PDA

View Full Version : The stat-less thread- brought to you by Kstat.



Kstat
01-13-2007, 06:45 AM
This thread is for anyone and everyone that is sick and tired of people explaining away every part of basketball with stats. I realize they have a place in the game, but they are not the end-all or be-all of basketball. At its core, I'll always believe that basketball is an art and not a science. You can't just plug in a forumla of stats and churn out a winner 10 times out of 10. There are intangibles, variable you can't account for. Yeah, a point guard might get 13 assists, but how good is he at feeding the post? A SG might score 30, but how good is he with his off-hand? A PF might grab 15 rebounds, but can he set a pick? Can the center who blocks a shot throw an accurate outlet pass after securing the ball? Often, these are the things that decide games, but we never notice because they don't show up in the stat sheet.

Basically, this is a thread where any and all statistical references are banned as they pertain to basketball discussion. I don't say this because I believe they have no place in the game, but I think it will promote people to actually put original thought into their comments about a player, instead of just looking up his stats on NBA.com and pasting them here.

I'm not sure how this thread will turn out, but at least if people try and stay true to it, it will provide an interesting challenge.

Tank31
01-13-2007, 08:37 AM
Yeah, a point guard might get 13 assists, but how good is he at feeding the post? A SG might score 30, but how good is he with his off-hand? A PF might grab 15 rebounds, but can he set a pick? ...

Basically, this is a thread where any and all statistical references are banned as they pertain to basketball discussion.

I'm not sure how this thread will turn out, but at least if people try and stay true to it, it will provide an interesting challenge.


Kstat, I know I'm knew here, but I've been reading the Digest for about 2 years and respect your opinions. But did anyone else find it funny and/or ironic that Kstat used a couple of stats in his opening remarks in a stat-less thread? ;)

Roferr
01-13-2007, 08:40 AM
Stats are kept for a reason. There is a definite use for them. However, true fans don't solely rely on the stats. They know a lot of what you speak of aren't reflected in stats. A prime example; how many extra offensive chances do we get by Jeff tipping the ball out and keeping it alive? True fans know this. As far as the others, you weed out their posts.

indygeezer
01-13-2007, 08:50 AM
This thread is for anyone and everyone that is sick and tired of people explaining away every part of basketball with stats. I realize they have a place in the game, but they are not the end-all or be-all of basketball. At its core, I'll always believe that basketball is an art and not a science. You can't just plug in a forumla of stats and churn out a winner 10 times out of 10. There are intangibles, variable you can't account for. Yeah, a point guard might get 13 assists, but how good is he at feeding the post? A SG might score 30, but how good is he with his off-hand? A PF might grab 15 rebounds, but can he set a pick? Can the center who blocks a shot throw an accurate outlet pass after securing the ball? Often, these are the things that decide games, but we never notice because they don't show up in the stat sheet.

Basically, this is a thread where any and all statistical references are banned as they pertain to basketball discussion. I don't say this because I believe they have no place in the game, but I think it will promote people to actually put original thought into their comments about a player, instead of just looking up his stats on NBA.com and pasting them here.

I'm not sure how this thread will turn out, but at least if people try and stay true to it, it will provide an interesting challenge.


That IS an interesting challenge. I dislike stats anyway, worked with too many statisticians and saw how numbers can be crunched and then smooshed.

Anyway, I thik this is a good place to enter this comment my broker made yesterday (and I completely agree with). He played college basketball and has held Pacer season tix for the 30+ years I've known him. His opinion is that the P's draft or trade too many athletes and not enough basketball players. Players like Bird, West, Havlicek, Workman, Parrish, McHale, Keller, Mount, etc etc would not have been drafted had their teams been looking for the best athletes. Instead they were drafted because they were "gamers"...they just knew how to play and win and they did what it took to make it happen rather than relying on being the fastest or jumping the highest. They played and got themselves into position to be successful, whether it was position for a rebound, the outlet, the screen, the steal, whatever...the used their savvy to succeed at basketball.

We have too many athletes and not enough "players". Whether that is strictly a function of draft position or the team philosophy only TPTB know.


(and it's good to see ya again KSTAT...missed ya)

Quis
01-13-2007, 09:05 AM
As far as basketball is concerned, stats tell us in a very fundamental way things that we would not notice otherwise.

First of all, no one can watch every minute of every game. Second of all, human beings are notoriously bad at seeing patterns just by watching things--I'd wager you can't tell the difference between a 50% shooter and a 47% shooter by watching them over a season, but one is significantly more valuable than the other over that period.

In this way, stats can give us an agreed upon framework for aguing upon: we all know that Dirk scores more ppg than Duncan, now what does this say about one compared to the other, etc...

Stats tell us a lot about what's going on, that we otherwise woudn't pick up. They may not tell us everthing, but that's no reason to disregard them. That's merely a reason for thinking about them more, so we can refine them until they give us more and more information. That's how we've progressed from crude ppg/rpg/apg/fg% to more advanced stats like PER and TS% which, while still nowhere near perfect, take more factors into account and tell us more about the game. These new insights can in turn help us make actual decisions about the game--where to play players, what teams/players are better than others, which players play well with others etc. Stats, contrary to popular belief, don't exist in some made-up fairy land. They measure and have a tangible effect on the game we watch.

Quis
01-13-2007, 09:12 AM
From now on, for every game, they should stop keeping score, and after roughly 48 minutes (or whenever our human spirit feels the game has come to completion), everyone in the arena should vow to take an honest poll on who played the better game. The scoreboard is just full of numbers, but our human spirit tells us who really won the game.

People don't understand context. All numbers do is tell you exactly how many times something happened. What you do with them is your choice.

BlueNGold
01-13-2007, 09:53 AM
I think there is a close correlation between stats and true performance. BUT, they are not perfect because some things really don't show up there, or at least directly, particularly on defense.

For example, when your PG is a poor defender the man he is guarding will get into the lane more easily. This will force your interior defenders to take action, resulting in more shot blocks, charges. This might also create a better chance for 3 pt plays either fouls on the PG, or worse on the poor C who has to cover for the weak perimeter D. Another aspect of poor D is when your PG drives to the bucket for another missed FG attempt and does not get back on D. This might result in a transition basket and momentum change for the opponent. This is particularly true when your PG is a poor shooter.

The problems are not limited to defense.

On offense, you want your PG to make good, mature decisions. You want him to be a leader. He is like the quarterback to your team, making the most critical decisions on offense. You want him to make your other players better, play to their strengths. Get them involved. Avoid the one-on-one game, which directly takes away what your team mates can contribute, particularly when you are a poor shooter.

Dribble penetration is a very good thing, BUT is not good in and of itself. It must be used to score points...even if it is a stat. The PG does not always need to take it to the bucket when he penetrates, particularly if he cannot convert well. Perhaps he could penetrate and dish to another player. If your PG cannot convert well, his dribble penetration is worth nothing...and actually takes away from you because it creates opportunities for the other team to get out on transition. This is particularly true because the quickest player on the other team is often a guard the PG is responsible for guarding...one who will fly down the floor and have to be guarded and maybe fouled by one of our bigs.

Your PG should be able to hit the majority of shots when left completely open. This will result in the ability to space the floor and create better passing angles, more opportunities for assists, more player movement, less likelihood of doubling JO...and obviously a better convert rate by your PG.

You PG should be a deadeye shooter at the foul line, or at least a very good one. He will be on the floor at the end of most games as a ball handler. He might even need to hit important free throws that could win a game against Dallas in the NBA finals someday.

ALF68
01-13-2007, 11:27 AM
This thread is for anyone and everyone that is sick and tired of people explaining away every part of basketball with stats. I realize they have a place in the game, but they are not the end-all or be-all of basketball. At its core, I'll always believe that basketball is an art and not a science. You can't just plug in a forumla of stats and churn out a winner 10 times out of 10. There are intangibles, variable you can't account for. Yeah, a point guard might get 13 assists, but how good is he at feeding the post? A SG might score 30, but how good is he with his off-hand? A PF might grab 15 rebounds, but can he set a pick? Can the center who blocks a shot throw an accurate outlet pass after securing the ball? Often, these are the things that decide games, but we never notice because they don't show up in the stat sheet.

Basically, this is a thread where any and all statistical references are banned as they pertain to basketball discussion. I don't say this because I believe they have no place in the game, but I think it will promote people to actually put original thought into their comments about a player, instead of just looking up his stats on NBA.com and pasting them here.

I'm not sure how this thread will turn out, but at least if people try and stay true to it, it will provide an interesting challenge.

Finally, a topic where I can't be accused of something! I may be wrong, but I don't believe that I have made one single post in which I used stats as a main theme of the post. Stats are just what they are, past history of a team or players PAST performance, now that is not to say that they aren't usefull, quite the contrary. Stats taken over a long enough peroid of time can predict a player or teams future performance based on those stats.


As a fan, I prefer not to get too bogged down with the statistical part of the game, I leave that up to the REAL EXPERTS and not wannabe coaches on a message board. I may be hanging around the wrong type of fans, because none of them talk like some here on this board. Bottom line, some posters get so obsessed with stats, that they are missing the joy of just watching the game for what it is, a GAME.

JayRedd
01-13-2007, 12:13 PM
Stats tell us a lot about what's going on, that we otherwise woudn't pick up. They may not tell us everthing, but that's no reason to disregard them. That's merely a reason for thinking about them more, so we can refine them until they give us more and more information. That's how we've progressed from crude ppg/rpg/apg/fg% to more advanced stats like PER and TS% which, while still nowhere near perfect, take more factors into account and tell us more about the game. These new insights can in turn help us make actual decisions about the game--where to play players, what teams/players are better than others, which players play well with others etc. Stats, contrary to popular belief, don't exist in some made-up fairy land. They measure and have a tangible effect on the game we watch.

First of all, I do really enjoy looking at stats. And while none of them tell us everything, I really enjoy the efforts like those made by Roland Beech over at 82games.com and John Hollinger and all the others working to elevate the way we use numbers to understand basketball.


But there is one fundamental problem with basketball stats to me: They are often complete lies. I don't mean that score-keepers right things down wrong, I just mean that the final tally in a stat category is often not really what happened---or not what we should care about anyway.

This is especially true on defense. Many times in every game, an offensive wing player will try to make a pass only to have another perimeter defender deflect the ball. Then the ball randomly caroms over to a second defender and that defender gets credited with a steal. Well, it was actually the first guy that played the good defense and the second guy just had a ball hit him in the hands.

Of course, no one really cares about defensive stats. But this similarly happens with rebounds all the time. I'll use Pacer players for the example just so its easier. Say we're on D and a perimeter guy takes a jumper and misses. Lets say Jeff and JO both have good inside position on their men for the rebound and are boxing out and waiting for the ball to come off the rim. But out of nowhere, the guy Al was guarding and not boxing out leaps up enough to get a hand on the ball, but can only get enough to try to tap it back to a teammate. But the tap actually goes right to Al. Chalk up a rebound for Al. Yes, he wound up with the ball. But in reality, he was the one front court player who did absolutely nothing to elevate the Pacers chances of securing that rebound.

Things like this happen all game long. Also, assists are easily skewed. Many times, it's the guy that makes the pass to the guy that actually gets the assist that really opened up the spacing on the court and allowed the play to occur. There are certain guys who were masters at doing this---it's like they're playing chess and see two moves ahead. But they never get credit for these "Gretzkys". Larry Bird, Scottie Pippen and Lebron all instantly spring to mind. Mullin and McKey used to do it all the time. Shane Battier was doing it all last night for HTown.

Los Angeles
01-13-2007, 02:17 PM
Well, I am not really a stat fan. I think the assist to turnover ratio of a point guard is a very telling stat. I also get a lot from FG% and FT%. I hate stats like blocks and steals. I really would prefer that the turnover stat be split into "forced" and "unforced" to cover the steals category.

But other than that, I've finally reached the point where the only stat I care about is W-L.

As I get older, I watch basketball with less and less of a analytical eye. I don't really know how many rebounds one guy got vs. another guy, but I can tell who uses the right spacing and has a "nose" for the ball.

Further, there isn't a single decent stat to reveal a player's contribution on the defensive end. For example: towards the end of regulation Danny was doing all the right things in guarding Dirk. Active hands down low to prevent a pass, great foot position to prevent a drive to the basket, fantastic effort. But Dirk quick release a 3 and drained it. What do the stats say about Danny's excellent job in guarding a 7 foot super-star on the perimeter? Dirk N: 3 pts.

Gut instinct is all I have left for guiding my topsy-turvy life right now, it only seems appropriate that it's pretty much all I use anymore when watching basketball. Frankly, I don't have much time for much more than watching the game as it is playing.

Naptown_Seth
01-13-2007, 04:46 PM
but I think it will promote people to actually put original thought into their comments about a player,
You are living in a dream world. The biggest reason stats get brought into battle is to combat the insane level of hyperbole that is associated with fandom, especially on the internet.

By the way, comparitive words like GREATEST or BETTER are a ranking system, and that is itself a stat. For example, "according to KStat player X is the #2 PG in the NBA". Stat. Perhaps not a good one, but still.

We are just swapping out the measuring device. You watch something, you give it some mysterious scoring in your own head (that's good, that's bad, he's smart, he's quick) and start amassing data. Semantically you dismiss it as such by just chalking it up as opinion, as if its formed formed in the netherworld and brought to you in a dream.

It's not. It comes to you just like all stats, from SUMMERIZING the events you witness.

Some people have an aversion to seeing things written down or spelled out explicitly, but that doesn't mean it isn't happening inside their own head all the time anyway.

And what LA descibes is just an example of where someone has yet to keep a written down stat on defensive effort. From what he says you could easily do it though. "Hands down to prevent a pass" - check, "Great (must be defined) foot position to prevent a drive" - check. Okay, put 1 in the "great defensive effort" category for Danny.

Now do it for every defensive effort of every game and you have a stat. Does that observation suddenly become crap because I took the time to add up every time LA (or anyone) thought a player was playing great, good, average, bad or awful defense?

What I'm saying is that LA is doing it already in his own head. Obviously he put a check in his internal "great defense" box for Danny. That will be used to form his opinion of Danny as a defender.


I know it sounds esoteric, but I really think it's much more cold, hard reality. We like to differentiate "feeling" and "truthiness" ;) from the "lies" of stats, but it's not actually different except that by spelling them out you remove the bias where a person might let emotion give more credit to one play than another identical one for whatever reason.

In stats a FG is a FG, but emotionally it can be "I like Danny, he made a FG proving he is good" vs "I hate Jackson, sure he made a FG but it was another bad shot like he always takes".

Stats say "what's the difference?"

Just don't confuse weak stats or ambigious stats with the idea that keeping record and account of your opinion is unproductive. Basic stats DON'T TELL THE WHOLE STORY. I totally agree. But that's a weakness with the stat itself, not the concept of statistical record.



BTW - great post by Quis.

Kstat
01-13-2007, 06:10 PM
I never said stats were totally meaningless.

THe point of this thread was to take the mindless stat-whoring out of the equation, and force people to actually articulate as it pertains to a particular player.

And yes, absolutely i'm changing the measuring device. That's the challenge. Support your opinions WITHOUT the use of numbers. It helps people to see things they didn't see before.

If you don't like it, you don't have to post in this thead. Nobody's forcing you.

Will Galen
01-13-2007, 06:17 PM
What gets me is a guy that uses stat in his user name is tired of stats!

(giggle,snort)

Kegboy
01-13-2007, 10:13 PM
Did this make anyone else think of the great Dajuan Wagner / Chauncey Billups debate, or is it just me? :chuckle:

Seriously, anybody who remembers Writerman can tell you the use of statistics alone does not make a good argument. And anyone who remembers my "Mehmet Okur = Austin Croshere" thread knows that stats do not predict future performance.

NPFII
01-14-2007, 09:18 AM
I love the idea in this thread, as I see it more as "Which stats would YOU like to keep?" thread.

Of course, the regular points, rebs, assts, fg%, ft%, blocks, +/- etc. are good stats. But as mentioned - there seems to be a feeling that there could be so many more stats, and that they are missing.

By the way - although I like 82games.com & Hollinger, they haven't invented any new stats - they just accumulate & crunch differently, but their basis are the same old stats that we all know.

So here's my short list of stats I'd like to see being kept on a regular basis:
Defence:
1. Defensive assists. I know coaches keep this stat individually. Count in deflections, a trapper that cuased a TO, charges taken, etc.
2. Defensive points allowed. The number of points the man you're guarding scores directly on you, including FTs by your fouls. If there's a switch on D then the new defender will recieve these "bad" points. If it's a zone then it goes to the defender who got broken down (if any).
3. Defensive pressence. How many of your man's shots were contested/uncontested vs made/missed. How many were fouled. How many times were you picked off vs broke the screen.

Offense:
1. Offensive decisiveness. How many seconds a player holds the ball before passing/shooting.
2. Screen efficiency. How many times were your screens effective/broken
3. Secondary assists. How many times did your play allow the made basket although you weren't the shooter nor the assister. For instance the back-screener up top for the alley-oop, or the effective screener (see 2) who allowed for the open made shot, or as mentioned before, the passer who made the extra pass to the assister.

And those are just off the top of my head. I'm sure there could be many more.
I really wish there was an automated computer application that counts whatever you define it to count, so that anyone could make up his own "stat-definitions" and get results. Anyone up for the task?

JayRedd
01-14-2007, 02:40 PM
I love the idea in this thread, as I see it more as "Which stats would YOU like to keep?" thread.

Rebound rate is one of the most important stats that still isn't mainstream, IMO. Essentially, it measures the % of rebounds a player gets out of the total number of rebounds that occurred while he was on the court.

Rebound Rate = (Rebounds x Team Minutes) divided by [Player Minutes x (Team Rebounds + Opponent Rebounds)]

http://insider.espn.go.com/nba/hollinger/statistics?sort=reb&qual=true&pos=all


Effective Field Goal % is another good one that factors in the added value of the three-point shot. True Shooting % takes this another step further and adds in free-throw shooting as well, but my personal preference is to keep free-throw shooting separate from what happens from the field.


Not sure why these two stats don't get more play.

Hicks
01-14-2007, 06:00 PM
Somehow, I don't think it was Kstat's intention for this to be a "let's debate which stats are best" thread, given the fact that it explicitly said it was about discussing the game without the use of stats.....

but I could be wrong. :shrug:

Kstat
01-14-2007, 06:02 PM
Yeah, I thought it would be challenging and original to make people think outside the box for once.

But like I said before, I think some people really have a lot of difficulty talking basketball without the use of stats.

BlueNGold
01-14-2007, 07:28 PM
Yeah, I thought it would be challenging and original to make people think outside the box for once.

But like I said before, I think some people really have a lot of difficulty talking basketball without the use of stats.

I agree that it is more challenging to make a point without stats. It takes more time and thought, something not all of us have time to do. It also takes more knowledge of the game.

I use stats pretty often because I'm a numbers guy and think it is the most objective measure out there. I also don't consider myself John Wooden. Reality is too complicated IMO even for some of the best basketball minds. IMO, if Dallas is using WinVal, stats are good enough for me.

Israfan
01-15-2007, 04:38 AM
As an experimentator in physics/chemistry I must say that everything is art, not just basketball. In many purely scientific fields there is no adequate model or as Kstat said you cant "churn out a winner 10 times out of 10".

However, many multi-variable models for things such as stock market have being successful to some extent - people use them. So why shouldn't we use models and statistics in basketball?