With horse breeding ratios are not going to be that applicable for many characteristics that will be important. Most organs and muscles (including the heart) are going to be the product of many if not thousand of genes and it is not possible to disentangle the effect of individual genes (yet that is, eventually this will not be so. So while characters like yellow and green can be counted with a series of intermediates between two extremes, so there are no distinct groups, these need to be measured.

There are two types of traits:

1.  Traits influenced by one or very few genes -called simply inherited traits and;

2.  Traits influenced by many genes called polygenic traits.

Classic examples of simply inherited traits include coat color and polledness in cattle, whilst reproductive performance, milking and mothering ability and growth traits are polygenic in inheritance. As a rule, simply inherited traits will be of little importance to horse breeders and the more important traits affecting racing ability would be classed as polygenic. There is the possibility though that heart size is affected significantly by one gene and more on this later. For polygenic traits quantitative genetics is important.

Basic observation of quantitative genetics; a cross between a very tall person and a very short one does not give a 3:1 Mendelian ratio, of say 3 tall to 1 short, but a continuous distribution of individuals from one parental extreme to the other. 

Galton (of Galton Laboratory fame) noticed "regression to the mean". Two very tall parents, or two very short certainly had tall or short children, but they were usually closer to the population average than were the parents themselves. Most people are near the average. This is regression to the population mean. There are decreasing numbers that may be tall or short. This can be explained if many loci are involved, giving scope for a large number of genotypes and if there is an interaction between genes and environment. That is, someone with a particular gene will grow tall, but only if given a lot of food. 

Genes are greatly influenced by all sorts of environmental factors. Steve Jones in one of his lectures gives the following example:

Different inbred rat strains show great differences in "intelligence. Bring them up in normal conditions and then run them through a maze to reach a goal. "Bright" strain makes an average of 120 wrong turns, "Dull" strain - 170. However, bring them both up in a "restricted environment" - plain white walls, no excitement - both strains make an average of 170 wrong turns. Bring them up in a very exciting environment and the difference returns, but smaller than before. The bright strain makes 110 wrong turns, dull strain 120. The dull strain in a challenging environment does just as well as the bright strain in a normal environment.

It is meaningless to say that the difference is only genetic or only environmental.

Genotypes may have different Norms of Reaction and the phenotype they produce may depend on the environment in which they are placed.

Phenotype = Genotype + Environment.

Most breeding will be based on looking only at phenotypes, not directly at genes.  Since phenotypes are affected by many loci and also by the environment there will be a continuous distribution from one extreme to the other with most in the middle.  It is worth bearing in mind that while heritability is a ratio a high heritability can be due to shared genes, shared environments, or both.  So most things with a high heritability will not be insulated from the effects of the environment.

Keep the following in mind

A small number of offspring can show results that are more unreliable than a large number

1.  Probability decreases as variables increase

2.  Most important traits affecting horses will include many variables and will be the result  of many genes inter acting together

3.  You can expect most offspring to be average for the gene pool they represent

4.  The type of horse you get will be the result of both genetic and environmental factors