Dosage and Breeding In the Game
This article is based on a series of e-mail exchanges between myself and Mike Wallace, the game’s creator. Portions of it are reprinted here with his permission. It is intended to promote better play and understanding of the game.
A lot of what is written isn’t news for many of us as I wanted to ensure that newer players could get something out of this as well.
Each point, unless stated otherwise, is taken directly from Mike’s explanation of the game. Any opinions by myself are clearly stated.
Before we get started with what’s in this article, let’s discuss what it’s not. This is not a “Hey, I broke Mike’s code!” article. What I originally started was a study to see if dosage calculations that are used in real life are present in the game. I used as a case study the top ten male horses in the turf long, sprint dirt and classic dirt divisions. I calculated their dosage exactly the way it’s done in real life by using all 16 sires in the 5x pedigree. (the game only uses 15 if you breed from scratch) It was only after I had done this and came up with all the charts and graphs that I decided to see if Mike would answer an email with a couple of questions in it. What I got in return was almost an entire article in itself as Mike went above and beyond what I had asked for.
This new information completely took my work in a different direction as I was what I had originally hypothesized was killed with Mike’s explanation of the game. However, the results of my study are still stunningly realistic and I wish to share both Mike’s view and my own.
I hope you enjoy it and feel free to continue this discussion on the message board.
Dosage
First, we need to separate truth from fiction: Chef-de-Race dosage as it’s calculated for real horses does not exist in the game. What Mike uses instead is historical racing data such as average winning distance, high and low winning distances and the variances across them. Although it’s hard for the average person to find historical racing past performances for free, (There are a few sites that you pay for that will have the information such as Equibase) many sites like www.bloodhorse.com have them or at least the notable races.
When asked about dosage Mike states, “I believe in the theory of dosage, but not necessarily in the Dosage Index.” The concept to him is pretty simple: Distance runners are more likely to produce distance offspring than non-distance runners. Mike admits that DI works to some degree but is too much a simplistic approach. He adds, “I completely ignore DI when using horses in the game.” Instead he relies on the above mentioned racing data for determining what distances a sire might be attributed to. By using that data he believes that the player can get a much better idea of the usefulness of a sire rather than a “Well, he’s a Chef-de-Race in this or that”.
However with that being said, it’s my opinion that because Chef-de-Race sires seem to produce like runners, the data is strangely similar. When you take the 5x pedigree of the virtual horses used in the study and compare their dosage index to runners in real life you get almost similar data. Is this mere coincidence or is there more realism here than we first thought?
What can be assumed is while Dosage Index is not present in the game’s coding; there is some reliable data that can be utilized in Chef-de-Race dosage. Again, if great sires produce similar running style offspring, then what you can conclude is that sires who are not Chef-de-Race are still similar to their parents.
I probed Mike further on the similarities and he added, “I would expect the way I calculate a horse’s potential to have a lot of correlation with the Chef-de-Race method. Although I don’t use DI directly, it’s really kind of implied as they are trying to measure the same thing: the distance a pedigree should run”. He further explained that the main reason he shies away from DI is the fact that there are a very limited amount of Chef-de-Race sires, and the player has to rely on what he has. “I try to rate each sire in a pedigree individually”. What he concluded with was the fact that DI ignores non-Chefs all together. His calculations take a lot less into “assumptions” as the game has thousands of sire ratings versus only 205 Chef-de-Race sires that are simply lumped into a handful of categories.
Taking Mike’s comments to heart, the best way to decide what type of runner you have is to review each sires historical racing data but this may take some time especially if it’s not one of the popular sires in the game. Also, the game only uses the sires that are on the horse page and leaves off many 4th generation sires but includes two 5th generation sires. Even so, you can still remain fairly certain that the Dosage Index can still you give you some indication of what distance your horse can best perform at. This is especially true when a Chef-de-Race is very close up in pedigree. Keep in mind though that virtual horses and their real life counter parts will “under-perform” or “over-perform” their pedigree.
Breeding
There are two way to breed a new horse: From scratch or breeding to a game mare/he-mare. If breeding from scratch, the horse uses only the sires listed on the horse’s main page. “It uses them to varying degrees but the closer in pedigree a sire is, the more it’s utilized so it does pick up some 5th generation attributes but very very little.”, Mike said. What was a surprise to me is the fact that the dam-sire is “used very heavily” in breeding from scratch. Although percentage of use can vary, “the dam-sire in scratch breeding can be every bit as important as the sire”. This of course explains the BP cost structure in scratch breeding and also the many breeding mistakes I’ve made in an attempt to save BP cost such as: top sire x bargain sire x top sire type breeding. Sadler’s Oasis, who I bred, is probably a very good example. In my best opinion, I would suggest picking two average priced sires plus a good DDS as opposed to breeding a premier sire to a bargain sire combination. You could probably have either for the same amount of BP’s/credits.
When breeding to a game dam, the game takes on a new perspective. Mike varies how much to use “pedigree” vs “performance. Some mares pass their personal attribute readily, while others tend to pass their pedigree attributes. Most pass both to varying degrees. In this case, the outer pedigrees are getting used a little more since they are already built into the mare. In my opinion, one of the advantages of breeding this way, if you plan to stay in the game for a long time to come, is that you have a set of breeding pedigrees to work with and that those pedigrees are consistent. You have the opportunity each year to play with different sires to see what result you can get. Of course one of the best resources is Yoss’s site where you can sample different sire combinations to see how they’ve ran in the past. This is true for scratch breeding as well. I encourage you to play with unique sire combinations as I think the game sometimes rewards this.
One of the popular breeding theories is inbreeding. This is breeding the same sire line in both the Sire and Dam-Sire sides. Mike isn’t as hard on inbreeding as most people thinks he should be. His reasons for this are simple as he states, “There are just not a lot of hard concrete stats about inbreeding as they relate to horses that actually make it to the track.” He goes on to point out that every horse bred is a “runner” although I think that term is loosely defined especially when it comes to my horses. There are examples of horses in real life that have performed well. Mike gave five time Graded winner Skimming as an example, who is 2×3 inbred to Northern Dancer. He admits though that he has been a bit unkind on the “Super Inbred” (2×2, 2×3) lately. The argument could be made that it’s not realistic but he states “As far as the simulation goes, there aren’t enough real-life horse racing examples out there to make a case either way”. For that reason he has to “guess” to some degree. In this case he says that it’s better to “guess” in favor of the player, which is not to punish them for trying out things.
Finally the question remains, what surface is my horse best suited for? Surface can be a bit tricky to figure out especially if a sire never races on one or the other. Mike explained, “In the case of newer sires, I have to pull from the horse’s pedigree and make guesses on my own”. A good example he gave was Smarty Jones who never touched the grass. However he states, “As good as he was on dirt, his pedigree “looks” like he could produce turfers….I have to assume that his offspring can handle it to some degree”. Sadler’s Wells immediately came to mind as another example of a horse that in real life produces turf runners almost exclusively but in the game can produce either type of runner to the same degree of performance. Again, the best source might come from Yoss’s site or sire statistics in the game. This works out better for sires that have plenty of offspring and have been in the game for a while as opposed to new sires. On a side note, he did mention that at this year’s gathering, many players may have heard him speak of Polytrack, a synthetic racing surface which is considering being used at some tracks. “Who knows what sires will produce polytrack lovers? Once we get some good stats in a few years, I’ll have to start adding a Polytrack rating to the sires I guess and make some polytrack courses”.
In all, it seems that many of the tools we have out there already are capable of measuring distance and surface preferences. What is missing is the ability to calculate it on a consistent basis for all sires with one method to measure it all. Perhaps someone with more knowledge in the coding of the game can come up with this.
He left me with one note that I felt obliged to pass along. He said, “The game benefits more from players and the things they add to the game than it does me. I just do what I can to keep the wheels from falling off”.
This article will be permantly housed on my website at: http://maverickstables.awardspace.com/index.htm
However, what is there now is the rough draft so give me a day or two post the final copy.