Here is the basis for my claim  to possessing that sort of expertise.

A circular basis for the same point is my growing list of clients. The idea here is that if more and more notable companies are relying upon my expertise in assessing patent quality, then I must possess sufficient expertise in law, business, and technology because the work demands it.

Actually, if you click the first link above, you’ll see that the areas of my expertise go well beyond law, business, and technology. Hence the term “Polymath” in the name of my company.

But “Jack” in the company name arises from the answer to the following question: How does one acquire expertise in such a wide diversity of fields at such a relatively young age (48 as of this year)?

Answer: By being the Biggest Loser.

Explanation: I was a pretty good computer scientist. Graduated in the top 1% of my undergraduate class. Did a Masters, and, recently, developed a few websites. (I was even a Rhodes Scholar finalist back in the day.) But no company today would hire me as a head of development. (And I didn’t get the Rhodes.) In terms of technology, I’m a hack.

I was pretty good law student and lawyer. Graduated in the top 10% of my Stanford Law class, clerked on the 9th Circuit, and worked for Fenwick & West in various IP legal matters in various capacities. But no law firm today would recruit me as the head partner of their IP department. In terms of IP law, I’m a hack.

I was a reasonably useful business guy. Held various strategic and business development roles within startups, big companies, and R&D units. But no company today would recruit me as CEO. In terms of business, I’m a hack.

I was a pretty good basketball player. Played on my national team. Played professionally in Europe. But I played only one year on each, and I never played in an Olympics nor in the NBA. Meaning? In terms of basketball, I was a hack.

I might well be the healthiest 48-year-old that you know. I measured my resting pulse the other morning upon waking — it was 38. My body fat is 12%. Blood glucose, blood pressure, and cholesterol readings are all optimal. In fact, virtually all 50 of my annual blood test readings are excellent. None of this is an accident of my genes. But nobody today would hire me as their health guru. In terms of personal health, I’m a hack.

I’m the inventor of a model linking the Enneagram theory of human personality with emergent findings in neuroscience. Personally speaking, I believe this work is my greatest yet, in any domain. Put simply, I think I’ve stumbled upon a core, eternal, truth about humans (and all vertebrates). But nobody wanted to publish my book, and very few have exhibited interest in this work. In terms of the neuroscience of personality, I’m a hack.

I could go on here. But you’re starting to get the picture, I hope. The answer to how one becomes a passable polymath is by attacking new fields long enough to achieve reasonable mastery, but then leaving the field well before the brass ring looms on the horizon.

[Digression: I re-read this list and wonder if I suffer from lifelong brass-ring-o-phobia.]

Sometimes, my clients ask me: “Do you know anyone else who can do this work that you do for us?”

I answer that I don’t know any. I haven’t met any other losers who are as big a loser as I am. I mean, I’ve met losers who are proficient in a couple of fields. That’s pretty common.

But rare is the loser with a list the likes of mine.

In 2010, NPEs represented approximately 18% of patent litigation in the United States, with 47 such cases in 2000 rising dramatically to more than 550 cases in 2010. NPE litigation has increased almost 500% since 2001.

Many people complain about NPEs (a.k.a “trolls”). But few get to the heart of what the problem is and why it exists in the first place.

My belief is that the fundamental problem is the stretching of low quality patents and that this problem is fostered by secrecy.

Secrecy is a fact of the legal world. But although necessary in many circumstances, it is also overused, and serves as a cover for various flavors of malfeasance, like incompetence, corruption, and thuggery.

Personally speaking, I don’t have any problem with trolls or NPEs, per se. The “problem” I have is with patent litigation where the patents are of low quality.

I don’t care for patent litigations relying on patents of low quality  that must be stretched beyond reason to reach the targets. Bringing such cases is quite feasible, however, because courts pay little or no attention to the patents at issue for at least a year of litigation, during which a great deal of attorney fees will have been spent.

So these patent-stretching lawsuits are filed, but the low quality nature of the patents is kept secret. The press doesn’t know how to read patents. And the court process is broken.

Basically, IMHO, the root of the problem is that the cops (courts) are asleep.

Now, I don’t see a problem with an NPE suing over a quality patent with solid reads on the targets. If I have any problem with that, then it’s a problem with the whole idea of the government granting patents in the first place. But so long as we live in a world  in which the government does that, then I’d like to see that world act rationally.

But secrecy defeats rationality.

So what we see repeatedly in the patent world is companies initiating patent lawsuits using low quality patents that were never envisioned to reach the targets. While dominated by the NPEs, this problem is not limited to the NPEs.

Companies, even “blue chip” ones, do this too. They “do this” when they take a patent in their portfolio that covers some feature of technology/market X, and stretch it to read upon technology/market Y of a target — even though the patent was never written to contemplate technology/market Y.

When such plaintiff companies are not even active in market Y, then they are, for all intents and purposes, “trolls” — even if in their own markets they are blue chip companies.

What makes these patent-stretching plaintiffs such a pernicious problem is that they have little or reduced incentive to settle, so long as they can continue to fund their litigations. Classic trolls do this by naming a great many defendants, and obtaining settlements from some of them. These settlements fund the litigation against the remaining defendants.

Real companies fund their “troll” litigations by diverting some of their revenue to pursue the questionable litigation.

Such patent-stretching plaintiffs are typically not subject to counter-claims because they are not practicing the business they are targeting with their patents. That makes them appear like ghosts to the targeted companies.

To me, the best solution to the patent-stretching problem would insert a patent quality check in the process. For my money, the best stage at which to insert this check would be the motion to dismiss stage immediately following the filing of a complaint.

I’ll have more to say on this notion in later posts.

The short answer is: marketing.

As I noted in the previous post, the work of assessing patent quality is esoteric in nature. Not many people can do it well because it demands polymath expertise in three areas: law, technology, and business.

Because the work is esoteric, it’s hard to describe it in the sort of pithy terms one would employ in Google AdWords. Moreover, because it’s such a rare skill, it seems not widely known that the work demands this tripartite expertise.

This is why, I believe, one regularly find lawyers or engineers tasked with the job of assessing patent quality. Such people usually possess only one or two of the necessary skills, but not the third. This makes their work deficient. But they don’t know it.

This is also why, I believe, one regularly finds news stories discussing a noteworthy patent litigation without so much as mentioning patent quality.

The final kicker concerning marketing my skills is that my work is done in secret for companies, some of whom are more or less paranoid about secrecy. The last thing they want to do is market my skills to others.

Ergo blogging examples of my work.

The other reason for blogging this work is that I’m a blogger. I have a socio-political blog. A health blog. A Twitter account I use to record personal health data. I’m the author of an hypothesis linking a theory of human personality to the brain.

The question isn’t why I should blog about my work. The question is: What took me so long?

The answer is what I’ve said above. My work is confidential in nature; not something to be blogged about. And I didn’t figure out this marketing angle until recently.

Let’s go back to the NBA analogy. How do we sports fans know which players are of the highest quality? We watch how the players perform in the crucible of the playoffs — particularly in the Finals.

For patents, the analogy is patent litigation. The degree of a patent’s quality is ultimately determined by a court during patent litigation.

So the process of analyzing patent quality involves imagining how that patent will “do” in future litigation. This includes assessing the claims of the patent, seeing if and how well they “read” upon the products/services of “targets”, and whether those targets are of value.

Read the last sentence closely and you’ll understand the esoteric nature of this work.

First, legal knowledge is required. That is, the person assessing the quality of a patent needs to know the legal rules explaining how to read the patent.

Second, technical expertise is required. Most utility patents concern a particular technology. Understanding the claims of the patent requires understanding the technology.

This technical requirement alone explains why courts flub patent litigations as a matter of course. Neither the judges nor their clerks typically understand the technology underlying the patents in a case, so the case plods along in darkness.

The third and final type of expertise required for assessing patent quality is business knowledge. This need is particularly poignant before the litigation is filed. The problem of assessing the value of a patent requires understanding the business landscapes to which the patent might apply, and knowledge of the value residing in those landscapes. This is the terrain of business analysts.

How many people do you know who are experts in all three of law, technology, and business? The only such person I know is me. I’ll blog more on this later.

But just realize that there are few people who can do a superior job of assessing patent quality. Hence, esotericism.

Moreover, those people usually do this work under contracts with strict confidentiality provisions, and/or under attorney-client privilege.

In other words, secrecy pervades the patent world.

So for me to blog about patent quality in public litigation cases, I’ll need to pick patents and cases I haven’t worked on, and spend the time to do the work.

Now this isn’t light work. The work is esoteric and difficult. When I do this kind of work, I typically need my brain to be at its sharpest.

For me, this means doing the work first thing in the morning upon waking. Compare that with this blog post that I’m writing on a Wednesday afternoon, using my more sluggish post-lunch noggin.

So the question is: Why would I want to do this kind of “heavy lifting” work and blog about it for free, when that’s how I make my money? Answer is in the next post.

Patent litigation (with a cross-claim) is like a sporting contest, with the refs as the judge, and players as the patents. Can you imagine coverage of the sporting contest accompanied by no discussion whatsoever about the quality of the players? Spend 5 minutes on ESPN and you’ll notice that 90% of the discussion is about the quality of players.

But not in the patent world. In the next post, I’ll discuss why I think that is. But for now, understand what I’m planning to do.

Put simply, I’m going to take public patent litigation cases, look at the patent claims being pressed, and offer my opinion on the quality of those claims.

I believe this will be interesting because patent litigation, especially the so-called “troll” kind, seems to be on the rise. If that’s true, I’m suspecting it’s a result of the economic downturn — the notion being that if we can’t make money the normal way, then there’s always using the courts to shake down deep pockets.

Another reason this line of blogging might be interesting is that distribution of quality among a typical portfolio of hundreds of patents looks quite a bit like the distribution of talent among NBA basketball players. In the latter distribution, there is a small handful of superstars (e.g. LeBron, Kobe, etc.), a larger pool of good players, with the vast majority of the league being “replaceable parts.”

In the patent world, there are a few diamonds, a larger pool of patents of reasonable quality, with the vast majority being worthless pieces of junk (i.e, the rocks in the quarry).

A worthless piece of junk as basketball player will not even merit a “cup of coffee” in the NBA. But a worthless piece of junk patent might well make the front page of the Wall Street Journal.

The reason for the former is that any common sports fan can spot a piece of junk of a player a mile away. But very few people can do the same in the patent world. This is the next topic.

For each service provider, I computed the monthly fee that our business pays them. Turns out that 94% of these costs ($1027) goes to two health insurance companies (Anthem Blue Cross for medical; Blue Shield for dental). (Our company pays the health insurance costs for our employees).

The remaining 6% ($67) of this monthly total pays for the following 9 services:

1. bookkeeping (ClarityAccounting)
2. billing (FreshBooks)
3. payroll (Intuit Payroll)
4. domain names (MyDomain)
5. telephone (Skype)
6. web site hosting (Slicehost)
7. banking (US Bank)
8. web site statistics (Google Analytics)
9. email (Google Apps)

As our business grows, this $67/month total will certainly grow some. But it sure as hell won’t ever approach the cost of health insurance.

Of course, the above 9 services are all web-based — i.e. services delivered via the web. Dental and medical care will never be delivered solely or even mostly over the web since we humans aren’t digital.

So count me among the Americans hoping Obama will grow a pair over the next month and get a fully funded public health plan rammed through Congress over the screaming objections of the insurance companies, pharmaceuticals, and the AMA.

I guess since I come from Canada, having moved to the US for law school at 25, I don’t buy into the doomsday propaganda coming from these interested parties about how bad a public health system will be for us. To me, they’re like the mafia, to whom we are paying “protection” money, claiming that crime will go way up after they are booted out — when they are ones committing the crimes in the first place.

It’s fun running a business in the age of the web.

Take Ocean Tomo’s 2009 Spring Auction which it now calls “IP Market 2009“. That auction was held almost 5 weeks ago (on March 26-27). Yet on Ocean Tomo’s auction results page of today, there is no mention of this auction.

Perhaps you’re thinking that 5 weeks is not enough time for Ocean Tomo to process its data from the March auction. Maybe I’m being too hasty calling them out for hiding the results of that auction.

Well, think again. Look at their press release for the Fall 2008 auction (PDF). It was issued on November 3, 2008 — a mere 4 days after the auction concluded.

No. The reason why Ocean Tomo is hiding the results of their Spring 2009 patent auction is that that auction was an unmitigated disaster for the company. Almost no patents sold at the auction.

One reason I know this is that an industry colleague who attended the auction told me this. But you don’t need even that sort of “insider” knowledge. When a company can’t wait to tell you its results in the good times, you can pretty well deduce the meaning of silence.

Now, in my last posting – Trouble on the High Seas — I noted that Ocean Tomo’s own data from its past auctions indicated its auction business was in sharp decline, foreshadowing trouble for the upcoming Spring auction.

Hate to say I was right, but …

What’s most baffling to me about the demise of this business is that the problem with it seems so obvious (see my earlier posts). Over the past couple of months, I have tried to contact the Ocean Tomo auction staff to raise these issues with them.

What was most amazing to me is that the most responsive Ocean Tomo employee to the emails I was sending to the company was the CEO himself! People under him, to whom he forwarded my mails, simply didn’t respond.

I have never encountered a company in which the CEO graciously and promptly responds to his email, but his staff doesn’t.

Call me a cynic, but I don’t think this company is long for the patent auction business.

Ocean Tomo kindly provides us with the results of its past 6 Fall and Spring patent auctions. Its press releases tell us which patent “lots” sold at the auction, and for what price. What those press releases don’t tell us is what happened to all of the patents that had been listed in the catalog, but that do not appear as sold on the ensuing press release.

It doesn’t take Sherlock Holmes to figure out that the “unmentioned” lots did not sell at the auction. That would explain why Ocean Tomo doesn’t say anything about them in the press releases.

So, for each of these six auctions, I looked at the percentage of patent lots (out of all lots offered for sale) that didn’t sell. Here is what the data told me (subject to possible immaterial typos):

• Spring 2006: 66%
• Fall 2006: 65%
• Spring 2007: 49%
• Fall 2007: 51%
• Spring 2008: 38%
• Fall 2008: 58%

From startup of the auctions in Spring 2006, through to Spring 2008, this looks like the story of a business gradually getting its shit together. That is, figuring out the best practices for selling patents via auction.

But just look at that Fall 2008 number. Almost 3 out of every 5 patents went unsold at Ocean Tomo’s most recent auction. Ouch!

That sure looks like an ominous number last Fall. So I studied the press release for that auction looking for some indication from the company explaining just what the heck happened. But as I read the release, it sounds like the sort of “skies are blue, the sun is up, and oh what a lovely day” drivel that PR people pump out in their sleep.

Well, with no help coming from there, let’s go deeper into the numbers. Let’s say you were among the lucky 2 out of every 5 owners who actually managed to sell your patent at last Fall’s auction. How did you do?

In its auction catalog, for each patent lot, Ocean Tomo provides  a figure it calls “Expected Value”. For example, Lot 6 in the Fall auction sold for $192,500. But its “Expected Value” was “$300,000+”. I assume this means that the owner of Lot 6 expected the patent to fetch over $300,000. But as it turned out, the patent sold for much less.

Actually, the owner of Lot 6 was among the lucky Ocean Tomo customers last Fall because $192,500 is at least within earshot of $300,000. Most of the others owners fared even worse.

A little digression here. I believe that “happiness” is simply the difference between what we expect, and what we get. Moreover, the degree of happiness/sadness correlates with the size of the gap between expectations and reality. With me so far?

Assuming that those “Expected Value” figures in Ocean Tomo’s catalog are expectations held by the patent owners, then it stands to reason from the data that Ocean Tomo auctions are a machine for generating very sad customers.

For example, the Fall 2008 auction had 15 web-related patent lots for sale. (Web patents are the specialty of my business, Jack Polymath LLC). Of those 15, 9 sold, and 6 went unsold (i.e. the web patents did much better than the non-web patents in that auction in term of being sold). Of the 9 that sold, 8 provided publicly disclosed “Expected Values”. The lot summary for the other said only: “Please Inquire Regarding Expected Value”.

Now, for the 8 web-related patent lots that sold in Ocean Tomo’s Fall auction that disclosed an Expected Value, the average selling price was $160,875 (median $170,500); the average Expected Value was $350,000 (median was the same).

So, on average, the “lucky” sellers of web-related patents who actually managed to complete a sale at the Fall auction got a selling price that was 50% less than what had been expected. i.e. It was a sad day.

Of course, maybe I’m being too literal here.  I mean, maybe that “Expected Value” figure isn’t really expected by anyone. Maybe that figure is more like the prices thrown out by the New Delhi street sellers that I encountered on my trips to India in the late 1990s.

Maybe Ocean Tomo’s high-falutin’ “21st century” patent auctions are, in reality, more like a 17th century “Third World” bazaar in which hawkers shout out the sorts of prices that only fools would pay.

I don’t know. I don’t work there. I don’t know anyone who does.

I live and work in Silicon Valley, California. And I must say, having pored through the numbers that Ocean Tomo kindly puts out every half-year on its auctions, I’m baffled why that company evidently seems to think its doing a good job.

I don’t mean to crap all over Ocean Tomo here. Like I said, I appreciate that they are a leader in publicizing this kind of data. The brokers in this space, who seem to be kicking ass and taking names, are as secretive as Swiss bankers. Very little but a wall of silence is coming from those folks on the data. This is why I appreciate Ocean Tomo.

In fact, I appreciate that company so much that I’m going to explain exactly what is wrong with its approach. As I said at the start of this post, the problem comes down to its 2-3 page “lot summaries”. I went to the trouble of skimming through the web-related ones for the Fall 2008 auction — looking for some clue as to why this auction business seems to suck, big-time.

I noticed, reading these summaries, that they have it backwards. That is, the summaries try to sell the patent as a future vision. The standard argument boils down to: “In the future, lots of companies will be reading on this patent.” Summary after summary, that is the core message.

Memo to Ocean Tomo: That’s the wrong message, dude! The correct message isn’t who might read on the patent in the future. The correct message is who is currently reading on the patent right now.

For example, for the web-related patent lots for last Fall’s auction, those patents were filed in the late 1990s through early 2000s. By the time they were being sold, it had been 5-10 years after the original invention.

Memo 2 to Ocean Tomo: The web moves really, really fast, man! Inventions conceived a decade ago related to the web are either currently being practiced, or they never will be practiced. In other words, either these patents are the shit, or they ain’t worth shit. And which is which comes down to this divide.

You’re welcome.

However, the biggest shortcoming of that conference wasn’t its scattered nature. After all, that nature describes the overall patent market today. So the conference merely reflected the market in its current state. No shame there.

Moreover, I suppose I can’t slag the conference for missing the biggest story of all in patent valuation. This is because the press release that made this story the biggest one didn’t come out until one week before the conference, and did so without any fanfare. So, it’s not surprising that the conference missed it.

Nonetheless, there were some subtle hints of this biggest story at the conference. Those subtle hints concerned the “swagger” of the patent brokers during the conference. That is, the conference was filled with very different kinds of players. Academics, expert witnesses, corporate counsel, IP investment banks, etc. Brokers were just one species among many.

But among all the species there, the brokers stood out in terms of their confidence — a confidence that bordered on arrogance. And I don’t mean just one or some of the brokers. I mean all of them. If a non-English speaker had attended the conference, and thus hadn’t understood any of the words spoken, that person could still have picked up that, as a class, the brokers carried themselves like “they got game”.

Why? Is swagger and overabundant confidence merely a common personality trait of all brokers? Perhaps. After all, “broker” is just a glorified name for “sales guy”. But I think that, beyond the familiar sales guy bluster, there is a powerful reason justifying the demeanor of the brokers at the conference.

That reason is found in a single paragraph, in a January 29 press release (PDF) issued by leading broker IPotential, LLC. In this press release, among other information, IPotential announced that, over its 5-year life, it has handled 137 patent transactions, bringing in $265,000,000. That works out to an average price per transaction of $1,934,307.

Now, keep in mind that these are not audited numbers. These are numbers being put out by a private company. For all we know, these numbers could just be ones that IPotential is pulling out of its ass. Still, we have no reason to doubt these numbers. So let’s proceed.

The other caveat is that, to my knowledge, no other brokers have disclosed even these skimpy numbers. For example, Inflexion Point Strategy, LLC, tells us only that “We have handled the IP aspects of over $50 billion worth of high tech joint ventures, mergers, acquisitions, divestitures and investments.” As opaque as that is, Pluritas, LLC, is even less forthcoming.

But since the swagger of both the Inflexion Point and Pluritas people seemed to match that of the IPotential people at the conference last Friday, I tentatively suspect that the numbers of these “silent” brokers are closer to those of the “speaking” broker (IPotential), then to those of Ocean Tomo, LLC.

Ocean Tomo owns and operates a well-known patent auction. These are live auctions, held generally twice a year (Fall and Spring). Between 2006 and 2008, Ocean Tomo held six such live auctions. And it is kind enough to post the results of these auctions on its web site.

Now while these results are not audited either, they are already public in a way. This is because many people attend the live auctions, so many are already aware of the prices that individual lots of patents are going for at the auction, before Ocean Tomo announces those numbers on its site.

The other reason I suspect that Ocean Tomo’s numbers are accurate is that they are so shitty in comparison with those put out by IPotential. I mean, if Ocean Tomo was lying about its numbers, would’t it lie a little more beneficially to itself?

Anyway, on with the bloodbath. You can get the results of Ocean Tomo’s auctions from 2006-2008 here. Going through the six Fall and Spring auctions for those three years, we find 224 patent transactions going for a total of $82,429,600. This computes to an average price per transaction of$367,989.

This means that, in selling patents,  the broker IPotential is (allegedly) getting a price per transaction that is OVER FIVE TIMES what Ocean Tomo’s auctions are getting.

That’s not a typographical error. In fact, the truth might even be worse than that for Ocean Tomo. Ocean Tomo’s average is skewed by a single $16M+ transaction that is wildly beyond any other price fetched for the 224 transactions. If we drop this highest price and the lowest price from Ocean Tomo’s results, its average price drops to $296,970, which is 6.5 times less than IPotential’s number.

What’s going on here? This kind of difference in price is astounding. This question should have been the story of the conference.

Well, since it wasn’t, I’ll try my hand at answering this question. As I mentioned above, I suspect that the numbers for the other brokers are closer to those of IPotential’s than to Ocean Tomo’s. In other words, I think there is something systemic going on here.

Basically, I think the answer to this question came out of the conference. As I mentioned a couple of posts ago, there seemed to be general agreement among the brokers and corporate counsel that the most important thing that brokers do in selling a patent is to create a marketing package that features claim charts. Ocean Tomo doesn’t do this.

Why might claim charts account for the huge difference between the numbers of brokers, and those of Ocean Tomo? I think the answer comes from something IPotential CEO Ron Epstein said at the conference. He said that bidding was the key to getting value for a patent owner, just as bidding is the key to value in all markets.

How do claim charts help create a bidding dynamic? Well, the parties identified in the claim charts are motivated to bid because they allegedly infringe the patent being sold. In contrast, the two- to three-page “lot overviews” generated by Ocean Tomo tell us no such thing.

Could it be that simple? Could this simple difference in marketing approach explain the (justified) swagger of brokers? In a market full of very smart people, this explanation doesn’t seem likely.

However, parahrapsing E.O. Wilson, we people are sheep, after all, and there’s nothing like the “crowd mentality” to make even the smartest among us miss the most obvious of things.

Looked at close enough, and also from far back enough, everything is everything else

This fuzzy philosophical idea becomes concrete when we compare the topic of the last post — Patent Valuation Comes Down to Litigation — with the field of data compression. This post explores this correlation.

The previous post discussed the following three approaches to patent valuation:

1. Litigation
2. Manual Transaction
3. Data Analytics

Essentially, this list represents increasingly “lossy” representations of an original signal. In this case, the original signal is patent Litigation. That signal comprises all of the testimony at the trial concerning infringement, validity, and valuation, and the ensuing judgment.

In the world of music, the original signal can be the analog music playing on a legacy record, or the lossless representation of that sound on a compact disc.

Moving on now to lossy techniques, the claim charts used in Manual Transaction can be seen as a lossy representation of potential, future litigation. Although these claim charts are “lossy”, they still retain high fidelity to the prospective original signal. In other words, they are a rather accurate roadmap of the contours of future litigation.

In this way, these claim charts are like MP3 representations of music. MP3 is a lossy format for representing music. But the fidelity to the original signal is so strong that they suffice for most music listeners.

Finally, the third approach to patent valuation — Data Analytics — involves extremely lossy techniques. At this stage, the original signal is faint, if at all recognizable. But even so, this “highly compressed” data still reveals some sense of value. In the case of patents, this data concerns forward citations to the patent.

In the case of music, an example of an extremely lossy compression technique that still retains valuable information is digital fingerprinting. Just as a human fingerprint uniquely identifies a human without telling us anything more, so does a digital fingerpint of some music or a video.

And there we have it. We know which patent valutaion techniques are most useful because the sound ones line up just like increasingly lossy data compression techniques.