Do Pre-Season Games Matter?
During the pre-season, coaches frequently don’t even play their starters at the end of a game and often give plenty of time to marginal players, making it all seem somewhat inconsequential. But is that true? For those of us obsessed with hoops we still care about pre-season games ‘cuz we’ve been jonesing for some bball and we love to parse any little bits of data, trying to extract larger meaning from it all. We know that if the Knicks lose a game where they have Andy Rautins and Landry Fields playing the key final minutes of the fourth quarter that we shouldn’t put much stock into the loss. Although, god dang it, we sure wish we’d won.
In Preseason, Do Coaches Give Bill Murray’s “It Just Doesn’t Matter” Speech?
Even amongst coaches the pre-season is treated differently. Some care about it, others don’t. Some use it to fine-tune their teams, others look at as something that eliminates practice time (particularly if you have to travel overseas for a game, not only adding endless hours going there and back, but due to jet-lag your players are out-of-sorts for like a week while they try to quickly adjust going there and then coming back). Also, for some coaches, it’s less about their philosophies, and more about the team they’re given. Doc Rivers has to be very careful with his elderly, injury-prone crew of Celtics, forcing him to deal with the pre-season far differently than if he had a young, healthy exuberant bunch who needed to learn a lot.
Regardless, fans, players, coaches, media and even select Martians have been conditioned to believe that pre-season games don’t matter. So I found it intriguing when 48 Minutes Of Hell (a San Antonio Spurs blog) posted some research today showing that pre-season results are a pretty decent harbinger for the regular season.
First, let’s clarify what they mean by pre-season “results.” For those who haven’t read much about higher basketball stats, let me briefly introduce the concept of Scoring Margin, which is considered the best measure of how good a team is. One would think that, duh, win-loss is a pretty dang good measure of how strong a team is. Maybe not after a few games, but over the long haul, it’d seem logical. However, stat-geeks have discovered that a better predictor of success is how many points by which you win/lose.
Like we’ve all seen games where it comes down to the last possession, and maybe, as with one of the games in the Lakers/Oklahoma City playoff series, a player (Kobe) shot an airball, but another one (Ron Artest) scrambled to read it and got a relatively easy put-back. Yes, that player hustled, however the victory doesn’t show that one team is definitively better than the other, but more that in a sense they got a lucky bounce (even though that particular Laker shot hit nuthin’ so it didn’t bounce, but you all know what I’m saying). If we just looked at win-loss record, then that loss counts as much as say when Orlando was beating the Charlotte Bobcats by over 20 points a game. Of course, as with any stat, you need a large sample size for it to be significant. Like if the Lakers get bored during the regular season and only beat say the Nets by one point, it doesn’t indicate they’re only slightly better than the Nets. However, if Team A ends up winning say 55 games by an average of 10 points (a huge margin), while say Team B also wins 55 games but only by an average of 1 point, it makes sense to think Team A would be a bunch better despite having the same record.
All of which goes to say, that when 48 Minutes examined pre-season games, they correctly looked at scoring margin rather than win-loss to see what, if any, correlation there was between that and the regular season scoring margin. They also looked at the first 8 games of the regular season. That way they could compare whether those first 8 “real” games would predict how the team would be for the rest of the season, better than the 8 pre-season games. If you want to see the data, check out the great post. The non-stat-speak part of what they found is this:
Not only do good teams play better in the preseason, but, after accounting for the number of games played, the preseason isn’t much worse at gauging regular season success than the regular season itself.
[…] I suspected that teams with poor records last year would have preseason records that would be more meaningful than better teams. This is because poor teams would be more likely to spark personnel turnover and to rebuild their roster. Thus, they have younger players and higher draft picks, who play more minutes in the preseason.
In instances of significant player movement, such as what the Heat experienced this year, preseason performance can paint a much clearer picture of future expectations.
As such, I sorta feel like the Knicks, being in a far different position than the Lakers or Celtics, should be playing their starters more minutes during this pre-season. While it’s important to give like Andy Rautins a chance to show what he can do, it’s far more important for say Amar’e Stoudemire and Raymond Felton to get the pick-and-roll down. If we go into the regular season with those guys clicking, it’ll provide far more wins than if we confidently know that Rautins will be our ninth man in the rotation. Yes, Felton and Stat are probably spending tons of time doing that in practice, but they only go up against the second unit in practice who’s defensive schemes you know intimately (since they’re your own). Even if Kevin Garnett and Rajon Rondo are only in for the first few minutes of the first and third quarter, it’s a much better test going up against them.
We have a ton of new players learning a new system. And even those who do know the system intimately (like Amar’e, Gallo and Chandler) need to learn how to work with totally new teammates. Personally I’d be playing the starters near their regular season minutes (like Amar’e and Felton should be going say 30-35 minutes, instead of the 36-42 they might see later on).
And I’m not just saying that ‘cuz 48 Minutes has shown that if we win in the pre-season it’d “assure” us of being good when the games count.