DAVIS, Circuit Judge.

Essie M. Chennault was a secretary with a San Diego unit of the Navy. She was terminated in November 1983 for unacceptable performance under 5 U.S.C. Chapter 43. The Merit Systems Protection Board (MSPB or Board) finally upheld that removal in its Opinion and Order of September 17, 1985, 29 M.S.P.R. 122. The only issue now raised before us is whether there *466 was sufficient proof that the applicable Navy performance system, by which petitioner was judged, had been approved by the Office of Personnel Management (OPM). The full Board held that there was such proof, and we affirm.

Because of the limited question presented to us, we need not spell out all the details of the agency and MSPB proceedings. Suffice it that petitioner was charged with and removed for unacceptable performance. She appealed to the MSPB which held a hearing, after which the presiding official issued an initial decision affirming that' removal. That decision, referring to a Navy regulation (on appraisal standards) which so stated, included the statement that “the appraisal system under which [petitioner’s] demotion [sic] was implemented was approved by the Office of Personnel Management as being in conformity with 5 U.S.C. § 4302.”

On Ms. Chennault’s petition for review, the full Board (without considering the other aspects of petitioner’s appeal) remanded to the regional office for further consideration of the matter of OPM approval of the Navy’s performance system, and for a supplemental initial decision on that point. 1 MSPB Order of November 15, 1984. That Board order stated that the Navy regulation, cited in the initial decision, “does not evidence proof of an OPM approved performance appraisal system.” 2

On remand, the supplemental initial decision held that the record should not be reopened (because the agency had had full opportunity to make its proof on the point) but that the record contained only proof of the Navy regulation (on performance standards) reciting that OPM had approved the Navy’s system; the initial decision then naturally ruled, in view of the Board’s remand order, that the regulation was insufficient proof of OPM approval.

However, when the case came back before the full Board (see fn. 1, supra) it then held that the Navy regulation — which was already in the record made at the first hearing before the presiding official, and stated .that,OPM had approved Navy’s Basic Performance Approval Program — was in fact sufficient proof (in the absence of any rebuttal by petitioner).

The nut of petitioner’s argument to us is that, in its first order, the full Board held that the Navy regulation was insufficient proof of OPM approval and then, in its second order, improperly reversed itself and held, without explanation, that the Navy regulation did constitute adequate proof of OPM approval.

II.

A step-by-step description of the evolution of the Board’s position on proof of OPM approval (of performance appraisal systems) will help illuminate the course of the proceedings (on that point) in this case. When the first initial decision was issued (November 1, 1983) the full Board had not yet determined how proof of OPM approval should be made; without such guidance from the Board, the presiding official determined that the Navy regulation (with its statement of OPM approval) was enough proof. Later, on October 8, 1984, the full Board decided Griffin v. Department of Army, supra, which held that (a) the agency had the burden of proving OPM approval by specific proof, and (b) evidence of OPM approval “may consist of documentary and/or testimonial evidence,” specifically including a letter from OPM, or an affidavit from or testimony of an agency official with knowledge of the approval. The full Board’s remand order in this case came *467 slightly more than a month later (November 15, 1984) and pointed out that Griffin had been decided “[s]ince the issuance of the initial decision in this appeal.” 3 The supplemental initial decision, on remand, was issued on February 28, 1985. Meanwhile, on January 2, 1985, the full Board had promulgated Campbell v. Veterans Administration, 25 M.S.P.R. 556, which held summarily that a regulation entirely parallel to the Navy’s regulation here (i.e., stating that the agency’s performance appraisal system had been approved by OPM) was adequate proof; Campbell did no more than cite Griffin to support that proposition. Finally, when the full Board again considered the present case (on September 17, 1985), it ruled that the Navy regulation was sufficient proof of OPM approval, merely citing both Campbell and Griffin.

From this chronology, it is clear to us that when the full Board considered Ms. Chennault’s case for the first time — very shortly after the Griffin decision and at a time the Board assumed the issue had not been raised in the present case — it thought that OPM approval had to be proved by affidavit, testimony, or a letter. Later on, however, when directly faced in Campbell with that very issue, it concluded that an agency regulation on performance standards, incorporating a statement of OPM approval, was itself a proper showing of approval. In its subsequent second consideration of Ms. Chennault’s case, the Board simply followed Campbell.

III.

One of petitioner’s contentions is that the first full Board decision, ruling the Navy regulation insufficient proof, is the law of the case which had to be followed by the full Board in the second decision (now on appeal to us). But, all else aside, the doctrine of law of the case includes as an exception the situation where “controlling authority has since made a contrary decision of the law applicable to such issues.” Smith Int’l, Inc. v. Hughes Tool Co., 759 F.2d 1572, 1576 (Fed.Cir.), cert. denied, — U.S. —, 106 S.Ct. 87, 88 L.Ed.2d 71 (1985). Here, the full Board had decided Campbell diametrically contrary to the position stated in the first full Board order in this case, and Campbell then became the controlling authority for the MSPB. It cannot be the rule that the Board could decide Campbell as it did but later, in this current case, had to follow the first Chennault order even though the Board had itself decided in Campbell that that earlier position was wrong.

We are next urged to set aside the Board’s decision here because it departed from the former decision in the first Chennault order without explanation or rationale. True, there is a general principle that administrative agencies should not depart from their prior precedents without some indication of the reason for the change (see, e.g., Atchison, Topeka & Santa Fe Ry. Co. v. Wichita Board of Trade, 412 U.S. 800, 808, 93 S.Ct. 2367, 2375, 37 L.Ed.2d 350 (1973)) but the underlying ground of that principle is that the reviewing court should be able to understand the basis of the agency’s action and so may judge the consistency of that action with the agency’s general mandate. Id.

In this case, we have no difficulty whatever in assessing why the summary remarks made in the first Chennault order were not followed when the Board had actually to decide the precise issue in Campbell. The Board obviously felt that the earlier observations were hasty and mistaken, and that there was no reason to reject, as proof, a formal agency regulation stating explicitly and officially that OPM had approved that agency’s performance appraisal system. That would appear to be a top-level showing of OPM approval, fully probative and acceptable.

*468 Nor is there anything to the contrary in Griffin. That decision generally refers to proof by “documentary and/or testimonial evidence,” and then gives as examples a letter from OPM, or an affidavit to that effect by a knowledgeable agency official, or comparable testimony by such an agency official. 23 MSPR at 663. But those methods were merely illustrative and not said to be exclusive. In fact, any “probative hearsay evidence” on the point of OPM approval was specifically indicated to be acceptable. Id.

In a word, we think the full Board was quite mistaken, in its first Chennault order, in suggesting that Griffin precluded reliance on the Navy regulation. Campbell corrected that error and second Chennault followed suit. The Board could well have been more meticulous in spelling out why it disavowed its earlier Chennault observation, but that is not a valid reason for now overturning its current position. We have no trouble at all in understanding the basis for the change made in Campbell and second Chennault, and it would be a caricature even of mechanical jurisprudence for us to remand for such unnecessary spelling out, or to reject outright the Board’s present view on this issue because of the remark it made in first Chennault.

AFFIRMED.

Notes

. The Board also ordered that the supplemental initial decision, on issuance, be forwarded (with exceptions, if any) to the full Board which would then consider the whole case on Ms. Chennault’s petition for review.

. The Board added: “See Griffin v. Department of Army, 23 M.S.P.R. 657 (1984) at 8 (evidence of OPM approval of an agency’s performance appraisal system may consist of documentary evidence, i.e., a letter from OPM stating its approval)." Griffin has been reported at 23 MSPR 657.

. In that November 1984 order, the Board . seemed to think that the issue of OPM approval had not been raised before the presiding official in the present case. It was in that context that the order said that the Navy regulation (which the presiding official had cited) was insufficient. The full Board appeared to assume, in that order, that the presiding official made that determination sua sporite.