How do extreme auction results impact the PCGS Price Guide?

Just curious if anyone knows how "extreme" auction results in high-profile and well-respected auction venues (like Legend Auctions) affect the pcgs price guide. I have always heard that pcgs will utilize auction data in adjusting guide prices -- and I've seen some evidence of that -- but it appears to be done in a very inconsistent manner. It is just random how results are handled and in determining how or if a recent auction result will cause an adjustment in the price guide? This is a good example below, where a recent Legend Auction produced a result where the hammer price was close to 30x over guide -- in the weeks since, no adjustment to the guide price has happened (and I am guessing nothing will happen). Is there a man in the loop to determine whether or not to adjust? or is it an automated algorithm? By the way, I did not buy or bid on this coin below (so I have no skin in the game) but i've been curious to see if there was any ripple in the price guide because of it. So far, zippo. [Honestly I am not sure what methodology is used to come up with PCGS Guide Prices. Some of the current Guide Prices seem quite odd to me, but perhaps they make sense to some?]
Comments
I don't know how the people at PCGS price guide do things.
I reported verifiable eBay sales of three mint state US-Philippines Pesos that occurred in January at prices substantially above the price guide levels, yet no change to this day of current values.
?????
I always wonder this my self do they throw out the highs and lows going with the average after that or is it an average of everything? We know not all coins are average some are weak in some way that makes them less desirable then on the other side you have the cream of the crop or famous coins they are rare indeed. I would think you would have to take the ultra low and high out of the equation to get a true average for the guidance
https://www.pcgs.com/setregistry/quarters/washington-quarters-major-sets/washington-quarters-date-set-circulation-strikes-1932-present/publishedset/209923
https://www.pcgs.com/setregistry/quarters/washington-quarters-major-sets/washington-quarters-date-set-circulation-strikes-1932-present/album/209923
Perhaps, it’s similar to how the Olympics scores with multiple judges. Throw out the highs and lows and average the middles. Or something close to this.
Plot a simple bell curve. Find the center, take a certain number of auctions above and below and then average.
Throw auctions values in a hat and pull one.
My guess is that adjustments are made (or not) based on how likely PCGS thinks it is that the auction result was representative of the current listed value.
For example, if a wildly toned common date Morgan dollar brings multiples of current price guide value, it's unlikely that the result will result in a price adjustment for that assigned grade. On the other hand, if a particular very rare coin (which has not traded in several years) brings a price which varies considerably from the current price guide value, the result is far more likely to cause a price adjustment.
Mark Feld* of Heritage Auctions*Unless otherwise noted, my posts here represent my personal opinions.
"Extreme" results - called 'outliers' - should not present significant problems in valuation tables if correct statistical methodology is applied to the data. To determine that, you will have to ask the valuation publishers.
@WingedLiberty1957 said:
"Is there a man in the loop to determine whether or not to adjust?"
Might be a woman.
End Systemic Elitism - It Takes All of Us
ANA LM, LSCC, EAC, FUN
My wager is they come up with a consensus representative number not skewed by wild swings like some monster toner super giant bringing big money.
I think they would have a market analyst to keep sanity / market trend in the numbers. I look at last, high, average. Then may drill down to see the particular coin if possible seeking why so high or low.
For something like a 1927 MS64 Saint will take Pcgs auction app look at several coins compare their quality and see what they went for. Here I will look at strike, luster, marks, any toning no gos or spots, etc. Then if bidding will compare that piece to those.
I suppose my issue is not about extreme price valuations since MS64/+ U.S.- Philippines Pesos do bring high prices.
Good luck(!) in buying these at PCGS price guide levels.
But the listed prices do a disservice to those not in the know (especially those who possess one.)
This was addressed (Mike Sherman???) recently in the PCGS Rare Coin Market Report . . . . .
They handled the topic with quite perfect logic . . . . .
Drunner