Pythagorean Expectation

**Noleafclover** · 07-18-13, 08:01 PM

After two heads, should I expect two tails because I had a degree of expectation of an even distribution from the outside?

**matthew919** · 07-18-13, 08:09 PM

Originally posted by Noleafclover

After two heads, should I expect two tails because I had a degree of expectation of an even distribution from the outside?

This is not what he's getting at smartypants. Pythagorean expectation will tell you (in theory) which teams have inflated records due to "luck." I.e. a team with a negative run differential that is playing well over .500 in the first half would be expected to regress in the second half, because eventually a team that gives up more runs than they score will lose more games than they win.

So (in theory) PE will predict which teams should improve or regress, record wise. I've never seriously attempted using this approach, but for futures like win totals it is certainly a reasonable place to start. Don't expect it to be that simple though- be prepared to get creative.

**Noleafclover** · 07-19-13, 12:02 AM

My bad, I had it confused with a different theory used for futures predictions at the start of a season. Not my style, but sounds to at least have a basis in reason.

If I were to do something like that, I'd want to control out some of the input in blowout losses or wins.... at least in other sports. I don't really know baseball in the slightest.

**Mr_Bad_Guy** · 07-19-13, 02:11 AM

Originally posted by million2one

Now that we are midseason in MLB.

Is is safe to say that teams whose actual win percentage is greater than their expected win percentage will do worse in the second half and teams whose actual win percentage is less than their expected win percentage will do better in the second half.

For example the Phillies are at .500 but there expected win percentage (as determined by the runs scored to runs allowed ratio) is .443, so will they perform less than .500 in the second half?

Anyone ever try this?
I am thinking of trying this to place some futures.

Thanks

The performances of the 2nd half are really indipendent from th 1st half.

You can verify this using the stats from the previous years.

**Nick@SI** · 07-20-13, 01:34 AM

I think basing it off just runs scored and runs allowed is to basic. If was that easy would be filled with winning bettors need to get to the root of the talent and figure out how many runs should have been scored and allowed maybe have chance at something that way. While performance maybe independent overall skill to do so isn't nearly as much so.