|
JCapper Message Board General Discussion
|
| Home |
Register
Log In |
| By | The Human Anchor | |||||
| jeff 5/4/2008 11:03:47 PM | 5-4-2008 MNR R9 Watch the replay. The #8 horse Disco Sue stalked the leaders to the turn, took command at the top of the lane, and drew 5 lengths clear. In mid stretch she looked to be home and dry. Had I turned my monitor off 3/16ths out I would have been about 99 pct certain (barring a breakdown) she gets her picture taken in the winner's circle tonight. There's a reason I'm posting this. I used to have a term to describe riders who are... there's no nice way to say this... ahem... lacking in atheletic ability. I used to call them Human Anchors. Disco Sue was selected by two of my best performing UDMs. She looked great in what I was able to see of the post parade and warmups. She also fit a diligently maintained track profile for 5.5f at MNR that I have been keeping. But I didn't have the name of the rider on another list that I like to keep: Riders who have been riding well of late. So I pulled up the Data Window and ran the rider name for Q1 2008 and came up with the following:
Needless to say I was a little stunned. I wish I could tell you that I sat on my hands. Disco Sue simply looked too good in the post parade. I figured she was in condition and could beat this field all on her own. With about 2 mtp I pulled the trigger. 3/16ths out - despite a 4 length lead - I sensed my bet might be in jepardy. The rider's action was suddenly out of synch with the horse's stride. Watch the replay. Watch Disco Sue's rider. Compare Williams' action to that of the two riders going past Disco Sue nearing the wire. This is as perfect an example of the Human Anchor effect as I can ever find for you. Numbers from the Data Window aren't some cosmic accident. There's usually a reason behind them being what they are. In the grand scheme of things it's only one lost bet resulting from a situation that very seldom comes up for me. But still... I'm posting because hopefully you (and I) can learn a thing or two from my lost bet and this post: When you're about to back a horse ridden by a Human Anchor... DON'T. -jp .
| |||||
| busseb 5/5/2008 5:13:32 PM | Jeff = Now you know why I am so insistent about getting you to add the "shipped from" and the "avoid shipped from" catagories added. You chose a race that was a WV bred only - commonly referred to as the circus because of how much the horses that run in these races resemble clowns. The horse you bet on was from Charlestown. CT horses need to be about 5-10 points higher in gap catagories to be competitive and 1-2 rank catagories higher to be competitive. This same horse shipping from Beulah or River Downs or Finger Lakes would all run up the track - human anchor or not. The same horse shipping in from Turfway or Keenland would have won by 10 lengths. What I'm trying to say is that the inter-track numbers are way off. The reason I so desperately want to have the shipper catagories is that by being able to identify what track the horse is coming from, I can account for the differences of the numbers in my UDM. CT numbers must be higher than MNR numbers before the horse is worth a wager while KEE and TP numbers can be 5-10 points lower than MNR numbers and result in a win. Welcome to my world. ElPaso
| |||||
| busseb 5/5/2008 5:27:14 PM | Oh - and rhe owners and trainer are probably just as upset as you are - they shippped their horse 250 miles each way and brought their jockey because they believed that the numbers they were reading from the racing form showed they were going to win - even with a crappy jockey. Surprise - the numbers from CT to MNR are not consistent. What they thought should have been a slam dunk turned into a 500 mile loss. Those of us who do the UDM's came to realize the problem early and we still struggle with it manually. ElPaso PS - Next update to include the shipper catagories?
| |||||
| jeff 5/6/2008 2:06:27 AM | Shipped from track: if not the next update, an update in the VERY near future. Switching gears a little bit... Re: Track to track adjustments Bris says in the Library section of their site that their speed and pace figures are meant to be interchangeable across all tracks, surfaces, and distances. I think you and I (and everyone else out there too) knows full well that the figs aren't perfect. And yet in all the years I've been doing this I've never made a real attempt to adjust raw figures for those track to track, surface to surface, and distance to distance differences or idiosyncrasies. Instead I've focused my R&D primarily on the relationship between thousands of factors found in every horse's past performance record... including the raw figures (however imperfect they may be.) And depsite the flaws in the raw figures, I continue to use them as just one smallish piece of the handicapping puzzle. And in so doing, I've enjoyed a fair share of success at the windows - and for quite some time now. My own belief is that the figs themselves are over focused on by far too many players. My own R&D continues to show that an inherently flawed fig, such as the JCapper Weighted Fig, returns a higher flat bet win roi than more accurate figs. Why? Because it has some basis in reality and is DIFFERENT than what everyone else sees. I'm going to throw out a suggestion... But understand before you read it that the suggestion comes from a place of being DIFFERENT than what everyone else sees as opposed to being more accurate than what everyone else sees. That said, here goes... Would anyone have any use for a separate table in the JCapper2.mdb file (maybe name it the TrackToTrackAdj table) having the following fields? Active TrackCode Surface MinDist MaxDist PaceAdj FinalAdj The Active, TrackCode, Surface, MinDist, and MaxDist fields are self explanatory. I'm thinking users who've studied differences from track to track (and surface to surface as well as dist to dist at the same track) could enter point adjustments in the PaceAdj and FinalAdj fields. Suppose your R&D told you that Bris/TSN E2 pace figs were 3 points too high and final time based figs were 5 pts too low for turf routes at APX when compared to "normal" (whatever that is) observed in large data samples. My thoughts are having the ability to enter the following info into the table...
might enable the number crunching algorithms in the program to automatically subtract 3 points from all E2 pace figs and add 5 points to all final time based speed figures earned in turf routes at APX... in effect, allowing the user performing his own R&D to effect his own track to track, surface to surface, or distance to distance adjustments. In my mind this might be a nice addition. Let me know what you guys think. If there's enough interest I'll consider adding it to the program. -jp . ~Edited by: jeff on: 5/6/2008 at: 2:06:27 AM~
| |||||
| ryesteve 5/6/2008 9:51:53 AM | Interesting... I've always relied on figs to be interchangeable across tracks, but I suppose that's an assumption that isn't necessarily correct. But I'm wondering how you'd separate fig bias from fig differences that come about simply because the races are run differently (eg slow pace on poly and gun from the gate races out west). So even though this sounds like it could be interesting, I think that personally, I'd be too leery to try to mess with it. And a comment about those owners of the MNR horse: I have no sympathy for people who'd bring a 1/134 jockey with them specifically to ride their horse.
| |||||
| freddie 5/6/2008 11:20:25 AM | Jeff won't the speed figs that you are proposing from Woodside eliminate the adjustments for the tracks they cover???? I'm sure many of your clients play the cheaper tracks not covered.....would purchasing a good set of pars then allow for the proper adjustment with your proposed factors??......Freddie
| |||||
| busseb 5/6/2008 5:12:49 PM | Don't know if I'm the exception or the rule: I only download and play 2 circuits at a time. One of those is HOL (transitioning to SAX, FPX, DMR and OSA) and then MNR. The differences in these tracks means that every single UDM is unique to its own track. The way SAX and HOL play caused me to not use any SAX UDM's on the other track. I had a 20% difference in win percentage when going from one track to the other. MNR is so unique in how it plays that every UDM is segregated by class descriptor, distance, surface condition(fast or off) and shippers. NO shippers are included in any MNR UDM's because of the differences in the ship-in tracks. TPX and KEE are much higher class tracks, but the purses are lower causing severe errors in class adjustments and JPR numbers when they ship in. RDX,FL and Beulah are 1 mile tracks with low purses and crappy runners. Their numbers tend to be about 10 points higher than they are capable of running. Whoever does their numbers inflates them quite a bit. Charles Town is a 6f bullring with huge purses meaning 4 1/2 races look a lot better (class and speed) than they are. You need to segregate shippers by the distance switch coming from CTX. Also, MNR has a lot of off-track races run there. The BRIS numbers for MNR are inflated in the wet races. You need to check back on the actual running lines to see what kind of races produced those numbers. Please be careful when betting on Mountaineer. It can be a place where you get huge prices, but the regulars are very good on getting the chalks home. Also, the 2nd best horse seldom ends up 2nd. If you bet exactas, place and show can really be a crap shoot! I try to keep my win UDM's at 35-40% and most of the time the place percentage is 50-60%. That means that horse only comes in 2nd or 3rd 10-20% of the time. You will go broke quickly betting the best 2 horses in an exacta. MNR is first or last - nothing in between!! Finally, if you see a long price horse that can't lose according to your UDM's,be REAL careful if it isn't getting bet at least a little. ElPaso
| |||||
| jeff 5/6/2008 6:02:51 PM |
Fred, Realistically, I have no way of knowing whether or not Henry Kuck style figs require any type of track to track, surface to surface, or distance to distance adjusments of the kind being talked about in this thread. I've only recently gotten things far enough along to where I've been able to import them into a JCapper database. I'm trying to block out some free time this week so I can begin testing their effectiveness across a year's worth of data. I WILL say that in the few smallish samples I've looked at there have been some very promising results. Enough so that I want to create some JCapper factors based on what I've seen and make those factors available in the UDM Wizard and Data Window. Other than look for me to publish some results involving large data samples once those become available, that's about all I can report on the subject at the moment. Would a set of pars enable you to make the kind of adjustments being talked about in this thread? Maybe. I think anyone planning on doing so needs to do some due diligence and know beforehand where Bris/TSN figs are out of whack if at all.
Brian - Yes. You are probably in the minority. But if in fact Bris/TSN figs are out of whack your ideas make perfect sense. Being able to adjust to something with a better basis in reality than what everyone else sees could be a real difference maker. -jp . ~Edited by: jeff on: 5/6/2008 at: 6:02:51 PM~
|
| Reply |
|
Copyright © 2018 JCapper Software back to the JCapper Message Board www.JCapper.com |