A Closer Look: TS% and the Toronto Raptors

This is a guest post by my buddy Julian, who writes the blog Comedy Landfill. Like me, he’s a huge Raptors fan. Also, he’s fond of playing with numbers, so get ready for a barrage of stats the likes of which have never been seen before on this site! Anyway, enjoy. And follow him on twitter.

Since the Bryan Colangelo era began, the Toronto Raptors have been a team that looked to punish opponents with its long distance shooting and offense in general. During the 06/07 season, the Raptors put their excellent shooting on display and managed to win the Atlantic Division (albeit in a year where there wasn’t much competition for the honour) by tying the franchise high 47 wins in a season. Since that point, the Raptors have been mired in mediocrity to just plain not-goodness; this past season being a lot of the latter, mixed with a bit of the former. I think any Raptors fan might be wondering what went wrong. Why were we so gosh darn awful this past season? Are the Raptors going to be any better in the upcoming season?

Getting to know TS%

First of all, let me introduce to you one of my favourite basketball statistics. True Shooting Percentage (hereunto referred to as TS%), is a statistic that measures how efficiently you score the basketball. Most of you, if you’re slightly more than a casual fan of basketball, understand what FG% is. FG% is the number of shots a player makes divided by the amount of shots a player takes, put into percentage form. This was the statistic that was used by the NBA for a long time, long before the 3 point line was introduced to NBA Basketball in the 1979-1980 season. The problem with FG% is that it doesn’t account for 3 point shots, nor does it account for free throws.


For instance, if you take two players, both of them shooting 40% in FG%, you would assume that neither of these players are very good. However, if someone informed you that “Player A” shot all of his shots from the 3 point line and “Player B” shot all of his from 2, it would be quite easy to see that Player A is a more efficient shooter, because his shots are worth 1.5 times more than the other guy’s! But FG% just sits there judging both players as the same. “It’s not fair!” you’re probably yelling at your computer screen right now. I agree. Let me give you another example: Let’s say that there are two players that shoot 45% FG%. Now, looking at those stats, you would assume that both players are pretty average scorers. But if I were to tell you the “Player C” shot 10 free throws a game, and hit 90% of those free throws, and “Player D” went to the free throw line 1 time a game and shot 50% from the stripe, you would realize that FG% has failed us once again! “Player C” is a far more efficient player than “Player D”!

The solution to this problem is TS%. TS% accounts for two point shots, three point shots and free throws when gauging how efficient a player is from the field. “All right!” you’re probably saying with a fist-pump. I agree. TS% is awesome! Now TS% isn’t an end-all, be-all statistic. It doesn’t account for rebounding or turnovers, so the number of possessions must be accounted for as well, and if you’ve read the sports pages in any newspaper, you will know that the Raptors were a very poor rebounding team. However, TS% does allow us to examine the offenses and defenses of teams and players around the league.


Applying TS%

Now, back to the question of why the Raptors stunk so badly last season. As a team, the Raptors scored the ball at a 54% TS%. While this is not awful, it is nowhere near the efficiency that the top teams in the league can boast. Cleveland, for example, scored the ball at a 56% TS%. You may be saying to yourself “Hey, are you actually saying that the Raptors and the Cavaliers are only 2% apart in terms of shooting the ball as a team?” And the answer is YES! If you are surprised by that statement, the thing that you are probably not considering is that a basketball game consists of many, many possessions. Cleveland, for example, took 58 2 point shots a game, 20 3 point shots a game and 24 free throws a game. If they scored on 100% of those attempts, they would have scored about 202 points per game. Considering that, a 2% difference in shooting efficiency is going to account for around 4 points per game, which is actually quite a big difference. The 06/07 Pheonix Suns, for example, one of the greatest offensive teams of all time, had a TS% of 59%! Still “only” a 5% difference from the 08/09 Raptors in efficiency. This is why those seemingly small percentage differences actually do matter in the grand scheme of things.

Because a lot of you probably have an idea of what a good FG% is and what is a bad FG% is, I’m going to give you my analysis of how to rate TS% when it comes to players:

1- The Mendoza line: A TS% of 48% or below. If you are shooting a TS% of less than 48%, you are hurting your team every time you take a shot. Players who shoot this type of percentage are usually fringe utility players that play deep on a team’s bench, and are brought in (occasionally) for rebounding, defense or playmaking. Most of the time, this type of TS% means you will be out of the league soon.
2- Awful: A TS% of 48% to 50%. This is still quite bad.
3- Poor: ~51% TS%.
4- Not good: ~52% TS%.
5- Acceptable: ~53% TS%.
6- Fine: ~54% TS%.
7- Good: 55% to 56% TS%.
8- Very Good: 56% to 58% TS%.
9- Excellent: 58% to 60% TS%.
10- Outstanding: 60+% TS%. Anything over 60% TS% will put you near the top of the league for efficiency. At this point, you are either a guy who is an unbelievably efficient scorer (think Steve Nash), or you are a guy who scores only by way of dunking or laying the ball up via an assist (think Tyson Chandler — former center for the New Orleans Hornets — who got gift-wrapped dunks and layups from Chris Paul).

Now, this same logic doesn’t apply to teams, because teams of players usually incorporate not only very good scorers, but also defensive players, rebounders and playmakers, who may not be as efficient scorers as the star players on the roster. What happens is that while you may have a player that is exceptional at scoring the basketball, the TS% of the team maybe dragged down by other players on the team who shoot a much lower percentage.

TS% and the Raptors

Perhaps at this point you are saying to yourself, “These statistics are great and all, but how do they explain last season’s woes?” That’s a good question. As I just explained, the best scorers on a team in terms of TS% may be dragged down by the rest of the team. With that in mind, let’s take a look at some of the players on Toronto’s roster last season.

Below are the players who played for the Raptors (minus a few players who rarely saw floor time), with the amount of shots (FGA) and free throws (FT) they took, as well as their true shooting percentages.

-Chris Bosh: 1263 FGA, 617 FTA, 56.9%
-Andrea Bargnani: 958 FGA, 266 FTA, 55.9%
-Anthony Parker: 754 FGA, 145 FTA, 52.4%
-Jose Calderon: 644 FGA, 154 FTA, 61.3%
-Jason Kapono: 604 FGA, 42 FTA, 52.5%
-Joey Graham: 480 FGA, 160 FTA, 54.2%
-Jermaine O’Neal: 456 FGA, 142 FTA, 52.6%
-Shawn Marion: 342 FGA, 62 FTA, 52.3%
-Roko Ukic: 324 FGA, 60 FTA, 43%
-Jamario Moon: 317 FGA, 65 FTA, 56.2%
-Will Solomon: 181 FGA, 24 FTA, 51.2%
-Pops Mensah-Bonsu: 96 FGA, 41 FTA, 42%
-Kris Humphries: 90 FGA, 48 FTA, 51%

A few things on the above list may pop out at you. One could be “Wow! Jose Calderon is outstandingly efficient!” or “I thought Jason Kapono was a really good shooter, why is he only shooting a 52.5% TS%?” or maybe even “We had a lot of really inefficient scorers on our team last season.”

First, let me tackle Jose Calderon. Yes, Jose Calderon was incredibly efficient for the Toronto Raptors last season, even though he was injured. He has also been a very efficient player over the past 3 years as well, boasting a 58.8% and 60.7% TS% mark in his previous two seasons with the club. From my own experience, this is because he takes a lot of good shots, and hardly ever takes a bad one. He shoots when open, and when he isn’t open, passes the ball. This also explains why a guy with such a high TS% doesn’t take more shots. I think every Raptors fan would like to see Jose take more shots, however.

Secondly, Jason Kapono’s TS% is not an aberration. In fact, if you watched most of the Raptors games last season, you’ll know that Jason Kapono, while a good 3 point shooter, was absolutely awful whenever he was not shooting a 3 point shot, which was actually quite a bit. While his 3 point average of 42.8% is quite impressive, his two point percentage of 43.4% was not, and he took about 150 more 2 point shots than 3s. Also, if you take a look at the above graph, Jason Kapono averaged an anaemic 42 free throw shots for the season! Jason played 1,831 minutes that season, which roughly translates to 0.02 free throw attempts per minute, or one free throw every 45 minutes of playing time. In comparison, Pops Mensah-Bonsu played only 263 minutes and got virtually the same amount of free throws. It’s no surprise that he didn’t score very efficiently when you take those factors into account. On top of that, Kapono was by far the worst defender on the team, but that’s a story for another day.

Lastly, and most importantly is that the Raptors DID have a lot of sub-par scorers taking a lot of shots last season. Kapono, Parker, O’Neal, Marion and Ukic all did not impress on the offensive end. But surprise surprise! If you look down the list, virtually all of the players apart from Bosh, Bargnani and Calderon are either no longer apart of the team, or figure to have a much smaller role with the club next season. So, let’s take a look at the players that will either be gone, or have their minutes marginalized next season:

-Anthony Parker: 754 FGA, 145 FTA, 52.4%
-Jason Kapono: 604 FGA, 42 FTA, 52.5%
-Joey Graham: 480 FGA, 160 FTA, 54.2%
-Jermaine O’Neal: 456 FGA, 142 FTA, 52.6%
-Shawn Marion: 342 FGA, 62 FTA, 52.3%
-Roko Ukic: 324 FGA, 60 FTA, 43%
-Jamario Moon: 317 FGA, 65 FTA, 56.2%
-Will Solomon: 181 FGA, 24 FTA, 51.2%
-Pops Mensah-Bonsu: 96 FGA, 41 FTA, 42%
-Kris Humphries: 90 FGA, 48 FTA, 51%

In all, those players took 3,644 shots last season, accounting for well over half of the Raptors’ 6,673 shots total. A curious mind such as my own wondered what the TS% of that group of players was. I did the calculations, and found that that group of players averaged a TS% of 52.6%; rooted somewhere in between “not good” and “acceptable”. Not exactly an offensive juggernaut, that group. I think this plainly shows the “drag down” effect, which mitigates the accomplishments of Bosh, Bargnani and Calderon on the offensive end.

With that in mind, why don’t we take a look at their replacements? While the roster may not be totally completed as of yet, we now have a idea of what the Raptors roster will look like come tip-off time. Below is a list of players that we have acquired this summer, with their number of FGA a game (I’m using per-game metrics because some of them were injured and missed time), FTA a game and TS%.

The Replacements

Hedo Turkoglu: 13.3 FGA/G, 5.1 FTA/G, 16.8 ppg 54.1%
Jarret Jack: 10.4 FGA/G, 3.2 FTA/G, 13.1 ppg 55.4%
Marco Belinelli: 7.5 FGA/G, 1.2 FTA/G, 8.9 ppg 54.7%
DeMar DeRozan: X, X, x%
Reggie Evans: 2.3 FGA/G, 2.2 FTA/G, 3.3 ppg 51.4%
Rasho Nesterovic: 6.1 FGA/G, 0.5 FTA/G, 6.8 ppg 52.4%
Antoine Wright: 6.6 FGA/G, 1.5 FTA/G, 7.3 ppg, 50.1%
Amir Johnson: 2.6 FGA/G, 0.6 FTA/G, 3.5 ppg, 60.8%

Running the calculations on that group of players, their average TS% was 53.8%. Notice that I didn’t even attempt to extrapolate DeMar DeRozan’s stats, because unlike some statistical experts (*cough* Hollinger *cough*), I have absolutely no faith whatsoever in the college-to-pros numbers game that people like to fool around with, especially when it comes to unfinished “project” players that DeMar figures to be.

While these guys already project to be better than the group of players that they are replacing, I’m going to make a couple of guesses about these stats to paint what I believe to be a more accurate picture of what will transpire next season. I think that Antoine Wright’s minutes are going to go down, and as a result, his shot attempts too, because of increased competition at the 2 spot from Belinelli, Jack and DeRozan. I also think that Jack’s minutes and FGA are going to take a hit from alternating with Calderon. I also believe that Belinelli’s minutes and FGAs are going to increase, now that he’s not in Don Nelson’s doghouse. I think that Johnson will compete with Evans for the backup 4/5 spot, and should get more minutes at the 5 (he’s 6’10”) if Rasho Nesterovic continues to decline.

So now let’s have a little fun and see what the ultimate TS% projects to be for our team, using the non-adjusted stats from last season.

Chris Bosh: 16.4 FGA/G, 8.0 FTA/G, 22.7 ppg, 56.9%
Andrea Bargnani: 12.3 FGA, 3.4 FTA/G, 15.4 ppg, 55.9%
Jose Calderon: 9.9 FGA, 2.3 FTA/G, 12.8 ppg, 61.3%
Hedo Turkoglu: 13.3 FGA/G, 5.1 FTA/G, 16.8 ppg 54.1%
Jarret Jack: 10.4 FGA/G, 3.2 FTA/G, 13.1 ppg 55.4%
Marco Belinelli: 7.5 FGA/G, 1.2 FTA/G, 8.9 ppg 54.7%
DeMar DeRozan: X, X, x%
Reggie Evans: 2.3 FGA/G, 2.2 FTA/G, 3.3 ppg 51.4%
Rasho Nesterovic: 6.1 FGA/G, 0.5 FTA/G, 6.8 ppg 52.4%
Antoine Wright: 6.6 FGA/G, 1.5 FTA/G, 7.3 ppg, 50.1%
Amir Johnson: 2.6 FGA/G, 0.6 FTA/G, 3.5 ppg, 60.8%

First of all, this shouldn’t be taken 100% seriously as a real projection, because there are a lot of problems with doing this sort of calculation. One of the problems of course is that we score 107 points and use 84 possessions with only these players, a large jump up from last season, which isn’t likely seeing as we haven’t even included DeMar DeRozan or the scrubs yet, who figure to get around 5-10% of the minutes. What this means is that some of the players are likely to have their minutes and shot attempts scaled back. The TS% of the team works out to 55.2%; a 1 percent jump from last season. Seems like a decent improvement.


I think when looking at the upcoming season, you have to understand something. Future projections are always educated guesses that rely on data being the same, or growing in ways that follow a historical or statistical trend. But this is not always how things work in real life, rather, that’s just how things work MOST of the time. The Pheonix Suns of 04/05 are a great example of this. Prior to that season, they were an abysmal 29-53 (very similar to the 08/09 Raptors!) and had a rookie coach who came in mid-season and went 21-40 (extremely similar to Jay Triano!), and had just signed a 30-year old Steve Nash to a contract that everyone thought was insane (very similar to Hedo Turkoglu!), and everything was put together by Bryan Colangelo (very similar to… Well, you get it) and fans were gearing up for a disappointing season. But the Suns came in and blew the doors off, tying the franchise record in wins at 62 and bucking all of the expectations that were placed upon them by stat-head prognosticators such as myself.

But how was the Phoenix rebirth possible? I attribute it to something called synergy. Synergy is the state in which all parts of the team are working together smoothly, like a well-oiled machine. Synergy is when the system employed by the coaching staff fits the roster perfectly. Synergy is when disparate elements come together to become much, much greater than the sum of their parts, and synergy is something that every awful team that has done a bit of tinkering in the offseason can hope for. Before Steve Nash entered the equation, guys like Marion and Amare Stoudemire were putting up the stats, but were not even close to as efficient before he got there. Steve Nash, in turn, had a career year that propelled him to his first MVP trophy.

While the influx of new players seems like it will improve the offense of the ballclub a fair amount on paper, I think that every Raptors fan with a heartbeat hopes that Bryan Colangelo manages to catch lightning in a bottle twice, and the 09/10 Raptors will emerge a synergistic team with a knockout offense able to overcome the obvious shortcomings they have on the boards and defense, much in the same way Phoenix was able to five years ago.

Update from Vittorio: I’ve got to thank Julian again for posting the most-commented article on this site thus far. I like all the discussion. Perhaps I should make start making controversial claims like “Kevin Durant will be better than LeBron James” in my blogs now (not that Julian did anything like this). Anyway, this post and Khandor’s comments have inspired Tom Liston to do some statistical analysis of his own. Here’s his graph, showing the correlation between opponent’s TS% and wins:

Liston's Graph



Filed under A Closer Look, Free Agency, Guest Posts, Stats, Toronto Raptors, Trades

36 responses to “A Closer Look: TS% and the Toronto Raptors

  1. Pingback: New Article On Vittorio De Zen’s Fast Break « Comedy Landfill

  2. A few quick points for you to consider further:

    * The Raptors are not adding a new player to their roster that resembles Steve Nash in any way, shape, or form

    * Jay Triano does not resemble Mike D’Antoni, as an NBA head coach

    * Statistics like TS%, in isolation, have little correlation to games Won vs games Lost, in the NBA

    * Jamario Moon’s TS% was 56.2% last season, which would place him in the “Very Good” category … a designation with which I agree, whole-heartedly BUT one with which, IMO, MOST other Raptors fans would strongly disagree

    * re: “I think every Raptors fan would like to see Jose take more shots, however.” … FYI, count me amongst those who would NOT like to see Jose Calderon take more shots. Unfortunately, ‘MOST’ Raptors fans have a relatively poor understanding of what it takes for a team to succeed in a big way in the NBA. ๐Ÿ™‚

    • Tom L

      TS% does have a reasonable correlation. Plot the data. Defensive FG% is the best, but offensive TS% is still a good indicator.

    • comedylandfill

      I think you’re a little confused about my comparisons. I was comparing their situations, not the players or coaches themselves, i.e. an interim coach with a 21-40 record vs. an interim coach with a 25-40 record, a 30 year-old free agent being signed to a rich contract that people think is crazy vs. a 30 year-old free agent being signed to a contract that people think is crazy.

      “Statistics like TS%, in isolation, have little correlation to games Won vs games Lost, in the NBA”

      Offensive TS% and opponent TS% have little correlation to games won vs games lost in the NBA? I’m sorry, but you’re incorrect.

      I do agree with you about Moon, however. VDR and I would often talk at length about how he was being sold short by Raptor fans. I think that they had unrealistic expectations on Jamario, because they saw him as an athlete that could throw down cool dunks. Unfortunately, I think most intelligent fans realized that Moon developing a handle that could take him to the rim this late in his career was a longshot, so I was fine with him taking jumpshots as long as he was making them (which it turns out he was).

      “* re: โ€œI think every Raptors fan would like to see Jose take more shots, however.โ€ โ€ฆ FYI, count me amongst those who would NOT like to see Jose Calderon take more shots. Unfortunately, โ€˜MOSTโ€™ Raptors fans have a relatively poor understanding of what it takes for a team to succeed in a big way in the NBA.”

      While I do think that some of the time, “the great unwashed masses” get things wrong about the game of basketball, I don’t think they’re incorrect about wanting our most efficient shooter to take more shots. If he sacrifices a few % in TS%, but take a couple more shots a game, it’s still better for the team overall, which is what I was trying to illustrate in my article. I think that with Hedo Turkoglu on the team, Jose will get to play off the ball more, and as a result get more shot attempts from the perimeter.

  3. Tom L

    Great Julian – very well done.

  4. bt

    Awesome post!
    I don’t usually comment, but I took the time to do it now; that’s how much I enjoyed this.
    The only thing is %s are individual stats. Most of these players are on a new team now, so their stats might change (for better or worse). Synergy is key.

    Keep it up!

  5. Excellent analysis, as usual.

    That Suns team did something that I think more NBA squads should try- they pushed pace as hard as they could for every possession of an 82 game season. It made them exciting to watch and it inflated each player’s stats so that they became more valuable as trade pieces (and their agents were happy). Not sure it’d work for your Raps- Calderon, Turk and Bargnani are definitely half-court guys.

    • Vittorio De Zen

      Just to be clear, the analysis wasn’t mine – this is a guest post! An awesome one, IMO – this won’t be the last time Mr. Comedy Landfill writes something for this blog.

      But yeah, I don’t expect the Raps to resemble those Suns teams, basically because, as you said, their main guys are half-court players. DeRozan, Amir Johnson, and Jack are really the only rotation guys that would be best served by running, I think. Well, maybe Antoine Wright.

      Anyway, the point still stands – The Raps have upgraded roster spots 4-9 quite significantly, I think. The offense could be really, really good, though admittedly not Suns-good.

      As for more squads trying to push the pace… I’d love to see that too. If the Clips did that next year, I’d watch every game.

  6. Pingback: Toronto Raptors Linkage for August 17th from 15:18 to 22:48 :The AltRaps Blog

  7. Tom L & comedylandfill,

    If you can please state the numeric correlation that you claim exists between “Wins” in the NBA and “TS%” I’d be most appreciative of that fact.



    I understood the initial comparisons which you made between the numbers:

    * 21-40 vs. 25-40


    * 30 years of age vs. 30 years of age

    Are you quite sure, however, that you really understand what i meant when I said that:

    * Hedo Turkoglu does not resemble Steve Nash


    * Jay Triano does not resemble Mike D’Antoni?


    re: the need for Jose Calderon to shoot more shots

    Are you aware of the findings put forth from Jon Nichols concerning the correlation which exists between the 3PT-shooting percentage of PG’s and Team Offensive Efficiency in the NBA? And/or, the correlation which exists between the frequency of 3PT-shot attempts by a PG and Team Offensive Efficiency in the NBA?

    • Tom L


      I have worked through a lot of this data as you can see here:


      The data is there on opponent’s **FG%**, which, of course is not TS%, but its the quickest “support” I can give on short notice for everyone to see (at least I’m provide *some* support to back up my thoughts).

      And as you see, the correlation with wins for last season was -0.873302 > I think that’s not to shabby, if I understand anything about statistics.

      But, I have not published on a public website anything beyond that for now – I will not have the time in the near term. But this provides a bit of evidence of the work that I’ve done.

      Since you made the first claim, I’ll ask you to show your evidence. (Not simply refute mine)

      (BTW, I completely agree with your assessments of Moon and Calderon).


  8. Tom L,

    * If the initial claim here here had pertained to FG% instead of TS%, I doubt that I would have provided my food for further thought to this subject

    * re: the presentation of contradictory data … I do not think that you truly believe it wrong to support the null hypothesis until proven otherwise, although I could be incorrect in this assumption

    What I happen to be in search of, in this instance, is an actual data set which supports the hypothesis that TS% has a stronger and, therefore, greater level of statistical significance than Total Rebounds and/or Rebound Differential when using a Regression Analysis for Wins vs. Losses [both i. Straight-up, and ii. Against the Wagering Line] for NBA games.

    To this point, I have been unable to locate such a data set … but the effort continues. ๐Ÿ™‚

    PS. What is even more interesting to me, personally, is the relationship which exists between FG%, TS%, T-Rebounds or Rebound Differential and Wins vs. Losses in NBA games [both i. Straight-up, and ii. Against the Wagerline Line] when “projecting forward” to an upcoming contest … as opposed to looking back at already completed ones.

    • Tom L

      FG% should be quite correlated with TS% – some teams that rely on 3pt shots more will skew this a bit – but my evidence is VERY strong. TS% likely isn’t too far off. And you said “statistics like TS%…” FG% is not a totally foreign stat from TS%.

      You were confident that it was not correlated – at least provide something or even *anything*? Its only fair, since I have.

      If you’re going to talk the talk, walk the walk.

      Looking forward to your data analysis to back up your statement. “Statistics like TS%, in isolation, have little correlation to games Won vs games Lost, in the NBA” > you said it. What are you data are you basing this on then.

    • Tom L

      Oh and sorry – the data can be easily compiled from here:

      True Shooting Percentage; the formula is PTS / (2 * (FGA + 0.44 * FTA)). True shooting percentage is a measure of shooting efficiency that takes into account field goals, 3-point field goals, and free throws

      Good luck.

  9. Tom L

    khandor and Julian,

    khandor said “Statistics like TS%, in isolation, have little correlation to games Won vs games Lost, in the NBA”

    Julian noted you were incorrect and I submitted it does have a reasonable correlation.

    and you replied
    “If you can please state the numeric correlation that you claim exists between โ€œWinsโ€ in the NBA and โ€œTS%โ€ Iโ€™d be most appreciative of that fact.”

    It is -0.839494025 between Opponents’ TS% and Wins for the year 2008-09 in the NBA.

    To my surprise it was much more robust than I originally thought.

    I will send the graph to Vittorio.

  10. Tom L,

    Thanks for providing the correlation you found for TS% and Wins for last season’s NBA teams.

    Continuing a protracted exchange of this sort on Vittoria’s site was not my intention. If you’d like to follow-up with me directly, however, at my blog, I’d be happy to do so. [or, perhaps, via email, if you ask Vittorio to pass along my address]

    FWIW … your thoughts on the following study would be appreciated.


    • Tom L

      Thank you khandor.

      You had said Julian and I were incorrect and I think I owed it to the reader and the author to defend our position with actual data.
      If you did not intend to further the exchange then you would have originally supported your argument and you would not have insisted our argument was incorrect.
      You asked me to defend my position and thus it was natural for me to do so – else the readers would be left believing that perhaps there was not a correlation as you insisted.
      I did provide you with my actual name as well as email and Twitter account in my comment where I linked to my RR post – always welcome feedback (publicly or via email) as it was my first ever post and there is no doubt there are many areas where I could improve.
      I will look further into that study – but I do not see TS% mentioned at all (via a search). It is also quite old (data is at least 9 years old). And the source of the data is a dead link (hard to see what other variables the author may have looked at). Otherwise, it likely provides a nice basis for an update – perhaps introducing variables like TS% instead of FG% and see what yields the best fit. It would be interesting to use a wider data set as well as and the last 8 years.
      Perhaps we can work on an “update” piece together.

  11. Tom L,

    – Your original position was incorrect, from my POV

    – Terms such as “little” and/or “big” are comparative terms [rather than absolutes]

    – “in isolation” may have been the two key words in my original comment here

    – If you would have said something like “in conjunction with other factors”, then, in all likelihood, I would have passed this topic by

    – “protracted” is another key word in my prior comment

    – It doesn’t appear as though I’ve “insisted” upon very much here, at all; as opposed to just stating a different POV to the original claim made in this article

    – Like you, I do not see TS% mentioned in the study which I referenced for you; in part, some of my interest in your initial article lies in this fact

    – Although I could well be wrong in my next observation, I think there’s a pretty fair chance that you’d be prepared to acknowledge the increased strength of a .9076 correlation in comparison with -0.839494025

    – I missed your email address the first time through; thanks for pointing that out this time

    – FWIW, I choose not to use twitter; nor, facebook

    – collaborating on a project of this sort, isn’t something which I have a great deal of time [or opportunity] to do at the moment; but, at some point in the future, this might be a possibility

    – IMO, the power of “Rebounding Differential” in determining Wins from Losses in the NBA, is NOT to be “in isolation”, as well ๐Ÿ™‚

    – As I’ve mentioned already, how measures of this sort [both, individually, and in concert with others] perform when asked to forecast the W-L outcome of an upcoming game hold more interest [for me, at least] than does looking backward to account for the variables involved in a victory/loss

    All the best to you, sir … and, please feel free to follow-up with me directly, if you wish [using the means I’ve suggested above; and I will do, likewise].

    • Tom L

      Unfortunately with that last comment I have lost the respect I may have had for your opinions.

      As Barney Frank pointed out the other day:

      “Trying to have a conversation with you would be like arguing with a dining room table.”

      You make a simple statement and then when it is clear you are incorrect, you pretend you made a different statement. Please follow the thread above.

      We said *nothing* beyond stating there was a [if you believe “little” is – and yes its a relative term – -0.84 correlation, then we need to re-write the stats textbooks! All those poor kids are being misled!)
      (“TS% does have a reasonable correlation.” was my specific comment). On its own – on its own – on its own. Repeat after me.

      You sent the discussion into an entirely different area by introducing aged multi-factor model data – we did *not* make any claims that the correlation was stronger – based on one variable – than the robustness of a multi-factor model! This is comical. Even still – you don’t prove that the author even looked at TS% as opposed to FG% (as one example)! See why I’ve lost respect? You keep changing the criteria to attempt to divert us away from the incorrect statement you originally made.

      If you originally made the statement that there are multi-factor models out there that are robust, I would have completely agreed – but you targeted one variable and specifically mentioned it has “little” correlation. This is not a correct statement and you did not provide and *direct* evidence that it was.

      But you clearly said – and its in print above – “Statistics like TS%, in isolation, have little correlation to games Won vs games Lost, in the NBA” That is verbatim.

      And I have *shown* you that “Statistics like TS%, *in isolation*, have [a strong] correlation to games Won vs games Lost, in the NBA” (emphasis added and brackets show the changes). I did not say that this one stat was the “be all, end all” – I simply provided a *direct* answer to your statement. I have never had such a hard time answering such a simple question before.

      You say you have your own blog – I will look there for a previous (or upcoming) post on “how TS%, in isolation, has little correlation to games Won vs games Lost, in the NBA.”

      Vittorio, I apologize on behalf of both of us for the long posts – it should have evolved into an honest and interesting discussion, but unfortunately it morphed into something less interesting.

  12. Tom L,

    If you would have simply said, instead:

    “TS%, in isolation, has little correlation to games Won vs games Lost, in the NBA … in comparison with what, according to you?”

    in reply to what I wrote initially, then I would have told you … in comparison with the robustness of a multi-factor model [which includes Rebounding and/or Rebounding Differential, as well].

    Arguing like a dining-room table?

    LOL ๐Ÿ™‚

    It is always interesting to see what some consider stronger to be “strong”, in comparison with others.


  13. Sorry … that should read as:

    It is always interesting to see what some consider to be โ€œstrongโ€, in comparison with others.

    Typing has never been a strong suit of mine. ๐Ÿ˜‰


    PS. Vittorio, you have a fine blog.

  14. comedylandfill

    Thank you Tom, don’t worry, what you are writing is very clear, and very sound. Khandor used to be a member of the realgm.com forums under the name “Condor”, and when people started raising points that gave him a difficult time, he began to obfuscate his points, and change his arguments to make it difficult to show that he was wrong. You’ve done as good a job as possible showing why his original point (TS% has no correlation with W/L) was incorrect. Don’t feel upset if he doesn’t get around to addressing this point.

    Also, thanks a lot for doing the due diligence when it comes to TS%’s correlation to wins. I’m not surprised that there is a strong correlation between it and wins, because most good teams can put the ball in the basket efficiently.

    In any case, I think this should be argument should be concluded, because the original point has been settled, and Khandor is trying to drag you into other ones in order to not appear incorrect.

    Tom, send me a PM if you feel like continuing this discussion further. Thanks!

  15. comedylandfill,

    A former member of realgm.com forums?

    LOL, ๐Ÿ™‚

    Perhaps you shouldn’t be quite so gullible as to simply believe what someone else might have to say about me in an environment like that one.

  16. Tom L,

    Agreeing to disagree is never a problem for me; neither is disagreeing without being disagreeable.

    Any time you might care to discuss the relative ease involved with attempting to elevate a team’s FG%, or it’s TS% versus its Rebounding Differential … and the resulting effect such improvement has on games Won vs games Lost … feel free to contact me directly.


  17. comedylandfill,

    Btw …

    re: Youโ€™ve done as good a job as possible showing why his original point (TS% has no correlation with W/L) was incorrect.

    Those who take the time to read closely should be able to see that what you identified as “my original point” was not what I actually wrote, in the first place, at all [i.e. re: there being “no correlation”].

    IMO, that’s the sort of misinformation [whether purposeful or accidental] about me … and what I have to think about the game … that leads to factually incorrect hearsay and disharmony, in the first place.

    • Tom L

      You are correct – you actually said “little” which to any educated reader really means “strong” correlation. It was a gross oversight on our part. I was under the false impression “little” = “little”

      So, not being as educated as you I had to look it up.
      “not much” or “small in importance”

      • Tom L,

        1. Are you trying to suggest that a -0.839494025 correlation is in some way not to be understood as being a more “little” correlation than a .9076 correlation?

        2. re: “It was a gross oversight on our part” & “not being as educated as you”

        Why write this type of stuff, in the first place?

        Although both did succeed in giving me reason to chuckle …

  18. comedylandfill

    Well, I’m almost convinced that Condor was you, because his posting style was exactly the same, and your views on the Raptors were exactly the same. I’m not surprised that you are denying that, because you were banned from realgm.

    Your assertion that TS% had little correlation to wins or losses was incorrect (.84 is actually quite strong, statistically speaking), and as a result, you are attempting to say that it is not as strong as other stats that are composite statistics. That is what is known as “shifting the goalposts”.

    Furthermore, you are also trying to infer that correlation is relative to other statistics, in that if there are statistics that address correlation better, the other one can be judged as not being correlated enough. That is simply not the fact of the matter. This stat shows sufficient correlation between itself and wins and losses in the NBA to be judged as a worthwhile statistic when judging teams. Fact. The presence of other composite stats does not change this fact.

    Saying that there is little correlation between this stat and W/L doesn’t mean there aren’t statistics that are more correlated to wins and losses. You’ve invented that after the fact so that it won’t look like you’ve been proven wrong. Tom was more than generous with his argumentation, and your obfuscation tactics have caused him to simply concede out of frustration.

    This will be my last reply, because I think anyone who has read this back-and-forth that isn’t you will understand how it works.

  19. Fantastic work, as always, Tom.

    comedylandfill, quite the interesting article. Love to see how this type of article leads to intelligent, passionate discussion, something that is wanting (at times) in Raptors related forums. Hopefully readers of the comment section will not allow it to be taken away from the facts as they are presented and clearly backed up.

    Fine work.

    • Tom L

      Thank you sir. I was hoping the analysis would resonate with someone! I was concerned that I was not clear in stating my views and/or the data support wasn’t clear. Great to hear.

  20. comedylandfill,

    It’s a good thing that you included the word “almost” in the first sentence of your comment.

    What’s the reason you think I included the words “little” [rather than e.g. “no”] and “in isolation” [rather than e.g. “in conjunction with other factors”] in my initial comment?

    What’s the reason you think I did not reply to Tom L’s assertion:

    “TS% does have a reasonable correlation.”

    with a comment suggesting that Tom L might be “wrong” about this fact?

    I have tried to point out where along the way I seem to have a different perspective on this subject than you do, as well as Tom L.

    In contrast to this approach, however, it now seems as though you have adopted a different “agenda” with your comments concerning me, casting accusations in my direction unrelated to what I’ve written here about the subject of TS%.

    It is certainly interesting to see who here first chose to use the word “obsfucation” in his comments.


    Vittorio … as I said before, IMO, you have a solid blog.

    Hopefully, behaviour like this from comedylandfill … who I understand, may be a friend of yours … directed towards me isn’t something which I should expect to see again if I choose to visit in the future.

  21. Tom L

    I glad to have made you chuckle. Most of my conversations are very lighthearted – even when readers have disagreed with me. I’m just a simple guy that made a simple statement and have answered every challenge and question you have asked. You, however, have not answered some our most basic of questions and requests.

    Readers can decide on their own whether Julian and I have supported our comments or not.

    Otherwise, I’m off to enjoy my weekend and do not intend to comment unless you present to us real data on “… TS%, in isolation, have little correlation to games Won vs games Lost, in the NBA” – because I would doubt you would make a statement and then challenge ours unless you have specific and clear data to back it up – I would hope that you have not wasted the readers and our time otherwise.
    I have been generous (I believe) with my time in order to support my view. If my view ends up being incorrect, at least I will be able to say that I attempted to put in the work to defend my arguments all the way through.

    • Tom L,

      Do you think it’s been a waste of the readers’ time to find out that a “stronger” positive correlation actually exists between the six-part multi-factors model presented in the study by Walters and Staudenmeyer [referenced here, by me] than for TS%?

      If you do, then, I happen to disagree with that observation as well.

      As you’ve said already … I, too, believe that I’ve been more than generous with my time in these comments and stand by what I’ve written here.

      Have a terrific weekend!

      PS. When attempting to understand properly the factors which influence Wins and Losses in the NBA, those who choose to disregard such predictors as “Rebounding Differential” do so at their own peril. ๐Ÿ™‚

  22. Pingback: A Closer Look: Raptors Defense « Vittorio De Zen's Fast Break

  23. Pingback: A Closer Look: How Things Could Go Wrong For The Raptors « Vittorio De Zen's Fast Break

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s