THOMPSON, Circuit Judge.

Overview

This diversity case involves an insurance-coverage dispute, governed (the parties agree) by Massachusetts substantive law. On one side is SGI-Walpole, LLC and Stor/Gard, Inc. On the other is Strathmore Insurance Company. A federal magistrate judge (presiding by the parties' consent, see 28 U.S.C. § 636(c)) granted Strathmore summary judgment, and SGI-Walpole and Stor/Gard ask us to take a second look. Exercising de novo review, we affirm, though our reasoning differs somewhat from the magistrate judge's. See, e.g., RTR Techs., Inc. v. Helming, 707 F.3d 84, 92 (1st Cir.2013) (discussing some of the ins and outs of summary-judgment review). We will explain our thinking shortly. But first we summarize the relevant facts in the light most friendly to SGI-Walpole and Stor/Gard (the summary-judgment losers). See, e.g., Soto-Padró v. Pub. Bldgs. Auth., 675 F.3d 1, 2 (1st Cir.2012).

A Hard Rain Falls

The story begins in Walpole, Massachusetts, in March 2010, after a late-winter nor'easter blew through the area, dumping about seven inches of rain in three days. To give the reader a sense of just how bad the storm was, the Bay State's governor ended up declaring a state of emergency. Anyway, SGI-Walpole owns land in Walpole with self-storage warehouses on it. Stor/Gard manages the property, which, importantly, butts up against a retaining wall at the base of a steep slope. And, rounding out our brief description of the parties, Strathmore insured the two companies' interest in the property. Unfortunately, by the storm's end, a pile of soil had slid down the hill and into and over the retaining wall, damaging one of the buildings — to the tune of several hundred thousand dollars.

The Policy

Stripped to essentials, the policy between the parties — called, in insurance lingo, an "all risks" policy — covers all physical loss to the property unless "caused by or resulting from" an excluded peril.1 Among the many exclusions is one — found in the exclusions section's first numbered paragraph — barring coverage "for loss or damage caused directly or indirectly by... earth movement," which includes a "landslide." "Such [earth movement] loss or damage" is noncompensable "regardless of any other cause or event that contributes concurrently or in any sequence to the loss."2 Known in the insurance world as an "anticoncurrent cause" clause, what this provision does is deny coverage whenever covered and excluded perils combine to cause the loss. See, e.g., Boazova v. Safety Ins. Co., 462 Mass. 346, 968 N.E.2d 385, 393-94 (2012).

For clarity's sake, a word or two more about anticoncurrent-cause clauses might be helpful. Understanding how all-risk policies work is fairly easy, at least at a certain level. If a peril is excluded, there is no coverage. See 3 Stephen A. Cozen, Insuring Real Property § 48.03[1], at 48-19 (2009). If a peril is not excluded, there is coverage. See id. It gets a bit more complicated when excluded and covered perils combine to cause the loss — i.e., when there is concurrent causation. Courts have adopted a few different approaches for dealing with this very situation. See id. We need not delve into them any further than to say that Massachusetts courts follow the "efficient proximate cause" approach: Skipping over nuances not relevant here, coverage exists if "the predominant cause of the loss is a covered peril," Boazova, 968 N.E.2d at 394 n. 4, or if a covered cause sets in motion a "train of events" leading to the loss, Jussim v. Mass. Bay Ins. Co., 415 Mass. 24, 610 N.E.2d 954, 955-56 (1993).3 Looking to contract around the concurrent-cause doctrine, the insurance industry has come up with anticoncurrent-cause clauses, which, as we just said, bar coverage for damage caused by an excluded cause, regardless of whether a covered act also contributed to the damage. See Boazova, 968 N.E.2d at 394-95; see also 7 Lee R. Russ & Thomas F. Segalla, Couch on Insurance § 101:57, at 101-78 (2005).

Now back to our policy. Another key exclusion — found in numbered paragraph 2 of the exclusions section — excludes coverage "for loss or damage caused by or resulting from" a "collapse, except as" set forth in the policy's "additional coverage for collapse" section.4 And another exclusion — found in the exclusions section's numbered paragraph 3 — precludes coverage "for loss or damage caused by or resulting from ... weather conditions ... if weather conditions contribute[d] in any way with a cause or event excluded" in numbered paragraph 1 (e.g., landslide) "to produce the loss or damage."5 A few pages later the "additional coverage-collapse" proviso appears — a proviso that says that Strathmore "will pay for direct physical loss or damage" to the property "caused by collapse of a building or any part of a building" insured under the policy. But there is a catch. The collapse must have been "caused by" a listed peril — one of which is something called "specified causes of loss," which (most pertinently for our purposes) means "water damage," which in turn means "accidental discharge or leakage of water ... as the direct result of the breaking apart or cracking of a plumbing ... or other system" on the property.6

The Denied Claim — Lead-Up and Fallout

SGI-Walpole and Stor/Gard filed a claim with Strathmore, and Strathmore sent AEGIS Engineering, Inc. and GZA GeoEnvironmental, Inc. to determine what had caused the damage. AEGIS concluded that rain from the nor'easter had soaked into the soil, causing a landslide that caused the loss. Acting on an idea floated by a representative of SGI-Walpole and Stor/Gard, AEGIS checked whether water leaking from a drain pipe had caused or contributed to the landslide. AEGIS did detect some water leakage. But after running some tests, AEGIS flatly said that that leakage "was not a cause or contributing factor." The pipe-leakage amount was "negligible" compared to the rain amount, AEGIS noted, and besides, the "majority" of the leakage "flowed ... away from the slope and retaining wall failure area...."

Similarly, GZA's investigation led it to blame the "slope and retaining wall failure" on "a combination of heavy rain infiltrating into the slope and the inadequate subsurface drainage behind the retaining wall." GZA also spotted some water leaking from the pipe. But, like AEGIS, GZA rejected the idea championed "by others that the failure was caused by leakage of stormwater from the subsurface drainage system...." That theory, GZA emphasized, "is not supported by our analyses," which "indicate[s] that such leakage represented merely 2 percent of the water [that] infiltrated the failed area."

Armed with reports from both firms, Strathmore denied coverage, citing (most relevantly here) the landslide, collapse, and weather exclusions. Unwilling to take this lying down, SGI-Walpole and Stor/Gard sued Strathmore in federal court, alleging breach of the insurance contract and violation of the Massachusetts consumer-protection act, Mass. Gen. Laws ch. 93A. Not missing a beat, Strathmore counterclaimed, seeking a declaration that the policy did not cover the claimed loss. Both sides moved for summary judgment, and ultimately, the magistrate judge denied SGI-Walpole and Stor/Gard's motion and granted Strathmore's.

The policy excludes damages from landslides, the magistrate judge noted. And, she said, the AEGIS and GZA reports characterize the slope movement here as a landslide. Plus SGI-Walpole and Stor/Gard presented no evidence that the "accident" was not a "landslide," she added. Then she took on their elaborate coverage theory, which went something like this: The exclusions section, they said, has an anticoncurrent-cause clause in numbered paragraph 1, but, they insisted, the additional-coverage section does not. And so, in their view, the additional-coverage section provides coverage even if an excluded peril (think landslide) contributed to the loss. Quoting a snippet of the GZA report saying "2 percent of the water which infiltrated the failed area" came from a leaky pipe, they contended that that leakage constituted a specified cause of loss per the additional-coverage section, which combined with the rain water to trigger the landslide that caused the collapse. In the end, because the pipe leakage is a covered concurrent cause and because there is no anticoncurrent-cause clause in the additional-coverage section, Strathmore is obligated to pay up despite the landslide exclusion — or so their argument concluded. The magistrate judge was unpersuaded. Causes of loss in the additional-coverage section (like water damage), she stressed, "are still subject to the exclusions in the exclusion[s] section" (one of which is landslides). Also, a commonsense reading of the policy, she believed, shows that the anticoncurrent-cause clause in numbered paragraph 1 of the exclusions section (excluding all loss caused "directly or indirectly" by a landslide, "regardless of any other cause or event that contributes concurrently or in any sequence to the loss") bars "coverage for any damage caused in whole or in part" by losses covered in the additional-coverage section. And, she ruled, because the loss here "was caused at least in part by" the excluded landslide peril, Strathmore's denial of coverage was proper as a matter of law, which, she added, doomed SGI-Walpole and Stor/Gard's chapter 93A claim.

So the magistrate judge entered final judgment against SGI-Walpole and Stor/Gard. Unhappy with this outcome, SGI-Walpole and Stor/Gard appeal.

Our Take on the Case

A few legal principles (some of which we mentioned earlier) guide our decision. They fall into two groups. The first deals with some of the inner workings of the summary-judgment standard. The second deals with some of the basics of Massachusetts insurance law.

As always, we give fresh review to the grant of summary judgment, affirming only if the record — read as required (here, in the light most agreeable to SGI-Walpole and Stor/Gard) — shows both the lack of any genuine issue of material fact and the moving party's entitlement to judgment as a matter of law. See, e.g., Candelario Del Moral v. UBS Fin. Servs. Inc. of P.R., 699 F.3d 93, 99 (1st Cir.2012); Rodríguez v. Municipality of San Juan, 659 F.3d 168, 175 (1st Cir.2011); see also Fed. R.Civ.P. 56(a). And of course we are not straitjacketed by the magistrate judge's reasoning — quite the contrary, we are free to uphold her order on any basis present in the record. See, e.g., Loubriel v. Fondo del Seguro del Estado, 694 F.3d 139, 142 (1st Cir.2012).

Moving on to Massachusetts insurance-coverage decisions, we see that the insureds (here, SGI-Walpole and Stor/Gard) have the initial burden of showing that the case involves a generally covered risk under the policy. See, e.g., Boazova, 968 N.E.2d at 390. Should the insureds accomplish that task, the burden shifts to the insurer (here, Strathmore) to show an exclusion applies. Id. And if the insurer satisfies that burden, the burden shifts back to the insureds to show an exception to the exclusion holds sway. Id. Making life a little easier for us, the parties just argue about whether SGI-Walpole and Stor/Gard shouldered this last burden. Finally, when it comes to causation, remember the applicable standard here is efficient proximate cause, using the predominant-cause or train-of-events tests highlighted above. See id. at 394 n. 4; Jussim, 610 N.E.2d at 955-56.

SGI-Walpole and Stor/Gard's argument to us — a variation on the one they made below — involves (bear with us) seven steps. Step one: Landslides and collapses are excluded causes of loss, and (they concede) their losses here would be excluded by these provisions but for one thing — the additional-coverage section. Step two: That section provides coverage for a collapse caused by "water damage," defined as including water that leaked from a cracked plumbing (or similar) system. Step three: GZA found that water had saturated soil on the slope, helping to cause the landslide that caused the collapse. Step four: 2 percent of that water came from a leaky pipe, according to GZA. Step five: The presence of any water from a cracked pipe — "no matter the percentage," they write — means "water damage" helped cause the collapse. Step six: The magistrate judge got it wrong in construing the policy, because the additional-coverage section cannot be trumped by the landslide, collapse, or weather-condition exclusions, given that it has no anticoncurrent-cause language. Step seven: So the summary-judgment ruling cannot stand. Though we give SGI-Walpole and Stor/Gard points for creativity, their theory does not hold together.

"The simplest way to decide a case is often the best," a distinguished judge once wrote. Chambers v. Bowersox, 157 F.3d 560, 564 n. 4 (8th Cir.1998) (R. Arnold, J.). Thankfully, there is a simple way here. Let us put aside SGI-Walpole and Stor/Gard's complaints about the magistrate judge's policy interpretation (step six of their seven-step argument, for those keeping track). We need not (and therefore do not) decide those tricky issues (and the anticoncurrent-cause issue, for example, is nothing if not tricky), because to get anywhere SGI-Walpole and Stor/Gard must convince us that the summary-judgment evidence shows that water from the cracked pipe "caused" the collapse, to paraphrase and quote the key parts of the additional-coverage section. Recall that it is SGI-Walpole and Stor/Gard's burden to make that showing. See, e.g., Boazova, 968 N.E.2d at 390. And this they have not come close to doing. We explain.

Properly understood, SGI-Walpole and Stor/Gard's reversal theory turns on concurrent causation (just like they agreed, at least implicitly, at oral argument) — i.e., that a covered peril (water damage, to be precise) combined with one or more excluded perils (weather conditions, perhaps, in the form of rain, or a landslide) to cause the collapse. But even on their own view of things, Strathmore can be on the hook for their losses only if the water leak was the efficient proximate cause of the collapse — in other words, only if the leak was the predominant cause of the collapse or set in motion the cause or causes of the collapse.7 See, e.g., Boazova, 968 N.E.2d at 394 n. 4; Jussim, 610 N.E.2d at 955-56. SGI-Walpole and Stor/Gard argue that a sentence from the GZA report makes that required showing. Not so.

Yes, GZA found that "2 percent" of all the water that had "infiltrated the failed area" had come from the leaky pipe. But pulling no punches, GZA thoroughly rejected the theory (pushed by someone on SGI-Walpole and Stor/Gard's behalf) that that leakage had caused the failure — such a theory "is not supported by our analyses," GZA declared. Elsewhere GZA stressed that "the heavy rain" and "the inadequate subsurface drainage" had combined to "cause[]" the slope and retaining-wall failures. Note that GZA did not say that water from the rain and the cracked pipe had combined with the poor drainage system to cause the failures. And GZA was not alone in rejecting the leakage theory. AEGIS did not buy it either. The leakage, AEGIS wrote, was "negligible," in comparison with the amount of rain water that had soaked into the soil, which is a pithy way of saying that the leakage was "so small or insignificant as not to be worth considering." See Oxford English Dictionary Online, http://www.oed.com (quick search "negligible") (last visited May 15, 2013). The leakage, then, "was not a cause or contributing factor," AEGIS bluntly stated. For what it is worth, SGI-Walpole and Stor/Gard submitted no expert reports supporting their hypothesis or contradicting GZA's and AEGIS's conclusions.

The upshot of this is that SGI-Walpole and Stor/Gard cannot prevail. After all, it would take a topsy-turvy view of causation to say that a leak not listed as one of the combined causes of the collapse (GZA report) and dismissed as too insignificant to matter (AEGIS report) really was the predominant cause of the collapse or put in motion events that caused the collapse. See, e.g., Boazova, 968 N.E.2d at 394 n. 4; Jussim, 610 N.E.2d at 955-56. And because Strathmore acted well within its rights in denying coverage, the magistrate judge properly denied SGI-Walpole and Stor/Gard's summary-judgment motion and properly granted summary judgment for Strathmore on all claims. See, e.g., Timpson v. Transamerica Ins. Co., 41 Mass.App.Ct. 344, 669 N.E.2d 1092, 1098 (1996).

Final Words

Our story over, we affirm the judgment below in all respects. Each side shall bear their own costs on appeal.