Weighting Issue -- PSA response requested...

theBobs

From the Hank Aaron thread...



<< <i>In my opinion PSA should come forth and explain the changes, their rationale, and how they calculated the weights....at the very least as good faith gesture to the high-end collectors (Vargha and others) who have promoted PSA and the Registry by displaying their amazing accomplishments. >>



I felt this deserved an appropriately titled thread due to the subject matter. Hate to see these ideas get lost due to the title... Excellent points MS and John.

HalleyGator

I know that I may be in the minority ... but it seems to make sense that a card that would cost someone $1,000 to find in PSA 8 condition should be weighted 100 times higher than a $10 card.

After all, if you took two 1972 Topps sets -- both of which were 90% complete -- and sold them at auction ....

but ONE set was missing 50 commons and the OTHER set was missing the 50 biggest Stars ...

would anyone in their right mind think that the collections were "Equal" ???

theBobs

I see your point, and I don't know the answer. I also don't know what the exact purpose of weighting really is. Is it to rank $ invested? Is it to rank the entire setbuilding skill? Heck, the SMR is laughable in some years. Should it be linked to the pop report and total number of registered sets? I mean if 24 people are building a set, and the pop of a card is 5 -- shouldn't that be weighted accordingly?


I do know that moving a card from 10 to 300 should be explained. When a PSA 8 of a single card is weighted equal to 240 PSA 10 commons something is strange.


I don't think there is a perfect solution. I just want a well thought out solution that is communicated to us folks builing sets...

mikeschmidt

Halley:

I do understand your point -- and I do think that there should be some sort of compromise.

But this revision in theory has basically made the Set Registry Awards/Finest Ever solely a function of money. If you have the most money, you win. Period, in most cases.

It simply places too much emphasis on high-dollar cards in baseball card sets. Any set collector will tell you that the most frustrating part of a set is completing it in PSA 8 -- it has *never* been about finding the Mantle cards in PSA 8 or better. But the way it is set up now, a 1951 Bowman Mickey Mantle PSA 8 by itself would have a higher set weighting than an entire run of commons and minor stars in PSA 8. Performing the first is simply a matter of spending a ton of money in an auction (and those cards turn up year in, year out). The latter, however, will probably cost more money and take years. There simply are way too many low population commons for anyone to just complete a 1950s set in PSA 8 -- it will take years, competition, trades, etc. to put together the last few cards.

Your 1972 set example is a bit erroneous. Sure if one set was 90% complete with no stars, and one set was 90% complete with all stars, it is very clear which one is the better set. But the example we are talking about is this:

1964 Topps Stand-Up Set: 4% complete: Has Mays and Mantle in PSA 8
1964 Topps Stand-Up Set: 96% complete: Has every card but Mays and Mantle in PSA 8

With the new re-vamped system, it is quite possible that the first "set" will have a higher rating than the second set.

Or, here's a better example:

1951 Bowman set: 100% complete, All PSA 8's with a PSA 9 Mickey Mantle card
1951 Bowman set: 100% complete, PSA 8's of all the major stars, and 275 PSA 9 MINT cards, the vast majority being 1/1's.

In that example, why should the first set be worth more? It just proves that you have six-figures to throw at a Mantle rookie at will.


Perhaps my points don't make sense to everyone. But this re-weighting really seems to me to be a shift to "reward" people for having lots of money to throw at cards, and not to reward collectors who build sets. If the goal is to massage the egos of the Charles Merkles, John Brancas, Marshall Fogels, Don Louscious, Tom Candiottis, etc. of the world, each of whom has million(s) in their card collection (with a few having a single set with an SMR > $500,000), than that's great. Because I think that that is all this re-weighting serves to accomplish.

HalleyGator

Schmidty:

I hear you loud and clear, and I thought this is why PSA was going to use a PSA 8 based scale that was linked to the value of the card in PSA 8 condition ... but which was STILL on a 0 through 10 scale!!!

In other words, I thought they would say something like: Any card with a PSA 8 price of $10,000 or more is worth 10 points... NOT 500 points if the card happens to be worth $500,000.

HOWEVER ... would it really be any different if they made the 1951 Bowman Mantle worth 10 points, the Mays worth 8, and the commons worth .03 ??

It would technically then be on a 0 - 10 scale ... but the weighting would still be skewed tremendously.

I can see where people will complain because the lower priced rare cards are worth less points....

But since the weights are directly linked to the SMR prioces, isn't the REAL COMPLAINT of those people that the MARKET PRICES of those low pop commons ???

If set collecting keeps driving up the prices for high grade commons on EBAY, then eventually the market prices and the SMR will reflect those increases, and presumedly the WEIGHTS will increase as well when that happens.

HalleyGator

Let me try again in ENGLISH since I can't type:

But since the weights are directly linked to the SMR prices, isn't the REAL COMPLAINT of those people that the MARKET PRICES of those low pop commons are WAY TOO LOW???

If set collecting keeps driving up the prices for high grade commons on EBAY, then eventually the market prices and the SMR will reflect those increases, and presumedly the WEIGHTS will increase as well when that happens.

zardoz51

I have to wholeheartedly agree.

Far too much emphasis is placed on the star cards.

Basilone

Regardless of our thoughts...I would still like to hear a official explanation from PSA.

John

69topps8

I think once the populations for commons get to sufficient levels (however that is defined), then tough commons should be weighted higher. For most sets (1960's on up), there haven't been sufficient quantities of commons graded to establish if a card is truly scarce. A good example is the '72 Larry Stahl. It had a population of 3 or so for a long time. Lately 8's have been auctioned on an almost weekly basis. I don't know what the population is currently but I'd guess there are at least 10 8's or higher now. I think it would be difficult to weight commons for sets that still have some 0 population cards. Over time, I think the populations will mature enough to establish the true rarities. I think the pre-1960 sets could probably be assigned higher weights to the low pop commons based on the current population numbers.

mikeschmidt

Halley:

You are basically correct. Unfortunately it is a very slippery slope in weighting/grading/pricing commons for any major vintage set.

"Commons" from 1955 Bowman, in PSA 8, routinely sell for $30 for the easy ones up to $300 for the low pops, popular teams, etc.
Unfortunately, population dictates the value of a common card, and every set I can think of from 1952 Topps to 1968 Topps to nearly every other major set has huge variances in the common card value. Thus, the commons will continue to have a lower SMR (you don't want to have an SMR over prevailing prices for too many cards), and the situation will remain stagnant.

I'll send Joe Orlando a link to this post, perhaps he can help us out.

MS

FB

Guys,

I've been watching this thread (and the Aaron PlayerSet thread) and I've been playing a little bit in EXCEL. Depending on how far we want to take this we could probably keep a team of statisticians at work for several weeks and still not everyone will be happy with the outcome.

To me, it looks like the missing component is an additional factor of current average population of cards in the set (in that grade or higher) divided by the population of the specific card (in that grade or higher).
It would be more volatile than what we see today (your set weight would change when the POP Report updates), but as John mentioned regarding the 72 Larry Stahl, the relative weight for that card would drop as more of them are graded compared to others.

Right now, the AVERAGE number of each card graded PSA 8 or better in the 72 set is roughly 19 (PSA 9 or better is 2.8). So, for the Yastrzemski which has had 37 cards graded 8 or better it winds up being multiplied by (19 divided 37) or roughly a factor of .5. So instead of a factor of 40 (grade of 8 * weight of 5) you'd get roughly a 20.

Or it might be better to "reward" commons with an extra multiplier based upon pop versus average.

Commons with populations of less than 25% of the average pop for that year give them a weight of 4.
Commons with populations of between 26 and 50% of the average pop for that year get a weight of 3.
Commons with populations of between 51 and 75% of the average pop for that year get a weight of 2.
Commons with populations of between 76 and 100% of the average pop for that year get a weight of 1.

Bottom line is that these solutions will help in some situations and not in others. But I honestly believe that relative population versus other cards and grades in the set need to be taken into account in order to more accurately reflect the relative weight of a set.

The downside is - that a 1 of 1 PSA 10 Hank Aaron Rookie in the 54 set is still going to carry a lot of weight - and it should.... But, the tougher to find commons will get more value as well...

Just my two cents....

BJ

Hello everyone,

As many of you know, the weighting on some of the sets has changed in the last week. What Joe Orlando is trying to accomplish is a more proportional breakdown in the weighting. But from this thread and some email we have received, it is clear that this issue is controversial and one that requires some more thought. Therefore, we have put a hold on any re-weighting for a few more days, so that Joe and others can discuss this some more.

The Set Registry is ever evolving and extremely important to us at PSA. Your wonderful participation has made this Registry what it is today and the participation is sure to increase as more and more collectors find out about it. That being said, all of you have a voice here and all of your opinions matter to us. Nothing is set in stone. The Registry weighting will never be an exact science, but we want it to be fair and make as much logical sense as possible.

In the end, maybe there is a "happy medium" to the weighting. Collector feedback is what helps steer all of our projects at PSA. We want you to be happy with our products and service and, if we share ideas and communicate, everyone wins. We'll submit updated weighting plans to you by email as soon as we can.

Thank you for helping make the Registry a huge success and please share your feelings with us.

FB

Leave it to BJ to be perfectly reasonable in the middle of a nicely developing rant!

waitilltheytrytosell

I don't think the pop on commons that are hard to find at the moment need to be taken into consideration. Rare commons have a frequent tendency to catch up. People tend to hold back commons and send in only stars. In many cases they are rare in graded form because nobody has bothered to send them in - 1964 Giants for example. Stars, highs, SP's - those type of cards are consistently in demand, and will consistently command higher prices as pops even out. Say you are the first one to complete a mixed 7, 8, and 9 sort of set, and then stop with say an average of 8.3. If you don't ever change a card, it should stay at 8.3. If you add a living pop factor, as more cards get graded and the pops increase, your number will continue to drop. It doesn't seem right to do that.

Vargha

So is that what happened to the 1951 Bowman set ratings? I thought that maybe PSA's programmers were smoking crack. Because one of the PSA 7 cards I have is Mays, I now have a 7.90 rating on a set that is a total of eight PSA 7's short of being PSA 8 or better, even though I have 34 PSA 9's! Whatever . . . I'm not collecting the set for the rating. But this method is ridiculous since it is weighting based on SMR and not on any marketplace reality. Was Branca ticked off because he had a PSA 9 Mantle and PSA 9 Mays yet still wasn't the top rated set? I believe that my #308 Beard PSA 9 (1 of 1 with only six PSA 8's) is worth far more than my PSA 7 Mays in terms of what it would sell for. I'm sure that the current rating system doesn't even have a clue about how to deal with such a discrepancy.

bosoxphan

Good thing you don't have the Mantle in a 7. Talk about an incredibly high card weighting. In one fell swoop the registry went from underweighting high priced superstars to overweighting them.

FB

waittil,

I think I have to disagree with your point on hard to find commons catching up. Of course we won't know till the loads of commons that are locked away all bubble to the surface, but I think that the gap will even widen on certain low pop cards as time goes on. Maybe some day there will be 65 PSA 8 1972 Larry Stahls. But, I think that you'll find that at the same time there will be 395 1972 Denny Lemasters.

Certain cards are inherently well centered. Whether thats because of placement on the sheet, the way that the machine cut the particular cards or the way the planets were aligned at the time they were created there is a certain consistancy for most cards in most years.

I have a 5000 count box of 72's here. I have roughly 30 Lemasters (#371) and I don't believe that any are centered worse than 65/35. On the other hand, I have 22 Paul Schaal In Actions (#178) and the best one of the bunch is 85/15 with most being 90/10 or worse. As I rifled through every 72 that I could find at the Fort Washington show looking for low pop commons to send in to PSA, my experience was just about the same - certain cards are just much tougher to find centered well enough to get the higher grades.

Also, by including an extra factor for lower pop commons that diminishes as the gap to the average is closed - the additional multiplier would also diminish and become less of a factor and then a non-factor as the gap narrows - while giving credit to the tougher - undervalued cards.
Again - just my two cents

Buckwheat

I believe that both of the prevailing arguments here are right. Weighting from 1-10 grossly underweights the stars. Grading strictly by the SMR grossly overweights the stars. A rational compromise is needed. Here are some thoughts.

Trying to value commons by population, while a laudable goal, is probably impractical. the numbers are constantly changing, and you would have to factor in whether they are yankees, etc. The commons will have to have a common value, whatever it is. If after some time, we find that some commons always sell for much more, they in fact are o longer "commons", but probably will earn their place among the semistars, with their own individula value in the SMR. SP's and high series certainly can be valued higher as a class.

Weighting from 1-10 is not particularly fair. An example of why is that the cost and effort to upgrade a Star from 8 to 9, far outweighs adding two PSA 5 commons, but adds less to the weighting. This is obviously not fair.

Weighting strictly on SMR is also inherently unfair. Mostly this is because cards sell for more than SMR on commons, particularly low pop, while often for much less than SMR on Stars. Also, as others have noted, they are often much more difficult to find than stars.

A compromise could be to take the differential ratio between commons and a star card, and dividing that number by two. For the 1961 set I collect, a low series common 8 is $22 in SMR, while the highest card, Mantle is $895. This makes a ratio of 40. A common 9 has SMR of 75, vs $3800 for the Mantle, for a ratio of 51. Taking half of these numbers leaves a scale of 1-20, or 1-25. Both seem like reasonable numbers. On a scale of 1-25, bumping the 25 card from 8 to 9 can be matched by adding 4 common 7's, or 3 common 8's. That seems reasonable.

In answer to what should the set rating attempt to measure. Perhaps we could all agree that eventually all top sets will have 100% completion. With that assumption, I would argue that the highest set should be the one that would sell for the highest amount at an auction. Since people are valueing a set they buy at auction in dollars, I would have a hard time if a lower ranked set would bring more dollars, and thus is valued more highly by collectors, than a higher ranked set. I don't think you can get around the fact that someone who spends big bucks on a set is going to have a better set than someone that spends less. I am constantly outbid on auctions for PSA 9's by several high dollar bidders. I would expect than when those people post their sets, they will be rated higher than mine. While I wish that I could have bought those cards, I can't say that I have nicer ones! So as egalitarian as we many want to be, the top sets will always be high dollar sets, and rightfully so.

That is my 2 cents, for what it is worth.

Buck

bosoxphan

I think that's a very reasonable solution.

Buckwheat

Thanks Bosox,

Perhaps on the really high ones, like the 334 that MS mentioned, there should be a cap of 50 or 100 or something.

theBobs

<< <i>I'm not collecting the set for the rating. >>



This is what it really gets down to. Nicely said.


As for the weighting monster, keep the idea machine churning. I am sure that BJ is listening...

vorthian

More emphasis needs to be placed on the "weights" of individual cards. Linking the SMR, as others have pointed out, is going to present problems because 1) some of it is way off and 2) the emphasis seems to be who can throw more into the hobby.

I have weighted many sets and I didn't even look at the SMR once. I used several factors, including the scarcity of the card itself. For instance, if you were building an 1910 e98 set - you would find Kling a tough common... in fact, you will find more Cobbs, Youngs, Wagners in the marketplace than you will Kling. Should Kling be weighted as much as Cobb? No, of course not. But it should have more weight than an average, easy-to-find, common. As should low pop 1915 CJ's in PSA 8 that are rarely encountered.

Commons are vital to set-building. For the most part, finding them in high grade is going to take more energy/time than simply buying a high grade HOF. The former should be rewarded, not the latter. As should a 1915 CJ Mordecai Brown in PSA 8 - because of the scarcity (in high grade) over an average HOF.

Careful weighting is the key here. Personally I would like to see an "average grade" placed somewhere in the listings so the "weight" isn't the driving force. Those with the highest average grade should also be given some notice, along with the one who has the highest weight. People throwing gobs of money into this hobby doesn't impress me, as it does many people - money is not everything. I am also not impressed by those people that constantly suck up to the "super collectors", having some kind of an agenda. Pathetic.

I just received word that Joe Orlando is in charge of the vintage "player sets" and their weights. I also disagree with this as it should be COLLECTOR driven. Those that are proud of their player sets and have invested years of their life chasing and attaining their favorite players should have the control.

In closing, I suggest a "panel" of several people who were in charge of the weighting system. This would eliminate the weighting of a set at will, as the emphasis is on the careful construction to the weights.

brucemo

I think the set registry is mostly stupid, because it is difficult to agree upon how the sets should be rated, and once we agree upon how the sets should be rated, it would probably be very difficult to rate them properly.

Once we get past that, the set registry is still mostly stupid, because the set registry competition diverts people from card collecting and gets them into the ego comparison game, which is a moronic waste of time.

Grading was supposed to be a means by which people can get what they pay for. It is to be expected that people will take it a step further and collect the grades -- this is what I do, when I collect sets in PSA-7 or whatever. Many people do this.

I think it is a perversion to use grading as a tool to measure the size of one's penis, which is in large part what some people have turned the registry into, and which is a practice which is encouraged by PSA. It does PSA no credit that it provides a place where people can primp and strut like this, and encourages them to do so.

bruce

zardoz51

Gee

C'mon Bruce, tell us how you really feel, don't hold back!

Wayne8348

Bruce,

I understand what you are saying and that probably is correct for some but not for everybody. I don't think it's an ego thing for most people on the registry. I enjoy seeing some of the awesome collections others have put together. I enjoy watching friends find key cards and seeing their collections grow after a lot of very hard work. I like the encouragement and help that I have received from other 1965 collectors with my set that I wouldn't have gotten without the set registry. I like keeping records of my collection on line where the records are safe. I really like looking at the sets of those who have scans. (I will hopefully figure that out one of these days myself.)

You are correct that it is difficult to agree on house sets should be weighted. There is no perfect system. But I don't see any problem with discussing that here and striving to come up with a more accurate system understanding that it will never be perfect.

I guess my overall impression of most of the people I have come into contact with through this registry is quite a bit different. Jay (Quality Cards) emailed me yesterday and said that he would sell me 1965 commons for $11 apiece to help me complete my set. He is building this set also. I would want to pay market value but what an incredible offer! I have many other examples. Anyways, some may thump their chests but others get a lot of enjoyment out of this for the right reasons. Is competition bad - I don't think so. I just haven't experienced the competition in the same light as you presented it.

Wayne

thegemmintman

Vorthian - the panel is an interesting idea. The panel would be composed of experts from PSA, correct? I hope you're not suggesting people outside PSA as I personally would not trust such a group no matter who they are.


Brucemo - "I think the set registry is mostly stupid."

Shame on you Brucemo! That's like saying mom, baseball, and apple pie are stupid. Shame Brucemo, shame. Now I want you to think about what you said, and get back to us after you've cooled off.

jaxxr

Would not the unique age of a set be a factor in weighting ?

A 1987 Topps BB set ( which has at least two active ) certainly can't have any real population differences. Only value/price can affect weights.

A 1978 BB set has some population history, but who sent in "commons" to be graded two or three years ago? It should probably be mostly value/price driven.

A 1965 BB set has large, proven population history. Star cards and Hi Nos. are fairly well set. The weighting should probably be equal between value/price and population.

A 1953 Topps BB set has a small number of cards relative to 1959 and later . Sets from this time frame, roughly, have a long history of collecting and thus a strong, reasonably accurate population data. The weighting should be mostly population versus value/price.

"Population" info might be more valuable if the other two major grading companies' data was combined. It would be quite sophomoric to think there are no other cards worthy of an 8 or 9 grade, just because they aren't in a PSA holder.

HalleyGator

The more I think about, the more I think that the SMR-based weighting is really only needed on sets that "span time."

In other words, in a set that spans only ONE year (such as the 1951 Bowman set or any other annual set) ... the 1 - 10 weighting works just fine because EVERY CARD in the set is just as old and just as rare and just as "weathered" as any other card. There is NO WAY that any "star" card is that much harder to find that any common.

BUT ... on the player sets that span generations (like Aaron from 1954 to 1976) and the HOF set (from 1870 to 1997) ... there really IS a NEED for some cards to be weighed MUCH MUCH higher than others. A scale of 1 - 10 just isn't going to work.

A PSA 9 1887 "Four Base Hits" card of King Kelly is unequivocably a ONE OF A KIND card. It simply can NOT be worth only 10 times what a PSA 9 1995 Topps card of Kirby Puckett is worth (since there would be MILLIONS JUST LIKE the Puckett card)!!!

Hopefully BJ will try to implement the SMR-based weighting into the "multi-year" sets and see how it works there.

Vargha

It looks like the 1-10 scale is back in operation.

brucemo

To those of you who replied to me, I still fee the same way.

I think that competing to create the best set is perverse. Collecting is about collecting, not about crushing the competition.

Some of you won't agree with my previous sentence, which is fine. For those of you who don't agree with it, I argue that assigning "rating" to sets is undesirable because:

1) Doing it by totalling up the SMR prices is crass. Assuming that raw money can be applied to the problem, all you are doing is figuring out who is richest and most committed to applying cash.
2) It will be difficult or impossible to agree upon what "hard" means and figure out exactly which cards are "hard", if you want to do it some other way.

I think that a beautiful set is a beautiful set, and can be appreciated without getting too crazy about assigning numbers to it. The sets should be enjoyed, not statistically analyzed.

Why am I reading the set registry forum if I hate the set registry? I don't hate the set registry. What bothers me is that people are treating card collecting like it is a video game, the goal of which is to get the high score.

Listen to this. There is a segment of the PSA collecting/following community, that when these people see a card in a PSA slab, they glance at the card and dwell on the grade on the slab. The point of the card is not the card, it is the number on the slab, and that is all that they see. We should be trying to get away from this, rather than trying to perfect the algorithm used to create the final rating.

I think that the final set rating should be HIDDEN, and the condition breakdown, by series and by keys, should be given somewhere instead. Completion should count for more, too.

bruce