只要文本格式良好，并且总是以相同的顺序打印出来，我根本不会使用正则表达式，我会逐行使用函数解析文本。

class CallCenterActivity
{
    public CallCenterActivity(string callActivity)
    {
        AgentStistics = new List<AgentStatistic>();

        using(var reader = new StringReader(callActivity))
        {
             ActivityDate = ParseActivityDate(reader.ReadLine());
             MaxAgents = ExtractInt(reader);
             MinAgents = ExtractInt(reader);
             //(Snip)
             AvarageTimeBeforeAbandon = ExtractInt(reader);

             if(reader.ReadLine().Trim().Equals("Per agent statistics:") == false)
                 throw new InvalidDataException("We where not on the line we expected to be for \"Per Agent statistics:\"");

             string currentLine;
             //This loops till we break out of the agent section
             while((currentLine = reader.ReadLine()).Trim().Equals("Queue related statistics:") == false)
             {
                  var agent = new AgentStatistic();
                  agent.AgentId = ExtractInt(reader);
                  agent.DirectCallsAnswered = ExtractInt(reader);
                  //(snip)
                  agent.QueueLongestTimeInCall = ExtractInt(reader);

                  AgentStistics.Add(agent);
             }

             TotalCallsPresentedToQueue = ExtractInt(reader);
             //(Snip)
             CallsAnsweredByVoiceMail = ExtractInt(reader);
        }

    }

    //These parser methods are small and kept static so you could easily write unit tests against each parser.
    private static DateTime ParseActivityDate(string activityDateLine)
    {
        throw new NotImplmentedException("Here you would turn your \"Fri 11:00 - 12:00\" in to a DateTime");
    }

    //ParseInt and ExtractInt are separated to ease Unit Testing.
    private static int ParseInt(string line)
    {
        var split = line.Split(':')
        return Int32.Parse(split[1]);
    }

    private static int ExtractInt(StringReader reader)
    {
        return ParseInt(reader.ReadLine());
    }

    public DateTime ActivityDate {get;set;}
    public int MaxAgents {get;set;}
    public int MinAgents {get;set;}
    public int TotalCalls {get;set;}
    public int AnsweredCalls {get;set;}
    public int AbandonedCalls {get;set;}
    public int AvarageTimeToAnswer {get;set;}
    public int LongestTimeToAnswer {get;set;}
    public int AvarageTimeBeforeAbandon {get;set;}
    public List<AgentStatistic> AgentStistics {get; private set;}
    public int TotalCallsPresentedToQueue {get;set;}
    public int CallsAnsweredByAgents {get;set;}
    public int NumberOfCallsInTheQueue {get;set;}
    public int AvarageTimeToAnswerQueue {get;set;}
    public int LongestTimeToAnswerQueue {get;set;}
    public int NumberOfAbandondCalls {get;set;}
    public int AvarageTimeBeforeAbandon {get;set;}
    public int CallsForwaredToVoiceMail {get;set;}
    public int CallsAnsweredByVoiceMail {get;set;}
}

class AgentStatistic
{
    public int AgentId {get;set;}
    public int DirectCallsAnswered {get;set;}
    public int DirectCallsAverageTimeInCall {get;set;}
    public int DirectCallsLongestTimeInCall {get;set;}
    public int QueueAnswered {get;set;}
    public int QueueAverageTimeInCall {get;set;}
    public int QueueLongestTimeInCall {get;set;}
}

不知道你想做什么，但这里是如何使用Regex解析字符串的部分。

Regex ex = new Regex("[0-9]*"); //generic regular expression -- look for numbers
Match match = ex.Match(searchString, startAtIndex);
while(match.Success)
{
    string currentMatch = match.Value;

     //do something

    match = match.NextMatch()
 }

你在正则表达式中使用了很多.，我建议你避免使用它。相反，通过匹配 [任何不是我**感兴趣的] 来匹配一大组你不感兴趣的字符。在这种情况下，你想要数字，对吗？
我会这样做（\D与大写D的意思是[任何不是数字的]）：

Agent:\s+(?<AgentNum>\d+)(?:\D+)(?<DirectTotal>\d+)(?:\D+)(?<DirectAvg>\d+)(?:\D+)(?<DirectLongest>\d+)(?:\D+)(?<QueueTotal>\d+)(?:\D+)(?<QueueAvg>\d+)(?:\D+)(?<QueueLongest>\d+)

我假设数据集总是像你的例子一样格式化，行顺序永远不会改变，数字也不会省略（而是设置为0）。
我没有C#或任何东西来测试我是否正确使用了(?<NameSubgroup>matchPattern)，我也不熟悉C#必须迭代匹配并提取命名子组的方法，但这应该会让你走上正确的道路。
(?:\D+)中的?:确保非数字组不会被捕获到我使用的语言中的子组中。