TGJB,

The current model assumes a typically fast pace in the Derby and thus discounts the chances of front-runners. If you somehow deduced that this year's pace would not hurt the front-runners and ran the model under that assumption (i.e., assume a neutral pace), you would get significantly different results (see below).

Not surprisingly, War Emblem and Medaglio D'Oro would have moved up quite a bit and War Emblem would have been flagged as the biggest betting "edge" according to the model.

Perfect Drift still would have been predicted to be the most likely winner and a good value.

Proud Citizen still would not have been given much of a chance but keep in mind that the model does not have a "Lukas in Kentucky" factor.

Chris

P.S. Thanks for being curious.

*****

The Model's top 10 in order of estimated likelihood of winning

Actual Finish, Horse, Post-time odds, "break even" odds, and Betting "Edge"

3, Perfect Drift, 7.9, 3.0, 127%
4 - troubled trip, Medaglio D'oro, 6.9, 6.2, 10%
1, War Emblem, 20.5, 6.7, 179%
7, Harlan's Holiday, 6.0, 9.8, -35%
10 - injured, Saarland, 6.9, 11.0, -34%
5, Request For Parole, 29.8, 11.2, 152%
6, Came Home, 8.2, 17.8, -51%
11, Blue Burner, 24.2, 34.5, -29%
2, Proud Citizen, 23.3, 35.8, -34%
9, Essence of Dubai, 10.0, 37.3, -71%