Attempts to drain a pool of oil too fast, or at too many places, may diminish net extractions. States therefore limit access to proven oil fields. "Prorate" systems credit owners with part of the production from pools under their lands to compensate them for inability to drill their own wells. Illinois does not use a prorate system but establishes minimum distances between wells. If the top of the oil-bearing stratum lies between 4,000 and 6,000 feet underground, 900 feet must separate the wells. Rule IV(1)(A)(1)(c) of the Oil and Gas Division, Department of Mines and Minerals.

Kendra Oil & Gas, Inc., drilled a well known as Dessie Matthews # 1 in southern Illinois. Homco, Ltd., later drilled a well known as O'Daniel # 5 some 760 feet to the north. O'Daniel # 5 struck oil more than 4,000 feet down in a layer known as the "Salem formation". If Dessie Matthews # 1, also producing oil from the Salem formation, found the top of the pool more than 4,000 feet deep, then O'Daniel # 5 is illegal under Rule IV(1)(A)(1)(c). Kendra filed this diversity action seeking damages on account of O'Daniel # 5. The parties consented to trial before a magistrate, 28 U.S.C. Sec. 636(c), and a jury returned a verdict of more than $260,000 in Kendra's favor.

Homco's principal contention on appeal is that the Mining Board of the Department of Mines and Minerals has "exclusive jurisdiction" to adjudicate a claim that one well has been sunk too close to another. Ill.Rev.Stat. ch. 96 1/2 p 5407 provides that the "Mining Board shall have jurisdiction and authority over all persons and property necessary to enforce effectively the provisions of this Act." State courts have not construed this grant as one of "exclusive" powers--indeed, courts have not construed this grant, period. Homco concedes that litigation would be appropriate after the Board had interpreted the rule. Its argument therefore is one of primary rather than exclusive jurisdiction. Federal courts will defer to the primary jurisdiction of a state agency under the principles established by Erie, if state courts would defer. See, e.g., Hines v. Elkhart General Hospital, 603 F.2d 646 (7th Cir.1979) (federal court will wait for completion of administrative steps required by the Indiana Malpractice Act).

Because resolution of the dispute between Kendra and Homco turns on both technical questions of petroleum geology (expert witnesses disagreed about whether the top of the pool under Dessie Matthews # 1 lies below 4,000 feet) and questions concerning the meaning of the state's rule (what is the "top" of the pool for its purposes?), it would have been desirable to obtain the views of the state's Mining Board. United States v. Western Pacific R.R., 352 U.S. 59, 65-69, 77 S. Ct. 161, 165-68, 1 L. Ed. 2d 126 (1957). No one suggested this course to the magistrate, however, until after the jury had returned its verdict. Having lost on the merits, Homco suddenly perceived the virtue of a second opinion. Too late, for no doctrine of "plain error" protects litigants in civil cases, Deppe v. Tripp, 863 F.2d 1356 (7th Cir.1988), unless the error concerns subject-matter jurisdiction. Fed.R.Civ.P. 12(h)(3). To try the case fully, discard a verdict, refer the case to an agency, and then retry the dispute is to waste everyone's time, and courts don't have time to waste.

"Primary jurisdiction" can't be shoehorned into "subject-matter jurisdiction". The diversity statute, 28 U.S.C. Sec. 1332(a)(1), provides ample jurisdiction over the subject matter of this dispute. A court that lacks subject-matter jurisdiction can't ever resolve a dispute, so throwing the case out as soon as the defect appears does not presage a wasteful return engagement. "Primary jurisdiction" involves questions of timing, not of judicial competence. Western Pacific R.R., 352 U.S. at 63-64, 77 S.Ct. at 164-65; United States v. Philadelphia National Bank, 374 U.S. 321, 353, 83 S. Ct. 1715, 1736, 10 L. Ed. 2d 915 (1963); United States v. Michigan National Corp., 419 U.S. 1, 4-5, 95 S. Ct. 10, 11-12, 42 L. Ed. 2d 1 (1974). Primary jurisdiction therefore may be waived or forfeited. Johnson v. Artim Transportation System, Inc., 826 F.2d 538, 547-48 (7th Cir.1987); United States v. Bessemer & Lake Erie R.R., 717 F.2d 593, 599 (D.C.Cir.1983). So Homco's failure to raise the question until after an adverse verdict is fatal, and we turn to other arguments.

Homco called Jack Morgan as an expert witness. The magistrate barred him from testifying on the ground that Homco had not listed him as a witness in the final pretrial order. It hadn't identified Morgan because it didn't need to. Kendra listed Morgan as an expert witness, and the order recites that "defendants reserve the right to call any witness listed by plaintiffs". The order governs the conduct of the trial unless modified, which may be done "only to prevent manifest injustice". Fed.R.Civ.P. 16(e); see Louis Vuitton S.A. v. Pun Yang Lee, 875 F.2d 584, 587-588 (7th Cir.1989); Smith v. Rowe, 761 F.2d 360, 365 (7th Cir.1985). Although Kendra claimed to be surprised by Homco's decision, it had no basis (other than its own unwariness) for shock. After all, Kendra had interviewed Morgan and put him on the list, stating in answer to an interrogatory that Morgan would testify about the "illegality" of O'Daniel # 5; Homco wanted Morgan to testify about just this subject. Homco's reservation of the right to call witnesses on Kendra's list is functionally identical to retyping all of the names under "defendants' witnesses". If Homco had used scissors and paste to repeat Kendra's list, Kendra could hardly have claimed "surprise". As this is what Homco did, but in fewer words, Morgan was eligible to testify. A party's unreadiness to meet its opponent's case does not justify excluding the evidence. Perhaps unpreparedness justifies a continuance; it does not justify exclusion.

Morgan, a petroleum geologist, had been named head of the Oil & Gas Division shortly before the trial. Homco made an offer of proof that Morgan "would give an opinion that the top of the pool of Dessie Matthews No. 1 is above 4000 feet; that he would be able, through his experience, to interpret each of the logs, which has been identified in evidence here; and that his testimony would be unequivocally that, with respect to each and every log, that the top of the pool is above 4000 feet." A "log" is a record of mineral strata made by lowering instruments down the bore. Several kinds of instruments had been used to generate logs of Dessie Matthews # 1 during drilling and after completion. These logs were introduced in evidence, and the principal dispute at trial concerned their interpretation. Expert witnesses for Kendra testified that the logs showed that the top of the pool lay between 4,002 and 4,014 feet underground. Homco's experts told the jury that these logs showed that the top lay between 3,994 and 3,998 feet. Morgan, an administrator as well as a geologist, might have been asked to give an opinion about the application of the regulations to these readings. Homco's offer of proof shows, however, that it wanted Morgan only to read the logs as a petroleum geologist. Because a gaggle of other geologists did exactly that, we agree with the magistrate that Morgan's testimony would have been cumulative, so that his exclusion is harmless error. Two geologists testified on Homco's behalf as Morgan would have done, and others listed in the pretrial order were not called. Nothing in the offer of proof suggests that Morgan would have added to these interpretations a new angle or argument, as opposed to the refrain "me too".

An error in the computation of damages is not so easy to pass over. The jury awarded Kendra the value of the oil Homco extracted, plus the loss caused by reduction in the productivity of Dessie Matthews # 1. This is double counting. As Kendra's loss logically is the greater of the two methods of measuring injury, Kendra is entitled to no more than that sum.1

Illinois forbids the completion of one well close to another in large measure because two holes in the same pool may reduce the gas pressure that contributes to productivity. Natural gas puts the petroleum under high pressure, driving it out of the rock formation and up the well. If a second well in the same pool vents the gas, the two together will recover less oil than the first well alone would have done. To simplify considerably: If a single well would have produced 100 barrels of oil, two wells might produce 40 barrels apiece. Twenty barrels are gone forever--or if not gone, obtainable only with expensive "secondary recovery" techniques. As the owner of the first well sees things, the second well reduces its production by 60 barrels, while the second well yields 40 barrels of oil. The first owner's damages are the amount of oil it lost: 60 barrels. An award based on the amount it lost (60 barrels) plus the amount the owner of the second well gained (40 barrels) would be duplicative. The owner of well # 1 would end up with the value of 140 barrels of oil: the 40 it extracted, the 60 it lost, the 40 the second well produced. Since by hypothesis the pool contains only 100 barrels of oil, the award exceeds the injury.

Over Homco's objection, the magistrate told the jury that it could award Kendra the sum of the production at O'Daniel # 5 and the diminution in the value of Dessie Matthews # 1, the latter being a measure of the value of oil not produced there. This would be a logical, if clumsy, way to compute damages if the two measures dealt with different periods--say, an award based on the value of production at O'Daniel # 5 for as long as it operated, plus the diminution of the production at Dessie Matthews # 1 for the years after O'Daniel # 5 shut down. Kendra sought, however, the gross value of all oil produced at O'Daniel # 5 (and O'Daniel # 6) between its opening in August 1984 and May 1, 1986 (when it was shut), plus the diminution in value of Dessie Matthews # 1. Kendra contended that until August 1984 Dessie Matthews # 1 had been producing 22 barrels per day, but that once O'Daniel # 5 opened production dropped to 10 barrels per day. It computed the capital loss as the value of 12 barrels per day for 900 days, at $29.00 per barrel.2

The lost production at Dessie Matthews # 1 is a superset of the gains Homco could have reaped at O'Daniel # 5 by tapping into the Salem formation.3 Kendra's losses cannot be greater than the diminution in production at Dessie Matthews # 1, measured from late August 1984 forward. Illinois gives the landowner the choice of recovering the value of what the wrongdoer took away or the diminution in the value of what remains; it cannot have both. First National Bank of Des Plaines v. Amco Engineering Co., 32 Ill.App.3d 451, 454, 335 N.E.2d 591, 593 (2d Dist.1975) (Moran, J.); Chicago Forge & Bolt Co. v. Sanche, 35 Ill.App. 174 (1st Dist.1889); cf. Associated Investment Corp. v. Laidlaw Waste Systems, Inc., 152 Ill.App.3d 279, 288, 105 Ill. Dec. 147, 153, 503 N.E.2d 1153, 1159 (2d Dist.1987) ("The measure of damages to real estate is the difference in value immediately before and after the occurrence [of the wrong]."). Neither Kendra's evidence and closing argument nor the magistrate's instructions to the jury required a separation of the damages theories, limiting Kendra to its loss or Homco's gain for any given period. We know from the jury's verdict what Kendra's losses are: $106,488.00. (The jury did not return special verdicts, but it awarded Kendra what it asked for to the penny, and this is what Kendra asked for on account of lost production.) Homco's other arguments, such as the contention that the lawful O'Daniel # 6 was the real source of the decline in production at Dessie Matthews # 1, were presented to and rejected by the jury. No further comment is necessary. The judgment is modified to award Kendra $106,488.00, and as so modified is affirmed.