There's a nice blog I recently found called basketballgeek.com, which focuses on numerical basketball analysis. One of the features is he maintains parsed play by plays which you can download. He doesn't have every game in there, unfortunately, but the majority are available. As of 3/8, he has 47 Rockets games play by plays available in csv format (about 75% of them). The missing games are spread out throughout the season, so I'd say what's available is a pretty good representative sample. http://www.basketballgeek.com/data/ If people are curious about any particular information that can be found in play by plays but aren't available at, say, 82games.com, let me know I can try to find it out for you. Here's a quick start, looking at field goal scoring from the three bigs in our rotation: Yao: made shots: 272 FG% on made shots: 56.3% made after off rebound: 19 (7% of made shots) FG% on shots after off rebound: 65.5% made and assisted: 168 (61.8% of made shots) made, not off rebound, not assisted: 85 (31.3% of made shots) Scola: made shots: 240 FG% on made shots: 53% made after off rebound: 8 (3.3% of made shots) FG% on shots after off rebound: 40% made and assisted: 166 (69.1% of made shots) made, not off rebound, not assisted: 66 (27.5% of made shots) Landry: made shots: 145 FG% on made shots: 55.1% made after off rebound 16 (11% of made shots) FG% on shots after off rebound: 66.7% made and assisted 98 (67.6% of made shots) made, not off rebound, not assisted : 31 (21.4% of made shots )
Stats have gone to another level now-a-days. In my opinion, their really shouldn't even be stats in the league. It really hurts everybody's opinion on players. I'm surprised that they have all these stats before they have players holding on to another player. Like when I watch on ESPN, They would usually show what's Kobe stats or another superstar when a certain player guards them. They should have that stat already. I like advanced stats, but It has gone way too advanced.
Hmm ... so what's wrong with it again? I honestly couldn't make out your argument. I was curious how often our bigs scored on assisted baskets, not including scores off offensive rebounds. You think there's something wrong with knowing that information?
No, I think it's fine that they are giving out these stats. I was just listing my opinion on stats overall. I thought those stats were available on 82games. It does show you how many times Player X gets assisted for score. http://www.82games.com/0809/08HOU16.HTM Yao gets assisted 59% of the time, which is weird, because It doesn't look like that on camera.
Just based on assisted baskets, Landry's baskets are assisted slightly less frequently than Scola's. One might look at that and think that he's therefore better at creating shots for himself. But I wanted to check how often his made baskets came immediately off an offensive rebound. It turns out, for Landry, a good number of them did. So, actually, a higher percentage of Scola's made baskets were unassisted and not off an offensive rebound (which, if you think about it, is almost like an assist from the guy who missed the shot ... Rafer led the league in those assists ). It was also interesting how much better Landry is finishing those shots off offensive rebounds. Over those 47 games, Scola was 8 for 20 when shooting a shot immediately after getting the offensive board. Landry was 16 for 24.
Nah man, stats are always great. For those that know how to use them, they can be either good or not very informative. Some people just don't understand them and unfortunately still try to use them for argument. I don't think there can ever be enough of stats. Of course, I do think only those that truly understand them should discuss them, otherwise, keep your mouth shut, but we know that will never happen. Anyways, I see no reason to hate on stats. Stats did nothing to us, is those that misinterpret them or just don't understand them that we should be "hating" on.
You got a pretty good point, I am just sick of everybody's opinion due to stats. Example: Dwight Howard vs Yao Ming, everybody just says well since Dwight's got the better stats, that must mean he is better? No, that's not the case at all. We know who the best center is and so does Shaq. Another example would be the current Kevin Garnett. He is having one of his worst stat seasons since his rookie year. People don't understand stats don't always tell the whole story, they really just tell half of it. They don't understand that their are a lot more to the game than stats, and their are a whole lot more to determine who is the better player.
This can lead to an interesting discussion. Simple question: Can you name one thing that is important to the story that can not be represented as a "stat" (i.e. quantified information)?
Agreed - stats actually provide some objectivity as to what is going on. Often there can be statistical anomalies - but when evidence is built up over a long enough period of time the anomalies are smoothed out.
the story of the houston rockets winning the champ has one unquantifiable phenomena that refuses to be represented by even something as powerful as stats, he is that which which holds the team together, they call him ... 'Batman'.
Pick n Roll defense, Leadership, Intensity, Skill, and desire to win. You cannot measure stats that way without watching the actually game. People just get it off reputation.
You can certainly measure pick and roll defense. I'm sure every NBA team keeps detailed information on how they're players defend against the pick and roll. Basketball skills, if you can define it, can also be measured. As for leadership, intensity, and desire to win ... I consider all those to be important, and you're right its more difficult to measure. But let's look at it more closely. Take leadership. How do you know leadership matters? Can you explain how that translates on the court in terms of how the players perform? If does not translate in any meaningful way, then we can discard it as a significant player attribute. But if it does translate, then we can attempt to measure it in terms of the ways it manifests itself on the court.
Leadership doesn't have a stat, which is why we have a GM and not just big computers. Pick n roll defense can be quantified easily. FG% against pick n rolls, do it for every team, rank yourself. I think durvasa has shown that skill can be as well - can you pass, will you turn it over, are your shots assisted, can you create your own, are you strong going to the line, etc... These can be quantified under skill. Intensity is worthless in basketball. A lot of it can be good or bad, and a little of it can be good or bad. See Tim Duncan. You can quantify "high leverage moments", clutch plays, and composure in high pressure moments. These are the most important. Also, see Hakeem's jump in stats from reg season to playoffs. That's intensity isn't it? It could even be argued that it's desire to win. Lots more can be quantified than we think. I'm sure if you sat in Morey's lab, you'd be shocked at the stats he monitors.
Something to think about is the stats for a team's record in games decided by three points or less. Well, that doesnt count games that are within 3 that get skewed when the losing team misses their final shot only to have a foul give the other team 2 more points making the spread a bit higher....
How the hell would we measure it as it manifest though? I highly doubt anyone can come up with a stat like. It would be ridiculous to attempt to measure when someone is attempting to be a leader and when they are letting someone else lead. I guess we could lose closely at those who we assume are the leaders and measure it in some way, but what about those we don't know are the leaders, they just do it different, how the hell would we measure that? I doubt we could ever attempt something like that. Just wanted to join the discussion, I know you aren't suggesting we measure it (or at least I don't think so).
What I'm saying, basically, is that if its observable then it's measurable. If it's not observable, then we should question its significance. Juwan Howard, I have no doubts, was a great leader off the floor. How did that manifest itself on the floor?
I remember back in the good old days (pre-espn) how when a player averaged 21 points per game, it was 21 gawdoggit.....not 21.4 or 21.8....
durvasa, I love stats and I love your analysis most of the time. I don't agree with your statement here. Just because we cannot find a way to "translate" something in a quantifiable way does not mean that it does not exist. For example, how do you quantify Mutombo's presence in the locker room or on the bench? Does his presence make a difference? Some people think so. I am not sure if I can deny it. But how do you support or deny the claim "The veteran leadership provided by Mutombo really helps the team" in an objective way?
defense and chemistry. 2 of the most important part of a winning team, other than putting the ball inside the hoop.
