Note: The above horse racing examples are fictional. Seems would need a lot of personal system programming to track trends like this. Just curious. Not sure if it would work in horse racing. — Jim Pommier

Great idea for discussion.

It has been a main part of my handicapping for over 20 years, but probably not quite what you'd expect - and possibly not what you mean.

Most people see "trends" as a synonym for "patterns." There's a lot of jumping at shadows with that - seeing patterns when there really isn't one.
___________________
Howard Sartin once told me something really profound:

• "Most changes in the BIAS are perceived rather than real."

That one was HUGE!
___________________
We were discussing keeping Winner models.
Specifically, the current model at SA, 6-furlongs, dirt.

One of his clients, Milt, a Reno dentist, had loudly proclaimed that the bias had clearly changed because 3 winners in a row had run from off the pace, when the bias had been early for a couple of weeks.

I explained how I kept 4 models at each track-surface-distance, based upon how many 7/8 point horses there were.:

• 3 or more (type 3+)
• 2 (type 2)
• 1 (type 1)
• 0 (type 0)

Those particular 3 races had all been types 0 or 1.

Sure, there is at least one front runner in every race - after all, somebody's got to be on the lead. But those wanna be closers who often find themselves on the lead are why you see turf miles in AUS go 52+ seconds to 4 furlongs: NOBODY REALLY WANTS THE LEAD!
(Aren't you glad you asked this question? LOL)
__________________________________________
__________________________________________
THE POINT
Trends are huge for me - but usually they really are PATTERNS that are attributable to SPECIFIC SCENARIOS that repeat themselves with great regularity.
__________________________________________
__________________________________________

Back to BASKETS.
Do you recall the angle called Win & Travel?
That's where a team wins at home and plays the next night on the road.

For YEARS that was a dead-bang winner back in the 80s as a bet against play.
(I actually wrote NBA & Hockey software for a private group back in the late 80s.)

It continued to work well into the 90s, even though it was well-publicized. Then, all of a sudden, some big players started to bet the angle huge and the line moved accordingly. Overnight it died. Maybe like 1993 or so.)

Another was how a team beats the spread for the first two games after a big star is injured or has limited playing time coming up. The belief was that it was caused by the other players get really pumped about having a game or 2 without playing in the shadow of the super star.

Typically the advantage went away after 2 games.
____________________________
Anybody else have thoughts?

Patterns to specific scenarios. I think that it's something that's been overlooked.
Yes, I do recall home-road dead-bang winner (or something to that affect). The one I remember most was a team with back-to-back road games with the first game in Denver. Because of the higher altitude in Denver, the players' shots would be off in the next road game. Take the home team. Today, very few 2nd road games after Denver are within 1-2 days.

It would be a very cool thing to have done by someone but I believe Jim Mazur already does that for meet-specific handicapper books that focus on the best 100+ trainers during the Last 3 or L6 meets depending on the meet

There are 5 different products that he sells on Saratoga alone for stats
1) Saratoga Handicapper- Covers L3 meets on Post Positions/Running Styles 40% Club (trainers w/ stat that is 40% winners+) Profiles on the top 20 trainers and 80 mini profiles on the rest of the trainers plus more

2) Blue Chip Trainer Plays- These are angles specific to the Saratoga meet that trainers have shown but there are specific criteria to follow:
The trainer must have maintained a 12% winning average at the last three meets.
• The angle had to score at least 3 times with 3 different horses.
• The angle had to score at least once at the previous meet.
• The angle had to produce a 33% win average and a 100% ROI.

3) and 4) Saratoga Handicapper Stats L3 Years and PLUS Handicapper L6 years- Breaks down 100 trainers who were shown in the Handicapper book that covers all the stats you need to know about that trainer and how that trainer does in a specific category SPECIFIC TO SARATOGA NOT OTHER TRACKS! For example, did you know that Jorge Abreu during the L3 years at Saratoga with horses off 90+ days has hit a 12% clip (2 for 17) with an Itm rate of 29% BUT during the L6 years he is actually hitting at 17% (5 for 30) 37% Itm? You wouldn't know if you didn't have both books on hand!

5) Saratoga Handicapper Trainer Detail- Lists the 100 trainers (I believe) from the books above with each trainer's horses that ended up winning with their specific info (LR raced days off surface joc, etc)

It would still be a very cool idea for someone to do but if you wanted to take the guesswork out of the stats I would highly recommend Jim Mazur's products!

He covers Saratoga/Churchill Downs (Spring)/Del mar/Oaklawn/Keeneland Spring-Fall/Gulfstream Park/Belmont Spring-Fall as well as info for the Triple Crown/Breeders Cup and his Pedigree information!

I agree about bias. Much is explainable. Much of the rest is coincidence. One notable exception was the old Aqueduct Inner Dirt... inside was for real.

For trainers, fertile for looks is how a trainer performs early meet, mid-meet, late meet. And, where the horse came in from because that might be a change of assistant trainers.

There are a lot of possible trainer and jockey studies, but sometimes you have to think outside the box.

Patterns to specific scenarios. I think that it's something that's been overlooked. — Jim Pommier

There's a fellow I know who was very statistical when it comes to his sports betting.
I wrote him some software, with an early AI engine (for me; 1988). They used it for NBA & NHL, the two sports that supported them.

He asked me if I had any unique ideas about it - especially on pattern of trends & tendencies that were not currently in use.

I pushed him towards "Pace Handicapping."
Specifically, that was to see the quarters of the games as calls in a horse race, and build the factors we pace handicappers know and love.

NBA Fractional Pace Factors

The Calls

• F1: Q1
• EP: (Q1+Q2) / 2
• SC (Q1+Q2+Q3) / 3
• FT: (Q1+Q2+Q3+Q4) / 4
• SR: (Q3 + Q4) ) / 2

The Specialties

• SP (EP + SR * 2) /3
• FW (EP * 2 + SR) /3
• FX (Q1 + Q4) / 2
• LP (Q1 + Q2 + Q3) / 3
• HF (Q2 + Q3) / 2

The Fractions
The individual Quarters

• F2 (Q2)
• F3 (Q3)
• F4 (Q4)

The Energies
Percent of All Points in the game

• F1%
• EP%
• SC%
• SR%
• SP%
• FW%
• FX%
• LP%
• HF%

A complete set for Overtimes

• OT1
• OT2
• OT3
• OT12
• OT123
• OT23

A few notes:
(Each factor contains...)

• Offense Points
• Defense Points
• Total Points
• Possessions
• Turnovers
• Fouls for
• Fouls Against
• In Bonus (Off)
• In Bonus (Def)

Wow. Imagine 1988 and you had 2022 sports betting on your iPhone with live in-game wagering. Making wagers during the game with this software would have been fun.

Funny thing... When they got the software, they said, "Now we have too much information!! What can you do TO HELP US?"

What I had them do was:

• Categorize each team on an ABCDE grading system for Offense & Defense.
• Then I built IVs for Win% vs Spread in all categories.

Imagine the Lakers were playing the Pistons in Detroit.
They'd query the system

• What is the IV for Lakers vs a team Like the Pistons on the road?
• What is the IV for Pistons vs a team Like the Lakers at home?

Considering offenses and defenses, there would be 4 IVs - 2 for defense & 2 for offense.
(Same for totals)

When both sides agreed... I think they were looking for 1.50 vs 0.67, (i.e. Lakers and pistons were opposites; one side says we lose and the other says we win) they had a play.

I recall the head guy saying to me, "Are you telling me we DON'T CARE WHAT THE SPREAD IS? How is that even possible?"

:100:

They kicked ass!
63% with 4 plays a day until the All-Star break. Then it flattened, which I warned them it would do.
NOBODY sustains 63% and 4x per day.

They finished the year at 58%, which is pretty remarkable.
(They never bet the playoffs, BTW.)

Wow. Imagine 1988 and you had 2022 sports betting on your iPhone with live in-game wagering. Making wagers during the game with this software would have been fun. — Jim Pommier

I've considered rebuilding it, but the data capture is an issue.

In 1982 I wrote a horse racing fractional times program with manual input and sold a few copies. A friend kept asking me what to do with it. My answer way back then… I don’t know.

Funny that I mentioned the old pattern-- "...... back-to-back road games with the first game in Denver. Because of the higher altitude in Denver, the players' shots would be off in the next road game. Take the home team." The Lakers played in Denver on Wednesday and are in Minneapolis tonight. The problem is the Lakers have been shooting horribly in their first 4 games. Not sure they can shoot any worse, altitude or no altitude. The line this morning was T-wolves -8.5, now it's -7.5. Not sure if the T-wolves want to put a hurt on LeBron, but for craps and giggles, laid the points-- T-wolves -7.5, $11 to win $21.19.

I've considered rebuilding it, but the data capture is an issue. — Dave Schwartz

A friend sent me 10 years of NBA data in exactly what I wanted.

Alas, not high on my priorities, but at least I can think about it.

A friend sent me 10 years of NBA data in exactly what I wanted. — Dave Schwartz

Does that friend also want to share it with a guy who may consider building a better model for NBA? :lol:

I'll ask him.