In this article Brian W Jones, President Merrick Bank, discusses how FRCP 68 could be used during litigation. Quick research with Open AI suggested there are theoretically around 100-150k cases per year where a Rule 68 offer could be made, whittled down to 50-80k per year where they would be ‘strategically meaningful’.
SettleIndex is programmed to model the effect of cost shifting offers, so if you have one (or a few thousand) relevant cases, we would love to hear from you!
No doubt our US readers will put me right, but I understand the sort of cases where the Defendant could use it to put pressure on Claimants to settle are:
Civil rights liitgation
Employment disputes
Personal Injury cases in federal court
Fee shifting statute cases
Cases where the parties have by contract given a disputes tribunal power to award attorney fees
The article is actually a much wider clarion call for the use of predictive analytics in litigation. In his article Brian recognises ‘optimism bias’, which we have written about ourselves as ‘partisan role bias’.
The article also canvasses how predictions could be made using the factual inputs of a pending claim and publicly available data about similar cases. This drives into focus the difference between trying to predict the likelihood of winning and trying to calculate the rational figure for settlement. The former can influence the latter, but to derive the optimum settlement figure you need to do the right probability calculations from two perspectives, including Award Ranges.
As things stand we believe the big data approach is not actually needed. Probabilistic calculations based on the extreme cases represented in the claim and defence documents provide the basic model, which the parties and their lawyers can then tune with their experience.




