OPINION

SILLS, P.J.

I

Laurence and Michaelyn Harper signed two contracts with Frank Ultimo and Ultimo Organization to stabilize the soil and re-level a pool on the Harpers' property. Both contracts were preprinted. Ultimo rejected the Harpers' attempt to add an "addendum" relating to start and stop dates, integration, and notification. The contracts also contained arbitration provisions, which provided that all controversies under, or related to the contract were to be settled "in accordance with the Uniform Rules for Better Business Bureau Arbitration." The Better Business Bureau's rules were not attached to the contract.

Allegedly, Ultimo broke a sewer pipe in the course of the work, with the result that concrete spread into the sewer system and soil, causing blocks to form in the house's plumbing system. Ultimo also is supposed to have caused considerable damage to the backyard drainage system. And he misled the Harpers as to the amount of work performed.

The Harpers' soon discovered, however, that the arbitration rules of the Better Business Bureau limit the damages and remedies available to dissatisfied customers. Customers cannot obtain compensation for "personal injuries" unless all parties otherwise agree in writing (and, after oral argument and as of the date of this writing, Ultimo has conspicuously not made an offer to so agree). Customer remedies are limited to full or partial refund, completion of work, costs of repair or any out of pocket loss or property damage, but "not to exceed $2,500, caused by provision of the service." Any additional remedies may be awarded "only if" the remedy is already included in a business's precommitment with the Bureau or, as in the case of personal injury claims, if agreed in writing by all parties. Customers are thus precluded under the Better Business Bureau arbitration rules from obtaining tort damages, punitive damages, or any other damages otherwise appropriate in a court of law.

The Harpers filed suit in superior court, alleging tort causes of action for negligence and fraud as well as breach of contract. The suit also seeks punitive damages as well as other tort and contract relief. Ultimo brought a motion to compel arbitration. The trial court denied the motion, concluding that the arbitration clause was unconscionable and therefore would not be enforced.

II

(2) Yet while both must be present, they need not be present in equal amounts. There is a sliding scale where the greater the evidence of procedural unconscionability, the less evidence is needed of substantive unconscionability. (See McManus v. CIBC World Markets Corp. (2003) 109 Cal.App.4th 76, 91 [ 134 Cal.Rptr.2d 446] ["a sliding scale is invoked whereby the more procedurally oppressive the arbitration clause is, the less evidence of substantive unconscionability is required to warrant the conclusion that the agreements to arbitrate are unenforceable"].) And vice versa. (See Armendariz, supra, 24 Cal.4th at p. 114 ["the more substantively oppressive the contract term, the less evidence of procedural unconscionability is required to come to the conclusion that the term is unenforceable"].)

Here both procedural and substantive unconscionability are so present that it is almost impossible to keep from tripping over them.

(3) Procedural unconscionability focuses on the factors of surprise and oppression ( Stirlen v. Supercuts, Inc., supra, 51 Cal.App.4th at p. 1532, quoting A M Produce Co. v. FMC Corp. (1982) 135 Cal.App.3d 473, 486 [ 186 Cal.Rptr. 114]), with surprise being a function of the disappointed reasonable expectations of the weaker party. (See Armendariz, supra, 24 Cal.4th at p. 113.)

Here is the surprise: The customer must inevitably receive a nasty shock when he or she discovers that no relief is available even if out and out fraud has been perpetrated, or even if he or she merely wants to be fully compensated for damaged property.

Here is the oppression: The inability to receive full relief is artfully hidden by merely referencing the Better Business Bureau arbitration rules, and not attaching those rules to the contract for the customer to review. The customer is forced to go to another source to find out the full import of what he or she is about to sign — and must go to that effort prior to signing.

But the oppression is even more onerous than that: As written, the clause pegs both the scope and procedure of the arbitration to rules which might change. And it is unclear whether an arbitration would be conducted under the Better Business Bureau rules as of the time of contracting, or at the time of arbitration. Thus even a customer who takes the trouble to check the Better Business Bureau arbitration rules before signing the contract may be in for a preliminary legal battle in the event that Better Business Bureau arbitration rules were to become substantively less favorable in the interim. Before the main battle commenced in arbitration, there would be a preliminary fight over which set of arbitration rules governed — something which, at the very least, would add to the customer's legal expense. (Cf. Armendariz, supra, 24 Cal.4th at pp. 110-112 [noting problem of forum fees in requiring party who imposes the arbitration to bear its costs].)

The arbitration "rules" of the Better Business Bureau are not just procedural ones, as Ultimo contended in oral argument. By limiting the scope of arbitral claims, the Better Business Bureau rules have the effect of substantively limiting the defendant's exposure.

(4) Substantive unconscionability focuses on the one-sidedness or overly harsh effect of the contract term or clause. ( Armendariz, supra, 24 Cal.4th at p. 114, quoting A M Produce Co., supra, 135 Cal.App.3d at pp. 486-487.) In the present case, the operative effect of the arbitration is even more one-sided against the customer than the clauses in any number of cases where the courts have found substantive unconscionabity. (E.g., Little v. Auto Stiegler, Inc. (2003) 29 Cal.4th 1064 [ 130 Cal.Rptr.2d 892, 63 P.3d 979] [either party could appeal any award of more than $50,000 to second arbitrator]; Szetela v. Discover Bank (2002) 97 Cal.App.4th 1094 [ 118 Cal.Rptr.2d 862] [arbitration clause absolutely barred class actions]; Saika v. Gold (1996) 49 Cal.App.4th 1074 [ 56 Cal.Rptr.2d 922] [arbitration award could be rejected if it exceeded $25,000].) As in Little, Szetela and Saika, the limitation of damages provision here is yet another version of a "heads I win, tails you lose" arbitration clause that has met with uniform judicial opprobrium. (E.g., Little, supra, 29 Cal.4th at p. 1072.)

In fact, unlike Little and Saika, the bare language of the arbitration clause here ("[A]ny controversy arising out of or related to this contract or the breach thereof"), the customer does not even have the theoretical possibility he or she can be made whole. At least in Little and Saika the plaintiffs could obtain full relief if they ran the gauntlet of a defendant's unsuccessful appeal. At least in Szetela the individual customer could be made whole, if not the rest of the class. Here, absent agreement from the defendant, there is not even the possibility of full relief.

The idea that since Ultimo is also limited in what he can recover from the Harpers, and therefore there is enough "mutuality" to render the clause not unconscionably one-sided, is specious. The relevant probabilities of the deal were these: The odds were far more likely that the customers would have claims in addition to just getting their money back than the business would have claims in addition to just getting paid. (See Little, supra, 29 Cal.App.4th at pp. 1064, 1072-1073 [looking to the probabilities as to which party was more likely to benefit in determining whether clause was unconscionably one-sided].)

III

So why isn't this a frivolous appeal? There are two things that not only give Ultimo's appeal sufficient color of rationality to obviate frivolousness, but implicate issues where the common law now provides only sketchy guidance.

A

The first stems from a comment made by the trial judge. He specifically rejected the idea that the Ultimo-Harper contract was an adhesion contract. There were market alternatives. "But in a case of this kind where we're talking about soils work, et cetera, you can go down the road and hire someone else instead, and that is different such that I do not believe that it is a contract of adhesion."

This comment is important, because beginning with Stirlen, supra, 51 Cal.App.4th 1519, most cases have routinely and overtly stated that the starting point in any analysis of unconscionability is whether the contract is one of adhesion. (See id. at p. 1530 ["The analysis commences with the determination whether the agreement to arbitrate is a contract of adhesion."]; Armendariz, supra, 24 Cal.4th at p. 113 ["Unconscionability analysis begins with an inquiry into the whether the contract is one of adhesion."]; Jaramillo v. JH Real Estate Partners, Inc. (2003) 111 Cal.App.4th 394, 400 [ 3 Cal.Rptr.3d 525] [quoting Armendariz]; Pardee Construction Co. v. Superior Court (2002) 100 Cal.App.4th 1081, 1086 [ 123 Cal.Rptr.2d 288] [quoting Armendariz]; Flores v. Transamerica HomeFirst, Inc. (2001) 93 Cal.App.4th 846, 853 [ 113 Cal.Rptr.2d 376] [citing Armendariz and Graham]; Bolter v. Superior Court (2001) 87 Cal.App.4th 900, 910-911 [ 104 Cal.Rptr.2d 888] [quoting Armendariz].)

It is therefore an easy mistake to assume that a finding of adhesion is a prerequisite for a finding of unconscionability. A few commentators have suggested exactly that. (See Looney Poole, Adhesion Contracts, Bad Faith, and Economically Faulty Contracts (1999) 4 Drake J. Agr. L. 177, 182 ["some courts suggest that the presence of standard form contracts as a near prerequisite of unconscionability"]; Stempel, Reassessing the 'Sophisticated' Policyholder Defense in Insurance Coverage Litigation (1993) 42 Drake L. Rev. 807, 831 ["many courts treat adhesion as a near prerequisite for deeming a contract term 'unconscionable'"].)

So let us be quite clear about it: Adhesion is not a prerequisite for unconscionability. To take the most obvious example: A person, free of debt, sound of mind, and without economic duress of any kind, could not voluntarily sell himself into slavery in our legal system for any period of time. Oh, it might be a wonderfully arms-length contract. The parties might each be represented by counsel. No preprinted forms would need be involved; the contract might, in fact, be set forth in a wholly custom-drafted manuscript, with each party initialing every provision. There might even be a "seller's market" for such contracts, supply and demand being what they are. (Notice we didn't mention the price.) But there is no way, adhesion or no adhesion, that any court in these United States is going to enforce that contract. Faust, depending on which version of the story you read, might or might not end up going to hell because he freely — and without duress — sold his soul to the Devil; he was pretty well off by the standards of his time. But someone who makes a faustian pact to become someone else's slave in our system is going to find that, among other things, the deal is unconscionable, period, end of discussion.

The source of the original citation in Stirlen for the idea that courts should "begin" or "start" with the existence of a contract of adhesion was the internal literary structure of Graham v. Scissor-Tail, Inc. (1981) 28 Cal.3d 807, 817-819 [ 171 Cal.Rptr. 604, 623 P.2d 165]. Graham did not explicitly say that a court should begin with the adhesion issue, it just did it. And that itself was not surprising, given that the appellant in Graham had framed the issues in the case also beginning with the idea of adhesion. (See Graham, supra, 28 Cal.3d at p. 817.)

Graham would set the pattern for Armendariz and Little, where the court began (or said that analysis should begin with) the question of adhesion, and a subsequent finding of adhesion. In those two cases, unlike the one before us, there was no occasion to confront the problem of unconscionability when there wasn't a contract of adhesion.

We could, of course, simply dismiss the trial court's comment that the Harper-Ultimo contract was not one of adhesion as mere dictum, and point out that the trial court need only arrive at the right result; it need not get all the math exactly right in arriving at that result. In support of that dismissal we might also point out that the contract before us now was a preprinted form contract offered on a "take-it-or-leave-it" basis, as underscored by the fact that the seller rejected the buyer's custom-drafted terms.

There is, however, something to be said for the trial court's observation. As this court remarked in Szetela, the available of market substitutes "may be relevant to whether the contract is one of adhesion." ( Szetela, supra, 97 Cal.App.4th at p. 1100.)

But, as we also said in Szetela, "even if the clause at issue here is not an adhesion contract, it can still be found unconscionable." ( Szetela, supra, 97 Cal.App.4th at p. 1100.)

Graham itself had obliquely alluded to the question of unconscionability's relationship to arbitration in a passage at that end of its unconscionability discussion, by stating that if an arbitration clause does not achieve "'minimum levels of integrity'" it would "be denied enforcement in any circumstances," but all the more so where it is in a contract of adhesion. (See Graham, supra, 28 Cal.3d at p. 828, emphasis added.) That is, adhesion was not essential to a finding of unconscionability.

What was oblique in Graham, though, has now become clear in Little. "The procedural element of an unconscionable contract generally takes the form of a contract of adhesion." ( Little, supra, 29 Cal.4th at p. 1071, emphasis added; see also Flores v. Transamerica HomeFirst, Inc., supra, 93 Cal.App.4th at p. 853 ["A finding of a contract of adhesion is essentially a finding of procedural unconscionability."].) That is, you can show procedural unconscionability by a showing of adhesion, but it is not the only way. There will be cases, like this one, where procedural unconscionability is obvious without the need to establish that the contract is one of adhesion.

Indeed. Very clever readers will have noticed in our discussion of procedural unconscionability above that we did not make any reference to the traditional criteria of adhesion. There was no need. Without becoming bogged down into the question of the degree to which market alternatives and available substitutes undercut the tendency to find a contract adhesive (see Szetela, supra, 97 Cal.App.4th at p. 1100 [rejecting idea that meaningful opportunity to obtain the goods or services from other providers was "the relevant test for unconscionability"]), this case certainly involves procedural unconscionability regardless of whether the contract is adhesive.

B

That leaves the other issue which puts Ultimo's appeal on the sunny side of possible frivolousness — the proposition that the trial court should have simply sent the case to arbitration while allowing the Harper's non-arbitral (i.e., tort) claims to remain in superior court.

Armendariz dealt directly with the question of the severability of unconscionable provisions. ( Armendariz, supra, 24 Cal.4th at pp. 121-127.) The standard of review on this precise point is a sort of heightened-scrutiny abuse of discretion: As the Armendariz court explained, the unconscionability statute (Civ. Code, § 1670.5, subd. (a)) "appears to give a trial court some discretion as to whether to sever or restrict the unconscionable provision or whether to refuse to enforce the entire agreement," but with the proviso that refusing to enforce the entire agreement is an option "only when an agreement is 'permeated' by unconscionability." ( Armendariz, supra, 24 Cal.4th at p. 122, emphasis added.) Armendariz further explained that the reasons to merely sever or restrict derive from the need to avoid parties from gaining undeserved benefits or suffering undeserved detriments. ( Id. at p. 123.)

(5) We conclude that the trial court's refusal to sever arbitral from nonarbitral claims was easily within its reasonable discretion, even subjecting it to heightened scrutiny. The trial judge did not throw out the contract. The overall contract between the Harpers and Ultimo may still be enforced, and — who knows? — the Harpers may end up owing Ultimo money after a trial in the superior court in this case. The severability issue comes to us here not in terms of the dichotomy framed by Armendariz — throwing out the "entire agreement" versus simply not enforcing a single unconscionable provision — but in terms of an attempt to split substantive claims, with some going to arbitration and some going to superior court.

But if one thinks about it, the idea that the trial court could peel off the Harpers' nonarbitral claims and allow those to be tried in superior court while the remainder were arbitrated makes no sense at all. The Harpers' misrepresentation, property damage, and contract claims are all bound up with each other. The arbitrator will be unable to avoid dealing with those issues, even if, theoretically, he is merely limited to giving the Harpers' their money back plus $2,500 while the tort and punitive damage claims are litigated in superior court.

Under such a scenario there is a risk of inconsistent adjudications. The arbitrator might find, for example, that Ultimo was truthful, had not misrepresented the amount of work he had performed, had not damaged any property, and announce a very low award (maybe Ultimo didn't do a good job replacing patches of the back yard he dug up), or no award at all. What will a jury do with that, particularly if they are inclined to find that Ultimo did misrepresent the amount of work and concealed significant property damage (like rendering the sewer system inoperable, not just failing to replace some divots) he caused? And would the arbitrator's findings on compensatory damages be merely advisory in the context of the jury's consideration of a claim for punitive damages? We could have the anomaly that the arbitrator might award little or no contract damages, find that Ultimo made no misrepresentations and fulfilled his contract, and the jury might find he did commit fraud, and try to assess punitive damages based on (perhaps hypothetical, perhaps not) compensatory damages inconsistent with the arbitrator's award of compensatory. In short, it would be a mess.

Ultimo suggests that the trial court acted unreasonably in not considering the alternative of sending the case to arbitration, including all claims made by the Harpers, sans the offending confines of the Better Business Bureau rules. Again, however, the trial court acted quite reasonably. This alternative would substantively rewrite the arbitration clause, as distinct from simply not enforcing it. At the time of the formation of the contract, Ultimo might never have agreed to an arbitration clause if he knew that it allowed tort and punitive damage claims to be arbitrated against him. When he signed the contract, he might have preferred, as a likely defendant, to have the benefit of the relatively slower machinery available in the publicly funded courts. That he may want arbitration now, with concrete claims against him, merely bespeaks an attempt to pick and choose forums after the fact.

IV

The order denying the petition to compel arbitration is affirmed. Respondents shall recover their costs on appeal.

Rylaarsdam, J., and Ikola, J., concurred.