New from KM / Aug + Sept

[km]

The double issue of the newsletter will be out at the end of this week.
August + September editions = about 24 pages.

Everything you ever wanted to know about the KLine and more.
SOR, Workout rating and Robot are covered too.
Spot Play examples
Bad Jockey "all burger" stats.

Please join us in our HTR Tournament Challenge at the end of this month. Rules and information can be found on the forum devoted to it on here now. If you were a prize winner last time or a carryover from a prior contest, your name already appears in a thread.

[Rick]

The HTR Monthly Report for Aug-Sep 2004 can be downloaded on the Members Site (http://htr.horsedata.com/)

The HTR Monthly Report for July 2004 is available in the HTR Library (http://www.homebased2.com/km/library.htm)

[Victor]

I read the Newsletter and I still don't know exactly what defines an overlay as seen by the Robot. Does it mean >=1.5 times the KLine?

[km]

Vic - the overlay in the robot is for any real odds higher than the KLine.

Example: KLine = 4.4 (4/1)

If the actual (tote) odds on this horse = 4.5 then it would get counted as an overlay in that test. If the actual odds = 4.3, it would not.

That's cutting it a little close for easy overlay identification!

In the next update I'll put in the 50% overlay so that user can identify them easily on the board. That would be the x 1.5 you mention.

Example of this --

KLine = 4.4 (4/1)

Overlay would be 6.6 or higher. Much more realistic.

Robot is a brand new baby. Exactly one month old. Growing pains are gonna happen, but I listen to feedback and suggestions and will make changes in future updates.

[Buckfan]

I love the robot and the potential that exists here !!

Suggestion for future Robot upgrade:

it sure would make it a lot easier (and flexible) for areas of the Robot like the "K" section, to have testing capabilities "from" and "to"...instead of the somewhat restrictive K1-K5 that exists now. This should also save space for you on the robot to put other functions. I know this K1-K5 area is where most winners come from...but allow researchers to research, and on various race types we may find nuggets from time to time.

Love the bells and whistles !!

[km]

You got it Daven, good suggestion -thanks

The robot gets real complicated if typed input is allowed. Instead I'll put in check-boxes for K 1-9 and you can tag any combination of them. If you want to test K3 with K7, that will be fine. Or K6 + K7 + K8 combined. Anything goes 1 thru 9.

[Victor]

K1=1
and nAODDS >= 1.5 * [nKLine]

July 10th through August 10th 2004

Total Bets 1229
WIN PLACE SHOW
Total Amount Bet 2458.00 2458.00 2458.00
Wins 187 400 564
Pct. 15.22% 32.55% 45.89%
Amount Won 2203.10 2228.60 2180.40
Profit/Loss -254.90 -229.40 -277.60
Pct Profit Loss -10.37% -9.33% -11.29%
Avg. Payout 11.78 5.57 3.87

BUT: same time zone, nAODDS >= 3 * [nKLine]

Total Bets 168
WIN PLACE SHOW
Total Amount Bet 336.00 336.00 336.00
Wins 17 37 59
Pct. 10.12% 22.02% 35.12%
Amount Won 334.20 305.70 281.60
Profit/Loss -1.80 -30.30 -54.40
Pct Profit Loss -0.54% -9.02% -16.19%
Avg. Payout 19.66 8.26 4.77

AND: nAODDS >= 4 * [nKLine]

Total Bets 60
WIN PLACE SHOW
Total Amount Bet 120.00 120.00 120.00
Wins 6 8 12
Pct. 10.00% 13.33% 20.00%
Amount Won 143.60 82.70 68.60
Profit/Loss 23.60 -37.30 -51.40
Pct Profit Loss 19.67% -31.08% -42.83%
Avg. Payout 23.93 10.34 5.72

!!!

[Victor]

Suppose I test a track with the Robot with a couple of conditions that show a positive return, then "My Plays" gives me all the tracks that meet those conditions. I would be more interested in the plays for the track I tested only.

[Lee Carson]

I agree with Victor. Need abiliity to check mark more than one track - now, its one or all
Lee

[km]

re: variable track selection in the robot
That's a complex programming task, even with Access. Easy to improvise though. Just copy the tracks you want into a separate folder. Suppose I wanted to test DMR and SAR only. I create a folder called c:\SARDMR and copy all my .dmr and .sar into it. Now I run HTR2 and change to the c:\SARDMR folder and run my robot with the ALL option. My test results with reflect only those two tracks.

re: overlay option
Maybe I can put in more than option for that. Nothing is set in stone.
Keep posting those results Vic and we'll know what to do in the future with KLine overlays.
Good work- thanks - I'll follow up with some tests later.

Typically I'm going to steer the user toward Access instead of implementing complicated robot changes. The beauty of it is in its simplicity. If the robot is not sufficient to meet someone's resarch needs then they need to step up to Access so they can improvise on the data parameters. But in the case of the variable track option methink that is difficult even in Access? But I'm not the expert.

MIkeDee, Ricks or Donnie or anyone? how would I pull data if I wanted to query SAR + BEL + AQU; just those three tracks together in my sample, ignoring all others. Does this require some VB code or could it be done with IF THEN ELSE statements?

[Rick]

It could be done easily in Access by listing tracks in criteria under tTRK in your query like

"bel" or "sar" or "aqu"

[km]

Thanks Rick, pretty easy - there is your answer about the multiple track selection Lee.

Use Access or simply copy the tracks you want to a separate folder and run the robot ALL.

If you want GET MY PLAYS from specific tracks in the robot - the output screen can be edited.
After the list pops up with your plays on it - use the mouse and highlight the entire text from the tracks you don't play and right-click "delete" or "cut" those off the list.

[km]

re: Testing KLine vs. Overlays

There are a couple of key issues with doing this =

1) There may be multiple plays per race and that always dilutes the results. The same problem occurs anytime you run a filter in the robot or Access that can produce multiples, such as using FTS as the only criteria. Some races have 6-8 FTS, others 3 or 4, many have none. It's hard to reconcile test results when there are a variable number of horses as qualifying plays. No one would bet them all.

2) The test results with KLine vs actual tote odds are not realistic for a horse bettor. We are testing after the final odds are known. That's almost impossible to discern at a minute or two before post-time when we have to make a bet. Some horses only become overlays after the race begins and final money is dropped in the pool. Many favorites that look like overlays drop below their KLine after the bell.

The above might prove Victor is absolutely right about using KLine x2 as the criteria for overlays. In most cases, we would likely be able to clearly identify those horses during the betting and there rarely be more than one per race.

KLine is an intuitive tool and hard to quantify mechanically.

[hracingplyr]

has anyone run any queries on kens research in the july newsletter concerning



HTR=1

K=1

ML=1

MLO<3.1


just wondering if anyone can tell me how this is performing.?



thanks

bob

[Victor]

HTR=1

K=1

ML=1

MLO<3.1

July 10 though August 10

Total Bets 2370
WIN PLACE SHOW
Total Amount Bet 4740.00 4740.00 4740.00
Wins 857 1391 1672
Pct. 36.16% 58.69% 70.55%
Amount Won 3837.80 4229.50 4289.00
Profit/Loss -902.20 -510.50 -451.00
Pct Profit Loss -19.03% -10.77% -9.51%
Avg. Payout 4.48 3.04 2.57

In a word, Pathetic. Are you sure this is the query you're interested in?

[Victor]

Same conditions, since August 2003

Total Bets 20314
WIN PLACE SHOW
Total Amount Bet 40628.00 40628.00 40628.00
Wins 7821 12109 14463
Pct. 38.50% 59.61% 71.20%
Amount Won 35087.80 37225.80 37239.30
Profit/Loss -5540.20 -3402.20 -3388.70
Pct Profit Loss -13.64% -8.37% -8.34%
Avg. Payout 4.49 3.07 2.57

Average mutual is WAY too low.