Guide to Analyzing
Bid Histories
 
There has been a lot of analysis of bid histories to reach some conclusions of late on the Stamps Board.  The following has been prepared to provide some suggestions as to what can and what cannot be determined by reviewing an auction bid history.

The selections below are from a real lot.  The links and most of the graphics have been eliminated to simplify the presentation and to eliminate extraneous, unimportant details.

Below is the essential data from the auction listing.  The First Bid was $1, there were 17 bids before we got to the winning bid of $100, and the auction ended at 14:58:49.  The final data is that the lot was won by, of all people, winner!  Further data has not been included here so we can assume that the item is fairly assessed by the bidders.
#ZZZ Nice Block 4 og hr's Crisp $500 CV $1 SB
Item #NNNNNNNNN
Currently $100.00 First bid $1.00
Quantity 1 # of bids 17 (bid history)
Time left Auction has ended. Location Terra Firma
Started 05/14/99, 14:58:49 PDT
Ends 05/21/99, 14:58:49 PDT
High bid winner(xxx)

And now we have the bidding history extract.

eBay bid history for 183 BLOCK OF 4 og hr's FRESH $500 CV $1 RES (item #NNNNNNNN)

If you have questions about this item, please contact the seller at the User ID provided below. Seller assumes all responsibility for listing this item. 
Last bid for this item:                $100.00
Date auction ends:                     05/21/99, 14:58:49 PDT
Date auction started:                  05/14/99, 14:58:49 PDT
Seller:                                stamp seller (xyz) 
First bid at:                          $1.00
Number of bids made:                   17 (may include multiple bids by same bidder)
Bidding History (in order of bid amount):

winner (xxx) 
        Last bid at:       $100.00
        Date of bid:       05/21/99, 14:58:47 PDT

underbidder1 (xx) 
        Last bid at:       $99.00
        Date of bid:       05/21/99, 14:58:27 PDT

underbidder2 (xx) 
        Last bid at:       $77.52
        Date of bid:       05/21/99, 14:49:23 PDT

underbidder3 (xx) 
        Last bid at:       $70.01
        Date of bid:       05/21/99, 14:57:16 PDT

george (xx) (not a registered user)
        Last bid at:       $50.00
        Date of bid:       05/21/99, 13:01:16 PDT

secondbidder (xx) 
        Last bid at:       $33.55
        Date of bid:       05/21/99, 12:06:55 PDT
openingbidder (xx) 
        Last bid at:       $10.00
        Date of bid:       05/17/99, 13:23:42 PDT
Note: there was at least 1 preceding (earlier) bidder which is immaterial.

Remember that earlier bids of the same amount take precedence.
Bid retraction and cancellation history (if any):

    There are no bid retractions.

The first thing we note is that we have incomplete data.  There were 17 bids made and we have only the final bids of 8 bidders.  We know nothing about the missing 9 bids made by these same bidders.  (eBay may, but we surely don't.)  We will therefore base our observations on the assumption that the final bids were the only bids made by each of the bidders.

Now we will analyze the bids. The technique we will use is to follow the bids and the times to determine the existing maximum bid at each bid.
 
Date/Time Bid Made Bidder Apparent High Bid Before High Bid Afterwards Max Bid (Proxy)/by
05/17/99, 13:23:42 PDT opening bidder 0 1 10 opening bidder
05/21/99, 12:06:55 PDT secondbidder 1 10.50 33.55 secondbidder
05/21/99, 13:01:16 PDT george 10.50 34.55 50 george
05/21/99, 14:49:23 PDT underbidder2 34.55 51 77.52 underbidder2
05/21/99, 14:57:16 PDT underbidder3 51 71.01 77.52 underbidder2
05/21/99, 14:58:27 PDT underbidder1 71.01 78.52 99 underbidder1
05/21/99, 14:58:47 PDT winner 78.52 100 100 or + -- a 1 sec snipe!

Now what do we know about george?  First he is no longer a registered user (presumably for shilling).  Did this lot give any evidence of shilling in and of itself?  Note that george was outbid smartly by 4 bidders after his bid was placed about 2 hours before the close of the action.  Probably the only shilling, if any, was the added activity of george's bid(s) on a lot that received 17 bids.

Next, did the bid by george impact the price paid by winner?  It would certainly be a far stretch when the winner considers that the winning bid was twice what george was willing to bid.  The second high bid was also nearly twice what george bid.  Certainly any attribution of auction impact by george has to be minimal if any.

One other thing, there are 9 bids of which we know nothing.  Often people make second bids after seeing themselves outbid.  This would make a considerable difference in the history but none in the conclusions.

It is interesting to note that underbidder3 never showed as the current high bidder.  His bid was too little, too late.

Conclusions:  First, we don't have all the data -- we never do when there is even 1 more bid than bidders.  we need to read the data you have very, very carefully to make any conclusions at all.  In this case, there is no conclusion of any sort that can be made about this bidding history.