Herein the Tax Court has made the following preliminary statement and findings:

On review here, the petitioners’ first point is: How can the Commissioner of Internal Revenue and the Tax Court take from the sketchy Exhibit 3 all of the income items on the left hand side and disallow those items of deduction on the right hand side? Further, the items on the left necessarily are the end result of allowing certain deductions on those days where there was a profit. How can that be done, ask the taxpayers, and still disallow their losses on the days when there were net losses?

Of course, the purpose of justice is to ascertain the truth. But how, as a prac *152 tical matter can a fact trier ever be quite sure he has got the truth in a case like this?

(Reference now should be made to the Tax Court’s opinion, Showell v. Commissioner of Internal Revenue, 23 T.C. 495 . The text herein will assume knowledge of the contents of that opinion.)

Just as the Tax Court reasons, we see no objection, in the absence of better evidence, to the Commissioner using the left hand figures as income on the theory of admissions against interest. If the findings were a little different here, we would find no objection to the Tax Court reaching the result it did in allowing the taxpayers only $3,000.00 on the right hand side of Exhibit 3 for deductions on their wagering operations. 1 Similarly, under different findings, if the dissenting opinion of Judge Withey, who heard the evidence, had prevailed then we would be disposed to uphold that conclusion.

The burden of proof is on the taxpayer to sustain by competent evidence his claimed deductions. When the trier of fact disbelieves or is not satisfied that the claimed losses were sustained, he has a right to disallow the claimed deductions. Similarly if he thinks that the taxpayer did suffer losses much smaller than claimed, but did suffer some losses the taxpayer cannot complain if the fact finder selects a half arbitrary, half intelligent figure for the losses. And, the government should not complain, if the conclusion has been fairly reached. It has not complained here.

As to particular findings in this case, we have such dissatisfaction with them that we have determined to send the case back to the Tax Court to-make new findings. One will note that, until the findings reach the final punch line: “The petitioners sustained wagering losses of $3,000 in 1949 in addition, to wagering losses allowed by the respondent (the commissioner),” the findings-are a summary of the evidence. So indecisive are they that they really lack the-elements of decision. They are more of a reporter’s condensed report of the testimony. In a way, this may be partly-explained by the fact that all testimony was presented by the Showells and on. the face of it there are no substantial contradictions anywhere. However, the-fact triers had the right to disbelieve-Jack Showell and his close office associate, 2 Walsh. Similarly, they have the-right to remain unconvinced, to retain an abiding doubt, and to rule against the-petitioner.

But the Tax Court should not recite as a fact, as it seems to do, that Showell and Walsh put down their losses on the right side of Exhibit 3, and then find that, the losses were less. Perhaps, there was an implied finding that the testimony was not satisfactory. 3 One might even say that is clear from the opinions of' the judges. But still this Court of Appeals is not satisfied.

It may well be on the remand that the Tax Court may come up with the same result as it has already. It may reach properly another result either more or less favorable to the taxpayer. No suggestion is here made as to what the result should be.

*153 The only thing that justifies the conclusions reached by the Commissioner or the Tax Court is disbelief or dissatisfaction with the testimony. Yet the findings are not sharp enough to tell us this. This is not to say that the Tax Court need point out line and page of testimony doubted or disbelieved. The findings should follow the traditional findings of ultimate fact.

Of course, there is always the case of the innocent individual who, through stupidity or for some other reason such as casualty, without his fault, no longer has his records. He is not to be ruined simply because he has lost the records, if his testimony or remaining skimpy records import honesty. There is argument that the Tax Court’s work here would create no such overriding precedent for the future. But that should be made clear.

The remand therefore will be on the ground that the findings were not sufficiently definitive.

The petitioners have a second ground for reversal. They say, “The government sends people to jail on the ‘net worth’ theory. We started to prove by the net worth method that we couldn’t have made any money on wagering. We were cut short.”

The taxpayers went quite a ways without objection along this net worth road. Finally, counsel asked Jack Showell, “What was the cost of buying your food for your family for a month?” After short colloquy, an objection to the question was sustained. The record does not ■clearly show whether this line was cut short because of refusal to permit further development of net worth or on other .grounds. And, if the only refusal to receive testimony on the net worth method is a refusal to hear about the cost of food, it is doubtful if a case should be reversed. Further, while we do not invite contumacy of counsel, yet, assuming that the Showells’ approach was correct, there should have been an offer of proof. At least there ought to be one more question in the same field, before quitting the subject. The one question refused, the line of questioning was abandoned by the Showells. (There was no announcement by the Tax Court judge of any positive theory which indicated a further question would necessarily be rejected.)

In our view, the petitioners did not lay sufficient foundation in the Tax Court for a review here of the taxpayers’ contentions on net worth.

Remanded for further proceedings.

POPE, Circuit Judge.

I would reverse the judgment of the Tax Court and remand the case with directions to enter judgment for petitioners. In my view, the findings of the Tax Court require that result.

As the findings set out in Judge Chambers opinion disclose, the petitioner Showell dealt with winners and losers. He paid the winners and the losers paid him. On a day when he collected more than he paid, he entered the net amount in a column headed “Gain”; on a day when he paid more than he collected, he entered the net amount in a column headed “Losses”. The entries made on Exhibit 3 are described in the findings as follows: “After the foregoing procedures had been gone through, entries, as follows, were made on a sheet of columnar paper, entitled ‘Sports — 1949’, and submitted in evidence as petitioner’s Exhibit 3. If the total of the amounts of the bets by losing bettors exceeded the total of the amounts to be paid to winning bettors, the amount of the excess was entered on Exhibit 3 in a column under the heading ‘Gain’. If the total of the amounts to be paid winning bettors exceeded the total of the amounts of the bets by losing bettors, the excess was entered on Exhibit 3 in a column under the heading ‘Loss’. The entries made on Exhibit 3 from January 1 to December 7, 1949, were made by Walsh. The other entries made on the exhibit were by petitioner.” Exhibit 3 in its entirety was then set forth in this part of the findings.

What the Tax Court has said in its findings is simply that some times petitioner won; some times he lost; that the amounts of the winnings and losses were *154 noted on slips of paper; and that the net amounts of those winnings or losses, as the case may be, were entered on Exhibit 3. Since the law permits petitioner to deduct the actual amount of his losses from the total of his gains, Exhibit 3, read in the light of these findings, discloses that there was no basis for assessing any deficiency against the petitioner.

It is true that the findings also say that apart from Exhibit 3 the petitioner maintained no account or record with respect to the money received by him and the sums paid by him into betting operations. In Bechelli v. Hofferbert, D.C., 111 F. Supp. 631 , some restaurant operators had also used a simplified system of keeping records whereby there was noted each day on a sheet of paper their cash receipts taken from the tape on the cash register and an itemized statement of the expenses paid every day. These sheets were not preserved, but the receipts and the disbursements were recorded by a friend in permanent books. The court said at page 633: “The critical test as to the sufficiency of the books on their face is whether they are sufficient to calculate the net income. If they are sufficient in this respect then the simpler the books the better. There is no prescribed detail as to just what books or how many must be kept. The question in each case must be determined on its particular facts and in view of the nature, volume and complexity of the business. Here the books as kept do show day by day receipts and expenses. If the figures are correct the books are sufficient to show the net income.”

It seems to me that the rule there stated is equally applicable here. If the figures are correct, the petitioner’s books are sufficient to show the net income from these operations. The findings detail the precise manner in which petitioner, with the aid of the witness Walsh, transferred the amounts from the daily sheets to Exhibit 3. There is not an iota of evidence that this was not done correctly or accurately and there is no finding either that the losses did not occur or that the method of computing and transferring them to Exhibit 3 contained any inaccuracies. 1

The findings next disclose that the Commissioner allowed none of the loss items other than those for rent, telephone and telegraph, and a certain publication, but the findings, as I view them, conclusively demonstrate that in acting in this manner the Commissioner was wrong. He was wrong and his determination arbitrary and excessive first, because he failed to allow any of the losses mentioned notwithstanding those losses were transferred to Exhibit 3 in the manner found by the court; and second, because the Tax Court holds the losses were substantial — ($3,000, it says).

Once the Tax Court has determined, as it has done here, that the action of the Commissioner was wrong, it was the duty of the Tax Court to find the facts for itself. No part of the arbitrary determination of the Commissioner can give rise then to any presumption of correctness. “The fact that the Commissioner’s determination of a deficiency was arbitrarily made may reasonably be deemed sufficient to require the Board to set it aside.” Helvering v. Taylor, 293 U.S. 507 , 55 S.Ct. 287, 290 , 79 L.Ed. 623 ; Gasper v. Commissioner, 6 Cir., 225 F.2d 284, 287 . Under the rules stated in the cases just cited, the Commissioner’s de *155 termination must be simply disregarded, and the Tax Court must ascertain whether there be a deficiency and, if so, how much, on the basis of the evidence before it. Its findings as to the manner in which the losses occurred, and the mode in which they were transferred to Exhibit 3, shows that a finding that the wagering losses did not exceed $3000 is not only inconsistent with the other findings but clearly erroneous because it is based upon no showing whatever which would warrant a rejection of the proof of the losses disclosed on Exhibit 3.

. Appellants say the husband’s type of bookmaking in 1949 in Arizona was entirely legal. The government does not contest this. Without inquiry, we assume that is so. The case of Nellis v. Commissioner, 1955 P-H T.C. Memo 55,050; 6 Cir., 1956, 232 F.2d 890 is somewhat similar to Showells’ cases, except in Ohio the operations appear to have been illegal.'

. Dyer v. MacDougall, 2 Cir., 1953, 201 F.2d 265 .

. As one reads the testimony, both Walsh and Jack Showell said, in effect, that the entries on Exhibit 3 were made thereon on (or, at least about) the dates they bear, with certain indicated exceptions. A certain sameness in the entries from January 1 to December 7 might justify expert analysis as to the dates the entries were actually made, granted that they were made by one man. Such analysis-may prove the document to have been made as Showell and Walsh say it was made and thus fortify their claims.

. As indicated above, the Tax Court has found the precise manner in which Exhibit 3 was made up. It has thus found that the losses incurred were recorded on that Exhibit. This finding requires a conclusion that the claimed losses so recorded should have been allowed. The majority attempt to avoid this necessary conclusion by saying, without too much assurance, “Perhaps, there was an implied finding that the testimony was not satisfactory.” The statute provides: “The Tax Court shall report in writing all its findings of fact, opinions, and memorandum opinions.” 26 U.S.C.A. § 7459 (b). (Emphasis mine.) A decision which is contrary to the findings cannot be bolstered or saved from reversal by some doctrine of implied findings, and since the findings adequately recite just how Exhibit 3 was made up, there is no occasion for suggesting that the findings be fortified by expert analysis, as suggested by the majority in footnote 3.