Posted: Mon Feb 14, 2005 1:05 pm Post subject: Game Situations
I want to try spinning yet another thought off from the various threads about the value of high-efficiency / low [minutes,touches] players.
Are all game situations equal? Is the ability to score a basket in the middle of the 2nd quarter the same as the ability to score a basket in overtime?
It seems like most of our analysis makes the easy assumption that, because the score at the end of the game equals the sum of the scores in each minute of the game, production is equally valuable at any point in the game.
The only reason that I have to question this assumption is that the way that coaches manage players with fouls (at every level of the game) is absolutely counter to this. I've always wondered about the fact that coaches will pull a player who accumulates a second foul in first quarter. It seems to me that if you want to get maximum minutes from the star players you should just play them. They may or may not foul out and if they don't you get more minutes of play from them then you would if you benched them and if they do foul out it just means that you're playing their replacement in the 4th quarter rather than the 1st quarter.
I've never seen a coach do that.
There is a conventional wisdom that in both the closing minutes of close games and in playoff games that refs call foul differently, teams play more defense, and the ability to create one-on-one is more valuable.
This, I think, is part of why people tend to think of Paul Pierce or AI as more valuable players than Brent Barry.
[counter-example, AI's lack of success in hitting game winning shots has been well documented and, IIRC, in the 2001-02 season there was a sequence of games in which Brent Barry hit game winning 3-pointers three games in row. I may be misremembering, and he may have shot potential game winning shots in 3 consecutive games and hit 2 of them.]
My impression is that in baseball analysis has tended to show that there are no "clutch hitters" (or, at least, that they aren't statistically identifiable) I don't know if it's the same in basketball.
In the regular season this probably doesn't make a big difference. Even if it's true that the game changes in certain situations, most teams don't play *that* many close games over the course of the season. If a typical team plays 14 close games in a season an increase in winning percentage in close games of 7% will only add one win to the season.
So I'm curious, do people think that there are situations in the game in which player production could be affected by the situation, and has anyone done the research?
I'm raising the question but, as I think about it, I suspect that this is a bit of the conventional wisdom that has little or no basis in fact.
But that doesn't explain the way that coaches handle foul trouble.
Joined: 10 Jan 2005 Posts: 30 Location: Los Angeles
Posted: Mon Feb 14, 2005 6:52 pm Post subject:
Others with more direct experince can speak better to this than I, but some teams clearly are taking situation into account with their statistics. Specifically, Mark Cuban noted in his blog that the Mavs use a +/- system that is weighted by both time of the game and how close the game is in score.
Some basic season-long stats are not going to look at situations, but if I worked for a team I would think having some weight to time and situation of a score would be of interest to GMs/coaches.
Brent had three game-winning threes in a span of about two weeks in late November-mid-December 2002.
I'm not sure that the way fouls are handled necessarily says a great deal about what coaches are really thinking. I think there's a certain conservative bias here in that a coach will never get criticized for putting a player in foul trouble on the bench. If the team falls behind ... well, that's the nature of the game. If, on the other hand, they leave the player in and he fouls out early, the coach will be blamed if the team loses.
Joined: 14 Jan 2005 Posts: 1501 Location: Delphi, Indiana
Posted: Tue Feb 15, 2005 7:41 am Post subject:
I like Nikos' thinking on leaving players in the game -- if I'm playing a computer simulation of a game.
Real life has human elements. A player gets out of rhythm for a spell. A ref is ticked off at him. The team isn't working harmoniously, setting up one guy for foulish situations.
The coach's job is partly to break up the rhythm when something isn't going well; then try again later. Timeouts are like that.
My gut feeling on monitoring game-time situations is that in the course of a season, there's not sufficient sample size. A Brent Barry may hit 3 game-winners in 2 weeks in a simulation game, too. Maybe, given his prodigious efficiency, he should have done that regularly.
Joined: 03 Jan 2005 Posts: 660 Location: Washington, DC
Posted: Tue Feb 15, 2005 8:19 am Post subject:
To piggyback on Mike's point -- players play differently when they've picked up a few fouls. Every now and then, you'll see a coach leave a guy out there or fail to get a sub to the scorers table in time. The guy in foul trouble usually plays tentatively and passively.
I like Nikos' thinking on leaving players in the game -- if I'm playing a computer simulation of a game.
Real life has human elements. A player gets out of rhythm for a spell. A ref is ticked off at him. The team isn't working harmoniously, setting up one guy for foulish situations.
The coach's job is partly to break up the rhythm when something isn't going well; then try again later. Timeouts are like that.
To sidetrack a bit, I don't know if I agree with this blanket statement. Phil Jackson was famous for not using timeouts in situations like this, preferring the players work out their problems on the court. Sure there are times when a player is out of rhythm or a ref is ticked off, and a time out seems to calm things down. But can we attribute one to the other? Aren't there equal amounts of time when, as Bill Walton likes to say, "the moment shifts" without a timeout?
Take the first quarter of the Knick game yesterday. Iverson beat Marbury early & often and would get hammered going to the hoop. This must have happened at least 3 or 4 times. At one point instead of running up the court AI gives it to the ref, who T'd him up. No timeout is called. After the shot, play continues.
At that moment it looked like things were getting out of hand for the Sixers. However afterwards, Iverson got every single call up the court, some of which were ghost calls. Kurt Thomas would pick up 2 quick fouls helping a toasted Marbury, and get sent to the bench. By the half Iverson had 31 points, a dozen were from the line. Herb Williams called a few time outs in the first half, but it didn't seem to matter as the Knicks went down by 12.
When something isn't working for a team, we sit like Pavlovian dogs waiting for the coach to call a timeout. However is that because a time out actually does something, or because that's what we expect? I think it's just an excersize in control, where the coach feels he has to do something, and since he can't just run onto the court and block a few shots his only outlet is calling a timeout.
Marbury can throw an errant pass, Crawford can take a bad shot, Nazr Mohammed can miss a 5 footer. If they happen spread out during the game, it's no big deal. Put all 3 consecutively and all of a sudden it's time for a break. So is this the human element showing up, or is it the clustering illusion?
We want our coach to call a timeout when the other team is scoring at will, or our team is coughing up the ball. If after the timeout the problem goes away, we attribute it to good coaching. However how many times after a timeout does the same condition persist? Do we even acknowledge this as a failed time out? If the coach is calling a time out because his players are fatigued or he'd like to change the alignment to an offensive/defensive squad then I'm all for it. Otherwise I'd like to see a study that shows calling a timeout and putting the same 10 guys on the court makes a difference.
Joined: 14 Jan 2005 Posts: 1501 Location: Delphi, Indiana
Posted: Tue Feb 15, 2005 11:15 am Post subject:
Presumably, timeouts are called so the coach can talk to his players about what to do differently; and perhaps make substitutions.
But I also prefer the (unprovable) theory that momentum just Happens, and there's only so much you can do about it. In my version of the theory, the wind shifts for X minutes, and you live through it to the next wind shift.
A timeout just carves a chunk out of the prevailing trend. If it's a 5-minute trend, you might lose 10 points of a lead. With a timeout thrown in there, you might only lose 6 points.
Completely unprovable, and you'd likely find many coaches that agree. It's always going to be a gut thing. Coaches that have confidence in their team might just expect the curse to go away, and thus let them play thru it. Weak-willed coaches will tend to throw in the towel early on.
Real life has human elements. A player gets out of rhythm for a spell. A ref is ticked off at him. The team isn't working harmoniously, setting up one guy for foulish situations.
That theory argues for a flexible approach to decisions about players in foul trouble. Certainly, not all foul trouble results from "foulish" situations; sometimes the player is just foul-prone or unlucky. Yet coaches never (or at least very rarely) leave the player in in those situations. Coaches' reactions are equally mechanized as always leaving players in.
I like Nikos' thinking on leaving players in the game -- if I'm playing a computer simulation of a game.
Just for the record, I'm not the same person as Nikos.
Mike G wrote:
Real life has human elements. A player gets out of rhythm for a spell. A ref is ticked off at him. The team isn't working harmoniously, setting up one guy for foulish situations.
The coach's job is partly to break up the rhythm when something isn't going well; then try again later. Timeouts are like that.
I agree with the other people that have commented that this is an accurate description of the conventional wisdom, but that we don't know if it is an accurate description of reality. It may be accurate but, as far as I know, it is a statement that could be tested empirically and nobody has done that.
Mike G wrote:
My gut feeling on monitoring game-time situations is that in the course of a season, there's not sufficient sample size. A Brent Barry may hit 3 game-winners in 2 weeks in a simulation game, too. Maybe, given his prodigious efficiency, he should have done that regularly.
Agreed on the lack of sample size. As for Brent Barry my argument is just that he didn't appear to lose value in clutch situations. Given that Brent Barry was a topic of much discussion in the other thread I'm just using him as a proxy for the general argument, "Fred Hoiberg may be good, but when the game's on the line you need to get the ball in Garnett's hands." That statement has a certain initial plausibility, but I find myself doubting if it is true.
Joined: 10 Jan 2005 Posts: 30 Location: Los Angeles
Posted: Tue Feb 15, 2005 3:08 pm Post subject:
KnickerBlogger wrote:
Phil Jackson was famous for not using timeouts in situations like this, preferring the players work out their problems on the court. Sure there are times when a player is out of rhythm or a ref is ticked off, and a time out seems to calm things down. But can we attribute one to the other? Aren't there equal amounts of time when, as Bill Walton likes to say, "the moment shifts" without a timeout?
Jackson had the luxury of teams that were sure to make the playoffs, he could afford to sacrifice a game to teach a lesson. Not that it ever came to that, but if his Bulls/Lakers teams lost to Golden State early in the season because he wanted them to work through issues, it was not a big deal to him. Not all other coaches have that luxury, nor do they have the more veteran teams who will learn quickly from that experience. It's a good strategy with the right team, but it's not for everyone.
That said, your point about assuming calling a time out is always good coaching is well worth consideration.
Joined: 30 Dec 2004 Posts: 529 Location: Near Philadelphia, PA
Posted: Tue Feb 15, 2005 4:45 pm Post subject:
Kurt wrote:
KnickerBlogger wrote:
Phil Jackson was famous for not using timeouts in situations like this, preferring the players work out their problems on the court. Sure there are times when a player is out of rhythm or a ref is ticked off, and a time out seems to calm things down. But can we attribute one to the other? Aren't there equal amounts of time when, as Bill Walton likes to say, "the moment shifts" without a timeout?
Jackson had the luxury of teams that were sure to make the playoffs, he could afford to sacrifice a game to teach a lesson. Not that it ever came to that, but if his Bulls/Lakers teams lost to Golden State early in the season because he wanted them to work through issues, it was not a big deal to him. Not all other coaches have that luxury, nor do they have the more veteran teams who will learn quickly from that experience. It's a good strategy with the right team, but it's not for everyone.
That said, your point about assuming calling a time out is always good coaching is well worth consideration.
Quick comment. Jackson often did the same in the playoffs, letting his guys play out of bad streaks. Dean Smith did it all the time in Carolina. Drove me nuts at first, but I see the rationale. You want to make timeouts productive. If you're just calling them because the team is playing poorly but don't know exactly why, what are you going to do to fix it? If you see a match up problem, if you see someone hurt, if you see a tactic that you weren't prepared for, if you see guys are tired -- then you can actually do something in the timeout. When I was coaching, I actually did this but didn't realize it (I'm more of a maniac from the seats than the bench). Maybe just calming guys down can help, but it depends on the team. A mature team won't let runs bother them if it is just luck. And a mature team comes from advance prep, I think.
In general, I do think that timeouts, as fundamental a tool as they are for coaches, are one of the mostly poorly understood aspects of this game. My analysis in BoP was based on statistics but not all that advanced. Roland did something last summer showing what teams did well and poorly after timeouts and that study, as inconclusive as it was, has kept my mind going since. Timeouts are zero-sum with both teams getting time to "think about it". But we seem to think that they are an advantage for the underdog, which Roland couldn't really study. It calls for a different but not overly hard study, I think. _________________ Dean Oliver
Author, Basketball on Paper
http://www.basketballonpaper.com
Jackson had the luxury of teams that were sure to make the playoffs, he could afford to sacrifice a game to teach a lesson. Not that it ever came to that, but if his Bulls/Lakers teams lost to Golden State early in the season because he wanted them to work through issues, it was not a big deal to him. Not all other coaches have that luxury, nor do they have the more veteran teams who will learn quickly from that experience. It's a good strategy with the right team, but it's not for everyone.
That said, your point about assuming calling a time out is always good coaching is well worth consideration.
Jackson would do the "no-time thing" in the playoffs as well, when he didn't have the luxury of sacrificing games. I just think it was his style of coaching, whether it was because he saw it as a random string of luck or as a psychological tool to motivate his team.
But I also prefer the (unprovable) theory that momentum just Happens, and there's only so much you can do about it. In my version of the theory, the wind shifts for X minutes, and you live through it to the next wind shift.
A timeout just carves a chunk out of the prevailing trend. If it's a 5-minute trend, you might lose 10 points of a lead. With a timeout thrown in there, you might only lose 6 points.
I like to see it as a coin toss. The other team has just flipped the coin 4 times in a row in their favor. You call a time out because you want to change the momentum!?!? No, you call a TO, because you attribute the bad result as your opponent's ability to fix the throw. You want to make sure your opponent hasn't found a way to flip the coin in his favor.
As for the wind theory, when you call the time out, how do you know that the wind isn't going to shift in your favor for those 5 commercials?
Finally didn't they find calling a time out before a game winning FG in the NFL isn't effective?
I agree with Mike's general sentiment. There are a host of psychological factors that go into game situations like this that at least seem likely to be contributing factors to what happens on the court.
For instance, suppose a player gets 2 fouls in the first quarter. As noted, the coach will almost definitely sub out the player, even if it seems likely that the player might actually log more PT in the long run if left in. But this coach's thinking is, of course, shared by the opposing coach and all the players on the floor. As was also previously noted, the player will tend to play more tentatively and passively, even if there might be little reason to adjust play this way from a statistical standpoint. Following the general perception that this player is in trouble, opposing coaches and players in this situation will smell blood and will probably be more likely to attack the foul-plagued player and try to draw additional fouls (especially for star players). Even the refs might be cognizant that this is an exceptional situation (whether it really needs to be that way or not) and change the way they call fouls accordingly (though in some cases, they might actually be more hesitant to blow the whistle).
Another situation brought up here is when a team calls timeout to 'break up' a run. Again, on the face of it, it may be the case that there is little immediately apparent motivation for such a tactic. But consider the psychology of the players. If the downtrodden team believes that the opposite team is really dominating them, and it's not just a fluke, their beliefs will tend to perpetuate their ineffectiveness. A timeout could be useful in such a situation to regain players' composure. There is also the more obvious influence of the crowd. If the enthusiasm of the crowd plays any role in how teams play (and it seems as if it does, given home court advantage), then quieting the crowd is useful for the road team, both to boost their own morale and to undercut that of the opposition.
The basic point is that, even if a given situation appears indistinguishable from others when considered from a distanced perspective, what really matters is what tends to happen on the court. And what tends to happen on the court will be greatly affected by the attitudes and perceptions of the players and coaches, regardless of other considerations. If I'm playing a game of cards and losing a string of games, and if I believe that I'm losing due to lack of skill instead of lack of luck, then my belief will probably make me more likely to make poor decisions or otherwise be distanced from my optimal strategy, regardless of whether my belief is based in reality or not.
Of course, this is only a defense of why situational game adjustments such as the ones mentioned in this thread are at least plausible. It would be nice to see a detailed statistical treatment that comes to one conclusion or another. However, even the interpretation of such analysis seems contingent on what pre-empirical stance we choose to pick. For instance, if we believe that there are psychological or 'momentum' factors at work, then we might choose to weight the performance of the 'runned-on' team more heavily than that of the 'run-ee's, such that a break even performance after a run interrupted by a timeout would reflect positively on the runned-on team instead of showing no benefits.
I think the problem with the psychological conventional wisdom is that there's a lot of possible backwards relationships.
Do players perceive a run as a big deal precisely because their coaches treat it as such? It seems a lot of Jackson's theory is that his team will be able to play through runs because he shows faith in them by not calling timeouts.
Do players treat staying on the court with foul trouble differently because that usually means leaving the game? Would we see players play tentatively if they regularly played with foul trouble, or would they learn to play through it? (Side note: At what point is a tentative starter a better or worse alternative than a weaker backup?)
The problem with introducing pyschology into these situations is that the number of expanations you can create is virtually endless.
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
