HTR2001 Upgrade - Awesome!

[Ken Massa]

A terrific new set of tools for your handicapping in the newest HTR2001. Should be available on Thursday night for download. If not, then email me for a copy.

An export function is ready with lots of new features and data. I'll post a message on the Access forum regarding that.

New in the program
-------------------------
Qp = Quirin Speed Points - computed exactly as Quirin did it in his book. Highest is best. Values range from +8 to 0 (zero). The 0 is the worst rating and indicates zero early speed expected. Appears on multiple screens and in the header of pp's.

QT3 = Quirin top 3. Find the top 3 Quirin speed points horses and add their points together.
If the value is 20 or higher, the theory is that the race is top heavy with early speed. Access users will find out the truth about this! This item is found in the header of all HTR2001 screens now.

Qp% = By totaling the Quirin points for all the horses in the race we can then find out what % of the total points each horse gets. When a single horse posseses > 30 then it is probably going to be alone on the lead. Find this item in the Program screen and the third section of the Velocity Plus screen. Also in the export.

GL8 = Good efforts Last 8 starts. Using the definition from the December newsletter for a "good effort" this value ranges from 8 to 0 with 8 being outstanding effort in all 8 recent starts. If the horse has less than 8 lifetime starts, the number is extrapolated from whatever he has run. Found on the Program screen and in the export.

*Early = I have refurbished this item from the DOS version. It can appear on one or more horses now. If one horse receives the (*) it will probably run to the lead. If two or more get the (*) it is likely one or all will go for it. If no star is found, the race the computer could not figure out who is likely to lead. Found on all screens where the RS (running style) is located.
Access users will find this item in the latest export.

Full Card Printing - you can print full card on the Program and Velocity Plus. I'm still working on the other screens. Print them one race at a time for now.

"Early Speed Extreme" - new screen not ready yet.

Help! button = has some info on the new data.

Export - fully functional now (see the Access forum)

Bugs - fixed some, but one bug eludes me. Charts cannot be displayed past race 10. I'll figure it out, use a text viewer for now. There are some other harmless quirks in the program that people have found., if you see something odd, let me know. The program is much smaller than previous versions - leaner and faster.

Thank you for beta-testing my new baby!

[Ken Massa]

The newest version is dated "December 13, 2001" on the opening screen. Be sure you have that and not the previous version which shows November as the month.

[hurrikane]

Sitting here drooling like Pavlovs dog.....

how many races can we export at one time and what is the relationship to the old export tables I have now?

[Glen]

Thanks for a nice update Ken. The new early speed factors look really interesting. Looking forward to "chugging and plugging" all those lovely figures.

[Rick]

You could probably load all your files and export them all at one time.

But, I wouldn't recommend it.

At least wait. Ken is doing a modification to the exporter that will let us export to ALL only. It will run about twice as fast as exporting to track files and an ALL file.

My early test show that I am having problems with the exporter stopping. When that happens I have to restart from scratch.

I would suggest starting with bite size chunks, something around two weeks to a month. For your first test try running other programs at the same time to see if the exporter will continue to run.

Check out the HX1.txt file to see which fields are included in this export. I am not ready to replace the HTR_ALL file but I might me tempted to add the new fields to what I already have in my master table.

[MikeDee]

Don't know what the actual limits are but I have already exported all the files for a single year for a single track over 2000 files. Still have to do a year at time because of the file name.

What I am going to do for now is take the fields from HX1. and join them to my master table which I keep on a track by track basis. From the joined files I will create a new single table with all of the unique fields from my master table and the new HX1 extract. This will be my new master table for that track for the time being.

Next I will include the HX1 table in my joined query with msac and impact. I will use this query on a daily basis to update my master tables on a going forward basis. As Ken adds more fields to HX1 I will revise the joined query that I use replacing fields from msac with fields from HX1 until HX1 is complete.

For the time being I am only going to bring in pace line 4 from the new extract. My current tables are pace line 5. So in my db I will have pace line 5 and pace line 4 pace factors. I am putting all data for a runner on a single line. So I will have a fr1 for pl5 and fr1 for pl4. I am renameing the fr1 field to 4fr1 for pace line 4 to distingish it from the fr1 for pl5.

I am very interested in pl 4 and I want to see how it compares to pl5 data.

Doing queries with multiple pace line data is going to present some additinal complexity to the query process. I am thinking that when you bring in multiple pl lines of data you will have to do a query with each of the pace lines and then do a query design on each of the 5 queries joined in the design in order to do a side-by-side comparison of the factors. Just thinking out loud at this point

[hurrikane]

Interesting ideas Mike.

I am getting an overflow error trying to export 2 mo data for 1 track. That is all pacelines but I think I will rethink this all paceline idea. too much data.

[Carl]

I managed to get three games back built into my NBA table and like it.

We may have to import several times into ALL_MSA table. Primary key assigned by access may be of value in this step, I know they let me do things with NBA data I wouldn't known how to do otherwise (thinking out loud at this point too).

BTW, agree that we have to find a way to get all a single horses data (be it one race, two races, three or four or five races back) on one line for effective querying.

[MikeDee]

You might try to isolate the overflow error. I got one as well and I was able to isolate it to a single race at CRC in July. When you get the error note the track date and race and put it in a directory and run it by itself and see if htr fails. If it does let Ken know may be a very obscure bug. I have only gotten it once so far

[hurrikane]

Yeah Mike..that's what it was. 10-28-01 SA race 6 causes an overflow in both the export and HTR2001. seems ok in the dos programs

[MikeDee]

Thanks for the heads up HK

Ken based on your comment about Oct 1999 seems like the thing to do is to just discard all HTR data prior to that date. Would you agree?

[Ken Massa]

Mike and Greg =

Those files were not bad - they were races in which all the horses Qp totaled = 0. That is an almost impossible situation, but it happened. Remember, HTR-DOS did not use a Qp of zero.

Data prior to Oct 1999 is not worthless for HTR4MSA processing. You can use it to run query on velocity data or the static HTR factors like TRN or JKY ranks. In Oct 1999 Jim Cramer added variants and figures for fractional times to our file along with his final time data. I use those to assist in computation the IMPACT numbers.

I'll upload a fixed program to Rick on Saturday. People are still reporting bugs and I want to fix them all before re-compiling.

[MikeDee]

For my purpose I think I will use the oct 99 as the starting point. when I am modeling I look at all the numbers htr and impact. It will get to confusing for me to try and remember that certain factors can't go back further then 10/1/99 and others can. Better for to use one starting point for all.

Ken - just to be sure. In the new exporter pace line 4 is the new pace line 4 from HTR2001 and not pace line 4 from the old dos version. Is this correct?

Wow how about that 0 Q points in a race. I guess nobody got the lead:D

Are you going to filter for that one?

[Ken Massa]

MDee =

Paceline 4 in HTR2001 and the new export therein is definitively NOT the same as HTR-DOS. (4) old <> (4) new. Nov newsletter has details.

Because new PL-4 is very strict it is not fair to judge it straight up against the others. This is because it blanks a horse that doesn't have a race at the today's distance and surface within the last 6 months. About 20% of horses are wiped out with such a method. But when handicapping races with PL-4, at least you get a solid line on everyone. Solve this problem by doing your comparative PL query on fast-dirt-non-maiden-claiming races. That is a large pool of races that won't have too many eliminations in PL-4 and can be compared fairly with other modes.

[MikeDee]

That's what I was thinking. I'm going to eliminate any races where one or more runners does not have a pl4 pace line, then use that race set for comparison to pl5.... should be interesting.

[hurrikane]

Ken,
I was wondering if you did any analysis of the horses without a PL4 paceline and if there is any way to use it as an elimination tool.

[Carl]

Good post Hurrikane.

I would like to bet that the "not 4's" set have a higher ROI than the "4's", albeit a lower win percentage, because these are horses conventual handicappers would not touch with a ten foot pole.

And they too, like every dog, "have their day....."

[Ken Massa]

Good idea on the blank outs on PL-4. Since the majority are layoff horses, they might show a profit. Others would be turf to dirt or route-sprint, sprint-route. Those are often good overlays.

One benefit of PL-4 is exactly that - immediate appraisal of the race by noticing the number of blanks. I switch between 4-5 while watching the Velocity Screen and notice the changes in rankings are often dramatic. You can assume the public is handicapping the race based on a perception similar to PL-4. A PL-4 race with many blanks is likely to be more volatile and produce an unexpected outcome.

A few years ago on my Stardust sheet, I listed brief comments like "good early speed", or "trainer high%" for each horse. When the horse looked like a dog and was ranked 9 on all Consensus factors - I (brilliantly) gave it the comment: "No Chance Today".

One day I get a call from a Stardust regular that was using the sheets. He had tracked the "No Chance Today" horses and they were showing a flat bet profit! He thanked me! TomCat and friend Bob from Lexington had similar good results with "lame", "dnf" and "vanned off" from the last start. Not many winners either system, but you don't need many when they are all 50/1.

The logic of illogic. This is a game about out-smarting your fellow man, not finding the best horse.

[tomcat]

He still plays the dnf...Jim Cramer once looked it up and it does have a positive ROI...if you live long enough. I don't ususally play it, but boy does Bob ever crow when one comes in.

[Carl]

Funny story about the guy playing your "no chance" horses and showing a profit Ken.

One of the guys on Gordon Pine's board ( http://www.netcapper.com ) swears by the http://horseplayerdaily.com "loser of the day". They pick one horse a day "guarantee to lose". They limit it to horses with a MLO of 6-1 or lower.

Anyway, this guy at Netcapper bets them, thinks the guy picking losers is great, because as he puts it, "he doesn't have a clue"........