History of the Methodology

[Bill V.]

A great big THANK YOU

I would suggest to Tom to build his record keeping SLOWLY
Try not to track too many factors at first but Slowly add additional factors, only after you are confident in what is happing with your beginning factors.


For me I only keep records for either Del Mar or Santa Anita*
I make my serious money betting into group tickets with sharp partners. at Socal .
When I got into making bets with big timers I realized I better know what I was bringing to the table, saying I like a horse because its tier 3 on blbl ...well , I would just get blank stares, I also notice the winners I was with all kept RECORDS

Now Ray and Mitch you are so right. I track many things now but I constantly try to hone down my database into smaller bits.
Mitch I agree about the most recent 20 races being what Doc recommends.

I save everything so my model and profile always includes the date race
and race number as well as the winner, what it paid and any scratches
This way I can go back and start to add more factors to races in my database
by going back into the race , Lets say I want to add a factor like trainers who
win off long layoffs, since my model shows me the race info I just reopen RDSS and grab the trainers info , Since I already have the DSLR infor factor
Its a snap to add the trainer name ,
I record every race each day onto the main worksheet I then copy and paste
races onto other work sheets for example
My main Del Mar work sheet has 412 races
but after each race I copy the race and paste it to a specific type worksheet,
For Del Mar Turf races I have 4 separate worksheets
Turf Sprints, Turf 1 Mile, Turf Long ( 8.5 or longer ) and Maiden Turf
After awhile, I will break these specialized work sheets down again into
male of female, then by age maybe someday into purse value or claiming , non claiming ect.

This system helps me keep my sample sizes not to over whelming
I can also feel confident to single horses In Tims example above he mentioned a 45% trainer with 3 year old fillies,
I have this type info and if I can spot this pattern in a race were yes the 45 percent guy is sharp but I also can see what the other trainers in the race are capable of doing with 3 year old fillies Trainer info is a tool but any database is only as good as the quality of the paceline selection method
used to generate the numbers.

In summary.
Keep records but focus on pacelines and contenders and build your work slowly,

Good Sill
Bill

ps

*I have year's worth of Parx but being mostly on the west cost I am now spoiled. Now I love fast tracks with little DTV crap so I only "play" Parx and Monmouth and NYRA for entertainment and to help my dad

[Tim Y]

I have found that modeling by RANKING does not work since they do not represent objective numbers that the horse created by challenging the pace of the race they were in.

Let's say total energies of a group are 160.5, 160.45, 160.42, 160.33. 160.21 as compare to rankings at the same relative level, are 160.9. 160.45. 160.01, 159.55, 159.21. The first group's rankings are skewed as directly compared to the second. Rankings are too dependent upon the range of the GROUP and cannot be compared directly.

% median, early/late balance, ESP range for the distance, etc, are OBJECTIVE

[Lt1]

Tim I have found rankings do work very well for me in factors such as EPR,LPR,CPR,AND VDC. When it comes to %med rather then ranking by number I establish an acceptable range for the distance and class concerned.
Tim G

[Tim Y]

Some factors are UNIVERSAL with only slight variances, i.e. early/late balance from the line USED to evaluate the animal COMING INTO the race.

No matter if you follow the earliest of the earlies (inner Aqu, Fort Erie, Parx) where the 6 furlong e/l is high +4 to +12, or a more presser track (Oaklawn) there are all within a close range of one another. The OUTLIER here would be the Fair Grounds with the inordinately LONG home stretch. As a reminder, the RANGE of e/l's on the grass is huge and the better thing to monitor is deceleration: Most grass races are won by horses in the 98.5 to 105 range. THERE ARE EXCEPTIONS, like Woodbine with its very long stretch.

Also, it is a good idea, at least twice a month, to search out a NEW venue that you have never tried before to see how your brand of selections work there. Laurel would be a good place to start (big fields classier than average horses). Canterbury has nice fields, works with the methodology except the prices are not great due to smaller than average field size. STAY away from fields of 6 or 7 horses.

[Mark]

I assume that the Segment Screen is based on the Dream race of Thoromation fame.
With adequate consistent contender selection the best fractions from the group of horses will be used to establish the rankings of each contender by racing segment and compounded factors on the Energy Screen.

Applied consistently this should overcome your objections. The program applies the same rules to all pacelines and beaten lengths whether you agree with this or not, that is the Sartin methodology. I, personally, am skeptical of the 3rd fraction calculations and the adjustment process used. It makes absolutely no sense to me to apply final time adjustments to the 1st fraction but that is the way it is. It just puts the responsibility upon the user to pick accurate pacelines and avoid aberrant ones.

[Tim Y]

I rarely ever use the third fraction evaluations except on TURF or marathons. The hidden energy is a better one. Deceleration standards are even better for turf lines.

My bone of contention with that 3rd fraction is that many a race is OVER midstretch and a winner does not persist with an all out effort. The "race" is early, always has been as those are the movers and shakers in the field.

The "passive" horses, (FAR OFF THE PACE OF THE RACE), content to just try and catch those actually "racing" are UNRELIABLE to put it mildly.

[kpmats10]

Thank you for all your pointers, some great reads here as well. I welcome the opportunity to speak to anyone willing to work with me collaboratively or individually. Like I said, I'm here to learn.

Here's where I'm at in my handicapping -

I started tracking myself in 20 race cycles. Mentally, that's been a big help and when I look at my cycle, it's not overwhelming if there's a bad day or two in there because I feel like I can pick myself up. Losing feels less, shall I say, demoralizing? That said, my contender selection is 90%. In my most recent cycle, it was 90% with one of the two horses I missed was a 30-1 longshot. Don't feel too bad about missing out on that. I'm not chasing bomb winners, I'm chasing winners, period, and profit.

My winning percentage though is around 40% betting two horses to win. And not showing a profit. Problem.

My homemade software automatically makes adjustments for distance, surface, track speed based off the DRF variant. It then calculates velocity for the incremental fractions, AP, EP, SP, FX for sprints and rankings for each. Also, creates %E. At the end of the day, the data is stored into a Brohamer model for all the rankings. I'm not currently tracking by class, but am tracking by distance and surface. It also makes TPR ratings and these also go into a model.

I also won a bunch of old Sartin programs as part of an ebay package several years ago and just recently discovered what I have and how to get them to open, but have no idea how to use them - Phase III with the Match Up, Pace Launcher 3 and Synergism 1A (I think)
I also make a track profile everyday.

I'm not presently an RDSS subscriber and am hesitant to become one solely because the money I spent on my own software was large (and it is customizable to add/subtract/change anything I've already created) and the data files I use work with two other programs I use. I can't justify the expense of more data files, especially to make a daily model means being a daily downloader. I wish RDSS was compatible with Brisnet or other data files.

Where I THINK I'm struggling -

PROPER paceline selection. What I mean by that is, I generally will use the last race unless there are reasons to go back (wrong surface, distance, trouble trip, etc). Generally my pacelines are in the last three races. But I have questions -

One thing I'd like to do much better on is projecting today's pace. I've read PMTR and started using the fulcrum method a bit. It's pretty good in getting you at least close to what the projected pace of today's race could be. So let's say you have your fulcrum horse. When looking for horses to rate against the probable pace, do you select + or (+) races against that pace? If you have a horse who's last race was a + and no excuse to go back other than a + race 2 back is at today's fulcrum EP fraction, what line would you use? Or is it better to use the most recent + line for each contender, regardless of what probable pace be and bet off your model rankings?

I've been reading the Match Up forum, and I can't seem to get the 5 steps down. Look at a horse's overall record and pacelines to guess it's projected pace? Does that mean looking at all 10 lines? Even if recently it's not running back to those numbers?

Self reflection says I may be leaning to heavy on my models to make my wagering decisions, but maybe the models are wrong because of my improper paceline selection.

Or I'm playing races where I think there's a vulnerable favorite and losing to a low priced chalk horse instead of taking the easy money.

Does anyone go over losing races anymore like they used to in the Follow Up?

I feel like I might be all over the place in this post and may not be expressing myself coherently. I want to learn to be better, that's the bottom line for me.

[Tim Y]

Here is some help: NEVER go on a single pace line or you WILL get the scourge of SAMPLE ERROR.

Find the fastest pace of race that each ran against: HOW did they do? Could that pace of race be the same today? If an entrant could not keep up to a 45.3 and the rest of the horses can, ELIMINATE that one.

ONLY wager on opportunities that are OVERLAID, or if none are available, PASS the contest. THE WAGERING SIDE IS THE MOST DIFFICULT THING TO LEARN, not the handicapping.

Dick Mitchell did a great article years ago that simple stated: "Even in the best of evaluation systems, the odds are that you have a good chance of going 7 races in a row without collecting. THAT is possible. Do not let that change your approach."

Try to ONLY bet in fields of 7 or more so as to spread out the monies enough so that HUGE favorites can be defused throughout the odds board. ANYTHING at 6 or under, is USUALLY, a waste of time even to bother with.

[Tim Y]

Another thing: we get paid from the monies that the crowd does NOT BET on the horses that hit the board. LEARN to know booking points and how much remains in the field after you have made selections. Don't think of odds as much as HOW much of the field does that entrant represent. Remember the field is represented as around 125% to account for the take out.

A very good article on that was printed for the American Quarter Horse Association in an article explaining how the morning line is written.

Here is that reference. It shows how to understand booking points.

https://aqha.com/racing/pages/racing...-morning-line/

[Mitch44]

kpmats10:

To solve any problem the first thing one must do is recognize that there is a problem.

Your own record keeping has shown your pace line selection is a huge problem and greatly contributes to that low 40% win rate. I follow the "Doc's" teachings of using the best of the last 3 at a similar distance and surface. There are several reasons why the last race may not be correct such as coming from off a layoff and needing that race and not performing to its true potential. Perhaps that race was just a conditioning race up in class to sit out the jail time off a claim etc. I try to stay within 1/2 F of todays distance and no more than 1 F max in selecting a pace line. The least you have to project the better off you are and races are just run different as distance increases.

In your own program using the DRF variant is basically ok in that your using a consistent procedure to do so. Anyone could argue for a better one but as long as you are consistent is a key. Where you may have a problem is how you proportion that variant among your fractions and figures. And that varies among different distances because at 6F the SC is 66.6666 % and at 8F its 75% so it depends weather your using SC & 3rd FR or all 3 FR's etc. I would investigate this with correct selected pace lines for a 20 race cycle to determine if that's the problem and whether your variant distribution can be improved upon. Getting from 40% to 60% + is more than just pace line selection. You can't just treat all race distances as 66.6666 % at the second call.

I still do post mortems, even knowing "The Hat" didn't believe in looking back. They can correct errors and really build your confidence knowing you made no mistake.

Not selecting correct pace lines and distributing a variant incorrectly will make your model useless because a model reflects your handicapping.

I understand when you say you invested much time and money into your own program. But I would recommend cutting your loses if your not winning and only getting 40% winners or making what you have better. The stuff in your program isn't the best of the Sartin stuff and was improved upon in later programs. RDSS contains all the latest programs up until the "Doc" passed and it can help you to pick better pace lines as just one example. Ted gives a free trial period so the cost is minimal or nothing if you just use the examples that come with it. With a free trial it won't cost you any further investment and you can determine if it ups your game or not.

I'm not a believer in the Fulcrum and stated some reasons as to why in another post, just 2 were mentioned above in using the last line. The Match Up is a valid concept and was the springboard for later programs and concepts.

Mitch44