cozzene,

It would depend on these things:

1. How likely I thought the fastest horse was to bounce and by how much. I would determine that on a horse by horse basis (not a general rulebook) based on his overall record and typical range of figures, his age, time to recover since last race, if there were any special conditions that might have contributed to that big figure that aren't present today (loose lead or other easy trip, off track, track bias last out), whether he was a young generally improving lightly raced horse or one that was older, had peaked, and was on his way out etc....

2. If any of the trainer/equipment changes were clearly huge statistical positives like a switch to a so called super trainer I would consider it a much bigger threat without being able to quantify it very well.

If I were very confident the fastest horse was going to bounce big and the trainer change was no big deal I might actually bet both A and B horses to win and box the exacta. I have no problem betting multiple horses to win. I did it several times BC day.

On the flip side, if I was having trouble quantifying the probability or degree of bounce and/or the switch was to a super trainer etc... I would admit to myself I have no clue what the hell is going on. So I would just pass the race.

Quantifying the "value" of trainer moves is my weakest skill. So I tend to not try to make a profit by betting on them (unless I like the horse and get the new trainer for free on the board), but simply avoid betting against them.

I try to maximize my ROIC. So I tend to avoid situations where I can't quantify the probabilities very well - which is lot of things. :-)

My skillset may be "WAY" different than yours. It took me many years of keeping records of my bets (and losing a little and then breaking even) before even knew what I knew. Now I win steady.