joined by PHILLIPS, Chief Justice, in all but Part I, and by GONZALEZ and OWEN, Justices, dissenting.
I do not agree that State Farm Lloyds committed a tort against loan and Liana Nicolau. I therefore respectfully dissent.
I
The Texas tort of bad faith is — to borrow Judge Alex Kozinski’s observation in a related context — “so nebulous in outline and so unpredictable in application that it more resembles a brick thrown from a third story window than a rule of law.” Oki America, Inc. v. Microtech International, Inc., 872 F.2d 312, 315 (9th Cir.1989)(Kozinski, J., concurring, referring to the tort of bad faith denial of contract created and then abolished by the California Supreme Court, Freeman & Mills, Inc. v. Belcher Oil Co., 11 Cal.4th 85, 44 Cal.Rptr.2d 420, 900 P.2d 669 (1995), overruling Seaman’s Direct Buying Service, Inc. v. Standard Oil Co., 36 Cal.3d 752, 206 Cal.Rptr. 354, 686 P.2d 1158 (1984)). The only flaw in Judge Kozinski’s metaphor is the implication that bad faith liability is limited to hapless passersby. A more accurate comparison would be to an assault weapon fired into a crowd at random.
That, needless to say, is the defendants’ perspective. For plaintiffs, bad faith is more like Hollywood television’s Wheel of Fortune, or closer to home, like the Texas lottery: it costs almost nothing to play, you can play whenever you want, and if you win you hit the jackpot — tens, maybe hundreds, of thousands of dollars for the awful mental anguish *454that invariably seems to accompany denial of even the smallest insurance claim, and millions in punitive damages. And like the lottery, bad faith liability is paid ultimately by the public. Insurance companies have not been authorized to print their own currency; the money to pay successful plaintiffs and their attorneys comes from policyholders, and they obtain the money to pay premiums from wages or sales. In effect, bad faith is a levy on everyone to benefit a few — what some have called a tort tax.
But whether bad faith is seen from the recovery end or the liability end, its principal feature is the same: it is chancy, far more so than any rule of law ought to be. A legal cause of action should not be a game of roulette, either casino-style or the Russian variety. Individuals and entities, even insurance companies, are entitled to know before they act what the law expects of them, what behavior is culpable and what is not. A legal cause of action must be adequately defined by principles and standards. The bad faith tort for first-party insurance relationships in Texas does not approach minimum requirements.
This is not a small problem. Every lawsuit in Texas by an insured against an insurer almost always includes an allegation of bad faith — at least every one filed by a competent lawyer. Why? Is it because the insurers who venture to do business in this State are a uniformly sorry lot? No; it is because the odds of recovery are always decent and the stakes — unlimited tort liability — are always high. The threat of liability also increases the settlement value of any policy claim and may make settlement less likely. Why is the threat viable in every case filed? Because this Court will not define the limits of the tort.
Until today the Court defined the tort in a single sentence: “[A]n insured claiming bad faith must prove that the insurer had no reasonable basis for denying or delaying payment of the claim, and that it knew or should have known that fact.” Transportation Insurance Co. v. Moriel, 879 S.W.2d 10, 18 (Tex.1994); Aranda v. Insurance Co. of N. Am., 748 S.W.2d 210, 213 (Tex.1988); Arnold v. National County Mut. Fire Ins. Co., 725 S.W.2d 165, 167 (Tex.1987). Accord Republic Ins. Co. v. Stoker, 903 S.W.2d 338, 340 (Tex.1995); Union Bankers Ins. Co. v. Shelton, 889 S.W.2d 278, 283 (Tex.1994). Like any general principle, it requires working out in actual cases. Dozens of cases in the courts of appeals over the past ten years have begged for further guidance. One court even went so far as to write of a sister court’s decision, “[t]his is not only dangerous precedent, it is frightening precedent.” State Farm Fire & Cos. Co. v. Simmons, 857 S.W.2d 126, 135 (Tex.App.—Beaumont 1993, writ granted)(referring to State Farm Lloyds, Inc. v. Polasek, 847 S.W.2d 279 (Tex.App.—San Antonio 1992, writ denied)). This conflict should be resolved; instead, the Court adds to the confusion.
The Court is not oblivious to the problems with bad faith. For example, it has recognized almost from the beginning that not every erroneous denial of an insurance claim can be bad faith. More precisely, the Court promised in Aranda that an insurer “will not be subject to liability for an erroneous denial of a claim.” Aranda, 748 S.W.2d at 213 (emphasis added). There is, however, a catch, as today’s decision illustrates: an insurer will not be liable unless someone, anyone — even the insured’s own trial lawyer— will testify that it acted unreasonably. If at least one witness will testify that there was no reasonable basis for an insurer’s denial of a claim, and the jury agrees, an insurer will be subject to bad faith liability. As such evidence is not hard to come by — the witness need not be specially qualified, and plaintiff can testify himself if his lawyer is for some reason indisposed- — the promise of Aranda means no more than that an insurer will never be hable unless a jury finds it liable. This is not much of a promise. It is not very helpful as a rule of law.
In Moriel the Court offered another palliative: “A simple disagreement among experts about whether the cause of the loss is one covered by the policy will not support a judgment for bad faith.” Moriel, 879 S.W.2d at 18. Again, there is a catch, as today’s decision illustrates: a simple disagreement among experts does not prove bad faith unless the insured’s expert or the insured’s *455lawyer says the insurer’s expert was unreasonable. If the insured’s witness testifies only that the insurer’s expert was wrong, not that he was unreasonable, then the insurer cannot be liable for bad faith. Desperate indeed will be the plaintiff whose expert or trial counsel will testify that a contrary opinion was wrong but cannot be coaxed to add that it unreasonable.
The Court’s rationalizations are approaching doublespeak. The Court writes that “evidence showing only a bona fide coverage dispute does not, standing alone, demonstrate bad faith.” Ante at 448. I should have thought it obvious that evidence of a bona fide dispute is no evidence of bad faith. How can such evidence, “standing” with other evidence, “demonstrate bad faith”? It is like saying that evidence of no negligence cannot, standing alone, prove negligence. The Court’s position may be accurately summarized as follows: evidence that a claims dispute was in good faith does not preclude a finding of bad faith, although such evidence never proves bad faith, except in some circumstances, such as when the factfinder finds bad faith, unless the factfinder is wrong. Is it any wonder that one court of appeals has drily observed that this Court “has ultimately done little to provide lower courts with any guidance for conducting a legal sufficiency review” in bad faith cases? Columbia Universal Life Ins. Co. v. Miles, 923 S.W.2d 803, 810 (Tex.App.—El Paso 1996, writ denied).
I am not in favor of abolishing the tort of bad faith, although I would not have been in favor of creating it. The experience in Texas and throughout the country persuades me that the law should afford some private right of recovery of consequential damages to insureds whose claims have been wrongfully denied or delayed, but I would have preferred to derive that right from established principles of contract law, or to have relied on statutory remedies, rather than to create a new tort and, like Dr. Frankenstein, take our chances. That point of decision is long past for most American jurisdictions. About half the states recognize a bad faith tort in first-party insurance cases. Although the contract route certainly looks far more attractive now than it did when the law was at the crossroads, we are so far down the tort road it would be hard to retrace our steps.
Having decided to create a new tort, it is this Court’s firm responsibility to infuse it with the principles, logic and limits necessary for any rule of law. Bad faith, no less than breach of contract or negligence or any other cause of action, must be defined to balance all the interests involved and provide logical, predictable standards of culpability. Unscrupulous insurers are not the only problem. Insurance companies have not cornered the market on greed and dishonesty; insurance fraud costs billions of dollars a year. A just and workable legal right can distinguish, with acceptable degrees of certainty and predictability, between insurance claims abuse on both sides and legitimate differences of opinion. The bad faith tort in Texas does not meet this standard, and it is the Court’s fault.
Today’s decision in Universe Life Insurance Co. v. Giles, 950 S.W.2d 48 (Tex.1997), changing the liability standard from “no reasonable basis” to “reasonably certain”, does nothing to define the tort. The Court correctly states in that case and this one that the result in each would be the same under either standard. That is true. In fact, the result would be the same under either standard in any case imaginable. The sole reason for changing the liability standard is to align the common law with article 21.21, § 4(10) of the Insurance Code. The Court is unable to point to a ease that would be affected by the change in liability standard.
The Court cannot simply stand by and hope that juries will do justice when we cannot define for them what justice means. Juries are entitled to be told what the law is. They cannot be expected to fulfill their function as factfinders until the court gives them the applicable law. I doubt the Court would approve of the Insurance Commissioner, experienced in the conduct of the insurance business, routinely handing out $50,000 fines to insurers with no more legal and factual basis than exists in this case. A standardless decision is no better because a jury makes it.
*456This Court’s only responsible alternatives in this case were to begin to define the tort of bad faith or to abolish it. While I am not yet in favor of the latter, I am absolutely against adding rationalizations to generalizations.
II
A
For State Farm to have acted in bad faith, its denial of the Nicolaus’ claim must not only have been in error; its denial of the claim must have been without any reasonable basis. There should be some indication that State Farm was an “unscrupulous insurer[ ] [taking] advantage of [its] insureds’ misfortunes in bargaining for settlement or resolution of claims.” Arnold, 725 S.W.2d at 167.
If ever there were a dispute about an insurance claim over which reasonable minds could differ, this is it. Five years it took the Nicolaus, consulting with experts and making repairs, before they themselves came to believe that the shifting in the foundation of their home was caused by a leak in the sewer line beneath the family bathroom — and thus covered by their homeowner’s policy. State Farm relied on two highly qualified engineers employed by a firm with long-established, worldwide business to conclude that the shifting was caused by the nature of the clay soil where the house was built (which also caused problems in a great many other homes in the Nicolaus’ neighborhood), atmospheric conditions, and the way in which the foundation had been constructed — and thus not covered by the policy. At trial, two engineers, a general contractor, a foundation contractor and two plumbers testified for the Nicolaus; five engineers testified for State Farm — three from firms other than the one hired to investigate the claim, and one of them hired as a trial witness by the Nicolaus. The trial lasted five days. The transcript of proceedings is over 1,200 pages long. By any fair measure, the disagreement over what caused the shifting in the foundation of the Nicolaus’ house was serious and substantive.
Given such a disagreement, where does the Court find evidence to suggest not only that State Farm should have paid the Nicolaus’ claim but that it had no reasonable basis for doing anything else? The answer is, not in the totality of circumstances, but in a few sentences of testimony by plaintiffs’ hired experts and by the Nicolaus’ own trial counsel. It should come as no surprise that some of the experts called to testify by the Nico-laus believed that they were right about the leak in the sewer drain causing the Nicolaus’ problems, and that there was no reasonable basis for anyone to think otherwise. Indeed, it is awfully hard to prove a ease of bad faith with witnesses who think there is room for reasonable disagreement. It should come as even less surprise that the Nicolaus’ lawyer thought that State Farm’s refusal to pay their claim was unreasonable. What is surprising, especially after Moriel, National Union Fire Insurance Co. v. Dominguez, 873 S.W.2d 373 (Tex.1994), and Lyons v. Millers Casualty Insurance Co., 866 S.W.2d 597 (Tex.1993), is that liability for bad faith can rest on no more evidence than exists in this ease. If all it takes to support a claim for bad faith is for a witness professing expertise or the plaintiffs lawyer to opine that an insurer had no reasonable basis for acting as it did, then few will be the cases with no evidence to support the claim. Such opinions are not hard to procure. No-evidence review in bad faith cases becomes no-review review. That is the result of today’s decision, as the facts of the ease show.
B
Cracks began to form in the interior and exterior walls of loan and Liana Nicolau’s Corpus Christi home in 1984, some six years after they moved in and ten years after the house was built. They suspected foundation problems as the cause. Many homes in Corpus Christi have foundation problems because of the Beaumont clay soil, although the number of houses with problems in the Nico-laus’ subdivision was abnormally high.
The Nicolaus insured their home with State Farm Lloyds. Their policy does not cover damage resulting from shifting in the foundation, unless the shifting is caused by a plumbing leak.
*457The Nicolaus hired Thomas Krismer, a foundation repair contractor in business about three years, to stop the cracking in the walls of their home. On Krismer’s recommendation, the Nicolaus also hired Dexter Bacon, an engineer with a local company, Maverick Engineering. Krismer and Bacon inspected the Nicolaus’ home and concluded that the problems were due to four mesquite trees in the front yard that were desiccating the area in front of the house, causing the soil there to shrink and the foundation to subside. On Bacon’s advice, the Nicolaus paid Krismer to install piers along the front of their foundation to prevent further settling and damage to the house.
The piers Krismer installed did not stop the cracking and shifting in the house. In 1986 the Nicolaus called Krismer back again. This time he told them that the cracks that were continuing to appear in their home were normal. In 1988 the Nicolaus called Krismer a third time. He again recommended that Bacon be consulted. Bacon reported to the Nicolaus that the persistent cracking was due to a drought that was causing some minor deflection in the piers that had been installed.
It must be noted at this point that from 1984 to 1989 Krismer and Bacon never took core samples of the soil around the house, or checked for plumbing leaks, or examined the foundation to see whether it had been constructed properly. The significance of this fact is that these are the very things that the Nicolaus’ trial experts, including Krismer and Maverick Engineering’s representative, criticized State Farm’s experts at Haag Engineering for not doing. Krismer testified that his approach was justified. In 1984, he said, the Nicolaus’ foundation was sinking, not heaving as it might have been if there had been a plumbing leak. In other words, Krismer did not think that a leak could have caused the problems he observed. Similarly, Haag did not think that a leak under one part of the foundation could cause the opposite side of the foundation, some thirty feet away, to shift. Krismer also testified that many of the homes on the Nicolaus’ street had similar problems, suggesting common causes, such as soil and drought, rather than causes more likely to be unique to one house, such as a leak or faulty construction. Haag also thought the most likely causes of the Nicolaus’ problems were natural. In short, for five years Krismer and Bacon believed that it was unnecessary to take core samples, look for leaks, or examine the construction of the foundation because the most likely cause of the shifting were weather and soil. No one suggests that Krismer and Bacon were incompetent or biased, or that they acted in bad faith. It is equally difficult, however, to level the same charges at Haag. As Krismer testified, “foundations are very complex things”.
In 1989 the Nicolaus consulted a different expert, Gerald Paseador, before calling Kris-mer back for the fourth time. While inspecting the house, Krismer found a leak in the master bedroom shower that he thought might somehow be causing problems. Following Krismer’s advice, the Nicolaus fixed the leak, but still the cracking in their home worsened. Krismer continued to consult with Bacon, but because Bacon had left Maverick Engineering’s employ, Krismer contacted another Maverick engineer, Fred Hayden. Hayden had a bachelor’s degree in general engineering from Texas A & I University and had become manager of Maverick after working for the company for eighteen years. He had experience in designing foundations and investigating problems in them.
Hayden suggested that instead of the front of the house settling, the back might be heaving due to another leak under the house that was causing the soil toward the sides or rear of the foundation to swell. At Hayden’s direction, Reed Glendenning, a plumber for fifteen years and business partner of Kris-mer’s, tested the sewer line in early 1990. This line carried waste water, not under pressure, from drains in the house to the sewer. Glendenning found that the line was leaking in the vicinity of the hall bathroom, near the center on one side of the house, marked by the “x” in the following rough schematic.
*458[[Image here]]
A few days after this discovery, in February 1990, the Nicolaus filed a claim on their homeowners’ policy with State Farm.
About a month later, Hayden provided the Nicolaus a written report of his investigation. Although the leak still had not been located precisely, it appeared that it would be nearer the center of the house and not toward the side and rear where the foundation appeared to be shifting. Still, Hayden theorized in his report “that the extremely wet condition of the soils on the left side of the residence may have caused the water from the plumbing leak to leak along the plumbing lines and trenches and may have caused the kitchen area to heave as well”. (Emphasis added.) However, Hayden and Maverick Engineering remained uncertain whether the leak actually was causing the shifting in the foundation.
Meanwhile, State Farm’s superintendent, Ralph Cooper, contacted Haag Engineering to investigate the Nicolaus’ claim. Haag had been in business since 1924 and employed about twenty engineers in all disciplines. Haag had offices in Dallas, Houston, and Corpus Christi, and clients around the world. Haag often worked for insurance companies. Haag assigned two engineers to the project: Jim Wiethorn, who had a bachelor’s degree from Baylor University in math and physics, and a master’s degree, cum laude, from the University of Texas in architectural engineering; and David Teasdale, who had bachelor’s and master’s degrees from the University of Texas in civil engineering. Wiethorn and Teasdale both specialized in failures in various kinds of structures, and Teasdale also specialized in problems with soils.
The Nicolaus argue that State Farm never intended to pay their claim and hired Haag to supply the expert opinion it could use as a pretext. The Justices who comprise the today’s majority accept this argument. There is no evidence — not even from the Nicolaus’ expert witnesses — to support it.
At trial, Cooper testified that he knew before he hired Haag “that they were of the general opinion that a localized leak beneath the house would not cause foundation damage as a general rule. Sometimes they felt it would, sometimes they felt it wouldn’t.” Wiethorn and Teasdale relied on a published study by the University of Texas at Arlington which concluded that localized leaks in an unpressurized sewer line under a slab foundation rarely caused significant shifting. They had also developed, in consultation with *459members of the engineering faculties at the University of Texas and Texas A & M University, a set of criteria for assessing the effects of an unpressurized sewer leak on a slab foundation. Following these criteria, Haag engineers in general, and Wiethom and Teasdale in particular, rarely found that sewer leaks caused foundation problems. Krismer testified that he knew of only one instance in which any Haag engineer (not Wiethorn or Teasdale) had reached a different conclusion.
Still, Krismer also testified that while he knew the Haag engineers relied on the UT-Arlington report, it would not be true to say that State Farm hired Haag simply to get an opinion that a leak beneath a house did not cause foundation problems. Cooper testified: “I hire Haag because to me they’re the IBM of the engineering profession. They do work worldwide. They have studies in this area. They’re well-respected. They’ve been all over the country, and they’re one of the leading engineering firms of the country, and I rely on them as an expert.” While some of the Nicolaus’ experts at trial were critical of the conclusions Wiethorn and Teasdale reached and the manner in which they performed their work, no one disputed their credentials or expertise, or Haag’s reputation, and no one accused them of being biased against the Nicolaus or of having preconceived ideas about the validity of their specific claim. Krismer expressly refuted the accusation when he testified:
Q If somebody were to say that State Farm hires Haag, and Haag only—
A That’s not—
Q —finds that there’s no relation between the leak and foundation movement, that would not be true, would it?
A No, that would not be true.
Wiethom and Teasdale examined the Nico-laus’ home in April 1990. At that time, the leak still had not been found, but Hayden had knocked a hole in the foundation near the hall bathroom and found the clay soil there extremely wet. The Haag engineers looked at cracks in the exterior and interior of the home and examined the hole Hayden had made in the foundation. Wiethorn and Teasdale thought it unnecessary to try to uncover the leak as they were willing to assume from Glendenning’s test the existence, nature and general location of the leak. After reviewing all the work that Kris-mer, Bacon, Hayden and Glendenning had already done, Wiethorn and Teasdale concluded in May 1990 that the leak could not be causing the shifting in the foundation of the Nicolaus’ home.
The Haag engineers reached their conclusion for several reasons. First, since the sewer line was not pressurized and was embedded in the clay soil beneath the house, the amount of water that could escape was minimal. As it turned out, there was a small air pocket in the soil at the point of the leak, but Haag eventually concluded that it did not allow significantly more water to flow from the leak. Second, Wiethorn and Teasdale did not believe that water leaking from an unpressurized line in the middle of the house could migrate to the back of the foundation where much of the shifting had occurred. The only possible conduit suggested for this migration, other than for the water to seep through the soil in all directions, was the plumbing trench that had been dug and refilled when the house was built. The Haag engineers did not believe that the soil where the trench had been, or the elevation of the area (the lot sloped upward from the front), would allow water to move along that area. Also, Haag saw no evidence of shifting in the foundation near the trench area as there should have been if it were moist from the leak. Third, there was no other way to connect the shifting near the edges of the foundation with a sewer leak in the center of the house. Finally, given the abnormal number of problems with other houses in the same area and the nature of the soil on which those houses were built, the Haag engineers concluded — as the Nicolaus and their experts had thought for five years — that the foundation problems were due to natural causes.
Based on the Haag engineers’ report, State Farm advised the Nicolaus in May 1990 that, although its investigation was continuing, it was possible their claim would not be covered by their policy. The Nicolaus nevertheless insisted that the leak be uncovered and repaired, and that soil samples be *460taken around the house to find whether they would indicate moisture coming from beneath the foundation. State Farm agreed to both requests and paid for all the work to be done. In fact, State Farm either paid or offered to pay for all the work Krismer, Maverick Engineering, Glendenning and others employed by the Nicolaus did in late 1989 and in 1990 to determine the cause of their foundation problems and repair the sewer leak. But in June 1990, when the Nicolaus submitted a claim for $102,200 to correct the problems caused by the shifting foundation, State Farm refused to pay it.
In August 1990, several months after Haag and Maverick had each issued their reports, Gary Prichard, another plumber hired by the Nicolaus, found the leak beneath the hall bathroom, exactly where it had been predicted to be, and repaired it. The sewer pipe was torn in two places, and a joint had come unglued. The nature of the leak was just as Glendenning’s test had indicated.
Chíen N. Fu, an engineer with Trinity Engineering Testing Corporation (“TETCO”) specializing in soil analysis and foundations, who held a master’s degree from the University of Texas at Arlington in geotechnical engineering, took soil samples from beneath the foundation and around the house. All the samples were extremely moist, and in one hole Fu actually found water, something he had never seen in any other situation. From the core samples, TETCO and Maverick concluded that the sewer leak had increased the moisture below the house and caused the foundation to shift. They believed the slope of the lot was irrelevant because moisture might have migrated along the plumbing trench because of evidence that the permeable sand beneath the house was deeper toward the rear.
The Haag engineers remained unconvinced, however, because the core samples showed a uniform distribution of moisture around the foundation, rather than more moisture near the leak and less farther away. The Haag engineers could not offer a satisfactory explanation for the accumulation of moisture around the house, but they still believed that a leak in an unpressurized sewer line buried in clay could not account for the volume of moisture beneath the house that the tests indicated.
To test their belief, the Haag engineers constructed an elaborate reproduction of the sewer line in a plexiglassbox filled with clay, and pumped water through it. They attempted to replicate conditions existing below the Nicolaus’ foundation, including the void spaces around the sewer line at the point of the leak. This test, which Haag charged State Farm more than $4,000 to conduct, tended to support Haag’s conclusion. Glendenning, the Nicolaus’ plumber, questioned the manner in which the test was conducted and set up his own model. Glen-denning’s test indicated that more water would escape from the leak than Haag estimated, although it is not clear whether, even in Glendenning’s test, the larger amount would have accounted for TETCO’s findings.
Based upon Haag’s supplemental report of its conclusions, State Farm again, at the end of August 1990, denied the Nicolaus’ claim. Fourteen months later the Nicolaus sued State Farm to recover benefits under their policy and for mental anguish and punitive damages for bad faith and violations of the Deceptive Trade Practices — Consumer Protection Act, Tex. Bus. & Com.Code §§ 17.41-.63 (“DTPA”), and the Texas Insurance Code. The Nicolaus’ central contention was that State Farm never intended to pay their claim, that it hired Haag to supply an expert opinion to be used as an excuse for denying the claim, and that Haag complied, conducting no real investigation of its own.
At trial, the Nicolaus’ experts, including Krismer and Hayden, criticized Haag for not doing more work on its own, specifically, for not taking its own elevations and core samples and attempting to uncover the leak. However, no one disputed Haag’s willingness to rely on the work done by the Nicolaus’ experts. Moreover, Krismer had not ordered core samples or suspected a leak for more than five years. He testified that his approach was justified because “foundations are very complex things.” Maverick acknowledged that the leak discovered in 1990 existed in 1984, and that its failure to find it sooner was “embarrassing”.
*461After the leak in the sewer line was repaired, the Nicolaus’ house continued to shift and crack. Krismer testified that this was due to a drying of the soil in the center of the foundation where the leak had been, causing the foundation to settle there. However, Hayden ultimately concluded that not all the problems could have been caused by the leak, and that some were inherent in the way the house had been constructed. Specifically, Hayden believed that the foundation was too light to be stable. Thomas Petrie, an engineer and the associate director of the Construction Research Center at UT-Arlington, who was hired by the Nicolaus but called to testify by State Farm, attributed some of the Nicolaus’ problems to changes in the weather and to the piers constructed by Krismer in 1982. Krismer suggested that there might be other leaks beneath the house that had not been located. Even Mr. Nicolau himself acknowledged at trial that all the problems could not have been caused by the leak.
Ill
The Court states that “the Nicolaus presented evidence from which a fact-finder could logically infer that Haag’s reports were not objectively prepared, that State Farm was aware of Haag’s lack of objectivity, and that State Farm’s reliance on the reports was merely pretextual.” Ante at 448. I examine all the evidence the Court cites. I do not weigh the evidence, as that is the province of the jury; but I must consider each bit of evidence cited by the Court to determine whether it shows that State Farm acted in bad faith.
The Court first points out that Teasdale admitted that eighty to ninety percent of his work and a substantial amount of Haag’s was for insurers. The Court notes that this is no evidence of bias, and of course, it is not.
The Court then says that “the evidence supports a logical inference that State Farm obtained the reports from Haag Engineering because of Haag’s general view that plumbing leaks are unlikely to cause foundation damage.” Ante at 448. With all due respect, that statement is simply not true. The only evidence the Court cites is the testimony of State Farm’s superintendent, Cooper, that he knew before he hired Haag that its engineers “were of the general opinion that a localized leak beneath the house would not cause foundation damage as a general rule. Sometimes they felt it would, sometimes they felt it wouldn’t.” But knowing what Haag’s general views were, and hiring them to assure a denial of coverage, are quite different things. The Court concedes that Cooper’s testimony alone is no evidence of bad faith, but it insinuates, without any support whatever from the record, that State Farm hired Haag to provide a basis for denying the Nicolaus’ claim. The Nicolaus’ experts admitted that State Farm did not always hire Haag to investigate foundation problems, and that it hired witnesses who found leaks to be the cause of the problems. Most importantly, Krismer, who was perhaps more critical of Haag and State Farm than any other witness called by the Nicolaus, and who could recall only one instance out of eighty or ninety when Haag had found that a leak caused foundation problems, testified:
Q If somebody were to say that State Farm hires Haag, and Haag only—
A That’s not—
Q —finds that there’s no relation between the leak and foundation movement, that would not be true, would it?
A No, that would not be true.
The Court then points out Krismer’s testimony that of the eighty or ninety foundation problems in other houses he knew Haag had investigated, he was aware of only two in which Haag had attributed the cause to a sewer leak. Ante at 449. But Krismer did not suggest that Haag had been wrong in its assessment of any situation other than the Nicolaus’, and he testified that the validity of an expert’s opinion in a particular situation could not be determined by his views of other situations in the past. Hayden, the Nicolaus’ engineer, testified that it was entirely legitimate to approach a situation like the Nico-laus’ believing the UT-Arlington study that leaks usually do not cause foundation problems. Petrie, the engineer hired by the Ni-colaus but called to testify by State Farm, stated that the differences in the opinions of the various witnesses were due, not to bias or *462prejudice, but to differences in training, specialty and experience.
The Court then mentions “evidence that State Farm, and the engineers on which it relied, failed to conduct an adequate investigation.” Ante at 449. Specifically, the Court cites testimony that Haag Engineering did not examine the leaking pipe itself or take core samples of its own. Unquestionably, Haag did not do either of these things, but equally unquestionably, they were unnecessary. One might as well criticize Haag for not testing for radon or asbestos. Haag assumed the existence of a leak as predicted by the Nicolaus’ other experts and as eventually located. Haag and State Farm were aware of the exact nature of the leak that was repaired. Their view was that the leak could not allow enough water under the foundation to affect it thirty feet away where it was shifting. Krismer and Brock Thomas, the Nicolaus’ general contractor, did not suggest what Haag would have gained by looking at the leak itself when it was willing to believe their reports of what they found. As for the core samples, Haag and State Farm reviewed the information TETCO reported from the samples it took, and there is no evidence that Haag would have gained anything by duplicating TETCO’s work. Again, Haag assumed TETCO’s tests were accurate, but did not agree that they showed damage caused by the leak.
As the Court notes, Ralph Cooper, State Farm’s superintendent, denied the Nicolaus’ claim without seeing the leak or taking core samples. But none of the Nicolaus’ witnesses suggested what Haag would have gained by staring at the leak or taking additional core samples. State Farm did not ask Maverick Engineering for additional information after Maverick concluded that the leak was the cause of the problems, but the Nico-laus do not claim that Maverick had any additional information that State Farm did not already have. Inasmuch as State Farm paid for Maverick’s investigation, if Maverick had any additional information, it should have provided it to State Farm without being asked. In short, State Farm and Haag reviewed all the information that the Nicolaus’ experts produced in their own investigations. There is no evidence that doing anything further would have affected their conclusions.
The Court then says that “[s]ome evidence indicates that State Farm knew, when it denied the Nicolaus’ claim for the second time, that the Haag reports did not constitute a reasonable basis for denying the claim.” Ante at 449. The only evidence the Court cites in this category is the TETCO report on the moisture content of four core samples taken at the Nicolaus’ house. As she notes, State Farm did not ignore this evidence but forwarded it to the Haag engineers, who reviewed it and issued a supplemental report. Haag did not itself take core samples because it accepted those done by TETCO.
The TETCO report is evidence in support of the Nicolaus’ policy claim, but it is not evidence of no reasonable basis for denying the claim, and it is certainly not evidence that State Farm knew there was no reasonable basis for denying the claim. While the Nicolaus believed that only the sewer leak could account for the level of moisture in the core samples, Haag explained that the leak was not a likely explanation because the moisture level was roughly the same near the leak and farther away. If all the moisture were coming from the leak, the soil should be wetter nearer the leak. There is no evidence in the record that Haag’s view was wrong or unreasonable. All the evidence shows, again, is a disagreement among the experts.
Finally, the Court says that “[ojther evidence also called into question the credibility of the Haag reports on which State Farm relied.” Ante at 450. The Court refers to Haag’s conclusions that the moisture levels found by TETCO were not high and could have been caused by natural condensation. Several witnesses, including some called by State Farm, testified that these conclusions were wrong. Assuming they were, the error does not call into question the credibility of Haag’s entire reports. As noted above, even if the moisture content of the soil was high, Haag offered an explanation why the leak was not the cause, which was not only plausible as a matter of common sense but was uneontradicted in the record. A leak might raise the moisture content of the soil around *463it, but the effect would necessarily be greater nearer the leak. The TETCO report indicated relatively uniform moisture levels. Haag’s ultimate conclusions did not turn on whether it considered the moisture levels to be high.
Not only was this dispute relatively incidental, it was no more a matter of credibility than the dispute over the cause of the Nico-laus’ problems. Both involved disagreements among experts, which, as the Court said in Monel, do not prove bad faith. The fact that two engineers testifying for the Nicolaus believed that the leak caused some of the problems does not necessarily mean that the five experts who testified to the contrary for State Farm were not credible, or vice versa. Hayden’s and Mr. Nieolau’s concessions that the leak could not have caused all the problems, and Krismer’s belief that there might have been other leaks, do not mean that the Nicolaus’ claim was not credible. In each instance, the experts were struggling to account for all the facts. The Nicolaus’ experts believed that the Haag reports were wrong in their entirety. Singling out one part does not make the dispute anything other than a disagreement among experts.
The Court also points to testimony by the Nicolaus’ attorney on cross-examination as evidence that State Farm handled the Nico-laus’ claim unreasonably. Even if bad faith could be established by the conflicting testimony of experts or by plaintiffs own testimony, surely it cannot be established by plaintiffs lawyer’s testimony. I would not hold her testimony to be evidence of bad faith in this case.
IV
State Farm could not evaluate the Nico-laus’ claim without the assistance of experts. No one disputes that the two engineers it hired were highly qualified, credentialed and experienced associates in a firm with worldwide business. No one denies that the two engineers believed — not for reasons related in any way to the adjustment of insurance claims, but solely for reasons related to science — that a leak in an unpressurized sewer line embedded in the soil beneath the foundation of a house cannot ordinarily cause the foundation to shift. No one challenges the legitimacy of this view as a general matter or the UT-Arlington study on which it is based. There is no evidence that the Haag engineers were predisposed to recommend denying the Nicolaus’ claim irrespective of the evidence; there is direct evidence to the contrary. There is no evidence that State Farm hired Haag to provide an opinion that would justify denying the Nicolaus’ claim; the direct testimony, from the Nicolaus’ own witness, is to the contrary.
Undisputedly, the Haag engineers reviewed all available evidence in their reports — evidence from their own investigation and from all the Nicolaus’ experts. Nothing the Nicolaus wanted examined was ignored, and State Farm paid for it all: for itself, two Haag engineers and two other engineers who testified at trial; and for the Nicolaus, a foundation repairman, two engineers, two plumbers and a general contractor. The only stone left unturned was the foundation itself.
All of the Nicolaus’ complaints about State Farm’s handling of their claim reduce to two: Haag’s conclusion was unreasonable, and State Farm hired Haag solely to concoct an excuse for denying coverage. The first is no more than a disagreement among experts, which we have held is not evidence of bad faith. Moriel, 879 S.W.2d at 18. The second is unsupported by any evidence. Viewed in the light most favorable to the Nicolaus, the evidence does not support the jury’s finding of bad faith.
The evidence in this case in no way depicts an unscrupulous insurer arbitrarily denying a claim to take unfair advantage of its insureds. Rather, the evidence depicts an insurer who, like the seven engineers and other experts who examined the Nicolaus’ predicament, tried in good faith to determine the cause. If this dispute was not bona fide, it is hard to imagine one that would be.
The arbitrariness in this case is not in the insurer’s conduct but in the cause of action. Our opinions today demonstrate that bad faith liability is as unpredictable as injury from a brick thrown out of a window or a gun *464fired into a crowd. For the reasons I have explained in Universe Life Insurance Co. v. Giles, I would adopt a legal standard of liability that would prevent such arbitrariness. 950 S.W.2d 48 (Hecht, J., concurring).
Accordingly, I dissent.