we BUILD DECISION TREES to VALUE LAWSUITS and facilitate Settlements
There’s a settlement offer on the table. Should your client accept or continue litigating? This pivotal question can arise at any stage of a lawsuit. To properly address it, litigators need a reliable method to determine (within a range of reasonableness) what a case is worth, and compare that value to the offer on the table.
Putting a dollar value on a lawsuit, however, is no easy task. It’s not like buying a car or a house. Every lawsuit is unique in terms of the facts, legal issues, evidence, witnesses, judge and jury pool. Yet, valuing lawsuits is a task that litigators must regularly and competently perform if they are to negotiate effectively with opposing counsel whether through a mediator or face-to-face.
Fortunately, there is a well-accepted quantitative technique for valuing lawsuits known as decision tree analysis that calculates the probabilities of different legal issues in the case resolving in a certain way (e.g., negligent vs. not negligent), and then combines those probabilities and applies them to damage estimates to derive an expected value for an entire case. Decision tree analysis has been used for decades by businesses to make investing decisions. When utilized correctly, decision tree analysis can help litigators derive a reasonably accurate case value to help inform settlement negotiations, whether face-to-face or in mediation.
The key issue that decision tree analysis is designed to address is the vague qualitative assessments that litigators frequently offer to clients when asked to evaluate the fairness of a settlement offer.
For example, consider a recent wrongful death lawsuit in Texas involving a bicyclist who hit a truck parked on the side of the road. Am. Guarantee & Liab. Ins. Co. v. ACE Am. Ins. Co., 413 F. Supp. 3d 583, 595-97 (S.D. Tex. 2019), aff’d, 990 F.3d 842 (5th Cir. 2021).
On the issue of the driver's negligence, the evidence showed that the truck driver was parked legally, which indicates the driver was not negligent. On the other hand, the truck driver testified at his deposition that even though legal, the position in which the truck was parked posed a risk. That testimony would support a finding of negligence.
There was also the issue of the bicyclist's comparative fault. Texas law bars recovery if the comparative fault of the plaintiff is greater than 50%. Here, the evidence (in the form of cracks on the helmet) showed that the bicyclist was riding with his head down and not paying attention to the road when he crashed into the truck. Such evidence seemingly created a strong likelihood that even if a jury found the defendant was negligent, the comparative fault of the bicyclist exceeded 50%. And even if the jury found that the bicyclist's comparative fault was less than 50%, the evidence suggested the percentage comparative fault would likely still be high, most likely in the range of 40%.
Finally, there was the range of possible damages, which appeared to be between $6-$8 million.
Using those facts, let's create a hypothetical. The plaintiff makes an offer to settle for $2 million at the close of discovery. If asked by the insurance company defending the case to evaluate the offer, the attorney handling the litigation may respond with something like, "given the evidence, there's a good chance that the jury will find that the truck driver was not negligent, and even if the jury finds the truck driver was negligent, the evidence suggests a reasonable possibility that the jury will find that the bicyclist's comparative fault exceeded 50%, and even if the jury finds that the bicyclist's comparative fault did not exceed 50%, there's a high likelihood that the jury will still find that the bicyclist's comparative fault was substantial (in the range of 30-50%). The range of damages of $6-8 million, but given the evidence, I like our odds of winning at trial either because the jury finds the defendant was not negligent or the comparative fault of the bicyclist exceeded 50%."
As the bolded words above show, the problem with this evaluation is that qualitative phrases like "good chance," "high likelihood" and "reasonable possibility" don't mean the same thing to the same people. For example, if a "good chance" means 40% to the attorney, but 60% to the client, then the client perceives a greater likelihood of success than the attorney intended to convey. Additionally, even were the client and attorney to agree on the odds that a "good chance" represents, how does one combine all the phrases ("good chance," "high likelihood" and "reasonable possibility") to derive a value for the entire case?
What decision tree analysis does is convert qualitative assessments into quantitative probabilities that can be applied to each legal question that will be posed to the jury and then combined and applied to different damages estimates. Consider the following revised version of the attorney's response:
"Given the evidence, we think there's a 40% chance that the jury will find that the truck driver was not negligent, and even if the jury finds the truck driver was negligent, the evidence suggests a 65% chance that the jury will find that the bicyclist's comparative fault exceeded 50%, and even if the jury finds that the bicyclist's comparative fault did not exceed 50%, there's a 70% chance that the jury will still find that the bicyclist's comparative fault was substantial (in the range of 30-50%). Given that the range of damages is $6-8 million, the case is only worth about $1 million, and so we think $2 million is too high to pay."
These percentages can be used to construct a decision tree that would allow the attorney to illustrate to the client the basis for his or her recommendation that $2 million is not an acceptable offer (because the case is only worth $1 million), as well as to adjust the probabilities (and the case value) should new developments emerge as the case progresses:
The percentages in the decision tree above are not taken out of thin air, but rather are based on considered evaluation of the evidence in the case and the applicable law. For example, the 40% chance of the jury finding that the defendant was not negligent would be based on the evidence that the truck was parked legally. But the percentage is not higher because the drive testified at deposition that he parked the truck in a manner that still posed a risk. If the case is being tried in a county where jurors tend to favor injured plaintiffs, an attorney may conclude that the evidence cutting both ways is more likely than not to result in a jury finding of negligence, and hence a 60/40 split is a realistic assessment.
But that assessment could change, for example, based on the judge's rulings on various motions in limine in advance of trial that either increase or decrease the likelihood of a negligence finding (which would require the attorney to update the tree and derive a new expected value for purposes of evaluating any new settlement offers).
Read Our White Paper
There is much more that can be said about decision tree analysis. To learn more, download our white paper, "Using Decision Tree Analysis To Value Lawsuits and Negotiate Settlements." (to be published post-webinar).