Brohamer Model Results
 

Has any active member reported on their findings (modeling) of the
key factors. Such as VDC, BLBL RX+ etc. ?
Anxiously awaiting any such report.

Hi gillb. I use Brohamer models for all the tracks I play but dist and surface. Since the model is based on the handicapping of each person no one elses' work will help you. I only
play 1 track a day so it's easy for me to update my models in real time.
Tim

Pacelines
 

Hi


As Tim says The factors you are looking for are pace line specific.
My models on VDC BLBL and RX are going to be very different then somebody who picks pace lines differently

Bill, is that model you made once (2017/2018?) which had ~900 races in it based on automatic paceline selection?

If so, it counts as 'paceline-independent' since thee is no user judgement or selection involved: you just let the software choose the line and bet and make models accordingly.

If not (i.e. you made paceline judgements), then your comment above applies.

Ted

That
 

That is correct Ted I had forgot about that model. Yes, I was on full automatic never changing a paceline with many tracks included. I will try and dig it up

I am tracking only Belmont at this time. One thing I have found interesting is that I add four factors together to get a line score. I have found that by adding the ranking of v/dc, csr,pl and bpp i easily eliminate horses. with less than 8 horses in the field, use a line score less than 20. With a field of 8 or more use a line score of less than 20. this will get the winner around 85% of the time. I use the automatic pace selector.

I'll look for it too and post if I can find.

Ted

Hello All my Pace and Cap friends

Here is a RDSS model combining a few models I have gathered for awhile
I have over 1600 races I would say 90 percent are " AS IS" using the RDSS default settings with the best of last 3 perceptor. If a line were changed it was almost always to a more comparable distance or surface

I have merged all my models into Dorianmodes nifty spread sheet with the filters on top from this post:
http://paceandcap.com/forums/showthread.php?t=12493

Here is my spreadsheet:
Attachment 48507

Earlier this year I modeled SA by distance for both dirt and turf and I also did GP for just the sprint distances of 6f, 7f and 8f. Its in a similar format as Bill V's info with best of last three info, black boxed from RDSS. The SA info stops once they took some time off to redo the dirt track.

There are two files that are rather large and probably cannot be linked to this email, so if you PM me with your interest I will send them via regular email. I also broke down the data by maiden and non maiden, but you will see that once you starting slicing and dicing the data, sample size becomes an issue at times.

In doing these studies I found, in discussions with Ted primarily, that Vdc and all of the Pace of Races stats are arbitrarily high due to ties in the rankings (Vdc especially) and in cases where horses in todays race are coming from the same race, their POR numbers are the same so sometimes you have 4 "best" numbers for 1F, 2C, etc. which makes a top 4 rank unuseable.

In general, the CSR info is better for dirt races. Top 3 or 4 data from final odds can be a little suspect, because in all cases these are the final odds from the race chart and those odds or ranks ARE NOT necessarily the same as what you might bet on, one or two minutes prior to PLANNED post time.

Like I said PM me with your regular email and I will send them on.

Bill

Bill, we can link them here, I can adjust file size settings if required, as with Bill V's files above. Remind me which ones they are, or a date you emailed them to me and I can upload and link them - if you like.

Ted

Ted,

Email to you on 3/28/19, also sent via regular email. One file is 2,000 and the other over 4,000 MB

Got your re-mailed copies today. Will get them uploaded this afternoon.


Thanks!


Ted

Hi Bill, (and all)
I took a look at this sheet you kindly posted. It's great work ... I'm glad it makes sense to you to be able to use it. There are some obvious improvements I see now I should make to the original sheet I posted, (for example) I should have added the filter function for the track, distance, surface and race type columns. I'll get that done today or tomorrow and post an updated spread sheet. That way you can filter a lot more things, and easily look at more specific information.

Thanks
 

Please, I look forward to your improvements

Ok, here it is ... sorry it took so long. (heh heh) ... :D
So ... I expanded the formulas to allow for 10,000 races.

Also I fixed the filter function, and took out the "dummy data".

The nice thing about the filters (which may not have been working properly in the sheet you uploaded ... but they are now)...is that as you did here, you can put ALL your data in one sheet, (ONE PLACE) and then use filters to look at what you want.

There is a filter applicable now to EVERY column. That means you can apply multiple filters at the same time ... for example if you were to first filter to chose just Belmont races, then second, add another filter to pick just turf races, ... look at what the "small function" tells you ! It tells you that 85 % of the Belmont turf races fall generally in the #1 position in most of the RDSS calculations. That might be very useful for one to know. There are endless variations of these filter combinations that will come to mind for the excellent creative astute handicappers here.

Always save an extra copy somewhere, before changing anything so if you or your computer messes up the sheet, you can restore it easily.

Also be careful about leaving "on" filters you don't want. Below is a little screen shot of what a filter looks like, when its "on". The little funnel is up.

If you see other changes or additional functions you want, let me know.
Tom
(Dorianmode)


Attachment 48532

Attachment 48533

Here are Bill Lyster's Excel Model sheets:

http://rdss2.com/models/Model-GP.Dirt.6-7-8.xlsx (2.5 Mb)

http://rdss2.com/models/Model-SA.2018-2019.xlsx (5.3 Mb)


Thanks very much for these Bill - very kind to share them!

(Bill, when I opened the SA workbook, Excel gave a message '...found a problem with some content ...' requiring 'recovery as much as we can'. Perhaps take a look at that one to be sure it's the same as what you sent me.)

Ted

Thank you Tom! I will be sure to incorporate your improvements to the model sheets when I release the next version of it (and send you some samples to test before hand).

Ted

For those wondering where this came from ... there is some additional information in this thread about how this began. Post #9
http://paceandcap.com/forums/showthr...ght=Dorianmode

I just opened that one up and did not get any error message. I got a message from the open source file telling me of new improvements, but no errors.

Bill

Tom:

I just tried to filter the surface column for turf only. When I did that the summary info in the top few lines said a sample of 605 races, but when I went to look at the sorted data, every race was marked "T" for turf.

So when you sort you should only look at the results in the top 17 lines? Is that correct?

Also the percentages of top two, three, and four show numbers in excess of 100% for each column. Surface column sorted for "T", 605 races in APV, and horses in the top two show 102, top 3 140. The formula covers lines down to line 1775 which would mean that any race whether turf or not would record if the rank was top 2, 3, etc.???

Also it appears that only the column formula for APV goes to Q10000, the other columns cover through Q 1775

Bill,

Yes. That's the point of the filter function. When you filter for "T", it selects only turf races. If you want all the data back, you have to undo the filter.

With the "T" surface filter on, the numbers in 1-16 have results ONLY for "T" races.

The % search for 1 and 2 was not changed to 10,000, as it should have been. The others were OK. I changed it now. Also the % of the totals for 2,3,4 and 5, is wrong, as you say. I have to think about a fix for that. I'll post an updated/corrected sheet this afternoon.

Thanks
Tom

Turns out the size of the file with the very complicated formula for filtering and counting the SUBTOTAL formulas is too big to upload. I'll think about a way to work around that.

Here is Bill's file with the filtering functions corrected, I think.
The formula for counting only visible filtered rows is very complicated. I'll keep my eye open for a simpler one that takes up less memory.
In order to upload it, I had to lop off all the extra rows at the bottom. But if you add rows in the normal way, it will still work, and the programming ranges will update automatically with the extra row numbers.

Bill, if you want me to fix the red and green and yellow colors in all the races, if that's helpful, it would be very easy to do.

Tom

No, the red and green was built into the template, it goes away after the first few programmed cells. NO problem with adding cells as long as when we do, we expand the range before getting to the last named cell in the range.

So you add row A1785 above the named range, which here might be A1790. The range expands by as many lines as you add, as you know.

There is a lot to check when making this spreadsheet. I spent a lot of hours during my research debugging formulas for all the columns all the way out to the D and E columns. Thanks for doing the leg work on the rest of the formulae.

Bill

So, sorry about this, but there is an error in the programming here. In the row which says "% ranked #5+", it's actually giving you the % ranked 6 and above. The number in the formula (towards the end) should be ">4", not ">5". >5 gives you 6 and above, not 5 and above. You can fix this easily by changing the first of these columns, in a column with data, replacing the 5 with a 4 .... and dragging it across, or ... if anyone is actually using this sheet, and this information, and wants that done and posted, I can post a corrected version.

Thanks again to Bill V for sharing all this great information and doing all the work of putting it in one place. The information here is (literally) worth a fortune.

Tom

Thanks Bill and Ted. I finally had time to "digest" these files.
An amazing amount of information here ... thanks Bill ... a huge contribution.