Steve Novak's 2008 Preseason

Wow, all the math crap is crazy....

 

That's true ... were you disagreeing with something I said?

It's not arbitrary. Using a .50 factor for free throws might make sense in some cases in a small stretch of games, but on average it won't be as accurate as .44 for capturing "total shot attempts".

TS% is an approximation, but a very good one.

 

It is true that the .5 factor does not change the leaderboard. However, it is not completely meaningless. This is supposed to be an adjusted field goal percentage, not a point per shot percentage. A normal field goal is worth 2 points. That's why you divide the point total by 2. In this, 3pt shots and free throw attempts are viewed as exceptions that needed to be adjusted.

That's exactly my point. Again, too bad they don't have the and-1 statistics (and technical foul shot, for that matter). Technical foul shots are rare enough to be negligible. But and-1 should be a significant statistical measurement.

 

Now you know why you should have paid attention in that Jr. High math class.

 

somebody needs to come up with some statistics to let us know if these statistics are meaningful!

I like the advanced stats, but do find they most often tell you what you pretty much already know. So and so is efficient. So and so is inefficient.

 

oh, I just pointed that out, because you said something about "where the .5 factor comes from." I was just saying that it comes from pure math...nothing arbitrary about it.

I disagree. It is arbitrary.

There is nothing that says PPS is not a better measure than TS%. If you miss a free throw, your PPS does not change, whereas your TS% goes down. Conversely, when you make a free throw, your PPS goes up more than it does under TS%. Why is one better than the other?

Look at it this way...I'm sure someone can make an argument based on historical facts to say Slugging % is not accurate because the value of singles, dbls, tpls and HRs is not linear. Furthermore, someone can argue that OPS is not accurate because Slugging and OBP should not be a simple addition. Enter the Hollingers of the world to seduce us with "advanced statistics."

PPS inherently takes into account missed FTs by not considering FTA as a shot. It doesn't have historically-based coefficients that will change over time.

but really, my whole point is the mere idea that the volume of FTs at the same FT% should have more value in player efficiency than increasing your volume of 2s or 3s at the same FG% is flawed. A player with the sheer ability to make a higher volume of Field Goals (ie. effective FG%) makes the whole team more efficient. Case in point is Maggette, Iverson, Hamilton and Peja. The epitome of getting to the FT line vs simply being a great shooter. Who had the highest TS% last year out of those players. Maggette. That's flawed. Peja and Hamilton clearly make their teams more efficient by just making their shots.

 

oh, and the other thing is...if you need to add historically-based coefficients to a stat to claim it is better for applying to boxscores...you might as well just chart the damn games like 82games does, and be accurate to what happened rather than applying averages to boxscores.

 

What is PPS? I was confusing it for something else, I think.

And the usefulness of a stat like TS% is it closely approximates the efficiency, but can be calculated very simply from the box score. If precise information was readily available from charting, then we'd use that. But the fact is its not readily available, so we make do with metrics that approximate what we're looking for.

Also as a matter of practicality, the purpose of any statistic is to help project future performance. So, in that sense, even precisely descriptive stats can only help us approximate what the player will do in the future. That extra precision you're looking for probably isn't going to make much of a difference in improving projections.

 

PPS is points per shot. Free throws are exactly that - free. Free points added to your PPS. Free points without taking a shot. I'm not a big fan of this stat either. If you're real accurate at threes or get a lot of calls, you will have a high PPS.

To be honest, the only sport I think where stats actually matter is baseball because it's so individualized. Basketball and football is way more team oriented, so the stats become more subjective and invite too many different interpretations.

 

durvasa,

I think both PPS and TS% are flawed. They will both make out Maggette to be more efficient than Hamilton and Peja, as if he makes his team better simply because he gets to the line a lot. The extra flaw in TS% is it uses a .44 coefficient that is based on a calculation in time. That number will change over time. In 20 yrs, that number won't be the same.

How can you advocate a stat that changes over time, such that a player now uses a different TS% calculation than a player in the future. Do you go back and adjust things, later?

It's like, oh we won't know who had the best TS% this year until the formula is readjusted.

I'll end with a quote from the original post in the link you provided regarding the definition of TS%:

Don't tell me .44 is not arbitrary or doesn't change over time.

 

PPS is TP/FGA. very simple. If you include the .5 multiplier (which is divide by 2 for value of 2ptrs, ie...divide by the value of a regular shot made), the formula becomes a percentage like TS%, ie

(TP/2)/FGA which equals TP*.5/FGA

including that .5 multiplier/divider will not change who the league leaders are in PPS. It just turns the number into a Base 2 percentage. The .44 coefficient merely attempts to gain accuracy at the expense of simplicity. However, changing that number will change who the league leaders are.

DPballer,

Who was it in the "sine/cosine" thread that you needed to be upgraded to member? haha.

 

Just because it can fluctuate over time does not mean its arbitrary. The number wasn't just picked out of a hat. Recent historical evidence shows it to be a good approximation. No one claimed it to be some fundamental constant of basketball.

 

But free throws aren't free, unless they're technicals. So that doesn't sound very good to me.

If a player gets fouled and misses two at the line, that's hurting his team and should be reflected in his efficiency.

If you miss a free throw your hurting the efficiency of your team (so your individual efficiency should get worse), and if you make a free throw you're helping the efficiency of your team (so your individual efficiency should go up). That seems pretty obvious to me, and TS% at least satisfies those minimal criteria whereas PPS clearly does not.

Huh? PPS ( points / FGA ... right?) doesn't take into account missed FTs at all. One player can shoot 4/8 from the field and 0/4 from the line, and he'll have the same PPS as a player who shoots 4/8 from the field and doesn't take a free throw. Clearly, the first player hurt the efficiency of his team more, because he used up maybe 2 possessions at the line without producing a single point.

A player who is capable of using a lot more possessions at only a slightly less efficiency can improve his team's net efficiency more. I think this is what you're trying to get at, and you're right. The reason is that players tend to be more efficient if they have less to do.

The ultimate goal for any team offense is to be as efficient as possible -- i.e. score as many points per possession as possible. But at the player level, you don't just look at efficiency (e.g. TS% or ORtg). You have to look at how active they are in creating shots for themselves and teammates. I think you're saying that because TS% doesn't take that into account, it's flawed. Well, no. TS% is specifically measuring scoring efficiency. You have to combine that with how prolific or active the player is in his team's offense to get a complete picture of his value on that end.

 

Look, do you know what the price of gas is going to be during the NBA regular season. No. Nor do you know whether .44 is the correct coefficient for this year. Yet, ESPN publishes .44 as THE number. Do you think they will change that number at the end of the year? doubtful. That makes it arbitrary.

bottomline: show me one stat in MLB that uses a time sensitive formula. TS% is time sensitve. PPS is not.

these quotes by durvasa are meant to say TS% is better than PPS. Does anyone see the flaw in these statements, whereby PPS does indeed factor in misses. HINT: I've already said it previously.

 

Ok ... we disagree on the meaning of the word arbitrary. You apparently think that because this formula is an approximation and not a precise measurement of particular events on the court, it's "arbitrary". My understanding of this term is very different, but whatever.

Based on PPS, a player who shoots 5/5 from the field and 0/100 from the free throw line has scored 1 point per shot, the same as a player who shoots 55/55 from the field. It would be obviously absurd to say these players are equally efficient, but that's what PPS would tell us. TS% would provide a more reasonable assessment of their efficiency: the first player gets a .10 TS%, and the second player gets a 1.00 TS%.

I guess I missed it.

 

Each FT made increases PPS faster than TS%. Thus, PPS inherently factors in misses by rewarding made FTs more. TS% merely attempts to normalize things around a time-sensitive, unpredictble coefficient.

 

I question the logic of this statement.

The metric FGM (field goals made) rewards made field goals. Does it thereby "inherently" factor in misses? No. Now, if we assume the same number of field goal attempts for two players, you could say that FGM would be higher for the more efficient player and therefore misses are indirectly taken into account. But that's only true under the assumpion that FGA are the same.

Similarly, in your case, PPS only "inherently" factors in misses if it assumes their FTA/FGA are similar. That's not even close to being accurate.

Also, it's accurate say the coefficient is time-sensitive. But saying its "unpredictable" (i.e. arbitrary) is false. There's a specific methodology in calculating it. And you can repeat it for any time period. That makes it predictable.

 

Novak has a gifted shooting stroke. Too bad he can't add handles, strength passing or anything like that in his books. Novak has the height, but just isn't good enough to play at SF and not strong enough at the PF. Novak would be an expensive role player if he learns to play more defense and pick up his passing game. Although at best he is a benchplayer, he can still become a very very expensive and useful backup. His quick shooting stroke and accuracy is hell of a weapon that he didn't build around.

 

Steve Novak's leave is a pity to me