Neither Side Sees The Other List First
Nominations and strikes are blind until both sides have submitted. The second submitter cannot tailor names after seeing the first list.
Published Procedure
Umpire Select is used when two appraisers cannot agree on an umpire. Each side names 10 umpires, each side may remove up to 5 names from the other list, and the platform picks one remaining umpire at random. The pick is coded so a third party can check that nobody steered the outcome.
This is the sequence both appraisers follow on a claim. Each appraiser consents when submitting their ten umpires. The random draw runs after both strike lists are in.
01
Each appraiser submits exactly 10 names from their own list, or adds someone who is not on that list. Submitting that list is the consent step: each appraiser types their legal name and agrees to be bound by the random draw. The other appraiser cannot see those names until both lists are in.
02
Each appraiser may strike up to 5 umpires from the other list only. You cannot strike your own nominations. You do not have to use every strike. Strikes stay hidden until both sides submit. Using fewer strikes simply leaves a larger pool. The draw runs automatically when the second strike list is recorded.
03
Everyone who was nominated and not struck remains. Each of those umpires has the same chance. Nobody adds names, reorders names, or reruns the list after this point.
04
The platform converts the locked pool and the hidden random values into a single number: an index into that ordered list. The umpire at that position is the selection. Every remaining umpire has an exactly equal chance.
05
The result is written into a downloadable record. That file lists the nominations, the strikes, the surviving pool, the selected umpire, and the coded values needed to recompute the same index.
After strikes, the surviving pool is every nominated umpire who was not removed. If both sides nominate 10 and both use all 5 strikes, at least 10 umpires remain. If either side uses fewer strikes, the pool is larger. The draw then assigns each remaining name the same probability. It does not weigh experience, location, or who nominated the person.
Repeating the process on a new claim produces a new result even if the names and strikes look the same, because each selection generates new random values. Repeating the check on the same record always reproduces the same winner.
Nominations and strikes are blind until both sides have submitted. The second submitter cannot tailor names after seeing the first list.
When the request is accepted, the server creates a secret random value and publishes only a fingerprint of it (a SHA-256 commitment). After the draw, the secret is revealed. If the secret had been swapped, the fingerprint would not match.
When each party submits their ten umpires, the server creates a random value for that party. The appraisers do not type it, and they cannot change it later. That stops either side from shopping for a seed that favors a name.
The same inputs always produce the same winner. A court, carrier, or opposing appraiser can recompute the fingerprint, the combined seed, and the index from the record and confirm they match.
The selection does not use Math.random(), a coin flip in the browser, or any value a client can edit after the fact. It uses SHA-256 and rejection sampling so every surviving umpire has an equal chance.
The record uses this method. You do not need to run it yourself to use the product. It is here so a lawyer, carrier, or independent reviewer can see exactly how the winner was chosen.
After a draw, either appraiser can download the record from the completed claim. That PDF includes the ordered surviving pool, the selected umpire, the commitment, both client seeds, the revealed server seed, the final seed, and the selected index.
A third party can recompute those hashes and the index from the values on the record. If they match, the recorded winner is the winner the procedure required. If they do not match, the record is not authentic.
Umpire Select does not vet umpires, check conflicts, or decide whether a court will accept the selection. The appraisers remain responsible for confirming qualifications, disinterest, availability, and conflicts of interest. The platform is a selection tool. It is not itself a party to the appraisal.