Merge bias is the extra schedule risk created when several parallel paths converge on one activity or milestone. The milestone cannot start until the latest path finishes, so the chance of meeting its date is lower than the chance of any single path finishing on time. Critical path scheduling ignores this; Monte Carlo simulation shows it.
Merge bias. Each path is at its median, yet the milestone has only a 12.5% chance of starting on time.
Worked example
Three independent paths feed mechanical completion: equipment delivery, piping installation and electrical installation. Each has a 50% chance of finishing by its planned date.
P(all three on time) = 0.5 × 0.5 × 0.5 = 0.125
The milestone has only a 12.5% chance of starting on time, even though every path is at its median. With five such paths the chance falls to about 3%. In practice paths are rarely fully independent, so the effect is smaller, but it is real and it grows with every parallel path.
Why critical path scheduling misses it
The critical path method uses single durations, so only one path is critical and the others show float. In reality, near-critical paths with little float regularly overtake the critical path. A QSRA reports this through the criticality index: the percentage of iterations in which each activity was on the critical path.
Where merge bias hits hardest
Mechanical completion and commissioning, where many systems must all be ready.
Brownfield shutdowns, where every tie-in package must be ready before the window opens.
Multi-contractor programmes in the GCC and UK, where several packages feed one interface milestone.
How to reduce it
Decouple paths so a late package does not hold up the others, for example with partial handover.
Buffer the merge point with explicit schedule contingency ahead of the milestone.
Protect near-critical paths identified by the criticality index, not only the deterministic critical path.
Common mistakes
Managing only the critical path shown in Primavera P6.
Reading float as safety on paths that converge.
Overstating it: ignoring correlation between paths exaggerates the effect.
Merge bias is the added delay risk where parallel paths converge: the merge point waits for the latest path, so it is less likely to be on time than any single path.
How do you calculate merge bias?
For independent paths, multiply each path's probability of finishing on time. Three paths at 50% each give a 12.5% chance that the merge point starts on time.
How does QSRA show merge bias?
Monte Carlo simulation runs every path in each iteration, so the merge point takes the latest path each time. The criticality index shows which near-critical paths drive it.
Need a defensible P80 on a live project?
IQRM runs independent QSRA and QCRA for owners, EPC contractors and PMCs across the UK and GCC, and turns the results into decisions a board can act on.
IQRM is a specialist institute for Quantitative Risk Management, delivering
Quantitative Schedule Risk Analysis (QSRA),
Quantitative Cost Risk Analysis (QCRA) and
Monte Carlo simulation training and consulting to major capital projects
across the United Kingdom, Saudi Arabia and the UAE.
Approved Saudi Aramco vendorADNOC approved consultant15+ years applied practice
