# Flipping a coin example • ### Question

• I looked at the example given in the user guide (A simple example), and extended it to try and further my understanding of Infer.Net. I am a beginner when it comes to statistics, so if this is an absolute newbie question, my apologies.

```Variable<bool> firstCoin = Variable.Bernoulli(0.5).Named("firstCoin");
Variable<bool> secondCoin = Variable.Bernoulli(0.5).Named("secondCoin");
Variable<bool> thirdCoin = Variable.Bernoulli(0.5).Named("thirdCoin");
Variable<bool> fourthCoin = Variable.Bernoulli(0.5).Named("fourthCoin");

Variable<bool> fourHeads = (firstCoin & secondCoin & thirdCoin & fourthCoin).Named("fourHeads");
Variable<bool> fourTails = (!firstCoin & !secondCoin & !thirdCoin & !fourthCoin).Named("fourTails");

InferenceEngine ie = newInferenceEngine()
{
ShowFactorGraph = true,
NumberOfIterations = 1000
};

Console.WriteLine("Probability four heads" + ie.Infer(fourHeads));
Console.WriteLine("Probability four tails" + ie.Infer(fourTails));
Console.WriteLine("Probability four same" + ie.Infer(fourHeads | fourTails));Output:Compiling model...done.Probability four headsBernoulli(0.0625)Probability four tailsBernoulli(0.0625)Compiling model...done.Probability four sameBernoulli(0.1211)If I look at the FactorGraphs, they look like I would expect. Also, the result for four head and four tales (1/2^4) - I understand.What I don't understand is the or ('|') of the heads and tales. How does it get to 0.1211 instead of 0.0625+0.0625 = 0.125?Thank you. ```

Friday, April 5, 2013 5:53 PM

### All replies

• This because the inference is approximate.  See the end of the FAQ, and forum threads here and here.
Friday, April 5, 2013 9:47 PM