tells the whole story Kus?

Almost no one argues that any single statistic can tell the whole story. Those who do are misguided or selling snake oil.

But I think statistics are more than useful. They are vital.

“I don’t understand it” isn’t a valid argument against a stat though. Even though I don’t know why they use 2/3 of an assist, there’s probably a good reason for it. The ones I don’t trust are stats with proprietary calculations. With something like PER where you have the formula, if you were so inclined you could measure PER vs. traditional performance and improve the formula, but that can’t be done with proprietary stats.

If open stats like PER can be improved (which it certainly can) they will be. It’s been happening with baseball advanced metrics for years now.

If you look at baseball, there’s a stat called OPS (On-Base % Plus Slugging %). You calculate it just like it sounds (OBP + SLG = OPS), but that actually puts the emphasis on Slugging %.

Since OBP maxes out at 1.000 (percentage of times reached base safely per plate appearance) and Slugging maxes out at 4.000 (number of bases reached per at bat), it will always be easier to compensate a low OBP with a high OPS.

The debate of the actual values of these stats was made “famous” by Paul DePodesta in Moneyball, who stated a 3:1 value correlation (that is, 3xOBP to 1xSLG), but the actual correlation of value has not been decisively determined (that I know of, at least).

Getting back to PER, this “2/3” factor for assists could be a similar scenario. If “1” or “3/4” skewed all of the data significantly towards PGs, then the results would be even less useful as a metric. Hopefully this helps to explain what Hollinger means when he says that he uses specific numbers “because that’s what works”.

I haven’t looked too closely into PER, and I’m not even remotely educated in advanced statistical measurements, but this was the first thing I thought of when I read your comment, and thought I’d add my two cents.

you would have baba ganoush-ed Vosoin as well.

… we should pick Tyler Hansbrough No. 1, Darren Collison should be No. 2, we should trade loser Kevin Martin for winner Glen Davis and loser Spencer Hawes for winner Fabricio Oberto and loser Jason Thompson for winner Luke Walton.

I don’t mean to be a dick about this, but COME ON.

If I have a team of 9 of ARod in his prime, and you have a team of 9 Jeter in his prime, I’ll beat you WAY more often than I lose.

There’s truly too many variables that come in to play when it comes to rating someone’s performance. I’m sure that someone can come up with more than 30-40 valid variables that should be taken into consideration and that’s just too difficult to do.

I think stats are helpful when they are used for an specific purpose, like getting an idea of how good an offense is, or maybe how good of a rebounder a player is in the defensive end. In those cases, you can use some 5 variables, for example, and get good consistent results.

When stat tries to do too much, like rating how good a player is (very vague question), it just becomes too difficult to understand and its validity is easy to question. Nothing against PER but when Speights is your best rookie and Calvin Booth has a higher rating than Dwight Howard, you know you have a problem.

It’s a slow news week for Kings. Give TZ a break.

but I’ve often felt that people who cling too closely to statistics are the types that have never actually played, and, therefore, don’t have a good understanding of the intangibles involved in sport – especially a team sport. That, of course, is a raging generalization without any metrics to back it ;-)

I work in the commercial real estate development field with finance guys that deal with the cold hard reality of numbers, spreadsheets and statistical analysis on a daily basis. If I had to guess, those guys are about 90% Republican, and hard-core, Fox News-spouting, shit-talking Republicans at that.

I always thought lefties were supposed to think with their hearts, not with their brains – and righties were the corporatist, bottom line numbers oriented thinkers??

I’m so confused now.

but analytical bean counters (of whatever political stripe) if left to run the country will waive the flag with one hand and out-source your job with the other – All in the name of the bottomline.

(no problem)

Cuz you’re a true genius JC. :)

Absolutely? Proud of it? You bet.

"stat geek fans who wear their gym shorts north of their belly buttons "

And you also make solid points. But I love to laugh, so I had to call that out.

Yes, we’ll get so much data, that we’ll just look at the box scores in the morning paper and enjoy the game that way. Eventually, statistics will get so accurate that we won’t even have to play the games. We’ll just look at each players PER’s from preseason and calculate the winning team. No more road trips or the need for Arenas.

BJ is right, statistics will never replace the need for subjective observations. Statistics can’t define a player or a team. Stats identify trends, problems or successes. And, they define where and when these identifers happen, so we know where and when to look to find the causes.

When Hollenger was on Koz’s radio show, he said that most players don’t like the +/-, and that it would take 3 years of data to give the Adjusted +/- any value. And, he commented on PERS lack of defensive relevency. Tools are only as good as the craftmen that use them. You can use a tack hammer to drive in a spike, but I perfer to use a sledge hammer. Using what ever meterics you want to prove your point, doesn’t make you correct. And, your subjective observations don’t make your opinions, more correct than anyone elses.

Only playing the game determines who’s right or wrong, and who Wins. And, sometimes not even then.

Great comment. I think I understand where you’re coming from.

For me, A Too careful immersion in the numbers has, like gambling on games, always taken some of the personality, the emotion out of sports that draws me personally to be a fan.

I don’t want to overstate that. I do study box scores, stats, but as HT exaggerated for effect above, there’s a real danger for me personally when I let the numbers become too important that they diminish the game itself, take away the personal stake I have in the game.

Hmmm. Come to think of it – this year, that wouldn’t have been such a bad idea!