More Inventors Increases Likelihood of Patenting

I created the chart above by collecting a large set of published patent applications (1.7 million applications) that were all published 2005-2011 (applications filed prior to 8/1/2009) and then looked to see how many inventors were listed on the application and whether the application had issued as a patent. I did not consider family-member applications.

25 thoughts on “More Inventors Increases Likelihood of Patenting

  1. 6

    In my experience, the patents increase the investors. It’s so much easier to get money once you have the issued patent. Good thing we’ve halted all software investments, huh?

    1. 6.2

      It’s so much easier to get money once you have the issued patent.

      We’re trying to promote innovation, not wealth transfer.

      1. 6.2.1

        Because innovation happens out of the goodness of one’s heart and is unconnected to wealth transfer, don’tcha know?

        Come people, where are your priorities? Listen to Jane and everything will be just alright now.

        /sardonic bemusement

    1. 5.1

      That was my first thought.

      Differences this small could be explained by just about anything, but obviously the more inventors there are, the greater the likelihood that one of them (or their employer) has the money to prosecute the case to issuance (and then file a continuation). In addiition, it seems likely that more pro se cases (of the sort that should never have been filed) are filed by single inventors.

      1. 5.1.1

        Probably more inventors there are the greater possibility that it is a corporation.

      2. 5.1.2

        For a good example of that pro se filer, see the resurrection burial tomb disclosed by Rev. Izzo in 2005/0027316. Of the 36 drawing sheets, FIG. 2 probably is the best.

    2. 5.2

      does not equal

      Interesting thought – but compare that thought to Lemley’s writings behind the AIA push (on who and how prosecution of applications may proceed – with less impact to getting inventorship incorrect) and the view that in today’s world, it takes a village…

  2. 4

    Dennis — would it be possible to run a comparison between single inventors (and so more likely to be independent inventors and “more than one” (so more likely to be corporate) inventorship?

    Also; wasn’t there previous comments by some that many times with corporate inventions, many folks are “thrown in” to the “inventor pile” for self-aggrandizement / keep the troops happy purposes?

  3. 3

    I doubt it. From an examiners standpoint we neither confirm they all contributed nor treat it differently (other than an additional search term). You could turn any single inventor invention into a double or triple and the office would neither know nor care. You could not include all the inventors of an invention and the office would never know.

    In all likelihood its either that when more people look at a problem they’re more likely to self-regulate out obvious things, or that multiple inventorship coincides with assignment to larger corporations who hire different counsel and have different monetary considerations. If you’re a startup you probably need a certain scope, if you’re a large corporation you might take a scope just to have a portfolio. i.e. The patent “effect” is far down the chain from the actual cause.

    1. 3.1

      Are you sure that you do not treat it differently? Isn’t there a part of examination related to an inventor’s own previous work that must be so considered? Isn’t co-ownership likewise a consideration that an examiner must be aware of?

      Not to tell you how to do your job… (but you should know your job better than you apparently do).

      1. 3.1.1

        Isn’t there a part of examination related to an inventor’s own previous work that must be so considered?

        nor treat it differently (other than an additional search term)

        Isn’t co-ownership likewise a consideration that an examiner must be aware of?

        If they’re represented by the same counsel (which they will be, because it’s assigned to a third party) there’s nothing about co ownership that isn’t covered by the additional search term.

  4. 2

    The numbers are surprising close, and the small difference might be due to the patents with plural inventors being more likely to be from larger R&D entitities that can afford better PTO representation? Including more care in adding dependent claims, which can increase co-inventorships for those claims. More originally filed dependent claims can also increase the odds of getting a patent by providing fall-back protection from the broadest claims being rejected on art or 101.

    1. 2.1

      good points paul – especially the aspect of better enabled to play the game of kings (through in that game the notion of the likes of Microsoft and IBM to be more likely to play the numbers game and aspire to obtain even silly patents just to inflate their numbers)

    2. 2.2

      It’s not just better PTO representation – it’s better R&D facilities, better earlier work (probably patented) to build on, and a better process for determining when to spend money on patent protection. Nobody’s inventing a new semiconductor manufacturing process based on something they did in their garage.

      I’d personally be interested to see data that further broke up these categories by entity size.

      On a side note, in my experience, true single-inventor applicants (i.e., not backed up by some large corporation) are somewhat more likely to file nonpublication requests, which would skew Dennis’s data by causing the abandoned applications not to be counted.

      1. 2.2.1

        Good point about the non-publication skew Apotu, one that likely will escape finer analysis, although an appreciation of the total number (or possible error size effect) may be gained given as application numbers are given in sequence.

        1. 2.2.1.1

          I agree with all of the above comments.
          I was even told that a surprising number of individual inventors do not even make it past the provisional application stage.

          1. 2.2.1.1.1

            Yes paul, but quite a bit of that was related to the invention-registration non-law firm sc am artists awhile back that promised the do-it-yourselfers “patent protection” when what you they were selling were provisional filings.

            1. 2.2.1.1.1.2

              the invention-registration non-law firm sc am artists awhile back that promised the do-it-yourselfers “patent protection” when what you they were selling were provisional filings.

              These are totally different from the firms that promised people that they could patent any claim that recited “on a computer configured to” because, “hey, you can put a computer on your desk”, but what they were really selling was t 0il et paper.

  5. 1

    What is the margin of error?

    I ask because the numbers are close – your headline might be accurate, but it also might not be accurate, or it might not be ascertainable as to an actual difference if the margin of error bars were to be shown.

    Lots of fun could be had with such a dataset…

    1. 1.1

      In terms statistical sampling (the usual reason for error bars), there is no margin of error here because I used entire population of published patent applications (during the representative time period) to calculate the results. If it were only a sub-population sample, the very large sample size (1.7 million) means that the average rates calculated here would have extremely small error bars and would not overlap one another.

      However, there are likely a number of confounding variables that were not considered by this simple calculation. In particular, you might ask whether there are certain technology areas that have a higher grant rate and that also tend to have a greater percentage of group inventors.

      As others have noted in the comments already, a higher grant rate for group-inventions does not necessarily mean that those inventions are more-inventive or higher quality. Rather, a good explanation for the difference is that groups (and those who work well in groups) have more political and perceived value within an organization with the result that those applications are given more resources in the prosecution process.

      1. 1.1.1

        Good response, I did not read “a large set” as meaning “every item,” and as Apotu points out, not every item has been accounted for.

        Other errors may exist as well. For example, items filed during that time period that are yet in process (you identify another type of “error” in noting that you have not accounted for family effects – as group activity naturally seems more likely to induce divisional applications, and any accompanying change in inventorship which should happen may not happen for a number of reasons.

        Of course, the one inventor=non-corporation presumption itself contains classic Type I and Type II errors.

Comments are closed.