@RanchWest
I wouldn't have thought to search that category.lol
I liked their perspective concerning how luck "co exists" with the probabilities and becomes more of a factor as the "game" gets more efficient. I'm of the view that we've probably reached very close to the limit of our efficiency in racing. So we need to find ways to enable "luck" to assist.

@RanchWest

I am watching a YouTube video called "Luck vs. Skill: Gambling Analytics". It's very interesting and talks, in part, about how as market efficiencies increase, the luck level increases. Something to think about. It is not about horse racing, but I see parallels. Time: 1:00:46.

This was an interesting and fascinating watch.

@RanchWest
2 things. The first regarding the "morning line" This line is not a "probability line " such as your own. It is a "guess where the money will go" line. To achieve this the line maker must,of course, do some handicapping but they must be careful not to guess too low. They get enough flak without making a legitimate 2/5 shot 2/5 , instead they'll make it 6/5. And they won't make a 100/1 shot 100/1 , instead they'll make it 40 or 50/1. To manage the whole field they'll use a "base" of 110% to 130% depending on the difficulty of the field.So comparing "your" line to "theirs " is "apples and oranges".
As to " real probability of failure " being expressed as a number, I think "maybe" the "global" win/loss percentages might prove useful in that regard. After all, when we strongly believe that a highly valued contender (favorite) is unlikely to win , the "winning" probabilities of our legit contenders must necessarily increase. I'll give it some thought to see if I can find a balanced equation that works.
.

@RanchWest
While thats not a "bad" way to start,I prefer to ignore the "morning line" altogether and I don't make a "true" odds line either. I'm trying to see how the money is likely to flow but not in what proportions. Then, if the real possibility of failure (for the top 2) exists , I can choose another option in the knowledge that I'll get "random odds +" on that option. (Your 19/1 shot only had "random odds" of 4/1 but because of the presence of that 3/5 "weak" favorite the odds were much higher.

@RanchWest
I think you're right about giving fair consideration to both favorites and nonfavorites according to their relative chances of winning. This is why we need to be able to, first, identify the likely top 1 or 2 in the betting and ,second, to identify situations where 1 or both are weak or vulnerable (or simply not deserving) thus raising the relative chances of "non favorites".

@Slim
There was a brief discussion of "Monty Hall" in part 2 of the Thread "What should an odds line be?'

@Steven
"Using Monty Hall to handicap" is listed in Dave's video. Under "Category list" choose "Dave's favorites" . It's about the 3rd video down. ( Be sure to carefully consider the logic and math involved)

@RanchWest
Perhaps that "odor" is unconscious "processor" sending a big "CAUTION" message. "PASS" would seem the prudent course..for me.

@RanchWest
Have you considered using "multi factor pattern matching" to test your idea that %median is a strong factor in finding winners in these "pace collapse" or "paceless" races?

@RanchWest
I'm always pleased to see your interesting observations.

@RanchWest
You're not alone .. LoL

I had hoped that this forum would be a place where players could exchange "different" ideas, perspectives and opinions.
It can't hurt to consider different ideas and there's always the possibility,however small, that something new will help.

The "process" is certainly not unique but it's objective is find selections which are "unique in that field". A white horse is not unique (there are plenty of white horses in the world) however ,a white horse in a field of horses which are not white is unique in that field. This is,of course, not a relevant handicapping factor ( unless we know that white horses run faster than nonwhite ones and a large portion of players can't see that it's white )

( the "order" of placing in the "set" is irrelevant, thus each set can present itself in 6 different "orders")

There's nothing "Wiz bang" or "super secret",
about the approach. It's merely a process of viewing the data from a different perspective . To use your example; imagine that my objective is to isolate a "unique" 3 horse "set" from your group using "best of last 3 speed ratings. I might set the parameters to be " best of last 3 is at or above par" . To find a usable set I must find 1 and only 1 set in which ALL 3 horses fit within that parameter, making it a unique set . I might not find such a set or I might find 4 or 10 (in a very competitive event) .In either case there is no play at all , other than using different parameters. I simply apply this method ,using multiple factors , to isolate a single horse which possesses a unique "set" of factors that fit the parameters I've set for that race.

LoL! (56) . Of course ..

Would this process of choosing which contenders to bet (or "not bet") be considered "unique"?

Of course ,I'm aware that the participation rate will necessarily fluctuate from pool to pool, which is why I suggested "average" participation. Am I correct in assuming that the ROI of +7.5% to 9% is factoring in the rebates?

Thanks for the clarification!

Tony;
Thanks for your comments! I'm hoping that everyone can see the benefits of ,at least, "considering" different perspectives!