Comments on: The Weekly Nichols: What the Regular Season Can Tell Us http://www.hardwoodparoxysm.com/2009/06/04/the-weekly-nichols-what-the-regular-season-can-tell-us/ Unbiased opinions from extremely biased people Mon, 21 May 2012 13:40:43 +0000 hourly 1 http://wordpress.org/?v=3.3.2 By: Brizzle http://www.hardwoodparoxysm.com/2009/06/04/the-weekly-nichols-what-the-regular-season-can-tell-us/comment-page-1/#comment-11391 Brizzle Sat, 06 Jun 2009 08:26:05 +0000 http://www.hardwoodparoxysm.com/?p=2813#comment-11391 "Regular season results could be skewed by injuries or odd circumstances." and the fact that the Lakers only give a full effort against the big name teams. Orlando was not. “Regular season results could be skewed by injuries or odd circumstances.”

and the fact that the Lakers only give a full effort against the big name teams.

Orlando was not.

]]> By: Mo http://www.hardwoodparoxysm.com/2009/06/04/the-weekly-nichols-what-the-regular-season-can-tell-us/comment-page-1/#comment-11379 Mo Fri, 05 Jun 2009 19:21:46 +0000 http://www.hardwoodparoxysm.com/?p=2813#comment-11379 Jon, Thanks for the response and good work. I'm always wary of questionable statistical analysis and it looks like this is not the case. No offense to you or your work, I'm usually wary if no mention of a p-value or significance is made. Good point re: binary vs net. Thanks again and great analysis. Jon,
Thanks for the response and good work. I’m always wary of questionable statistical analysis and it looks like this is not the case. No offense to you or your work, I’m usually wary if no mention of a p-value or significance is made.

Good point re: binary vs net. Thanks again and great analysis.

]]> By: Jon Nichols http://www.hardwoodparoxysm.com/2009/06/04/the-weekly-nichols-what-the-regular-season-can-tell-us/comment-page-1/#comment-11336 Jon Nichols Thu, 04 Jun 2009 18:29:54 +0000 http://www.hardwoodparoxysm.com/?p=2813#comment-11336 "What’s the p-value of those two tests?" The p-value of the first one is .0139. The p-value of the second one is .0079. "It looks like inter-conference series losers, the only ones that can have 2-0 results, the loser of the series wins 1/3 times. One would think that inter-conference series have less predictive value than intra-conference series due to there being only two data points. " The small sample size is indeed a concern for the inter-conference (Finals) series. That's definitely one of the cautions you have to have when looking at the data. "Finally, isn’t treating the win-loss of a series as the delta of wins misleading? Shouldn’t the Y be a binary variable, few care if the series ends in 4 or 7 as long as they win." I chose not to do a binary variable because I think there is some analytical value to how many games a series was won in. Obviously the end goal is just to win, no matter how many games it takes. But a 4-game sweep is a lot more convincing than a 7-game battle because the latter can sometimes be decided by a few breaks going either way. In a four game series, it is much clearer who the better team is. “What’s the p-value of those two tests?”

The p-value of the first one is .0139. The p-value of the second one is .0079.

“It looks like inter-conference series losers, the only ones that can have 2-0 results, the loser of the series wins 1/3 times. One would think that inter-conference series have less predictive value than intra-conference series due to there being only two data points. ”

The small sample size is indeed a concern for the inter-conference (Finals) series. That’s definitely one of the cautions you have to have when looking at the data.

“Finally, isn’t treating the win-loss of a series as the delta of wins misleading? Shouldn’t the Y be a binary variable, few care if the series ends in 4 or 7 as long as they win.”

I chose not to do a binary variable because I think there is some analytical value to how many games a series was won in. Obviously the end goal is just to win, no matter how many games it takes. But a 4-game sweep is a lot more convincing than a 7-game battle because the latter can sometimes be decided by a few breaks going either way. In a four game series, it is much clearer who the better team is.

]]> By: matt the jazz fan http://www.hardwoodparoxysm.com/2009/06/04/the-weekly-nichols-what-the-regular-season-can-tell-us/comment-page-1/#comment-11335 matt the jazz fan Thu, 04 Jun 2009 16:59:39 +0000 http://www.hardwoodparoxysm.com/?p=2813#comment-11335 interesting, but r² of 0.15 (or anything below 0.9) are usually about as reliable in predicting anything as... say, cardboard chairs. great website, though. will become a daily read! interesting, but r² of 0.15 (or anything below 0.9) are usually about as reliable in predicting anything as… say, cardboard chairs.

great website, though. will become a daily read!

]]> By: em-k http://www.hardwoodparoxysm.com/2009/06/04/the-weekly-nichols-what-the-regular-season-can-tell-us/comment-page-1/#comment-11333 em-k Thu, 04 Jun 2009 16:28:30 +0000 http://www.hardwoodparoxysm.com/?p=2813#comment-11333 Nice... thanks for doing this. There are a lot of Lakers fans out there displaying a little too much confidence. I'm a Lakers fan and I believe they can win, but I sort of wish everyone had picked against them. It's weird... last year they were the upstarts who beat the West and they got picked over a team with 1 certain and 2 possible (probable?) Hall of Famers and lots of veteran experience, not to mention a ridiculous regular season record. They had been swept in the season series v. the Celtics. Now this year they are the experienced team who everyone expected to be here, with a great regular season record, but they got swept in the season series, and everyone is picking them again. I don't get it how these "experts" do this. But it should be a great series. Nice… thanks for doing this. There are a lot of Lakers fans out there displaying a little too much confidence. I’m a Lakers fan and I believe they can win, but I sort of wish everyone had picked against them. It’s weird… last year they were the upstarts who beat the West and they got picked over a team with 1 certain and 2 possible (probable?) Hall of Famers and lots of veteran experience, not to mention a ridiculous regular season record. They had been swept in the season series v. the Celtics. Now this year they are the experienced team who everyone expected to be here, with a great regular season record, but they got swept in the season series, and everyone is picking them again. I don’t get it how these “experts” do this. But it should be a great series.

]]> By: Mo http://www.hardwoodparoxysm.com/2009/06/04/the-weekly-nichols-what-the-regular-season-can-tell-us/comment-page-1/#comment-11326 Mo Thu, 04 Jun 2009 15:07:39 +0000 http://www.hardwoodparoxysm.com/?p=2813#comment-11326 What's the p-value of those two tests? It looks like inter-conference series losers, the only ones that can have 2-0 results, the loser of the series wins 1/3 times. One would think that inter-conference series have less predictive value than intra-conference series due to there being only two data points. Finally, isn't treating the win-loss of a series as the delta of wins misleading? Shouldn't the Y be a binary variable, few care if the series ends in 4 or 7 as long as they win. What’s the p-value of those two tests? It looks like inter-conference series losers, the only ones that can have 2-0 results, the loser of the series wins 1/3 times. One would think that inter-conference series have less predictive value than intra-conference series due to there being only two data points. Finally, isn’t treating the win-loss of a series as the delta of wins misleading? Shouldn’t the Y be a binary variable, few care if the series ends in 4 or 7 as long as they win.

]]>