Is Brethren five lengths slower than Silver Medallion?

The Beyer figures are out for yesterday's three-year-old stakes races (Sam F. Davis, Robert Lewis, and El Camino Real Derby) and, for some of us, they conflict greatly with what took place from a visual perspective.  Below are the figures assigned for each winner (via DRF.com):

Silver Medallion - 93 (El Camino Real Derby)
Anthony's Cross - 90 (Robert B. Lewis Stakes)
Brethren - 83 (Sam F. Davis Stakes)

According to those speed figures, (and utilizing the beaten lengths chart in Andrew Beyer's Picking Winners) Brethren was five to six lengths slower than Silver Medallion, and about four to five lengths slower than Anthony's Cross.    

One issue with the Tampa number (and it's an issue that seems to come up a lot), is the fact that Tampa only ran one race at two-turns yesterday.  Based on things I've read in the past, the figure makers rarely use a split variant for one-turn and two-turn, although they will utilize split variants (see the big stakes days at last year's Saratgoa meet).  Furthermore, it appears that a split variant was used yesterday but how they decided what number to use for the Davis when there was only one race greater than a mile on the card is difficult to ascertain.  It's tough to get a read on the speed or slowness of a track when all the main track races are run at the same distance. 

Below is a snap shot of yesterday's main track races at Tampa, and their corresponding BSF from the Formulator interactive charts:

Many times a projection method is used when determining a speed figure to ensure the rest of the figures in a specific race (from 2^nd place to last place) "fit in" with the overall picture.  For example, if Brethren was assigned a 93 for yesterday's race, then every horse that finished behind him would have a larger figure; Too Experience would be an 86 (his highest fig. by 12), Watch Me Go an 84 (highest by 8 5), etc.  Since it's unlikely that every horse ran a career best, figures are adjusted to make sense.  It's part of the reason that figure making is as much art as is it science.  Based on that, I'm guessing that the number for Brethren was predicated somewhat on the fact that a higher figure for the winner would result in career highs for several of the colts running behind him. 

Here's the prime issue I have with Brethren's figure: it suggests that not only did he fail to improve in any meaningful sense from his last race (84 BSF -> 83), but he also ran slower.  That's the hardest hurdle to overcome, in my opinion - the suggestion that this colt didn't improve from his last race.

Here's another nugget from yesterday's races: Tapizar earned an 80 for his 5^th place finish in the Lewis and Comma to the Top earned an 86 for his 4^th place finish.  Furthermore, if you look at the figures of all horses in the three fields, the numbers indicate that Silver Medallion, Jakesam, Positive Response, Anthony's Cross, Riveting Reason, and Quail Hill ran faster than Brethren.  That's a pretty clear indication that the figure makers believe the Sam F. Davis was a much worse race, from top to bottom, than either of the races on the West Coast.  We'll find out in the future whether that's true or not.

The whole point of the speed figure is to have an "objective" number to validate (or invalidate) what we see with our own eyes.  Ask anyone that watched those three races yesterday which horse impressed the most, and I'd venture that almost all would say "Brethren".  But if you look at the figures, he was easily the least impressive of the three.  It's up to each handicapper to decide which of those two things they want to value more: their eyes or the numbers.

None of this is to suggest that a figure is "wrong" - the number is what the number is.  Instead, it's a reminder that the figure only captures a specific moment in time and it's heavily dependent on what other horses run in the other races on the day (and in the same race).  Like all data that a horseplayer has to consider, it's just one piece of the puzzle.