Eric,

You can get the Beta version of 4.0 from Nathan for $81. And like Will indicated it is well worth the cost.

Sounds like you have some questions about the program and if talking to someone on the phone would help, I could call any weekend after the races.

Feel free to email Nathan if interested and we can communicate beforehand by email.

By the way, great advice from the others!!!

Bill

Thanks for the help Hdcper

I need to get in touch with Nathan and find out about a few things regarding versions 3.5 and 4. I just purchased ver 3.5 a few months ago and need to come to grips with it. I'm supposed to get ver. 4 when it comes out of beta.

Eric S.

All of you fellas are.

I haven't seen one post where anyone took me up on the offer to make the phone call when I gave you the number.

202.456.1414 is a telephone number you pay for now. There ain't no free lunch around here.

BTW, that's the number of The White House. No fooling - call it and ask for King George.

He's selling Halliburton stock for Cheney this week,

Bob

Mailto:skyblue1@pobox.com

I read that this software has it's own trainer DB. Could it be queried? Say,trainer X in turf sprints at Lrl from x time to x time.Could this be done automatically based on what is desired for a particular race? Could the sire data be tracked to see how many 1st turf starters it has had in the last x days? Thanks.

Lou,

Yes, ES 4.0 has the ability to track "connection stats". In a nutshell you can track how well the trainers, jockeys, owners, and sires do in conjunction with the typical profiler "abilties" such as "Best Speed" as well as approximately 30 new abilities such as "turf to dirt", odds ranges, layoffs, etc.

The Profiler includes these values. You can also do analysis separate from the Profiler view using the Connection Analysis report tool.. That is - you can perform queries in the software.

That said, you can NOT do analysis on these stats using TIME as a filter. That is, you can not ask how well some factor works FROM TIME X to TIME Y. THe reason is a technical one that bummed me out when I had to make the choice to implement things the way I did. See, at first _every instance of a factor_ had its own record in the database. This allowed me to date stamp each instance, and provide other information about it. The problem was that the database got HUGE very, very quickly. It also was quite slow to operate. In order to fix these two problems I came up with a scheme the software uses to track each "factor" as it occurs in just one record (or, actually, one record per track). So instances of factors are _accumulated_ in one record. This saves a lot of space, and allowed me to make the database code very fast, but lost the flexibility to do queries over time. I do hope to bring this aspect of the connection stats back in the future, but its going to take some more serious engineering to make it happen.

Lou,

what I've been doing (at Hdcper's suggestion) is exporting from my main db to a new db for whatever time period/track/surface that I am interested in. Takes just minute or so.

Bill N

Bill --

Exporting/importing will work to achieve some things, but it will not work to achieve a breakdown of the "connection stats" over time. You could _build_ a database from files over a certain period of time (just use the "store by track", set it up, and come back later), but the export/import will not work.

Here's the reason explained with an example.

Imagine you have bill mott with Best E1 horse. THat's the stat you are interested in. Here's how I would have liked to handle this. You'd have a database table with records like this (hypothetical):

RaceIndex COnnection Ability Type Other Stuff
11/1/2003, R1 B. MOtt, Best E1, etc.
11/3/2003, R4 B. MOtt, Best E1, etc.
11/5/2003, R5 B. MOtt, Best E1, etc.
etc, etc..

The problem with this is that there are just too many records for what we are doing today -- the databases get HUGE and they get SLOW.

So, instead, it works like this:
Connection Ability Info (starts, wins, etc)
BMott, Best E1, info

NOw the software, when you store to DB, keeps track of whether it needs to update the BMott->BestE1 record (it knows if it has seen this race before). THen it just updates the existing record, instead of adding a new entry.

This makes the size of the database much more manageable, and the speed of the database queries very fast -- but does not allow one to query, or import/export, connection stats using time as a filter. You can also see how connection stats make the new datafresh stuff a real pain to deal with.

Thanks Nathan, I'd better check those db's again! I've been doing it both ways experimenting with different thing.

Bill