create a situation where multiple lots would be overlapping each other. Other options?
For the first or extending the auction this would be where the next lot does not start it's final closing countdown until the previous lot closed. This would be similar to HA internet auction that does not have a live auctioneer. Bid comes in and that lot is extended a dozen seconds or so and one gets the warning lot closing but another bid comes in and repeat until it closes. So instead of a 'hard close' every 3 seconds just one bid would extend that lot a dozen seconds or so and multiple bids longer. How long would it take to auction 5000 to 6000 lots?
For the second scenario, if each lot were to keep it's closing time of every 3 seconds unless it got a bid and was extended a dozen seconds or so, then there would be multiple lots 'in action' at one time due to prior lots getting extended due to bidding. Perhaps acceptable but I think some would not like it.
I will note that Legend auctions on their internet auctions (not their primary auctions) did use a system similar to this at one time but when a lot got a bid it was extended for a longer period of time (2 minutes? but don't recall). However, their lots were not coming up every 3 seconds but a longer period of time in-between which I don't recall and their total lots were much smaller. So LA could and did sometimes have multiple lots 'in action' at one time.
https://youtube.com/watch?v=wwmUMvhy-lY - Pink Me And Bobby McGee
.
https://youtube.com/watch?v=D0FPxuQv2ns - Ruby Starr (from 'Go Jim Dandy') Maybe I'm Amazed
RLJ 1958 - 2023
Most of the time I just bid my max first. I only try to snipe at the end if it’s a coin I really want.
Collector
75 Positive BST transactions buying and selling with 45 members and counting!
instagram.com/klnumismatics
A local auction house uses it through HiBid. They start closing in groups of 10. If there is a last second bid, it adds 2 minutes. Sometimes, lot 10 stays open while lot 300 is closing if lot 10 is active. They do 700 lots or so in 3 hours.
You may be right that their concern is overlapping closing. But that's already an issue. The 5000 lots do not have unique closing times where you could jump from lot to lot and still get a bid in.
As I said, I don't really care because in alnost all auctions, i place my max and walk away. But for something you really want, you have to go nuclear and hope someone else doesn't do the same thing. Even if you want to snipe, you better bid $5k on the $1k lot because if someone else snipes at $1100, you aren't going to be able to get back in.
For me it is best to set a happy price and let her go... there is always another coin.
It keeps everybody honest and I win often enough.
Almost discount shilling, but it keeps me out of bidding wars.
BST: KindaNewish (3/21/21), WQuarterFreddie (3/30/21), Meltdown (4/6/21), DBSTrader2 (5/5/21) AKA- unclemonkey on Blow Out
Actually, the literature suggests that sniping results in lower bids overall, which is why people do it.
So your theory is that GC knows best and every other numismatic auction hasn't figured it out yet?
Yup. Does eBay extend auctions until bidding stops? Pretty sure they figured it out.
I'm not talking specifically about sniping. I'm talking about forcing you to put your highest proxy bid in before the announced close, because you are not going to have an opportunity to increase it later. Because there will be no later. Snipes or not.
I'm betting GC and eBay have reams of data supporting the notion that the way they run auctions leads to better results for them and their consignors. Regardless of what a rigorous piece of undoubtedly peer reviewed research published 14 years ago in the International Journal of Nothing, and you, say to the contrary.
No such thing as "good for sellers or buyers." What's good for one is, by definition, bad for the other, since they have diametrically opposite interests with respect to pricing.
That's not true. The frustrations are different. You're only looking at price. The frustration for buyers isn't the lower price, it's the failure to purchase because the internet hiccups or because they can't get back in with a second bid.
There's this, from a 2023 paper: "By incorporating limited-attention into online auctions, we explain this phenomenon by demonstrating that the soft-close ending rule may fail to eliminate the strategic advantage of sniping. However, sellers’ expected revenues are higher than in hard-close auctions, snipers’ winning probabilities are lower and winning prices are higher, indicating that soft-close ending rule does in some sense solve the problem caused by sniping." -- https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4676973
"Failure to purchase" is a frustration that is easily solved by putting the best proxy bid in well before the published close of the auction. "Internet hiccups or because they can't get back in with a second bid" actually does relate right back to price, which is all that matters to the seller and the house.
If desire to win trumped desire to get the best price, the nuclear proxy bid would be the early and only bid.
As with @jmlanzaf's article, if true, your obscure Chinese researchers are far sharper than the minds behind eBay and GC. My vote still goes to those behind the highly successful platforms. Not the obscure authors publishing in the equally obscure journals.
As always, YMMV. The good news is that there is a platform for everyone.
(edited because i mis-read a statistic and have updated Ebays results based on merchandise volume, not ebays revenue)
Great Collections did annual revenue in 2023 of over 200 million. They use hard close timed auctions.
Stack Bowers in 2022 (last reported year I could find data on) was over 250 million. They are larger than GC. They use soft close auctions.
Heritage Auctions in 2023 did 1.76 Billion. They are much larger than GC. They do soft close auctions.
Ebay is over 10.1 Billion in 2023 revenue, with Gross merchandising volume of $73.2 billion. That figure includes all of their fixed price listings as well as the auctions. The info I see online (2022 data from eDesk) says 88% of their sales are buy it now fixed price. While we cant extrapolate for sure based on the % of listings because we dont know the price breakdown of those items, if we take the 12% of their sales that are auction and assume its 12% of volume, in terms of auctions, that makes them around 8.8 billion in auction revenue.
Sothebys (7.9 Billion) and Christies (6.2 Billion) are the biggest auction houses. They do soft close auctions.
So by your logic of back the format used by the largest most successful players....the larger share of the auction market seems to be soft close auctions.
Yup. But take a look at exactly what they are auctioning. The more valuable the material, the less likely people are to place nuclear proxy bids, and the more you want to give the auction time to play out.
The auctions of the most valuable merchandise tend to be live. And, as I said at the outset, they do not lend themselves to hard stops. Lies, damn lies, and statistics. Your numbers are interesting, but shed no light.
How do you want to measure "larger share"? By average hammer price of individual lots? Clearly that would be one of the prestige auction houses. In our world, SB or HA. By number of lots auctioned? Again, in our world, that would be eBay by a mile.
At the end of the day, Christie's does what makes sense for Christie's, and GC does what makes sense for GC. As I also said before, there is a platform for everyone.
Sniping is best. Give it your best shot in the closing seconds. Otherwise, your high bid becomes a target, if placed too early.
Sometimes, it’s better to be LUCKY than good. 🍀 🍺👍
My Full Walker Registry Set (1916-1947):
https://www.ngccoin.com/registry/competitive-sets/16292/
Put in a low bid to start out to put in my bid list. iPhone and desktop are on different providers and have both on the auction teed up and ready to bid. Bid with less than 10 seconds to go with a higher and different bid on each device. If I am against a dealer, I win. Another collector, who wants it more will win. I usually win.
What is your return policy?
Ian Russell on Sun, Jun 30, 2019 12:05:01 PM
On all non-bullion items, GreatCollections offers a generous return policy (unless otherwise marked on the item information page) as a courtesy to Buyers, providing the item is paid for within seven days of the auction date (or in the case of a Buy Now item, when you confirmed to purchase the item) and no request to delay shipment is made by Buyer. Certified coins/banknotes must be in the original sealed grading service holders. To return an item, the Buyer must notify GreatCollections within 24 hours of receipt and receive a return confirmation number. Please mail the coin or banknote via Insured Mail to GreatCollections within 72 hours of receipt of coin. Original and return shipping costs are not refunded. Returns will not be accepted without the return confirmation number. Registered bidders are allowed one free return each month. For more than one return in a calendar month, a 5% fee will apply, based on the total purchase price of the item. Clients found to be abusing our return policy will be notified in writing that they will no longer have any return policy privileges whatsoever. If you viewed the item prior to winning the auction or offered the coin for sale to any dealer, collector and/or marketplace, there is no return privilege.
Good strategy thanks for sharing. But 10 seconds is cutting it close, no room for any errors human or technical
Except the internet hiccup works against the seller, who you claim wins in this format.
And you can't solve the frustration problem by sinply placing a bid earlier because you could get outbid. This can not happen in a soft close where you can choose to do the Pittman Statue of Liberty if you wish.
And you continue to ignore all the snipers and all the literature that says that hard closes favor the buyer not the seller on price. I know those are inconvenient facts for you, but you have zero counter arguments other than GC knows better than all the bigger auction houses. Maybe Ian will send you a check.
We forgot that we're arguing with someone with whom facts are irrelevant.
Someone should call Heritage @MFeld and tell them they are doing it wrong.
I'm not conveniently ignoring anything. Of course, extending an auction when someone would have come in after the close with a higher bid would benefit the seller.
But that is not mutually exclusive with the seller benefiting in other cases, where the sniper is not a factor because he has already been outbid by a proxy bidder who would not have been motivated to enter an early nuclear proxy bid if he had been able to wait until the auction was scheduled to close to see how things were going, because the auction doesn't close until bidding stops.
Yes, I am assuming that @ianrussell is smarter than you, and that he knows what he is doing. He doesn't need to send me a check. The fact that he has money to send out calendars is fine.
The bigger auction houses have a different business model based primarily on live auctions. GC and eBay seem to know what they are doing in the internet auction space. The fact that you are occasionally frustrated when you don't want to show your hand early and don't have time to increase your bid later does not change that.
Rest assured, if that was consistently costing @ianrussell, and more importantly, his consignors, money, he would change things up or not be nearly as successful as he is.
Or not. Do you honestly think HA has the same business model as GC and eBay? Or that they are all not extremely successful, even with different models, because the only correct one is the one that benefits you the most?
No, "HA is not doing it wrong." Neither is SB, Christie's, Sotheby's, etc. By same token, GC and eBay also seem to be doing just fine, even if you think they could be doing even better if they put you in charge.
For GC I go thru line up bid targets (for retail stock) and put on watch list days b4 auc close. The final seconds can be a dogfight. Just work your angle. Bagged some CACG coins right 4 retail recently.
To win buying it right I need to bid as closest to auc close as possible. Had some really good wins there. Avoid bid wars.
U can do test bids see if some kind angle there.
Do your homework way before….then whup them near auc close.
As one of the comments by someone else above pointed out, one of the problems of placing a very high max bid early on GC, is that yes, you might end up the high bidder, but because of all the time left, people will “eat away” at your lead, and you may then end up paying more than you otherwise would have paid. In this scenario, other people who don’t snipe and don’t bid nuclear for fear of being the high bidder at a crazy price, but place their “max” bids with a minute or two left on the clock, once they see their max bid was not good enough, they may then place a new “max” bid, and do that several times, getting “more bites at the apple”.
Here’s a real life example that took place on GC about four weeks ago. A friend of mine wanted/needed a certain very high grade coin. It turns out only he and one other bidder were serious at a high price level (at that level the bid increment was $500). In advance, I tried to convince my friend to snipe, but he was nervous doing so, so he placed his VERY high max bid with about 30 seconds left on the clock. That made him the current high bidder at $18,000 hammer. Ten seconds later (with only 20 seconds left), the other serious bidder placed his “max” bid at $20,000 hammer. But lo and behold, that bidder was shocked to see that my friend was still the high bidder, but now at $20,500. With 13 seconds still left, that other guy now had time to place yet another max bid, which he did at $22,000 hammer. But my friends very high max bid placed at 30 seconds was still higher, so my friends bid was still the high bidder, but now at $22,500 hammer, and then the auction ended. Had my friend only placed his initial max bid closer to the end, his competitor would not have had time to place his second “bite at the apple” of $22,000 hammer. My friends very high max bid would have ended up as the high bidder at $20,500 hammer, instead of his actual ending high max at $22,500. My friend ended up paying $2,000 (plus an extra $200 bp) more than he would have had to pay had he properly sniped!
Steve
My collecting “Pride & Joy” is my PCGS Registry Dansco 7070 Set:
https://www.pcgs.com/setregistry/type-sets/design-type-sets/complete-dansco-7070-modified-type-set-1796-date/publishedset/213996
GC is not eBay. GC is closer to Heritage than eBay. But that's not really the point.
I don't know why GC chose a hard close. And neither do you. You've assumed it was because it yields higher hammer prices despite ALL THE EVIDENCE to the contrary. And, as is your style, you are running with your assumption as though it were a fact.
Hard closes lead to sniping. Sniping leads to slightly lower hammers in many studies. There are no studies that demonstrate the opposite. I've seen some that claim minimal effect.
Ebay didn't adopt it for higher prices, they basically invented it. Prior to eBay, the only auctions that had hard closes were mail bid auctions which operate differently as you cannot see any of the bids. There were also essentially hard close "by the candle" auctions in certain settings but almost all commercial auctions were "soft close".
The whole sniping industry developed in order to provide a way for bidders to game the system to the benefit of the bidders. Some entities actually make money by selling sniping software or services. So you have to assume that all of those bidders are irrational, including the ones in this very thread who advised sniping to save money not to maximize the price paid.
Hard closes are more predictable from a timing standpoint and there's nothing wrong with that for either eBay or GC. The reason they both adopted it is probably related to the number of simultaneous auctions as pointed out by a poster above. Heritage et al do not run soft close auctions with more than 1000 items.
I would just like you to provide one scintilla of evidence that it yields higher hammer prices, contrary to all evidence. But, don't worry, I'm not really waiting around for it.
I agree, and believe, that GC’s and eBay’s hard closes saves snipers like me money, working in the buyers favor. Those with soft closes, like Heritage, DLRC and Stacks, works in the favor of consignors, as sniping does not work! And yes, I use sniping software for my occasional purchases on eBay. Unfortunately, I’m not aware of any sniping software that works with GC. But if there was such a thing, I believe that would significantly increase the number of GC bidders that would then snipe, once they obtained confidence in the reliability of that software!
Steve
My collecting “Pride & Joy” is my PCGS Registry Dansco 7070 Set:
https://www.pcgs.com/setregistry/type-sets/design-type-sets/complete-dansco-7070-modified-type-set-1796-date/publishedset/213996
GC auctions often results in the highest prices in the world. Likewise, HA or Stacks often results in the highest prices in the world as well. There is no right or wrong answer here, because there are simply too many factors to consider that can alter the results. For example, bidder fatigue is a HUGE negative factor in the “soft” auction closes. I simply am too busy to wait around for many of these prolonged auctions resulting in “bidder drop off”. I simply put in a weaker bid than I otherwise do when “sniping” and if I win, I win. And, if I lose I lose. Alternately, with the “hard” closes, I have been known to put in a higher snipe bid and on occasion get “stuck overpaying” for a coin when someone else does the same. Not to even mention losing coins all together due to “technical” problems which can be a heartbreaking situation with sniping.
Conclusion: All of the auction companies are doing just fine with their approaches and we, as serious collectors or dealers, need to learn how to become most successful bidding under both formats (usually by trial and error).
Wondercoin.
The recent auction at S&B of the Gold Flowing Hair Privy coins was a course in ‘pre live auction bid’ worthlessness to me. Every single one of the early bids was outbid within a second…….and showed what many of us were willing to pay. Sadly, 99.9% of us that put in max bids of $15-18K for a DCAM 70 THOUGHT we had a fairly decent chance at snagging a coin, but quickly found out our bids were quickly quashed by the big money club. I prefer GC’s hard closing which gives me a chance to win..vs an auction that drags on as I watch my chances dwindle with each higher bid. While I like many of the coins/currency on HA auctions, their buyer’s premium seems a bit rich for my taste. To prepare for these auctions I put the coins I’m interested in on my WATCH LIST, and add values and grading populations in the “NOTES” to give me an idea of what I might want to bid as my max when the time comes. I also find if I wait and put in my MAX bid in near closing in a hard close auction it prevents me from overspending in a bidding war. If I win great, and if I don’t I just look at the other coins I would like to have and bid on them.
I think generalizing from the Flowing Hair results is a bit of a mistake. Early bidding is not final, of course. But the same could be said fir GC. The hard close doesn't change the fact that no one wins until the auction ends.
You could have won every single Flowing Hair coin with a bid entered 10 days before the auction closed by simply bidding higher. The problem with the Flowing Hair auction was more that no one new what the value was going to end up being.
The GC hard close does not in any way increase your chances of winning. It may alter your strategy in bidding since sniping is possible. However, i win hundreds of auctions every year with bids entered in the pre-bidding period under both formats. The highest bid wins, no matter when it is entered.
And, of course, the near irrelevance of the BP had been discussed before. All bids, at any venue, should be made considering the total cost including BP, shipping and sales tax (if applicable). That is, of course, why Heritage or SB are not players in the $50 and under market. But, costs to the seller in that price range are prohibitive at GC as well. Such coins really should not be sold at any of the auction houses.
Just don’t wait TOO long! I once had a winning bid with literally not a second to spare (I compared bid history time to auction ending time)! My bid was very strong, so not only myself, but also the consignor were VERY lucky. Had I blinked or been a second longer we both would’ve been on the losing end. It was just sheer luck. Make sure you allow yourself enough time for the system to process your bid and place it. It takes more than a few seconds, as it is a two-step procedure, which is good as it ensures no accidental bids.
Sometimes, it’s better to be LUCKY than good. 🍀 🍺👍
My Full Walker Registry Set (1916-1947):
https://www.ngccoin.com/registry/competitive-sets/16292/
I almost always snipe on GC but how do you know your friend would have payed less if he sniped? It’s quite possible if your friend and his competitor both sniped with nuclear bids that the selling price may have actually been higher. This is of course speculation on my part and we’ll never know and why hard endings IMHO are not favorable to the seller.
I know my friend paid more than he had to because my friend gave his competitor time to place one extra bid, and the competitor chose to do a max hammer bid of $22,000 on that second one. This was placed by the competitor only after his first bid placed at 20 seconds remaining was at $20,000 max hammer. Had my friend sniped with only 10 seconds or so left, while the competitor would have still placed his initial bid of $20,000 at 20 seconds to go, he would not have had time to see that higher bid of my friend register on GC, and would not have had time to place a new bid at $22,000 max.
Steve
My collecting “Pride & Joy” is my PCGS Registry Dansco 7070 Set:
https://www.pcgs.com/setregistry/type-sets/design-type-sets/complete-dansco-7070-modified-type-set-1796-date/publishedset/213996
As always, you like to argue for the sake of arguing, and always seem to assume you are the smartest guy in the room. The only fact I'm relying on is that eBay and GC are both highly successful in their own way, with a different business and auction model than their traditional auction house competition. Maybe because they are not traditional auction houses.
Honestly doesn't matter what eBay invented. It's not like they have not modified anything since they launched. So I'm further assuming that if there was an easy way for them to achieve even better results, both for them and their consignors, but most importantly for themselves, that they would have figured that out without your help.
You might actually be making my argument, although you'll never be gracious enough to admit it. If "the reason they both adopted it is probably related to the number of simultaneous auctions as pointed out by a poster above," that's just another way to say they make more money by having hard closes and more auctions than by doing less business by letting each auction run its course.
AND, if it was costing them consignment business, because consignors were consistently achieving inferior results as compared with soft closes, that would also force them to make a change. It hasn't.
So why argue with me for the sake of arguing? Take it up with @ianrussell, or take your consignment business elsewhere. Because you also seem to be insisting that hard closes benefit you as a bidder, so what's your problem with them? A need to argue with yourself, in addition to me? 😀
I understand it now and thanks for the explanation.
Bid what you wish to pay for the coin, if you win then yaay and if not, you didn't overpay. I have never bid on a coin that I could not do without and have never bid a nuclear bid. I just find what $ I wish to be in the coin at and bid as required for that. It's an auction. If 3 or more have bid since day one, I will move to another auction, as I have found it will be a war at the end. JMO
Jim
When a man who is honestly mistaken hears the truth, he will either quit being mistaken or cease to be honest....Abraham Lincoln
Patriotism is supporting your country all the time, and your government when it deserves it.....Mark Twain
I'll take that as a retraction of your prior argument.
Take it any way you want.
snipes only work if you remember to do them - I was going to do one tonight but got started in a project and when I noticed the time, it was 3 minutes too late
Does that mean the coin was not a big deal you were bidding on so you forgot
I guess so
No bidding strategy works when you have bidders willing to pay 10 times the value
Mike
My Indians
Danco Set
There is the possibility that the other bidder was testing the waters at 20k to see if it was enough and would still have bid 22k at the end but your friend's chances of paying less did go down by bidding too early (although he won and didn't have to worry about the internet going out or clicking bid a little too late and losing).