Stats versus reality

wigglestrue · Post #1 » by **wigglestrue** » Tue Apr 7, 2015 6:11 pm

From ESPN:

"Overall, Boston is expected to win about 1.7 of its final 5 games (34 percent), tied for the fourth-lowest projected win percentage in the NBA."

"Overall, Miami is expected to win about 2.9 of its final 5 games (58 percent), the 11th-best projected win percentage in the NBA and 1.2 more projected wins than the Celtics (who currently hold a 1-game lead)."

Which is realer, the majority of more probable outcomes expected by statistics that just might not wind up happening, or the possibility of an improbable outcome that defies statistical expectations and actually happens?

giberish · Post #2 » by **giberish** » Thu Apr 9, 2015 5:19 am

I'm sure they could come out with a better summary. Like
5-0 : XX%
4-1 : XX%
3-2 : XX%
2-3 : XX%
1-4 : XX%
0-5 : XX%

That would really give a better idea of likely outcomes. Of course - especially at the end of the season - you don't really know how strong of a team is going to be out on the floor for a given game.

ohio · Post #3 » by **ohio** » Thu Apr 16, 2015 6:48 pm

people with perfect ways to calculate sport stats either a. don't exist. b. are keeping it to themselves for betting reasons.

It's a lot like investing, you can learn a lot that increase your probability of making money, but there's individuals/organizations that haven't lost money in decades with their picks due to the formula they go by.

Joao Saraiva · Post #4 » by **Joao Saraiva** » Fri Apr 17, 2015 8:03 pm

Well, Cleveland gave away 2 games against Boston, so there go the predictions.

Post #5 » by **Quotatious** » Wed Apr 22, 2015 11:21 am

Honestly, I think that stats do a much better job when we use them to evaluate what actually happened, not to predict what will happen in the future.

Anyway, Joao Saraiva is right - Cavs clearly didn't care about winning those two RS games against Boston.