EP 3 388 760 B1 relates to a regulation method for inverter compressors in refrigeration facilities.
Brief outline of the procedure
The OD rejected the opposition. The opponent appealed.
The board decided that claim 1 as granted lacked IS over D9=Technical manual of HEOS device “CAREL” with version number “Heos+0300078EN rel. 1.0 -30/10/14 + D13 or D12, two PCT applications not mentioned in the ESR.
The same applied to AR1 and 2 and the patent was revoked.
The case is interesting in view of the board’s consideration on the standard of proof.
The proprietor’s point of view
It was not sufficiently proven that D9 had been published on the internet before the priority date of the patent. D9 thus could not be considered state of the art under Art 54(2).
The evidence and affidavits presented were contradictory as to the publication date and only came from the sphere of the opponent. The evidence thus lacked credibility and was insufficient with respect to the required standard of proof, which was “beyond any reasonable doubt”
The board’s decision
D9 is a user manual entitled “Heos – High efficiency showcase controller” issued by the opponent. It is considered to form part of the state of the art under Art 54(2).
The board reminded that under Art 54(2), the state of the art comprises everything made available to the public, and includes “any other way“. Internet disclosures fall under this definition.
In G 1/12, point 2 of the order, the EBA held that proceedings before the EPO are conducted in accordance with the principle of free evaluation of evidence.
In G 2/21, Reasons 30, the EBA held that the principle of free evaluation of evidence requires the assessment of evidence according to a judicial body’s own conviction, without observing any formal rules.
In G 2/21, Reasons 31, the EBA stated that the decisive factor for the evaluation of evidence is “whether the judge is personally convinced of the truth of the factual allegation, i.e. how credible the judge classifies a piece of evidence. To do this, the judge must put all the arguments for and against a factual statement in relation to the required standard of proof.”
In G 2/21, Reasons 33, the EBA made a link between its considerations on the principle of free evaluation of evidence and decisions taken by the administrative departments of the EPO.
The EBA’s reference in G 2/21 to a judge’s personal conviction makes clear that it is the state of mind of the members of the fact finding body which is decisive in the evaluation of evidence. This is always the case, regardless of which standard of proof is to be applied.
In other words, referring to the deciding body’s conviction that an alleged fact occurred, does, strictly speaking, not yet say anything about which standard of proof should be applied by the deciding body to arrive at this conviction. The board referred to T 1138/20, Reasons 1.2.1 and T 1311/21, Reasons 3.2.1(d)(vi))
As to “the required standard of proof” referred to in G 2/21, different concepts have been developed in the case law of the boards.
The EPO standard of proof is generally the “balance of probabilities“. By way of exception, the standard of proof is that of “beyond reasonable doubt“, mainly in opposition proceedings where only the opponent has access to evidence concerning, usually, an alleged public prior use, cf. Case Law of the Boards of Appeal, 10th edition 2022, III.G.4.3.
The standard of proof refers to the nature or degree of personal conviction that the members of the deciding body must have in order to be satisfied that an alleged fact occurred.
If the applicable standard of proof is is the “balance of probabilities“, an alleged fact is proven as soon as the members of the deciding body are convinced that the occurrence of that fact is more likely than not.
If the applicable standard of proof is “beyond reasonable doubt“, the required degree of personal conviction of the members of the deciding body is higher.
Standards of proof relate, in legal systems based on the principle of free evaluation of evidence, necessarily to a mental state, namely to the nature or degree of conviction of the members of the fact finding body.
Accordingly, it is difficult to quantify the difference in the required degree of conviction between the “balance of probabilities” standard and the “beyond reasonable doubt” standard.
In fact, attempting to describe this difference in the form of numerical thresholds, for example as a certain percentage of likelihood that an alleged fact occurred, can even be misleading.
In regard to the “beyond reasonable doubt” standard, it thus seems more expedient to focus on the term “reasonable“.
In the present board’s view, this expresses that the “beyond reasonable doubt” standard does not require absolute certainty, and that it is sufficient if the majority of the members of the deciding body have no reasonable doubt than an alleged fact occurred.
In other words, even if there is some remaining doubt, the “beyond reasonable doubt” standard of proof can be met as long as the remaining doubt is not reasonable, which can be understood as overall insubstantial in view of the entirety of the available evidence.
In any case, if the higher one of two disputed standards of proof is met, it can be left open which of these standards must be applied when assessing the evidence in question. Hence, if the deciding body is convinced beyond reasonable doubt that an alleged fact occurred, there is no need to decide which standard of proof is applicable. The board referred here to T 2466/13, Reasons, 2.1.1.
Comments
Although G 2/21 related to a different topic, admissibility of post-published evidence to confirm a technical effect in order to assess IS, the present decision leads to the conclusion that it applies as well when it comes to assessing whether a document is part of the PA under Art 54(2).
This decision will most probably not change the practice of the boards or of the first instance divisions. It attempts to give a legal basis on the standard of proof, by stating that, whatever the standard of proof might be, it is the “conviction” of the deciding body which matters.
By the way, T 1138/20 was commented on the present blog.
A follow-up question
One question does however come to mind: when deciding on N, and even more on IS, it also boils down to assessing evidence.
When deciding on N or added-matter, should the decision be based on the “conviction” of the deciding body or not?
Should assessing IS be based on the “conviction” of the deciding body or not?
What is the degree of “conviction” needed for the deciding body to give a positive or a negative answer when assessing N and IS?
The problem-solution approach has been set up by the boards, and is now the standard way of assessing IS by the boards and the EPO. The aim was to establish a more objective assessment of IS. What has the “conviction” of a deciding body then to do with it?
I compare this question with a quote of a late, well-known French layer, Me Mathély, about assessing IS:
“Assessing non-obviousness does not imply any judgement of value, be it about the creative effort of the inventor, or be it about the importance or the usefulness of the invention.
The judge cannot simply assert obviousness by merely saying what he is “convinced” of.
He must formulate the technical reasons which, by comparison of the invention with the prior art, lead to an objective finding of obviousness”.
This quote predates by many years the PSA, but shows that the “conviction” of a judge should not prevail when it comes to assessing IS.
It is very nice for legal members of the boards wanting to bring decisions of the boards in relation with general legal principles, but there is a danger, especially when relying on the “conviction” of a deciding body to blur the clear picture set by the past case law of the boards.
That in criminal matters, the judge’s conviction can play a role when assessing the case in the absence of any precise evidence of guilt, but leading to the decision to condemn. In technical matters, when it comes to N or IS, it appears to me as going too far.
Applicability to first instance divisions
As the present decision refers to a “deciding body”, it applies as well to first instance deciding bodies, i.e. EDs and ODs.
Comments
2 replies on “T 832/20 – Free evaluation of evidence and standard of proof”
Daniel, I come from a background in English law, where there is a difference between issues of fact and issues of law. Obviousness is a question of law. but it can only be addressed after the relevant facts have been found by the court.
The court has to apply a standard of proof, when deciding whether the party adducing evidence has done enough to establish that any particular “fact” they are using their evidence to prove is or is not “proven” to the extent necessary to establish it as a “fact” which the court can rely upon in its judgement of the legal issue presented to it. In criminal proceedings the standard of proof is “beyond reasonable doubt”. Conversely, in civil proceedings (contractual disputes for example), absent special circumstances, the defult standard between the warring civil combatants appearing before the court is “more likely than not”. What other standard is there, to be fair to both sides?
Of course you know all this, inside out and backwards. But perhaps not every one of your readers. Anyway, what I am not seeing in your write-up is full recognition of the important diffgerence between issues of fact and issues of law. Whether or not any particular alleged fact shall be taken by the court to be “proven” is one thing. Quite another, independent of the fact-finding process, is the methodology employed downstream, to decide the legal question of obviousness, and how objective it is.
When all is said or done though, no judge ever writes a decision that contradicts their internal “conviction” of the rightness of their judgement, do they, not only on the issues of fact but also the issues of law? If so, is it not pointless, at least in the context of civil judicial proceedings, to worry about the notion of “conviction”
I wonder, do you agree?
Dear Max Drei,
I agree with you about most of what you say. There are indeed, as it has been made clear by numerous decisions of the boards, only two standards to be applied at the EPO, depending on the circumstances: balance of probabilities and beyond any doubt. I would equate balance of probabilities with common sense.
For example, it is more probable that a catalogue has been distributed to potential clients than it has ended up on shelves. When the facts are within the exclusive sphere of the opponent, common sense is not sufficient, and the facts relating to a public prior use should be such that it can be verified in an way as objective as possible.
It is not pointless in the context of civil judicial proceedings to worry about the notion of conviction. It reminds me of an old say: in front of court it is like on high sea, you are in the hands of god. We need judges in order to interpret the law, and they have an important task, but to link, in civil law, a judgement to the conviction of an individual or a group of individuals, goes somehow too far, and especially when it comes to patent matters. This is the reason why I have put the quote of Me Mathély in my comments.
In my opinion, the problem stems from G 2/21 which, in its quest of avoiding to use the word “plausible”, the EBA had no choice but to introduce the “conviction” of the judge. According to G 2/21, the effect invoked by the applicant/proprietor does not have to be expressly disclosed in the original application. It has just to be “derivable” fromand “encompassed” from what is stated in the original application.
In this respect, I would like to draw you attention to T 314/20, commented in my blog. In T 314/20 both G 2/21 and T 116/18, the decision of referring the board following the issuance of G 2/21, are criticised. In this respect, the EBA, might have spoken about the conviction of the judge, see G 2/21, Reasons 30 and 31, but this notion introduces some uncertainty when it comes about technical matters.
That the conviction might play a role in order to admit a post-published document, although the PSA has been created by the boards, can well be accepted, but it introduces a degree of uncertainty which should not be generalised to any peace of prior art. I am not convinced that the EBA had this generalisation in mind when it issued G 2/21.
To me, conviction might play a role in criminal cases, when there is no direct proof of a a tortious act or a crime, but not in patents.