BGonline.org Forums

Kit's source data

Posted By: Tom Keith
Date: Sunday, 28 March 2010, at 4:13 p.m.

In Response To: Kit's source data (Timothy Chow)

For the purposes of constructing a MET, I don't see why the concept of cubeless gammons is relevant. All that is needed, in principle anyway, is the actual gammon rate and not the cubeless gammon rate.

You're right. All you need is actual results -- if you have enough data.

From Chuck's post, it sound's like Kit had only about 3000 games worth of data to work from. That's not enough to produce a smooth table. So Kit used a program he had already written and adjusted the parameters of that program to make the results more closely mirror the results of the human games. Presumably one of the parameters of Kit's program was cubeless gammon rate.

Messages In This Thread

• Rockwell/Kazaross versus Woolsey
  Ray Kershaw -- Saturday, 27 March 2010, at 10:28 a.m.
  • Rockwell/Kazaross versus Woolsey
    Phil Simborg -- Saturday, 27 March 2010, at 11:55 a.m.
  • Kit's source data
    mfic -- Sunday, 28 March 2010, at 2:52 a.m.
    • Kit's source data
      Rich Munitz -- Sunday, 28 March 2010, at 5:22 a.m.
      • Kit's source data
        Chase -- Sunday, 28 March 2010, at 5:56 a.m.
        • date, reference, more info
          Chuck Bower -- Sunday, 28 March 2010, at 1:39 p.m.
      • Kit's source data
        Timothy Chow -- Sunday, 28 March 2010, at 2:32 p.m.
        • Kit's source data
          Tom Keith -- Sunday, 28 March 2010, at 4:13 p.m.
          • Kit's source data
            Daniel Murphy -- Sunday, 28 March 2010, at 5:20 p.m.