Sunday, September 26, 2010

2009-10 Player Review: David Andersen

We've only got a few more player reviews left to do - and they're reviews of bench players who aren't likely to play very many minutes or contribute many wins next season. Up this time is David Andersen:

As you can see above, Andersen wasn't very successful in his one year stint with the Rockets, who wisely shipped him out to the Raptors for cash considerations and a future second round pick (the Rockets nailed Arturo's management principles on that one!). Why was Andersen so unproductive last year?

One big problem: he couldn't shoot. He was okay shooting at the rim (53.3%) and on 3s (AdjFG% 51.9%) but not so good anywhere else (41.8% <10ft, 46.4% 10-15ft, and 43.0% 16-23ft). The deep twos in particular are what hurt his shooting percentage the most. Even though his shot attempts and points scored were above average, those numbers were achieved inefficiently and so actually ended up hurting more than helping. His rebounding was below average, although not excessively so; same with his net possessions and fouls. His assists were actually above average and his blocks were pretty bad for a centre. Really, it would be better if Andersen played more power forward than centre (although he still wouldn't be all that productive), but with the Raptors desperate for size this year it's likely he'll be stuck at centre.

Tune in next post for a review of everyone's favourite ball-hog, Reggie Evans!

 - Devin.

Thursday, September 23, 2010

2010 FIBA WBC in Turkey: Leandro Barbosa Stats

Here is the spreadsheet for the Brazilian team WBC stats, which includes Leandro Barbosa, as well as fellow NBA players Anderson Varejao and Tiago Splitter:

Let's take a look at Barbosa first:

Not bad, but I think everyone was hoping for a little more against international competition. He was close to average in most categories except for rebounding and steals, which were above average.  His FGAtt were way more than average, which lead to his scoring being way more than average as well - but his shooting efficiency was a tiny bit below average and so his shooting wasn't all that helpful (although he did manage to pass the 1 PPS threshold). He played particularly poorly in Brazil's three losses against USA, Slovenia, and Argentina; in fact, take a look at the difference in Barbosa's production during wins and his production during losses:

Barbosa played incredibly well in victories (Iran, Tunisia, and Croatia) and quite terribly in losses (USA, Slovenia, and Argentina). Compounding this was the fact that he played fewer minutes in wins than in losses, presumably because when his team was ahead they felt that they would rest him, and when his team was behind they thought they 'needed him' to help them win.

Who's that A. Garcia? That would be former Spur (2003-04) and Hornet (2004-05) Alex Garcia. Garcia produced the most wins for the Brazilian team (1.03) and had a EWP40 of 0.256 (third best on the team). He shot more efficiently, rebounded better, turned the ball over less...did better than Barbosa in pretty much every category other than FG attempts and points. Garcia was the best guard on the team.

Now let's take a look at selected forwards and centres:

Machado was the team's best forward and best player overall (by EWP40). Despite putrid rebounding numbers, he scored so much and so efficently that he still managed to produce at an excellent rate. It's a shame that he only played the sixth-most minutes on the team, and as he is 35 years old, his international career is probably near its end.

Varejao, Splitter, and Becker split their time at forward and centre, so comparing their numbers isn't perfectly straightfoward. A couple of things stand out; for example, Varejao couldn't shoot. Interestingly, he seems to have known that because he did not take very many shots of any kind. The main thing he did well (and he did it amazingly well) was rebound; other than that, his low foul and turnover rates were positives, but not much else was. Splitter, on the other hand, was much more balanced; he didn't really stand out in any one area, but instead was slightly above average in most. Both produced at a similar rate - Varejao at an EWP40 of 0.157 and Splitter at 0.149 - but due to Varejao's injuries at the start of the tournament, Splitter player more minutes and so produced more wins.

Becker performed like a better version of Varejao - his main skill was rebounding, but he managed to outperform in every category other than fouls and steals, and was thus significantly more productive. If I was a centre-starved NBA team (*coughRaptorscough*), I'd probably consider giving him a look to see if he could maintain at least a portion of his efficiency in an NBA setting.

So, with all that out of the way, let me add a few more thoughts about the stats from these games:
  • I find it interesting (but not totally surprising) that the three best players on an international team with three NBA players were non-NBA players
  • Brazil kicked Croatia's butt - 1.339 EWins Produced to -0.339 EWins Produced.
  • Brazil actually out-performed Argentina - 0.508 EWP to 0.492 EWP - but still lost a close game.
  • Despite a very close score, the USA was clearly the better team in their game (0.625 EWP to 0.375 EWP)
  • Huge games: Barbosa (0.553 EWP) vs Tunisia and Scola (0.522 EWP) vs Brazil
Lately I haven't been posting as often as I would like - blame it on moving. I hope to get back up to speed in the next little bit.

 - Devin.

Thursday, September 16, 2010

Nerd Numbers: Dan and Dre podcast

Andres of Nerd Numbers has recently started a weekly podcast in which he discusses the NBA, basketball, and other sports (mostly Denver teams) with his brother Dan. Andres, of course, analyzes the stats from a Wins Produced perspective; Dan is a stats "skeptic" and prefers his own original blend of traditional basketball thinking and idiosyncratic whimsy. Last week I determined that Dan was more convincing. This week's podcast has a few interesting topics:
  • Andres has declared me to be "the stats ambassador to Canada". I need to put that on my resumé.
  • Dan would rather have Rudy Gay and Marc Gasol than Luol Deng and Joakim Noah. Given that Deng is better than Gay and that Noah is better than Gasol, I'll have to go with the pair of Bulls.
  • Dan keeps seems confused about the definition of a "good player". Good players don't necessarily have the best skills; they don't necessarily win games of one-on-one against everybody. "Good players" are those who help their team win more than the average player; the greater the degree to which they help their team win, the better the player.
  • Apparently not even Dan likes Bargnani. Somewhere, somehow, someone in a position of power on another team is going to be impressed by Bargnani's PPG average and make a bid for him. At this point I can't imagine that there are many offers that I would turn down.
  • They talked about baseball and football (snooze).
This week's contest was a lot closer, but...I declare Andres the winner! Dan put up a good fight, but in the end, Andres gave 110% and just out-hustled him.

I also have some strategic advice for Andres:
  • Explain how Wins Produced determines "good" players and "bad" players. Explain which events on the court lead to "positive outcomes" (aka winning) and which lead to "negative outcomes" (aka losing). Use general situations and avoid using specific players at first. Once you've convinced him of the logic behind Wins Produced, he'll have to agree with its evaluation of players.
  • I think Dan is better at using rhetoric than Andres. Andres: practice makes perfect!
Dan's doing pretty well - given that he's on the wrong side. I guess my only advice to him is to keep up the style that made him successful in the past and deny, deny, deny.

 - Devin.

Wednesday, September 15, 2010

Would Dampier Help the Raptors?

Erick Dampier was waived by the Bobcats yesterday. Because his $13.07 million salary was not guaranteed, it didn't make sense not to waive him, as the Bobcats were above the luxury tax limit.

Doug Smith, the man when it comes to reporting Raptors news over at the Toronto Star, thinks that Dampier wouldn't help the Raptors because "Dampier is not an upgrade of any significance over what they have, he’s a big and slow on a team that wants to be big and quick and multi-dimensional. A horrible fit, actually, who’d do nothing substantial to improve the team."

Let's take a look at this, shall we? How good is Dampier, and is he an upgrade over any of the Raptors centres?

Oh my! Dampier has been a very good player! At 35, he's bound to show some decline, but he's still quite above average. Is there any centre on the Raptors that can beat that? Here's Bargnani:

Nope, he won't do. How about David Andersen?

No good either. Joey Dorsey?

That's not too bad, but Dampier's better, and Dorsey hasn't ever played a lot of minutes. How about moving Amir Johnson over to centre (a bit of a stretch, but okay)?

More like it, but still not as good as Dampier. And we're not even going to bother with Alabi, as he is a rookie who was below average in college.

So, if I was in charge of the Raptors, would I consider signing Dampier to a short-term contract at or near the veterans' minimum? Absolutely. Even if he falls off a cliff this year, at least he would should eat up some of Bargnani's minutes.

 - Devin.

Tuesday, September 14, 2010

2010 FIBA WBC in Turkey: Linas Kleiza Stats

After all that work putting together the team Canada stats I didn't really feel like doing any further analysis - even though I started working on the WBC in order to cover former and current Raptors players. However, I thought about it and realized that...I already had most of the work done! All I needed to do was replace all the Canadian stats with other teams' stats (and add a couple of pages...and fiddle with positions...and formatting...). It's still quite a lot of work, especially given that I'm generally only interested in one player on each team, but at least I'm not starting over from scratch.

Let's take a look at Linas Kleiza's numbers from the bronze medal-winning Lithuanian national team:

As you can see, Kleiza did very well over in Turkey. His shooting was efficient and frequent, and as a consequence he scored a lot of points. His rebounding was very good and helped to compensate for his below average performance with respect to steals and TOs. His blocks and assists were slightly below average, but he fouled far less than the average forward. Put it all together and Kleiza had a WP40 of 0.221 and produced 1.56 wins in 281 minutes across 9 games.

If you are interested in the other Lithuanian players or want to see more detailed, game-by-game stats, here's the spreadsheet data:

A couple of things that I found interesting:

  • All of the Lithuanian players produced in the positive range - except for Jonas Maculius, who was barely negative.
  • The four best players on the Lithuanian team (by EWP40) were all front-court players.
  • Kleiza got completely shut down against the US. Take a look at Kleiza's game by game performance:

To make it more obvious, you can also arrange the data like this:

  • Durant killed Lithuania; he produced 0.609 wins by himself that game, which is enough to win the game without any additional production from his teammates. Even worse for Lithuania, USA actually produced more than one win over the course of the game...which means that the whole Lithuanian team produced a negative number of wins. That is what we call an ass-kicking (although not as bad as the ass-kicking Spain gave Canada).
That's it for now - check back in the coming days for more on Leandro Barbosa, David Andersen, and possibly other interesting international players.

 - Devin.

Monday, September 13, 2010

2010 FIBA WBC in Turkey: Wins Produced for Team Canada

Thanks to Mosi over at Miami Heat Index, I have managed to put together the Estimated Wins Produced numbers for all of the players on the Canadian men's national basketball team. These numbers include only the five games played in Turkey (against Lebanon, Lithuania, France, New Zealand, and Spain) and none of the qualifying or exhibitions. Putting everything together was a long process and also taught me about some new excel functions - and I wouldn't have been able to do it without Mosi's help.

First I'll embed the spreadsheets into this post and then follow up with some analysis.

If you are too lazy to look through all those numbers, let me give it to you in summary form. The players can be sorted into three categories: the good, the barely positive, and the negative.

The Good:

Only four players on the Canadian team managed to produce at an above average level (greater than 0.100 EWP40). They are (in order of EWP40): Levon Kendall, Kelly Olynyk, Ryan Bell, and Joel Anthony.

Levon Kendall was not only the most efficient player with a EWP40 of 0.302, he also produced the most wins at 0.907. In fact, Kendall - by himself - was more productive than the entire team...which also includes Kendall. Yes, after Kendall, the rest of the team actually produced a negative number of wins. Despite being below the average forward with respect to all aspects of shooting, he was above average in every other category. Where he excelled was rebounding and net possessions; at 12.7 reb/48 and 12.0 possessions/48, he even crushed the numbers for an average centre. All those extra possessions are very valuable and so boosted his WP40 to a superstar level.

Kelly Olynyk was the second most efficient player on the team with a WP40 of 0.284; unfortunately he was also 8th on the team in minutes played and as a consequence only produced 0.46 wins. Olynyk was productive mainly because of his shooting efficiency, scoring, ability to avoid fouling, and, to a lesser extent, his ability to protect the ball. In all other areas he was below average, but not very much so. So Olynyk was successful mostly due to his scoring skills and not much else.

Ryan Bell was the third most efficient Canadian team member, with a WP40 of 0.255. Unfortunately he played even fewer minutes than Olynyk did, and so only produced 0.340 wins. Bell scored an amazing 1.75 points per shot! That rate is so high, he could've halved it and still had an above average PPS! Unfortunately he only attempted 3.0 shots per 40 minutes - so ridiculously low that it's also the same as his free-throw rate (which is still below average for free-throw attempts). Thus, despite his crazy PPS, he still scored substantially less than an average guard. He also rebounded at a rate slightly below average and was poor at getting assists. Other than that, Bell was above average with respect to steals, TOs, net possessions, blocks, and fouls.

The team's only player with NBA experience, Joel Anthony, was the fourth most productive and last member of the Canadian team who managed to produce at an above average level. He produced at a rate of 0.158 WP40, and because he played the most minutes on the team (149 MP) he produced the second most wins (0.571). Anthony did three things very well: block shots, avoid fouls, and shoot efficiently. He did two things quite poorly: free-throw conversion and rebounding. He was also poor at gaining possessions and getting assists, but not too far away from average.

Together these four players produced 2.277 wins, or 253% of the team's wins, in 38.6% of the team's minutes.

The Barely Positive:

Only two other players managed to produce a positive number of wins for Team Canada during the 2010 WBC. They are (in order of EWP40): Denham Brown and Jermaine Bucknor.

Denham Brown was not very productive, but at least he managed to produce in the posivitve range. He scored at an above average rate, but was not very efficient while doing so. Other than that, in every other area, with the exception of TOs and fouls, Brown was below average. Because he played the 4th most minutes on the team, he managed to produce 0.081 wins.

Jermaine Bucknor was slightly less efficient than Brown, posting a WP40 of 0.027. He shot very efficiently, but did not take very many field goal attempts, keeping his scoring below average. Other than his steals and TOs, everything else was below average, particularly rebounds and fouls. He ended up 7th on the team in minutes and produced 0.023 wins.

Together these two players produced 11.5% of the team's wins in 18.6% of the team's minutes.

The Negative:

The remaining six players produced in the negative range. They are (in order of EWP40): Jermaine Anderson, Robert Sacre, Andy Rautins, Aaron Doornekamp, Jevohn Shepherd, and Olu Famutimi.

Jermaine Anderson just missed being a positive producer of wins. His assists, steals, and FT attempts were all slightly above average, but other than that everything was below average. Especially poor was his AdjFG%. Anderson somehow managed to play the second most minutes on the team, and in that time he produced -0.026 wins.

Robert Sacre was the backup centre (behind Anthony) on the team. His steals, rebounds, FT%, and net possessions were above average; unfortunately everything else was below average (with the exception of FG attempts...and at his PPS, this actually hurt his numbers), particularly his PPS and PFs. In 44 minutes, Sacre managed to produce -0.123 wins.

Andy Rautins was injured and did not play well. His scoring and rebounding numbers were the only areas that were above average; particularly poor were his net possessions and (especially) his PFs. In 39 minutes across three games, Rautins produced -0.153 wins.

Aaron Doornekamp also played while injured, and like Rautins, did not play well in that time. His poor shooting and rebounding really hurt his productivity, but he managed to be above average wit hrespect to steals, TOs, and assists. Rautins was 5th on the team in minutes played and produced -0.464 wins.

Jevohn Shepherd accumulated steals, avoided TOs, and shot free-throws at above average rates. Unfortunately he couldn't shoot (0.69 PPS) and rebounded worse than the average guard (3.6 reb/40). Everything else was also below average, but his shooting efficiency and rebounding were what hurt him the most. He played the 6th most minutes, and in that time, produced -0.512 wins.

Olu Famutimi played very limited minutes over the course of two games, which is very fortunate, because he didn't manage to perform at an above average level in any category. In his limited time he somehow managed to produce -0.203 wins.

Together these four players produced -165% of the team's wins in 42.8% of the team's minutes.

Overall Team Performance

The fact that the four worst-performing players on the team managed to play 42.8% of the team's minutes suggests there were a few errors in player management. In fact, if the minutes were divided amongst the players by productivity, the team would've produced 2.43 wins instead of 0.900.

What's interesting is that the top 5 players consist of two forwards, two guards, and one centre, and the top 3 off the bench are a forward, a guard, and a centre. Of course, it's easy to say which players should've played the most minutes after the fact. The coaches of the team had the very difficult task of dealing with determining minutes on the fly, often with limited sample sizes (that is...if they used stats to determine minutes). In any case, this tournament should provide some good data going forward - and the good news is that the oldest member of the team is 28 year old Joel Anthony. If the coaches can fine tune their minute allocations, this team could be more successful in future (look out for Kelly Olynyk in the years to come).

 - Devin.

Wednesday, September 8, 2010

Introducing the Glossary of Terms

When I make images, for simplicity's sake, I don't like putting in formulas underneath my tables; they take up space and complicate formatting. The problem with this is that, if you are unfamiliar with how I put together my table, you may not know what the hell I am talking about.

So let me introduce the Glossary of Terms, located on the right-hand sidebar under the "Pages" section. On that page I'll be adding most of the acronyms, jargon, and formulas that casual fans may not necesssarily be familiar with. If you are used to basketball stat terms you probably don't need to visit the page, but if you are ever curious about anything, it's there for you.

 - Devin.

2009-10 Player Review: Julian Wright

I've already posted a little bit about Julian Wright already, but let's take a more in-depth look this time:

Deja vu, man

Wright was something of a disappointment in New Orleans. He was one of those rare rookies who managed to be an above-average player in his first season, only to regress rather substantially for two straight seasons. Let's look for some clues as to how he lost productivity:

He started off shooting the ball very efficiently, breaking 1.10 PPS. He followed that up with seasons of 0.94 PPS and 1.01 PPS; with such volatile shooting efficiency, it's difficult to predict where he'll end up next year. There are some clear trends, though - Wright is a poor free throw shooter, a poor scorer, doesn't take many FG or FT attempts; is a below-average rebounder, possession-getter, blocker, and passer; and is above-average with respect to steals, TOs, and PFs. For Wright to improve next season, he should try to improve his rebounding and increase his shot attempts (while maintaining or improving on the efficiency level he showed last year).

While his overall progression does not look good, he is only 23 years old and I have him penciled in for a WP48 of 0.050 next season - right around his career mark. It is entirely possible that the change of scenery will do him good and he will return to the rate of production we saw from him as a rookie; then again, it's also entirely possible that his downward trend will continue and he will enter into negative WP48 territory. Based on Wright's career so far, neither would be exceptionally surprising.

 - Devin.

Saturday, September 4, 2010

2009-10 Player Review: Linas Kleiza

Let's take a look at the Raptors' international man of mystery, Linas Kleiza:

Well actually...this is more of a 2008-09 player review, but we'll pretend that it isn't.

Following the 2008-09 season, Kleiza played over in Europe for the Greek team Olympiakos. He was the leading scoring in the Euroleague (17.1 pts/game [a game is 40 minutes over there]) and as such won the Alphonso Ford Trophy. After the season ended he terminated his contract with Olympiakos and signed a 4-year, $20 million deal with the Raptors. Before going over to Europe, Kleiza played 4 seasons with the Nuggets - the last two of which he played at a near-average level. The following table includes his more specialized statistics for these last two NBA seasons as compared to the average SF and SG (the positions he mostly played) as well as - for fun - his Euroleague stats from last year, although please note that these numbers shouldn't be directly compared to NBA numbers.

The numbers show that Kleiza shoots with above-average efficiency, scores, rebounds, and gains possessions at an above-average level. Other than that, he is at or below average with respect to steals, blocks, assists, and fouls. As long as the Raptors don't try to play him as a PF, Kleiza should produce at a near-average to above-average level this year; as a 25 (soon to be 26) year-old, he should also continue to show some improvement. Based on past performance, I'm expecting a WP48 of at least 0.083, and he'll probably reach a WP48 of 0.100.

 - Devin.

Friday, September 3, 2010

2009-10 Player Review: Leandro Barbosa

Let's let a look at the Brazilian Blur, Leandro Barbosa:

To say that last year was not a good year for the Brazilian would be an understatement; he went from a nice WP48 of 0.130 in 2008-2009 to a WP48 of -0.032 and only managed to play a career-low 44 games (and a career low 786 minutes played). Moreover, he continued his trend of alternating between good and bad years. Let's take a look at what changed over the past three seasons:

Barbosa gets most of his productivity from his scoring. It's not surprising, then, that when his shooting efficiency nosedived last year his productivity nosedived as well. His FG attempts were actually up, but his AdjFG% and FT attempts were down. Most of his other productivity mesasures were down and below-average (as they usually are for Barbosa).

Why did Barbosa's shooting efficiency suddenly drop off a cliff? Perhaps the answer lies in his shot selection:

Data provided by HoopData

Looking at these numbers, I notice a couple of things:
  1. Barbosa's percentages are down at every location on the court
  2. The percentage of his close-range shots (rim and <10ft) that were assisted last year increased
  3. The percentage of his mid-range and long-range shots (shots >10ft) that were assisted last year decreased
To me that suggests that Barbosa's shot selection was off last year (by off, I mean compared to his normal self). Out on the perimeter he tried to do it by himself too often, and then end result was a low-percentage shot. He had trouble getting closer to the basket by himself - likely due to injuries - and so most of his close-range shots were set-up by teammates. Of course, this is just a theory. An alternate theory is that he simply forgot how to shoot last year.

What's the outlook for this year? I've projected him to have a WP48 of 0.080, but if he has recovered from his injuries and maintains his pattern of alternating between bad and good seasons, he could improve on that mark. Then again, if he really has forgotten how to shoot, he could be significantly under that mark. Regardless, if he plays around 2000 minutes on the season, he will likely produce around 3 wins.

 - Devin.