Wow! So, there must be a lot of computerized wager placement on Betfair.

My odds line is not a predictor of tote odds. I try to make it a predictor of likelihood to win. Since it is purely mechanical, I still have to handicap the race, but at least I know what my favorite factors are telling me. I like what you're saying about real possibility of failure. I wonder if that can meaningfully be expressed as a number.

Good point.

I have started comparing my odds line (after scratches) to the morning line (adjusted to a base of 100 after scratches). This tells me which horses are not being evaluated the same by me and the morning line maker. So, a horse that is way off may be an exceptionally good bet or an exceptionally bad one. At least, that's what I am looking at.

Getting back to the original topic in this thread, every horse has some chance of winning. So, our task is to be reasonable in both directions... to select horses in each portion of the odds continuum in various races. We know that favorites win about 35-41% these days. But if you're selecting favorites more frequently or less frequently over time, then you're probably not making optimal selections. If we NEVER give any real probability to that 19/1 horse mentioned previously, then we're likely below optimum. Any thoughts?

I don't understand. Is that for horse racing? Update what? Odds? Parimutuel odds?

One thing they talk about is speed of information. Just think about that in terms of horse racing. We hear a lot about seeing the posted odds dropping after the gates open, but have you considered how much variance there is in internet speeds and the speed of the data source? If you are trying to bet based on the odds, how much difference does it make if your odds data is 1 second slower? 3 seconds? What if you are watching Finger Lakes on YouTube, where they tell you that on THEIR end there is a 10 second delay?

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.

If you hide your Easter eggs, you are cheating. lol!

Thanks, interesting answer. I follow my top 4 contenders and find that I have a lot of P5 sequences in my top 4 contenders, but there are usually some races with my 4th choice in there. It's mighty tough to single many races. I remember singling a P6 at FG many years ago. I missed the last leg. The consolation paid $12.30. lol Too many favorites.

No anecdotal evidence suggests it works very well except when there are legit heavy favorites.

I have several patterns of when the pace falls apart.

1) Nearly all of the horses have low %Median as their typical pattern.
2) Giles Speed Points is zero.
3) Giles Speed Points is very high, like above 6.

Now, I have taken the median %Median of each horse and then the median of all horses. The likely winners are usually near that race median, even if they show few other positive factors.

I continue to think that %Median is a very important figure. I see a lot of longshots that figure to like races that match their best energy... some of them don't seem to have much else going for them.

Most of my different ideas get poo poo'ed. lol

I don't get this at all. If Monty Hall always works, why aren't you making millions?

I have been considering different perspectives a lot.

That's not just for the Surface Pro. I can do that in Edge on my Dell laptop.

Yep, they don't know what to look for.

Nobody has it because sometimes it helps to think like a trainer.

There's also the horses that match what you know works and nobody has it.

Wagering is likely to consolidate around a few large tracks, much like Hong Kong.