BPAI Appeal Statistics: The Plummeting Reversal Rate

In an earlier post, I examined what happens to patent applications after the BPAI reverses an examiner rejection. [LINK] This next post looks at the BPAI reversal rate itself. Here, I define the reversal rate as the percentage of cases where the Examiner rejections are completely reversed (as opposed to partial reversals).

Over the past several years, the BPAI reversal rate has dropped dramatically. From FY1999-FY2005, the reversal rate remained relatively steady between 35% and 40%. Then, in FY2006 we began to see a smaller proportion of reversals. By the first two months of FY2009, the reversal rate had dropped to 20%. Part of the drop is likely attributable to new filtering mechanisms that keep low-quality rejections from reaching the Board. The pre-appeal brief conference program started in 2005 and allows a patent applicant to request an internal review of the examiner's final rejection at the Tech Center (TC) level before an appeal brief is filed. Even if no pre-appeal brief conference request is made, the examiner's case for a rejection is usually reviewed at the TC level before the examiner is allowed to file a responsive brief.

In recent years, these filters have operated to weed-out 80% of the cases where an appeal brief or pre-appeal brief conference request had been filed. Prior to implementation of these filters, the majority of appeal briefs led to BPAI decisions. [See Katznelson, Slide 15]

When the low quality rejections are eliminated from consideration, we should expect that the BPAI reversal rate would drop – as it has. The BPAI is no longer seeing many of the easy reversal cases where the examiner made a clear legal error. Rather, today's cases before the board tend to be more focused on arguable issues involving obviousness, enablement, and patentable subject matter.

The filters only tell part of the story, and the reversal rate masks the fact that - in absolute numbers - the Board is reversing more cases than ever. Over this same period, the number of appeals filed has increased dramatically. And, although the BPAI has increased its throughput, the backlog has grown five-fold since 2005. These pendency pressures give the Board an incentive to make appeals less hospitable – as we saw with the new (not-yet-final) BPAI appeal rules.

Notes:

• Prior Patently-O Posts: How long does a BPAI Appeal take; BPAI Appeal Statistics; BPAI Patent Judges

Comments

[JohnDarling]

6k,

That is good news.

Let me know when they're all gone. We'll go out and have a beer and celebrate.

[6]

"Wait ... the invention is about the optimization of an angle and the applicant refuses to limit the claim to a range of optimized angles?"

I have three letters that I'm thinking about right now. One is a vowel. Can you guess what those three letters are?

That said, I really don't understand why people use that inferior optimization of ranges crp. Woodruff ftw. I have used that case in no fewer than 20 actions (probably more like 50, or 100 inc finals), and I have never had even one attorney, amongst some of the better attorneys probably in existence, come back with the ridiculousness that pds says about that caselaw being used the way I do. They either try to establish criticality, or the subject doesn't come up. They fail at establishing it quite often, I will say that. And even if pds is right (slim chance), that sht is effective as he ll.

Woodruff: Ranges?!?! Where?!?! Take 5, 6, I've got this one.

6: Thx man.

"(1) When an examiner gets a deficient response from an applicant, his job is easy -- he files a quick final rejection. The applicant responds with an RCE and an Amendment and the examiner gets a bunch of counts. In other words, the examiner wins."

1. when an examiner gets a deficient response from an applicant, his job is (sometimes) easy -- he files a (sometimes) quick final rejection. The applicant responds with an RCE SIX MOTHER FIN MONTHS LATER and another deficient Amendment then the examiner gets 1 count. In other words, the examiner gets FED OVER for SIX MONTHS. Thereafter, the examiner addresses the next deficient response and recieves another count which in some cases makes the whole thing tolerable. Be clear here, the "abandonment" count should be given to me the second I sign the death sentence for your deficient response, not 6 mo. later. That's why the ooo, he just got 2 counts is merely an artificial appearance. The examiner got 2 counts for your RCE because you withheld the first of those 2 counts for SIX MONTHS. I should add, that even when you don't wait 6 mo. (rare) it is still ridiculous. Frankly, RCE's are just an id iot tax. If you don't like it then lobby congress to stop taxing you for being st upid (/unwilling to look at art yourself).

Let me ask you something mr. the examiner wins. If you did some work for someone, and they wouldn't pay you for 6 months, would you consider yourself having "won", or would you consider yourself "fed over"?

Yet again with the incredible. Every day I come here I'm amazed at the misperceptions of wtf is going on over here. Then again, I suppose we do have that mystique about us. Go to a bar, drop "patent examiner" and watch the girls flock. ;)

lolol.

[Carnac the Magnificent]

6 -- I seriously doubt if even most of the applicants wait six months to file rce's.

Also, after a few years you get into a system where you are getting credit for old cases -- everything is just shifted three or so months.

[Lionel Hutz]

I do not believe you would receive an antecedent basis rejection for amending the claim to recite a non-zero angle as long as that was implicit in your description and drawings and especially if you had a dependent claim reciting a particular angle or angle range.

I have had to make similar ludicrously unnecessary amendments in the past and not received an antecedent basis rejection (or at least not one I did not overcome in one action)

[Lionel Hutz]

Max,

Your hypothetical of 11:11 am on July 7 does not make sense. Your widget would only infringe if it infringed the claim. If you mistakenly thought the claim was anticipated by your previous widget, then you would have made an error. To the extent that the terms in the claim are read in view of the written description, and the applicant is allowed to be his own lexicographer, you should review the written description before arriving at a decision of whether you infringe. However, the file history will NEVER turn a claim you do not infringe into one you do. Prosecution history estoppel would NEVER play a part in your hypothetical.

[Lionel Hutz]

Max,

I am sorry. I see you whole focus was on validity. Yes, to determine validity, you do need to look at the file wrapper as a claim that may appear anticipated on its face, may be narrower based on FWE.

However, I also am unaware of any times a patentee has been able to rely on statements made during prosecution to turn an invalid claim into a valid one.

[6]

From this morn:

"Note the language "consistent with the specification." Like all bad examiners, 6K ignores language (in the claims or the law) that cut against his argument. Their credo -- "why respond to language when you can ignore it?"

I'm pretty fin sure that in my arguments above I specifically stated I would make a determination consistent with the spec.

I also stated that the mess from Phillips should be overturned on the grounds that it does not support the issuance of claims that by themselves accomplish the public notice function OF CLAIMS (as opposed to claims+spec). Somehow I get the feeling that the SC would see through that bs immediately. You just like it because it lets you make overbroad claims and then limit them in a way that does not put the public on notice except after significant trouble.

"Examiner's are experts at reading meanings out of terms and/or interpreting phrases in the ways that would never be interpreted by a PHOSITA"

This is true, but we have to take the contingency position as you well know.

"One last point, turning to the hypothetical at hand (i.e., the non-zero angle), odds are great that the examiner never even provided an explicit claim construction. Instead, the examiner probably pointed to a feature in the prior art and it was up to the attorney to divine that is how the examiner was interpreting the claim."

Of course. If you still have trouble "divining" such easy things, practice just might not be for you darlin.

"The worse part about dealing with stu pid examiner claim constructions is that you don't even realize how they were interpreting the claims until after final or during an examiner's answer."

This is because you choose not to display your ignor ance by making a 5 minute phone call. Other attorneys do not suffer such a disability. That said, sometimes I'll throw in a construction tidbit if I feel like they might contest the term.

Remind me, was gram v deer en banc? I bet it was, but I'm too lazy to look atm. From all indications I've seen, the fed circ has a propensity towards making ridiculous decisions en banc that usually require the SC to step in to fix. But then, I suppose they do the best they can, it is usually a tricky subject they are ruling on en banc.

"Speaking of examiner incompetence -- recently I had a case where the examiner, in a 102 rejection, cited to column 1, line 6 through column 23, line 54 (basically, the entire document with the exception of the field of the invention and the claims)"

I'm about to start doing that myself, this other mess of line to lining every st upid limitation (even blatant ones) is dmb and wastes way too much time. If you can't tell what an fin bicycle seat is when you're looking at one then we have worse problems in the patent system than "examiner laziness", specifically, attorney incompetence. I actually had an attorney come back and argue specifically that a certain piece in the prior art must be what is designated as one of the claim terms and that such a designation would make the claim not anticipated. Well, he was right about one thing, the certain piece had been designated as that claim term, but he was mistaken about the claim not being anticipated as a result. His subsequent amendment was rather funny.

[6]

"6 -- I seriously doubt if even most of the applicants wait six months to file rce's.

Also, after a few years you get into a system where you are getting credit for old cases -- everything is just shifted three or so months."

Doesn't change the fact that at any given time apps are fin me through the office's system. Tell you what, let's call it 5 mo. avg. response Then it only takes it one month to be moved to my docket. I get credit in 6 mo. Just because the app screwed me for 5 and the office for 1 hardly matters to me.

[Malcolm Mooney]

Hutz "I also am unaware of any times a patentee has been able to rely on statements made during prosecution to turn an invalid claim into a valid one."

Happens every time a statement in the prosecution history is successfully invoked to support a narrower claim construction than that requested by the defendant (who is trying to tank the claim on enablement grounds). You think this hasn't ever happened?

[MaxDrei]

Thank you MM. But is it enablement and nothing else. From an English point of view, Defendant puts patent owner in a "squeeze" as follows: Your claim is capable of being interpreted to have two meanings, one broader than the other. If we take the broader one, your claim is old or obvious. But if we take the narrower, we don't infringe, so it doesn't matter if it's valid. So, go away.

What I'm on about is the mischievous use the patent owner's lawyers will make of the file wrapper, to finesse out of the squeeze, into a claim interpretation that's both valid and infringed. I'm with you, in visualising it happening, all the time, but again I defer to readers at home in US patent litigation

[curious]

in addition to ksr and bilski, there have been some crazy ass 112 rejections the past ten years or so which are driving up the rejection rates

[William Morriss]

I'm curious about the reversal rate. The post said defined reversal as cases where the examiner was completely reversed. I'm curious about what the numbers would look like if the examiner was reversed as to the allowability of at least on claim.

NOTE: I am aware that the post said that a later article would look at the reversals, but I still haven't seen it.

[bierbelly]

"Wait ... the invention is about the optimization of an angle and the applicant refuses to limit the claim to a range of optimized angles?"

No, the case involves using an angle as opposed to not using an angle. The interesting/annoying thing about the Examiner's characterization of the prior art is that he's asking us to prove a negative...to discuss and dismiss what the prior art DOESN'T address.

I've seen more and more of this kinda crap from the PTO recently. Like I've said before, while the PTO may be obligated to read the claims as broadly as reasonable during prosecution, I don't think the same standard should be applied to the prior art. The examiners are now reading things into the prior art that aren't there.

[Malcolm Mooney]

"The interesting/annoying thing about the Examiner's characterization of the prior art is that he's asking us to prove a negative...to discuss and dismiss what the prior art DOESN'T address."

I think the Examiner is merely asking you to explain why using an angle is a non-obvious improvement over the prior art. Merely saying that "nobody used an angle before in this technique" is a response to an anticipation rejection, not an obviousness rejection.

Consider: a method of making a peanut butter sandwich, comprising making a peanut butter sandwich while listening to the 2009 iPod model.

The Examiner takes an old peanut butter sandwich recipe and says it would be obvious for someone to combine the method with listening to the new iPod, because the new iPod will make the old method more enjoyable to the sandwich maker.

You can't get anywhere by responding that "nobody discussed making a peanut butter sandwich while listening to **any** iPod, much less a 2009 iPod." It misses the point.

This is KSR. This is the correct application of the law. This is the way obviousness should work.

Deal with it.