An accident in 1974 involving a railroad tank car and a chemical storage tank has led to years of environmental litigation and, ultimately, to this court’s consideration of those events almost a quarter century later. In June 1974, an engineer of the Providence and Worcester Railroad Company moved a group of railroad tank cars while one car was still attached to a chemical storage tank at a manufacturing facility in Cumberland, Rhode Island. The facility is located on the banks of the Blackstone River. A hole was torn in the bottom of the tank and the contents, over 6,200 gallons of perehlorethylene (“perc”), gushed out, boring a four-foot hole in the ground. The fire department responded to an emergency call and hosed down the area of the spill. No further action was taken, and so matters rested until 1979.
In October 1979, the Rhode Island Department of Public Health tested drinking water wells across the Blackstone River, in the nearby town of Lincoln, for environmental pollution. The state’s decision to test in this manner was an advanced one for the times. Those tests and subsequent tests showed that the wellfields and the aquifer from which they drew water were contaminated with a variety of volatile organic chemicals *37(“VOC’s”), including perc, 1,1-dichlorethane, 1,1-dichloroethylene, 1,2-transdiehloroethy-lene, 1,1,1-trichloroethane, trichloroethylene, and triehlorofluoromethane. The wells, which supplied water to Cumberland and Lincoln, were immediately closed. Subsequent investigation by the United States Environmental Protection Agency (“EPA”) pointed to an area across the Blackstone River and east of the wells as the likely source of the aquifer’s contamination. That area was occupied by the Peterson/Puritan manufacturing facility, which produced and packaged various household and personal care products. Peterson/Puritan is a subsidiary of the plaintiff CPC International, Inc., now known as Bestfoods. This facility is where the tank car accident happened in 1974.
This suit demonstrates the immense cost, complexity and duration of environmental insurance litigation.1 In 1987, CPC sued its excess carrier, Northbrook Excess and Surplus Liability Company,2 for indemnification of costs incurred during the EPA ordered clean-up of pollution at Peterson/Puritan. This is the third appeal to this court. Judgments have twice been vacated, this court has certified a question to the Rhode Island Supreme Court, the Rhode Island Supreme Court has issued an opinion which clarified Rhode Island “trigger of coverage” law, and the ease has been twice tried to juries. We discuss that history later. In 1997, a jury awarded CPC $12,632,885.94 in damages plus over $5 million in pre-judgment interest. That award is the subject of this appeal.
As is common in these cases, the jury had two main issues to decide. The first was whether an “occurrence” causing property damage took place between July 1,1979, and July 1, 1980, which was the policy period during which Northbrook provided coverage. The second was to determine whether the property damage resulted from company activities that were excluded from coverage by the policy’s standard “pollution exclusion” provision or whether the relevant occurrences fit within the. also standard “sudden and accidental” -exception to that exclusion (and was thus covered). As evidenced by the verdict, the jury decided both issues in favor of CPC.
Northbrook attacks the verdict on both fronts. First, Northbrook says the evidence compels the conclusion that there was no occurrence during the policy period. North-brook says that the “property damage” at issue in the case is strictly in the area immediately surrounding the Peterson/Puritan manufacturing facility, and that CPC either knew or should have known of this property damage prior to the policy period. Therefore, there was no “occurrence” (as that term is defined in Rhode Island law) during the policy period and no coverage is available.
Northbrook buttresses this challenge by saying it was unfairly hampered in its presentation of its ease (that CPC knew or should have known of the property damage) when the trial court excluded evidence about environmental events before 1979 elsewhere in the CPC corporate family. Northbrook says that the proposed evidence, two prior judicial decisions in which CPC was a party, contained fact-findings relevant to the state of CPC’s internal knowledge and the state-of-knowledge in the industry about groundwater pollution. Northbrook’s says that the evidence demonstrates that CPC’s regular waste-disposal practices and the 1974 perc spill should have put CPC on notice of property damage long before 1979, and thus there is no coverage during the insuring period.
*38Second, Northbrook challenges the jury’s conclusion that the property damage was caused by a “sudden and accidental” discharge. Northbrook concedes the 1974 perc spill was sudden and accidental, but argues that the evidence compels a conclusion that it was Peterson/Puritan’s routine waste-disposal and polluting practices, not the 1974 perc spill, which caused the pollution around the Peterson/Puritan site. Northbrook points out that the comprehensive general liability policy at issue here contains a standard pollution exclusion for the discharge of chemicals, and the only exception to that exclusion is for “sudden and accidental” events. Thus, Northbrook says, even if there were an occurrence during the policy period, North-brook is still not hable because the clean-up costs were driven by CPC’s routine polluting activities, not the spill.3
Now, twenty-five years after the tank rupture, almost twenty years after the discovery that the wells were contaminated and eleven years after the suit was instituted, we affirm the jury award. While the evidence did not necessarily require the jury’s conclusions, it certainly permitted them. We do not reach the cross-appeal.
I
Background Facts
We set the stage for the parties’ arguments with a general outline of the actions and findings of the involved governmental environmental agencies as to the two well-fields and the Peterson/Puritan site. Much of this is undisputed. The appropriate inferences to be drawn from certain environmental findings are, of course, disputed, and were argued to the jury.
In October 1979, the Rhode Island Department of Health, Division of Water Supply, using testing procedures advanced for the times, tested the municipal water supplies of the Town of Cumberland and the neighboring Town of Lincoln and discovered VOC contamination. The Quinnville Wellfield, supplying Cumberland, is located on the west side of the Blackstone River, across the river and approximately three fifths of a mile from the Peterson/Puritan manufacturing facility. The Lenox Street Well, supplying Lincoln, is located on the same side of the river as Peterson/Puritan, over a mile away. Both wells were closed immediately after the contamination was discovered.
In 1980, the EPA hired environmental engineers Goldberg-Zoino and Associates (“GZA”) to conduct a hydrogeologic study of portions of the aquifer underlying and around the Blackstone River in order to establish the source and extent of the groundwater pollution contaminating the wells. In 1982, GZA reported its conclusion that the most probable source of the contamination of the Quinnville Wellfield was the Peterson/Puritan plant. GZA relied principally on three critical findings: (1) the highest levels of VOC groundwater contamination were observed in the industrial area where Peterson/Puritan was located; (2) the VOC’s found in the Quinnville Wellfield were the same as those found in the groundwater in the industrial area; and (3) Peterson/Puritan was the only operation in the area known to use and store the VOC’s found in the water supply.
According to GZA’s report, the Blackstone River typically acts as a groundwater flow boundary, meaning that groundwater on the east side of the river is generally unable to cross over to the west side. When the wells are pumping, however, the river is not an effective barrier and groundwater is drawn into the wellfield.
During sustained pumping of the Lincoln wellfield, ... a portion of the flow crosses the Blackstone from the Cumberland side of the river. Flow enters the wellfield via both induced infiltration from the river and direct groundwater flow beneath the river from the northeast corner of the site.
In addition, the levels of the contaminant concentration in the wellfield are directly related to the intensity of the pumping.
*39When the wells are turned off and the flow-field reverts to its natural state, as described above, contaminant levels decrease significantly. When the wells are turned on again, contaminant concentrations increase with pumping duration to their previous levels. This indicates that contaminant flow is being induced via pumping from outside of the normal recharge area for the wellfield, i.e. the Cumberland side of the, Blackstone River [where Peterson/Puritan is located].
GZA concluded that the pumping of the wells drew contaminants from the area immediately surrounding the Peterson/Puritan plant under and across the river into the wellfield. Based on this analysis, Peterson/Puritan became the EPA’s primary focus.
Following this report, Peterson/Puritan hired Malcolm Pirnie, Inc., another environmental engineering firm, to further analyze the VOC groundwater contamination of the wellfield. Malcolm Pirnie’s report, issued in June 1983, supported GZA’s conclusion. “Peterson/Puritan is responsible for the release of VOC’s to the aquifer sufficient to have contributed to the past contamination of the Quinnville Wellfield----” In addition, “the sustained pumping from the wellfields could draw contaminated groundwater from the east to the west side of the Blackstone River where it could be drawn into the wells.” The report stated that the wellfield was no longer contaminated, but that contamination would be renewed by the recommencement of pumping within the wellfield without prior interception of the contaminant plume.
In 1983, the EPA designated an area including both the Peterson/Puritan site and the aquifer east of the Blackstone River as “OU-1,” and placed the site on its National Priorities List. In 1987, following several years of negotiations with Peterson/Puritan, the EPA issued an Administrative Order by Consent, pursuant to the Comprehensive Environmental Response Compensation and Liability Act (“CERCLA”), 42 U.S.C. §§ 9601-9675. That Order identified Peterson/Puritan as the party responsible for the release of the hazardous substances migrating into the groundwater at OU-1.
As required by .the Order, Peterson/Puritan undertook a Remedial Investigation/Feasibility. Study (“RI/FS”) to investigate additional responsible parties and further analyze site conditions. This report was prepared by ABB Environmental-Services and issued in 1993. ABB, like Malcolm Pirnie, largely confirmed GZA’s finding that Peterson/Puritan was the principal source, of VOC contamination in the aquifer and the Quinnville Well-field.4
The EPA issued its Record of Decision (“ROD”) for the OU-1 site in June 1993. The ROD stated that the Quinnville Wellfield “was a drinking water source in 1979, when it was closed due to OU-1 contamination. Prior to its closure, the wellfield provided water that did not pose any health threats.” The ROD further stated that the 1974 perc spill, along with historical releases into manholes and catch basins, was the primary source of contamination of the Quinnville Wellfield. The ROD concluded that the wells could be reopened when the contamination was remediated, and ordered CPC to clean up the OU-1 area as a prerequisite for the reopening of the wells. Finally, because the well-field was “a receptor of OU-1 contamination,” and “the potential future use of the wellfield as a drinking water source is a realistic possibility,” the December 13, 1995, Consent Order between CPC and the EPA *40extended OU-1 to include the Quinnville Wellfield.5
II
Insurance Coverage and Terms
A. Primary Coverage and Town of Lincoln Suit
In October 1982, based on GZA’s report, the Town of Lincoln filed suit against Peterson/Puritan for damages resulting from the contamination of its water supply. In June 1984, the suit was settled when Peterson/Puritan agreed to pay the Town of Lincoln $780,000 and install and maintain engineering controls in exchange for the Town’s release of all potential claims. The settlement was paid by Northwestern National Insurance Company, CPC’s primary insurance carrier, under a policy with a coverage limit of $1 million. On April 10, 1987, Northwestern National informed CPC and Northbrook that the primary insurance policy was exhausted, thus bringing Northbrook into the arena.
B. The CPC-Northbrook Insurance Policy
From July 1, 1979, to July 1, 1980, CPC was insured by Northbrook under an umbrella liability policy. This policy, with a limit of $25 million, was CPC’s first layer excess policy, hence the next coverage in line after Northwestern National’s $1 million primary policy was exhausted. The Northbrook policy insured CPC worldwide, including CPC’s subsidiaries and all other entities financially controlled by CPC.
Under Section 1, “Coverage,” the policy stated:
The Company hereby agrees, subject to the limitations, terms and conditions hereinafter mentioned, to indemnify the Insured for all sums which the Insured shall be obligated to pay by reason of the liability
A. imposed upon the Insured by law,
or
B. assumed under contract or agreement by the Named Insured,
for damages on account of
A. Personal Injuries
B. Property Damage
C. Advertising Liability,
caused by or arising out of each Occurrence happening anywhere in the world.
The Definitions section contained the following definitions:
“Property Damage” shall mean loss of or direct damage to or destruction of tangible property (other than property owned by any Insured) and which results in an Occurrence during the policy period....
“Occurrence” means an accident, event or happening including continuous or repeated exposure to conditions which results, during the policy period, in Personal Injury, Property Damage or Advertising Liability neither expected nor intended from the standpoint of the Insured....
The Exclusions section provided:
This policy shall not apply to Personal Injury or Property Damage arising out of the discharge, dispersal, release or escape of smoke, vapors, soot, fumes, acids, alkalis, toxic chemicals, liquids or gases, waste materials or other irritants, contaminants or pollutants into or upon land, the atmosphere or any water course or body of water; but this exclusion does not apply if such discharge, dispersal, release or escape is sudden and accidental.
In a “reservation-of-rights” letter dated April 27, 1987, Northbrook advised CPC that it had no obligation to indemnify CPC for claims arising from the Town of Lincoln action or the EPA-ordered cleanup. The letter was based, inter alia, on the pollution exclusion and on the date of occurrence.
*41in
This Litigation
A.- Commencement of Suit and Initial Application of New Jersey Law
On July 21, 1987, CPC filed suit against Northbrook in New Jersey Superior Court. CPC alleged that all of the conditions precedent to insurance coverage had been satisfied or waived, and sought a declaration that Northbrook was obligated to indemnify CPC for Peterson/Puritan’s “entire ultimate net loss” in excess of Northwestern National’s $1 million coverage limit. CPC also sought a judgment estopping Northbrook from further denying coverage.
Northbrook removed the action to the U.S. District Court which transferred it to the District of Rhode Island pursuant to 28 U.S.C. § 1404(a). CPC then filed a motion for a declaration that the substantive law of New Jersey still governed the litigation. The motion was allowed on the basis that a New Jersey eourt would apply New Jersey law to the case because New Jersey was the state which connected all the parties together and so had the most significant interest in the outcome of the case. See CPC Int’l, Inc. v. Northbrook Excess & Surplus Ins. Co., 739 F.Supp. 710, 715 (D.R.I.1990).
Applying New Jersey law, the district court allowed Northbrook’s motion for summary judgment on the ground that the pollution exclusion clause in the policy precluded coverage for gradual pollution. See CPC Int'l, Inc. v. Northbrook Excess & Surplus Ins. Co., 759 F.Supp. 966 (D.R.I.1991). The district eourt predicted that the New Jersey Supreme Court, which had not ruled on the issue, would define the term “sudden and accidental” as referring to events which are unexpected and unintended and which occur abruptly or over a short period of time. See id. at 973. The court ruled that CPC had not shown that the contamination was within this definition.
On appeal, this court reversed and remanded, saying the district court’s prediction gave insufficient weight to the decisions of the New Jersey Superior Court’s Appellate Division (New Jersey’s intermediate appellate court), which had concluded that the term “sudden and accidental” was ambiguous and had to be interpreted favorably to insureds as providing coverage for gradual pollution. This eourt concluded that the New Jersey Supreme Court would more likely construe “sudden and accidental” to mean only unintended and unexpected, i.e., not requiring the event to be abrupt or immediate. See CPC Int’l, Inc. v. Northbrook Excess & Surplus Ins. Co., 962 F.2d 77, 97 (1st Cir.1992).
B. Round Two: Application of Rhode Island Law and First Trial
On remand, Northbrook moved for reconsideration of the initial choice-of-law ruling in light of an intervening change in New Jersey choice-of-law rules. That motion was granted on the basis that New Jersey law now dictated that the substantive law of Rhode Island governed the case. See CPC Int’l, Inc. v. Northbrook, 839 F.Supp. 124 (D.R.I.1993). This court denied CPC’s petition for mandamus.
The ease went to a jury trial on January 28, 1994. At the close of CPC’s evidence, Northbrook moved for judgment as a matter of law under Fed.R.Civ.P. 50(a). On February 16, 1994, the district court granted Northbrook’s motion, saying that CPC had failed to present evidence from which a reasonable juror could conclude that there had been an “occurrence” during the 1979-80 policy period, because the perc spill took place in 1974 and there was no evidence that the contamination reached the aquifer during the policy period.
C. The Second Appeal: The Trigger of Coverage Question Under Rhode Island Law
On appeal, this court affirmed the application of Rhode Island law, but concluded that Rhode Island law was unclear on'the trigger of coverage and certified this question to the Rhode Island Supreme Court:
What trigger-of-coverage standard would the Rhode Island Supreme Court use for determining at what point an “occurrence” causing “property damage” took place, within the meaning of the insurance policy *42provisions provided in the separate opinion in this case, where an insured alleges that a spill of hazardous contaminants in 1974 migrated through the groundwater, causing immediate injury to the pertinent property, which was not, in fact, discovered, however, until at least 1979.
CPC Int’l, Inc. v. Northbrook, 46 F.3d 1211, 1222 (1st Cir.1995). The Rhode Island Supreme Court answered:
[A]n “occurrence” under a general liability policy takes place when property damage, which includes property loss, manifests itself or is discovered or in the exercise of reasonable diligence, is discoverable.
CPC Int’l, Inc. v. Northbrook, 668 A.2d 647, 650 (R.I.1995). In light of that answer, this court vacated the judgment in Northbrook’s favor and remanded the case for a new trial. The second trial was conducted in June of 1997 and resulted in the verdict for CPC. CPC was awarded $12,632,885.94 in damages plus prejudgment interest in the amount of $5,333,283. In addition, Northbrook was obligated to reimburse CPC for all costs incurred by CPC in remediating the Peterson/Puritan site after February 28, 1997. After the jury verdict, the court denied Northbrook’s motions for judgment as a matter of law under Fed.R.Civ.P. 50 and for a new trial under Fed.R.Civ.P. 59. It is from this jury verdict that this appeal is taken.
IV
Discussion
We discuss the pertinent evidence in light of the particular claims on appeal. The facts are stated as the jury and district court could have found them. See Cambridge Plating Co., Inc. v. Napco, Inc., 85 F.3d 752, 756 (1st Cir.1996).
A. The Discoverability of Property Damage
Northbrook’s central argument on appeal is that its motion for judgment as a matter of law was improperly denied, because the evidence compels the conclusion that CPC knew or reasonably should have known of property damage before July 1979.
Appellate review of the grant or denial of a motion for judgment under Fed. R.Civ.P. 50 is de novo, applying the same standard that governed the adjudication of the motion in the district court. See Costos v. Coconut Island Corp., 137 F.3d 46, 48 (1st Cir.1998). All of the evidence is examined in the light most favorable to the nonmoving party, drawing all possible inferences in its favor. See Cambridge Plating Co., 85 F.3d at 764. We do not consider the credibility of witnesses, resolve conflicts in testimony, or evaluate the weight of the evidence. See Wagenmann v. Adams, 829 F.2d 196, 200 (1st Cir.1987). We will reverse the denial of a Rule 50 motion “only if reasonable persons could not have reached the conclusion that the jury embraced.” Coconut Island Corp., 137 F.3d at 48 (citation and internal quotation marks omitted).
Here, the initial issue is whether there is sufficient evidentiary support for the jury’s conclusion that there was an “occurrence” between July 1, 1979, and July 1, 1980. The terms of the policy are set forth earlier. The definition of “occurrence” is given content by Rhode Island law:
[A]n “occurrence” under a general liability policy takes place when property damage, which includes property loss, manifests itself or is discovered or in the exercise of reasonable diligence, is discoverable.
CPC Int’l, Inc., 668 A.2d at 650. Under this formulation, the term “occurrence” and “property damage” are closely connected:
Read together, the provisions of the Northbrook policy provide coverage to an insured that sustains an occurrence—that is, an event that results in compensable property damage during the policy period. In other words, there can be no occurrence under the policy without property damage that becomes apparent during the policy period, and property loss and compensable damages cannot be assessed unless the property damage is discovered or manifests itself. “Property damage” and “occurrence” are thus inextricably intertwined.
Id. at 649.
Northbrook focuses its attack on the “discoverability” prong of the Rhode Island definition of “occurrence”, saying that CPC *43should have known that its routine waste-disposal and polluting activities and the 1974 perc spill would cause property damage long before 1979.6 Northbrook argues that the jury, while properly instructed on Rhode Island law, misapplied that instruction.7 As to how the jury applied the instruction, our knowledge is only that the jury returned a general verdict in favor of CPC in the amount of $12,632,885.94, and thus necessarily found that the property damage did not manifest itself, and could not have reasonably been discovered, before July 1979. North-brook did not request special verdicts.
B. Exclusion of Evidence
Before addressing the merits of the insurer’s argument on this sufficiency issue, we stop to consider Northbrook’s argument that it was.erroneously prevented from painting a fair and complete picture for the jury by the exclusion of evidence. As with any argument addressed to the exclusion of evidence, Northbrook faces the challenge of meeting the abuse of discretion standard. See Rodriguez-Hernandez v. Miranda-Velez, 132 F.3d 848, 855 (1st Cir.1998) (citation omitted). If Northbrook were correct that the exclusion of evidence was an abuse of discretion and prejudicial, we would not reach the issue of whether there was adequate evidence to support the jury .verdict. But while Northbrook’s argument is far from frivolous, our conclusion is that the ruling was well within the court’s discretion.
The major component of North-brook’s defense was that CPC should have known about the property-damage resulting from the perc spill and other polluting activities well before 1979. Towards that end, Northbrook put on state-of-the-art and state-of-knowledge evidence as to CPUs knowledge before 1979 of the dangers of release of VOC’s and perc in an effort to show that CPC knew or should have known that the chemicals would contaminate the ground and groundwater.
We take it as given that such evidence may generally be helpful to the jury in determining what a party should have known at some time in the past. Findings about what another operation of the company knew and had been told about the danger of groundwater contamination of a similar type can help the jury in determining whether CPC exercised reasonable diligence with regard to this particular spill. See Chemical Leaman Tank Lines, Inc. v. Aetna Cas. and Sur. Co., 817 F.Supp. 1136, 1150 (D.N.J.1993) (admitting evidence of plaintiff .insured’s problems at other tank cleaning sites on issue of company’s knowledge).
Such state-of-the-art and state-of-knowledge,evidence is used by both sides in many contexts in civil, and criminal environmental and toxic tort litigation. For example, it is used by insurers and insureds in insurance coverage eases. See Mottolo v. Fireman’s Ins. Co., 43 F.3d 723, 730 (1st Cir.1995) (discussing the usual summary judgment burden shifting rules applied to this area of law); Chemical Leaman Tank Lines, Inc., 817 F.Supp. at 1149-50 (use by insurer against insured); Hatco Corp. v. W.R. Grace & Co., 801 F.Supp. 1334, 1376 (D.N.J.1992)(insured submits state-of-knowledge evidence to oppose insurer’s contention that insured intended and expected to cause contamination at site); New Castle County v. Continental Cas. Co., 725 F.Supp. 800, 803-04 (D.Del.1989) (state-of-the-art knowledge about leachate from landfills apparently introduced by -insured on issue of whether the pollution was expected). It is also used in the allocation of responsibility under the “Gore” factors in private party CERCLA contribution actions. See Gould v. A & M Battery Tire Serv., 987 F.Supp. 353, 363-64 (M.D.Pa.1997)- (successor operator presents state-of-the-art defense). Some state-of-the-art evidence was introduced here, but specific evidence was excluded.
*44Here, Northbrook sought to introduce (1) judicial decisions in two previous CPC coverage suits involving groundwater contamination at CPC facilities in New Jersey and Michigan, see CPC Int’l, Inc. v. Hartford Accident & Indemnity Co., Bergen No. L37236-89 (N.J.Super. April 15, 1996); CPC Int’l, Inc. v. Aerojet-General Corp., 825 F.Supp. 795 (W.D.Mich.1993); (2) trial testimony of CPC’s Senior Corporate Counsel and others as to whether certain facts found in those opinions were correct; and (3) certain related documents. There was no specific offer of proof as to what any of the proposed witnesses would say; their depositions had not been taken. Nor was this broad band of evidence narrowly tailored to the precise issues involved at trial.
Northbrook argued that such evidence would show that CPC knew long before 1979 about the harmful effects of VOC groundwater contamination. Northbrook wanted to introduce certain findings of fact contained in the judicial decisions establishing that CPC was cleaning up groundwater contamination in its facilities before the 1970’s and that one state court had enjoined CPC from using lagoons for waste disposal because of the contamination which resulted. CPC responded that circumstances at the Michigan and New Jersey sites were very different from those at Peterson/Puritan, and that the prejudicial and confusing effect of admitting the decisions would far outweigh their probative value.
The New Jersey ease was a coverage suit filed by CPC for indemnification of environmental remediation costs incurred during the clean-up of three facilities operated by a-CPC subsidiary. All three sites had VOC groundwater contamination from the use of underground storage tanks and lagoons into which aqueous chemical waste residues were deposited.8 The Michigan ease was a consolidated CERCLA action where numerous parties contested liability for the cleanup of a dormant manufacturing facility which had heavy soil, surface water, and groundwater contamination, principally from the use of unlined lagoons as sites for chemical waste disposal.9
After extensive voir dire, the district court excluded the decisions on Fed.R.Evid. 403 grounds, saying first that their introduction would require a “replay of the litigation of those two eases,” in that the parties would have to argue about the similarity of the previous lawsuits to the instant case. This would take “much time and considerable effort, and I think with little result.” The court also said that the opinions were not relevant to whether CPC could have discovered property damage in the exercise of reasonable diligence, because the facts underlying the two opinions were different than the facts here; moreover, neither opinion addressed the issue of “property damage” as defined in the policy. The court concluded that introduction of the prior decisions could prejudice the jury about CPC and its conduct in the instant case.
We start with Northbrook’s initial burden of showing that the proffered evidence was relevant. On this, there is considerable confusion. If the proffered evidence plainly showed that the other sites involved similar chemicals and similar methods of transport and contamination, we could easily find the evidence relevant. But the record is confused and confusing. It appears that the other sites involved largely chemicals which were not VOC’s at all and included phenols and chlorides. It also appears that the sites did not involve a- massive quickly-happening and quickly disappearing flood of a chemical, such as the 1974 perc spill, but slow seepage *45and spillage from constantly filled waste lagoons.
But recognizing that knowledge is often gained from analogous events as well as from identical events, we will assume the evidence was at least somewhat relevant. Under Rule 403, the district judge was then required to determine whether “its probative value [was] substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury, or by considerations of undue delay, waste of time, or needless presentation of cumulative evidence.”
.The dangers of unfair prejudice were certainly real, as the trial judge recognized. “Unfair prejudice,” as the Advisory Committee Note teaches, means an “undue tendency to suggest decisions on an improper basis, commonly, though not necessarily, an emotional one.” Northbrook sought to introduce evidence in the form of judicial decisions that two other courts had made findings that other CPC operations had contaminated groundwater. Citizens do not look fondly on industrial polluters, particularly when the industries have been found to be such by a court. There was a danger such evidence would lead to a decision based on emotion or a desire to punish. We have said exclusions under Rule 403 were appropriate where such dangers are strong. See United States, v. Aguilar-Aranceta, 58 F.3d 796, 800-802 (1st Cir.1995); LaPlante v. American Honda Motor Co., 27 F.3d 731, 739-40 (1st Cir.1994); cf. Nickerson v. G.D. Searle & Co., 900 F.2d 412, 418 (1st Cir.1990) (affirming exclusion of questions of medical expert on his work in abortion climes because of danger of an emotional reaction).
By focusing on judicial findings, North-brook ran into other difficulties, ones this court described in Kinan v. City of Brockton, 876 F.2d 1029 (1st Cir.1989). In Kinan, the plaintiff sought to introduce evidence of two previous civil rights actions filed against the same police officer whose conduct was the subject of the instant suit; the plaintiff sought to introduce those previous decisions as proof that the defendant had a custom or policy of depriving its citizens of certain constitutional rights. See id. at 1033. The district court excluded the evidence. See id. Kinan affirmed the exclusion on the basis that the evidence, even if relevant, would result in jury confusion, wasted time, and be unfairly prejudicial.10 See id. at 1034-35. Kinan based its reasoning in part on the fact that the proffered cases had settled prior to trial and were decided on the basis of negotiation, not adjudicated findings of fact. See id. While Kinan is far from controlling, the dangers warned against in Kinan are real:
[Introducing evidence of the two other cases would inevitably result in trying those cases, or at least portions of them, before the jury. The merits of the two other cases would become inextricably intertwined with the case at bar. The result would be confusion and consumption of a great deal of unnecessary time.
Id. at 1034. These concerns were echoed by the district court in excluding Northbrook’s proffered evidence.
Weighed against this is the seemingly weak probative value of the evidence. In order to strengthen that probative value, Northbrook would have had to introduce evidence perfecting the analogy of the other sites and chemicals to the ones at issue in this trial. That, in turn, would have led, as the district court aptly noted, “to a replay of the litigation of those other two cases.” And that, in turn, could have misled the jury and certainly would have caused delay. The call made by the experienced trial judge, who had the look and feel of the trial before him, was eminently reasonable and not an abuse of discretion.
*46C. Sufficiency of Evidence on Discoverability
We return to the picture as it was painted for the jury. CPC and Northbrook presented two conflicting views of what CPC actually knew and what CPC should reasonably have known about pollution around the plant and the effect of the perc spill. Here, a jury rejected Northbrook’s claim that the state-of-knowledge about the dangers of VOC’s was such prior to 1979 that CPC was on notice about the dangers of groundwater contamination.11
On appeal, Northbrook constructs its argument—that CPC knew or reasonably should have known of that damage before 1979—out of these major elements:
1. CPC had actual knowledge of the perc spill in 1974.
2. CPC should have known that the perc from the spill would enter the groundwater within a few days.
3. CPC knew in 1974 that perc was a hazardous substance from the Chemical Safety Data Sheets (“CSDS”), which stated that “[i]f swallowed, perchlorethylene liquid has toxic effects.”
4. In 1972, the Manufacturing Chemists Association, the distributors of the CSDS for perc, distributed a manual entitled “Guidelines For Chemical Plants in Prevention, Control, and Reporting of Spills,” which warned:
2.1.6 Pollution of Ground Waters While many spills result in detrimental effects on sewage treatment systems of surface water, some can affect the quality of ground. waters. Leaks from storage tanks ... can percolate downward or laterally into shallow waters.
The geology of the area is an influencing factor in determining the migration of spills of this type. Sandy soils are particularly subject to percolation of pollutants into ground waters
This statement did not, however, refer specifically to perc.
CPC opposed Northbrook’s account with evidence that there was little reason before 1979 to think that a perc spill, or even historic releases of VOC’s into manholes or catch basins at the plant, would descend into the aquifer, move laterally into the wellfields, and cause property damage. CPC put on expert testimony that the dangers known to be associated with exposure to perc in 1979 were of skin contact, inhalation, and possible asphyxiation. CPC’s experts said that these were the primary dangers that were known concerning VOC’s.
As to the pre-1979 state-of-knowledge about responding to a perc spill, CPC’s evidence was that the earliest literature on that particular chemical was information in 1978 from the National Firemen’s Protection Association. That information was that such spills should be washed down with a fire hose. Those who actually dealt with the 1974 perc spill at the Peterson/Puritan facility testified that the company followed the directions of the Rhode Island Department of Environmental Management and washed down the spill and then covered it over with soil or gravel. The primary risk people were then concerned with was that of fire. Witnesses testified that there was little or no contemporary awareness about the risk and effect of groundwater contamination. Indeed, even as of 1979, there were no state or federal standards for impermissible levels of perc in drinking water.
We conclude that the jury could reasonably find, as it implicitly did, that the earliest discovered or discoverable property damage was the contamination of the Quinnville Well-field which caused it to be closed in 1979.12 *47The evidence did not compel a finding in Northbrook’s favor.'
D. The Pollution Exclusion and the Sudden and Accidental Exception
Having rejected Northbrook’s challenge to the jury’s finding that there was an occurrence within the policy period, we turn to its related argument that the property damage resulted from activities within the policy’s pollution exclusion, i.e., that CPC did not meet its burden of proving any “occurrence” was caused by a “sudden and accidental” discharge.
Northbrook does not dispute that the 1974 pere spill was sudden and accidental as that term is defined by Rhode Island law. Rather, its argument is that the perc spill is beside the point because the property damage for which the indemnification is sought is from all of the pollution at the Peterson/Puritan site, and this pollution is largely the result of the regular and routine polluting activities which took place at the plant over many years. This pollution included washing waste-water containing chemicals down various sinks and drains at the plant which discharged into the septic system and leaching fields. It also includes minor leaks and spills that occasionally took place at the chemical tank farm, where the spilled chemicals would travel directly through a gravel floor into the ground. Northbrook contends that these activities were not sudden and accidental, and thus are not covered under the pollution exclusion. The existence of a single sudden and accidental event, Northbrook says, such as the 1974 perc spill, does not mean that CPC may avoid the consequences of its ongoing contamination.
Northbrook’s argument falters on the fact that the jury accepted an alternative view of the evidence which has adequate support in the record. This alternate view is that the indemnification is sought for clean-up costs mandated by the EPA for OU-1 in its efforts to protect and restore the Quinnville Well-field, that the pere spill caused the damage to the wellfield, and that remediating the contamination from the perc spill is the principal source of the costs—including, as explained below, costs for remediating contamination which originated from routine activities at the plant.
Evidence upon which the jury could rely included the ROD, which stated that the wells had not been contaminated before 1979 and that, but for the contamination, the wells could be reopened. Referring to the perc spill, the ROD stated that “this spill, along with historical releases into manholes and catch basins, ... is the primary source of contamination.” (Emphasis added). The ROD described the wellfield as “a receptor of OU-1 contamination” and said that “the potential future use of the wellfield as a drinking water source is a realistic possibility.” The ROD thus required CPC to clean-up the OU-1 area to drinking water standards—a significant point which emphasizes the EPA’s focus on restoring the wellfield to use, not merely cleaning the soil (although one is prerequisite to the other).
The jury could also credit evidence that in the period before the pere spill such contamination as existed at the plant was trapped in the soil and did not migrate into the groundwater, and that the perc spill was the vehicle which brought the contaminants into the groundwater. On this point, there was expert testimony that the perc spill was responsible: some of the perc volatilize, forming a large, spreading vapor cloud which moved through the soil slowly, dissolving contaminants in its path, carrying those contaminants along with it as it descended into the groundwater.13
It is true, as Northbrook argues, that under the 1995 Consent Decree, CPC must carry out the remedy in the ROD including: *48(1) the excavation of contaminated soil in manholes and catch basins at the facility; (2) the'capping of soil in the tank farm and paving of soil in the O’Toole property; and (3) the implementation of a soil venting and soil vapor extraction system. But the jury apparently believed, on the evidence, that these costs were necessitated by the clean-up of the aquifer so that the wells could be restored, and that it was primarily the perc spill which necessitated that clean-up.
The evidence does not compel a finding in Northbrook’s favor on this point. The jury could reasonably have found that contaminants released at the plant would not have caused any property damage absent the perc spill, and that the perc spill was therefore the real and ultimate cause of the environmental damage to the wells. And the jury could reasonably find that the costs incurred by CPC during the EPA-ordered clean-up are for remediating pollution so that use of the Quinnville Wellfield might resume.
In sum, Northbrook’s challenge fails. The issues here were for the jury to decide and there was sufficient evidence to support the verdict. See New Castle County v. Hartford Accident and Indem. Co., 933 F.2d 1162, 1192 (3d Cir.1991).
E. Microanalysis
The parties and the amicus have argued in their appellate briefs about “microanalysis.” Northbrook suggests two different concepts under the label of “microanalysis.” The first is that where there is a history of a pollution-prone operation, it is wrong to analyze each single polluting event to determine whether it was “unexpected,” and thus perhaps subject to characterization as “sudden and accidental.” See, e.g., Lumbermens Mut. Cas. Co. v. Belleville Indus., Inc., 938 F.2d 1423, 1427-30 (1st Cir.1991) (rejecting such an approach). Amicus joins this argument, echoing this court in arguing that such an approach would “eviscerate the exclusion for pollution.” Id. at 1428. Various courts have characterized the so-called “microanalysis” approach as attempting to “break down [regular polluting activities] into temporal components in order to find coverage where the evidence unequivocally demonstrates that the pollution was gradual,” Smith v. Hughes Aircraft, 22 F.3d 1432, 1438 (9th Cir.1993) (internal quotation marks omitted); see American States Ins. Co. v. Sacramento Plating, Inc., 861 F.Supp. 964, 971 (E.D.Cal.1994), or as “an ill-fated effort to distinguish between virtually indistinguishable occurrences.” Lumbermens Mut. Cas. Co., 938 F.2d at 1428; see also Charter Oil Co. v. American Employers’ Ins. Co., 69 F.3d 1160, 1170 (D.C.Cir.1995) (describing approach as attempt to “disaggregate” “an activity occurring over an extended period”). The facts of this case do not invoke any of those perils. This is not an instance of attempting to parse a sequence of events in regular polluting activities’ into component parts and then arguing whether each part is sudden and accidental. Rather, there was a massive, sudden and accidental event in the perc spill and it was up to the jury to decide whether that, or the ongoing pollution, led to the property damage for which indemnification is sought. The short answer to Northbrook’s first “microanalysis” argument is then that this case does not raise that issue at all.
The second concept Northbrook raises under the label of “microanalysis” is whether damages for sudden and accidental releases may be parsed out from damages caused by routine, regular pollution within the exclusion. Massachusetts law, for example, allows for this possibility where the damage from sudden and accidental releases are identifiable and are themselves appreciable (or not de minimis) and compensable. See Highlands Ins. Co. v. Aerovox, Inc., 676 N.E.2d 801, 806, 424 Mass. 226, 234 (1997); Nashua Corp. v. First State Ins. Co., 648 N.E.2d 1272, 1275-76, 420 Mass. 196, 202-203 (1995); see also Millipore Corp. v. Travelers Indem. Co., 115 F.3d 21, 34 (1st Cir.1997). In those cases it was in the interest of the insured companies to'try to identify and carve out a portion of the damages as being attributable to sudden and accidental events (for which there was coverage) when the general damage suffered was attributable to regular polluting activities which were subject to the pollution exclusion (for which there was no coverage). But the doctrine is a two-way street: here, it may have been in the insurer’s interest to carve out a portion *49of CPC’s claim as attributable to regular pollution activities (and thus not coverable) against a backdrop of an overall damages landscape a jury could have viewed as shaped by one sudden and accidental event. But, as a matter of trial tactics, the insurer chose not to raise the issue, and it is waived.14 We thus have no occasion to predict whether Rhode Island law, which governs this case, will adopt a rule similar to the Massachusetts rule.
The decision of the district court is affirmed.15