Extended Data Figure 4: Comparison between measured and ‘real’ (simulated) parameters.
From: No large population of unbound or wide-orbit Jupiter-mass planets
a, Ratio between the measured Einstein timescale tE,out and ‘real’ (simulated) timescale tE,in for simulated events versus the blending parameter fs = Fs/(Fs + Fb). Timescales of faint and highly blended (fs < 0.1) events are not well measured and are biased by a strong degeneration between Einstein timescale, blending and impact parameters. Timescales of events showing a high negative blending (fs > 1.5) are systematically underestimated, but the bias is relatively small and such events comprise a negligible fraction of all events. b, Distributions of tE,out/tE,in for simulated events passing selection criteria from Extended Data Table 3 (including the cut on the blending parameter fs > 0.1). Regardless of the timescale, there is no systematic bias between measured and real timescales within 1%. For 90% of simulated events 0.63 < tE,out/tE,in < 1.65. The MAD is the median absolute deviation from the data’s median.
