Been 17 mos since I posted this. Honestly I really have not been playing horses much at all since then. Now and again I will play a carryover somewhere or handicap Mohawk and bet a little bit but nothing like before. Came to the conclusion that I have to make sense of that database if I am going to invest time, line races and invest money in these races. I would love to say that I have great news and I found the golden ticket. Nah. Tough game. Seriously. I continue to look for stuff, for instance with tris, supers.....haven't found anything yet. I probably have enough ammunition that if I was getting some real sweet rebates I can make a little noise, but I am not getting sweet rebates and barring the occasional entertainment it is just not really a wise use of time. I will probably transition into a strictly pick 4/pick 5 style of play, but even that is very challenging. At least with that you handicap, you bet, your done. It seems to me that everything I have posted about rebates making it extremely hard for the non rebated bettor to make money I have lived out. Not a surprise, not a big deal, life goes on.
That being said, that was not the point of this post. I actually wanted to share some data I have. It is not far removed from my last post but I think it is interesting data nonetheless that as horseplayers(are they any left
) you might find interesting.
What I did was limit my data based on my line(100% oddline) to horses I made 13-1 or less(I actually made them 12-1 but in tinkering with my line I raised them a point to make the line add to 100%). This seems to be the point where they perform reasonably well. Horses I make 15-1 to 20-1 win now an again, but too seldom to be concerned with and horses I make 25-1 or higher are dreadful performers.
So this cut my almost 4800 horse database to 2691. The point was to put them into one big database with no elminations so I can measure the performance of horses that were over bet and under bet by the degree in which they were over and under bet. The variables are 3, my line, actual line and projected number of times they should win based off of the the publics closing line. Theoretically wins should match projected win. Below is under performing and above is over performing. I mentioned this before but my database has closing tote board odds, so if a horse is 6-1, I have him as a 6-1 even though he may be as high as a 6.9 to 1. So to separate the horses I am using public line divided by my line. So if I make a horse 10-1 and he goes at 5/2, 2.5/10=.25.If I make a horse 6-1 and he goes off at 7-1, that would be 7/6 or 1.667.
Here goes: the 4 numbers will be 1st's-2nds-3rds-4ths
up to .3 225 horses 92.8 projected wins-99-34-32-18
>.3 to .5 480 horses 140.8 proj. wins 119-98-58-57
>.5 to .75 539 horses 113 proj. wins 109-91-92-66
>.75 to 1 458 horses 75.3 proj. wins 70-75-68-69
>1 to 1.25 207 horses 25.4 proj. wins 34-31-29-25
>1.25 to 1.5 214 horses 20.7 proj wins 22-27-37-29
>1.5 to 2 269 horses 20.4 proj wins 20-37-42-44
>2 ro 3 193 horses 10.5 proj wins 8-21-31-26
>3 413 horses 3.26 proj wins 2-9-5-16
Now just for some clairification the horses betdown to .3 or below out perfrom there actual odds, but that is because of odds on favorites(for instance I make a horse 3-1 and he goes off at 3/5) not because I make a horse 10-1 and he goes off a 3-1. That part of the database is clouded by the uncanny ability of the caw getting it right. The worse performers are the >.3 to .5 range. Remember I posted that horses under 2-1 pretty win as expected by the public. So if the horse is bet from 8-1 my line to 3-1 final odds, poor performer. But from 3-1 down to 7/5 win as expected by public.
So now I have redone the database eliminating any horse that went off at less than 2-1(a little over 400 horses eliminated). So now I will revisit the low end numbers.
up to .3 57 horses 14.02 proj wins 9-7-13-7
>.3 to .5 288 horses 60.31 proj wins 48-50-40-42
>.5 to .75 436 horses 76.44 proj wins 73-69-76-56
>.75 to 1.0 430 horses 65.87 proj wins 57-71-66-68
This is where things get very interesting to me. When you eliminate horses under 2-1 then you see horses are over bet, some are vastly over bet. They only amount to maybe 1 in every 15 horses(1 in every 2 races they not only win 57 when they should win 74 but they also come 2nd 57 times when they should I assume come 2nd around 74 times. That means they win 16.5% of the time they are 2nd 16.5% of the time. So if you are looking for horses to beat, these are them. Way over bet and 2-1 or higher You can even get them out of the trifecta 53% of the time. You also don't need to make a line to see a horse that is over bet(You know who you like so when you see a horse who is your 4th or 5th choice at 5/2, you know he is over bet(I think a lot of folks run the other direction betting the horse thinking he is live-which he is to some point he will win a lot more often than I projected on my original line, but still he is a money burner). I think if you can transition from trying to beat favorites you don't like to beating vastly overbet horses that are 2-1 or higher there is something exploitable, at least at Woodbine Mohawk Harness. The problem is that I am looking at final lines. So if a horse is 2-1 at post time, if he goes down to 8/5 he is a bad play against and if he sits at 2-1 he isn't or vice versa. So you structure you play trying to beat a horse because he is at 2-1 and then bang, 6/5 and you watch him jog.
Horses that are half my line or less are now winning 74% as often as the tote board thinks they will. From 1/2 my line to fair value they do not perform as bad, they win 91.5% as often as the public thinks they should. Now I am going to take a look at the strongest performing aspect of my database, minimal value horses, fair value to 50% above fair value. Without horses under 2-1 they look like this.
fv to 50% above fair value 415 horses 44.36 proj. wins 54-57-65-54
as you can see only 6 horses were eliminated from the above, which shows how seldom you see value on the low end of the tote board. The curse of rebates.
Remember you have to hit about 20% above projected to make break even(crowd loses close to 20%(more like 15% with my dataase because I don't use actual odds) but even at that winning 54 times when you need to win 51 times to break even isn't exactly something to write home about. Also when I hypothetically bring up the odds of fair value horses the results are not nearly as good.
So all of this is a tool that might give you a better understanding of what you are dealing with. For the vast majority beating this game without significant rebates is pretty much a pipe dream. That being said, if you get those rebates, or if you just enjoy the game, do understand that the days of making a line betting value and printing money are long gone. You can make a line you can recognize value and focus on value and do better than if you don't, but beating this game is a tall order without real rebates, not the 2 percent kind.
By the way if anyone has any questions about any of this I am happy to respond and I can look up anything I did not cover in this post if you are interested. I just found this all pretty interesting. No golden ticket, but interesting nonetheless.