BRIS trainer figs as win seperator??

[andicap]

Has anyone combined their own methods and the BRIS trainer stats.

That is, get your top contenders (however many you generally do) and only bet a horse if the trainer has a positive ROI on a trainer stat listed? I'd use BRIS instead of DRF because it has more specific stats and I belive don't go back as far as DRF so it's more recent (although I could be wrong on that score.)

Alternatively, choose the top 2-3 horses and only bet the ones with positive trainer ROIs.

You could decide to choose a minimum of 2 or 3 wins to make the stat more reliable.

I'm not a big fan of using BRIS trainer stats because I don't believe you can use individual stats in isolation (i.e., just look at layoffs without looking at say class drop or a certain jockey or a workout pattern, etc.).

But just wondering if you have handicapped the race down to the top horses if this could be a decent seperator.

[GameTheory]

This does seem to work, actually. I investigated it a year or so ago when BRIS was giving away all their Ultimate PP's everyday.

Here are some tips, from what I remember. You'll need to check it out some more:

-- Ignore overall trainer stats, or any stat that is too general (all races for trainer, claiming races only, etc.)

-- Look for the "angle" stats -- class drops, layoffs, off claim, etc. Jockey/trainer combo is also good.

-- Just any positive ROI isn't good enough. Look for at least 20% wins and +$.50 ROI together. If the trainer or trainer/jock has two such stats, all the better. Positive stats override negative stats, unless the negative stats are ridiculous. (+w%/ROI on class drops, but 0 for 100 w/ layoffs.)

-- DON'T be scared off by stats with tiny samples -- 1 out of 3, etc. You'll hit some good ones with those, but in that case look for strong overall stats from the trainer (20% trainer overall, etc.)


Since I was crunching the numbers by hand due to the pdf format, it was a more or less informal test, but after looking at a few hundred races there was a big difference between horses (contenders) supported with such stats and those without. Also remember, positive ROI stats generally come from high-prices, not super-high win %, so you're looking for horses that are going off at mid-range odds usually, not favorites...

[Rick]

andicap,

Jockey stats worked better than trainer stats when I looked into it and they didn't have to be all that impressive, just eliminate the big losers. Jockey ROI >= -0.60 either on all starts or at the distance improved the overall results.

[Tom]

andicap,
I tried a method using Fromulator data about a year ago. I epxorted the trainer file to Excel than sorted by trainer win %. The idea was bet horses that had at least 30% winhners and a positive roi in conditions the same as todays - ie, dirt, turf, allowance, etc.
Absolute wore results I have ever tested. I was really surprised at how horrible the result were. I quit after two weeks.
Now that I think about it, it might be worth looking at as a
contrarian elimination method!

[Fastracehorse]

<I'm not a big fan of using BRIS trainer stats because I don't believe you can use individual stats in isolation

I whole-heartedly agree.

<But just wondering if you have handicapped the race down to the top horses if this could be a decent seperator.

I actually like low profile connections - I'll bet any low % trainer if I like the horse and see trainer intent.

Tomorrow I like a horse at Woodbine - trainer is 1 for 16 - that doesn't phase me - if I'm wrong..........next!

fffastt

[Fastracehorse]

<Look for the "angle" stats.

Tis' very true.

Some trainers are lethal at their one and only angle - but they are terrible overall so they are tremendous over-lays.

If you bet Doug O'Neill in California with blinker changes - blindly - you would make a very tidy profit.

fffastt

[Fastracehorse]

<Rick Ransom, author of nothing, not selling anything, and can't win playing two horses in every race.

Too bad you weren't an author - you are a great read.

fffastt

[Fastracehorse]

<Absolute wore results I have ever tested. I was really surprised at how horrible the result were. I quit after two weeks.

You were surprised to find you now know something many others don't -

:mad: :mad: :mad: :mad: :mad: :mad: :mad:

fffastt

[andicap]

Originally posted by Fastracehorse@DRF


I actually like low profile connections - I'll bet any low % trainer if I like the horse and see trainer intent.

Tomorrow I like a horse at Woodbine - trainer is 1 for 16 - that doesn't phase me - if I'm wrong..........next!

fffastt

I actually agree with you. These are the type of horses the public overlooks.
Pizzolla loves 0-for trainers because they get such huge odds. If you're a longshot player and willing to take a 12% win with really high mutuels you have to play this way.
My game is somewhere in the middle --- and I've found many $14, etc. horses with name trainers. I especially like the middling guys -- Dom Galluscio and such who are decent conditioners but don't have the stock of the big guys.

Dutrow paid $10.80 at belmont on Sunday with a horse who figured.
James Jerkens paid $24! on Sunday with a horse (he's 21% for the year)
(and the horse had the best Beyers within 45 days -- who says Beyers can't bring high prices these days?)

[Rick]

fast,

Actually I did write some technical manuals years ago when I was a programmer but my name never showed up on any of them. But software people and engineers are notoriously bad writers so I don't count that for much.

[Fastracehorse]

It's not that my game isn't in the middle, or on the left or on the right.

It's just that trainers are both over and under-rated. My main point is that I bet the horse.

Mark Glatt was trainining at 19% last Santa Anita meet when I hit him on a 58-1 shot - so I like those guys too. Doug O'Neill has been the real magician at this Hollywood meet - he's been good for a long time - but he still knows how to throw the public of his horses.

fffastt

[Fastracehorse]

I know an engineer at my local track here in Canada - he's doing tote-board profiling for Southern Califronia tb's. He's relatively new to the game - middle-aged ( now that's tough ) and still doesn't understand the meaning of value.

The funny thing is that his passion is politics - yuk - Who do you like in the 5th??

fffastt

[alysheba88]

A trainers total win % is completely and utterly irrelevant to me from a betting perspective.

I am more interested in how he does with the specific race conditions. If a trainer wins at 20% that tells me nothing of how he does with first time starters. I know most here realize this, just pointing it out. There is much more info available about trainer stats.

And again any win % is pretty much irrelevant to me (even on the more specific conditions). Its ROI that I am interested in. The trainers who are 24% in a certain angle are almost always under 2 (and dramatically so) on ROI. Where the 13-16% trainers are often over 3 on ROI (and not talking about a small sample either) on a specific condition.

For example, before Da Hoss Michael Dickson was often a great bet on the turf with layoff horses. He was a relatively unknown and this was before alot of detailed information was esily available. Well since Da Hoss everyone seems to know about Dickinson and his ROI has dropped precipituously. I don't have any # in front of me now, but I often still see him winning 26-28% of the time with a angle or category but his ROI is under 2 because everyone knows it.

Same with a Pletcher second out.

In short, you often end up taking the worst of it from a value perspective when betting on the best trainers.

[Rick]

fast,

A lot of engineers I've known thought it probably would be pretty easy to find an easy solution to the handicapping problem. They're used to things being more predictable in their line of work and are usually in for a rude awakening when they try the horses.

As to politics, my opinion is best expressed by a commercial I once saw in Arizona for an independent candidate. It started out with an old lady exclaiming "the Republicans and the Democrats are lying to us". D'oh. Yeah, I guess that just about sums it up.