What's new

Stats--which are the best guide?

Gyp Rosetti

Banned
There are a ton of different stats out there. For instance, GVC made a reference to TS% in another thread recently. And it got me wondering which stats are the best guide to measure a player? I suppose most stats simply measure one's offensive game. So perhaps if you can break it into two halves, offensive stats, and then offensive/defensive stats (like +/-), this would be better...Heck, if you feel they need to be broken into even more (sub)categories, take it from there. I ask because I'm truly curious and would like to learn more about them. Here are a bunch. I ask that you rank them in order, from most important to least important, for both categories, and then also comment as to why you value those stats. Hell, if someone could even post next to each stat what/how it measures a player, I'd appreciate it. If I forget any, just tell me, and I'll edit this post and add them in. So here's the list in no particular order.


FG%
TS%
PER
USG
VA
EWA
On court/Off court/Net
+/-
Production (Own/Opp/Net)
Simple Rating
eFG%
iFG%
Win%
 
None. Stats act as a way to check our judgement, but none of them are good guides in and of themselves.
 
You forgot Bagadonuts Quotient (BQ). It's the best one. You might as well just forget about the rest. To compute it all you do is take the lineup of the team that wins the championship and rate them as the best of the 30 combinations for that particular year, and then you go on down the line until you rate the players on the worst team as the worst. The result is a perfectly accurate retroactive rating. The ones you have all try to look into the future by imagining that it will be the same as the past and history has shown that simply doesn't work. Go ahead and edit your post.
 
I would be interested to see what player on the Jazz has the best net stat of himself vs his opposing player night in - night out. Scoring, rebounds, assists. I have a guess who that would be.
Of course while people will say that doesnt matter (I dont know how it doesnt matter), if each player duplicated that feat themselves, win.
Hard to say what stat is best, but I'd side with Production (Own/Opp/Net)
 
any stat only paints a partial picture, so it just depends what part of a player/team you want to put under the microscope.

for overall type of measures, PER is generally accepted as the least incomplete stat, and lately i've been putting more stock in the win shares and WS/48 stats, especially when it comes to differentiating role players from true difference makers.

but both PER and WS skew offensive, so then you have to balance what you know about a player defensively. blocks/steals alone aren't a great indicator of defense, so sometime you have to look at floortime stats and how a team performs defensively with that player out there... but then you're not isolating that variable, because the defense of four other guys on the floor at the same time can drive that player's floortime stats up or down.

so in other words, just as important as selecting the right stat is understanding that stat's drawbacks and what it DOESN'T tell you.
 
I would be interested to see what player on the Jazz has the best net stat of himself vs his opposing player night in - night out. Scoring, rebounds, assists. I have a guess who that would be.
Of course while people will say that doesnt matter (I dont know how it doesnt matter), if each player duplicated that feat themselves, win.
Hard to say what stat is best, but I'd side with Production (Own/Opp/Net)

i actually don't love net production, for a variety of reasons. first of all, it punishes a player who plays at a strong position. if you're a PF at a time when the western conference is loaded with talented PFs, your net production is going to be skewed.

also, net production doesn't take into account all the cross-matching and team defensive schemes. for example, when the jazz play lebron, they'll try a number of tactics to slow him down: they'll switch on screens, they'll gold on screens, they'll guard him with a bigger player, guard him with a quicker player, trap him at the elbow, trap him at the post, double team him off the ball, show the double team and then retreat ... in short, a couple dozen different combinations and defensive looks ... it seems wrong to me that one player get all the defensive credit for succeeding/failing at slowing him down. net production ignores the fact that defense is a collective effort.
 
It also penalizes players that are smarter than to think the game is just them versus their matchup and not a team game.
 
It also penalizes players that are smarter than to think the game is just them versus their matchup and not a team game.

exactly. the last thing you want is guys not helping off their guy to help slow a superstar because they're worried that it will make their C-vs-C or SF-vs-SF net production look bad.
 
I try to look at them all to get as much information as possible. Unfortunately, basketball, like soccer, is a true team sport, where virtually unquantifiable team interactions make a huge difference. I'd love to have access to sportVU and synergy sports data.
 
Defense statistics are hard to gauge, especially individually.

Rebound rate is a good stat. Total rebound rate is kind of meh, but ORR% and DRR% are really good.

I've soured on +- over the years. Doesn't take factors into account that I think would be important for a +-, like who else is on the court.
 
I like ratio stats personally...

things like pts/shot

pts/possession

rebound rate
assist rate

Eff rating
 
I've soured on +- over the years. Doesn't take factors into account that I think would be important for a +-, like who else is on the court.
There are adjusted +/- statistics available at basketballvalue.com, which do account for who else is on the court.
 
PER is the most widely-used stat for comparing players across teams and positions. if it's junk, why is it the one that most analysts use?

like with any stat, you just have to know its limitations when you talk about PER. specifically, that it does nothing to account for defense or intangibles.
 
A bunch of years ago when I was a kid I attended the Rick Majerus basketball camp. Rick Majerus said that the most important statistic in basketball is defensive field goal percentage. Thought that was interesting.
 
Back
Top