|
|
|
Monday, December 11, 2000
Sorting through e-mails & stats
By Joe Lunardi
Special to ESPN.com
|
Welcome to another edition of "Box Score Banter." With e-mail traffic
picking up here at Box Score headquarters, please remember to include
your full name and location in future submissions (jlunardi@home.com).
Here we go ...
Delighted to learn that ESPN will feature sophisticated analysis of basketball this year. As food for thought, here are a few basketball metrics that I have always desired to see:
ADJUSTED FIELD GOAL PERCENTAGE: When calculating field percentage, give players credit for 1.5 FG for each "made" three-point shot. Since a three-point field goal is worth one-and-a-half times as much as a
two-point field goal, this would seem to make perfect sense. Players who take lots of 3-pointers would have their value shown in a truer light with this metric compared to regular, unadjusted field goal percentage.
RATIO OF ASSISTS TO MISSED FIELD GOALS: I feel that this is the best way to assess a player's passing ability, both in terms of decision-making and physical execution. After all, when a player has the ball in his hands, there are two things he can do: pass or shoot. Assists-per-FG is not an optimal stat for evaluating passing, in my
view, because in effect it punishes a player for making a field goal. This is quite wrong; making a basket is the best thing a player can do on the floor, so a player should not be punished for it (even for the purposes of measuring something else). Assists-per-Missed-FG juxtaposes good decisions to pass versus bad decisions to shoot, which is exactly
what we want to know in evaluating a player's passing ability.
TURNOVER RATIO (steals + blocks/turnovers): Players make turnovers,
but they also force them. A player's net turnover contribution to the
team would be useful information. I have seen turnover ratio from time
to time, but it generally only counts steals as turnovers forced. It is
true that a blocked shot doesn't always result in a turnover, but it
usually does.
FOUL RATIO (fouls committed/fouls drawn): Some players are
foul-prone, but also draw a lot of fouls as well. Foul Ratio would
illustrate a player's net "fouling" performance.
PLUS/MINUS RATING: I love this stat in hockey and would like to see it
in basketball, as well. Simply keep track of the team's points scored
and allowed while each individual player is in on the floor. This could
then be compared to the team's point differential while that player is
on the bench. Over a long season, plus/minus rating might be an
excellent way to quantify the overall contribution of a player to his
team (particularly those players whose contribution is more defensive
than offensive).
Lastly, one other fun stat I would like to see tracked is the number of last-second, desperation buzzer-beater shots a player attempts, so as to be able to eliminate them from a player's field goal percentage. This is not a big deal in college basketball, but in the NBA, 82 games of four quarters apiece makes for a possible 328 buzzer-beater attempts in a team's season. It is not inconceivable that one player could attempt 30 or more such shots in a season, which could dramatically affect that player's field goal percentage. I could swear that Scottie Pippen used to try at least one buzzer-beater per game during the Bulls'
championship years. He seemed to be their designated, buzzer-beater taker.
Incidentally, two of my proposed stats (Assists-per-Missed-FG and Turnover Ratio) were featured in the STATS Basketball Scoreboard, 1994-1995. If you have access to that book you can get a glimpse of what those stats look like when they have been tabulated using real NBA data.
Tom, you're the leader in the clubhouse for the "E-mail of the Year Award." My thoughts, in order:
I prefer POINTS PER SHOT to ADJUSTED FIELD GOAL PERCENTAGE, although both do a better job of reflecting the value of the 3-pointers than any currently available statistic. I think it means more to the average fan to say, "Troy Murphy averages 1.62 points per shot" than "Troy Murphy has an adjusted field goal percentage of .685." There is more
context for one than the other. However, I would equate Adjusted Field Percentage with Slugging Percentage in baseball. It measures the basketball equivalent of "extra base hits" versus the simpler (and often misleading) "batting average." Later in the year, when we start running regular Points Per Shot leaders, I'll run a corresponding Adjusted Field
Goal Percentage list to see how they compare.
ASSISTS-PER-MISSED FIELD GOAL is interesting, and I'll certainly check out your Stats, Inc., recommendation. However, the fly in your ointment may be the assertion that "a player with the ball can only do two things: pass or shoot." From a box score standpoint, this is not correct. Said player most frequently passes to a teammate, with the ball then moving multiple times before either a field goal attempt or assist is recorded. How is this "non-statistical" event measured? Said
player could also turn the ball over to the other team, a negative occurrence no reflected in either FG attempts or assists. You're right; it's a good stat. All I'm saying is that it isn't perfect with regard to what it measures.
TURNOVER RATIO and FOUL RATIO are obscure, but fun. When national, conference-by-conference data is available, I'll run something on Turnover Ratio. Your "Foul Ratio," I think, is more fun, but one of many stats that would have to be charted by fans or on the benches by assistant coaches/managers. No one has mentioned foul ratio to me yet among the data tracked on the sidelines while a game is in progress.
Any volunteers?
PLUS/MINUS is fast becoming a "Box Scores and More" crusade. I still see no reason why the software currently in place for automated box scores and play-by-play sheets (the latter of which tracks all substitutions) could not be adjusted to add this statistic to every single player's box score line. We're gonna' keep beating this drum until we get what we want. In terms of value vs. ease of calculation, Plus/Minus is the most important thing we could add to the basketball metrics discussion.
Thanks, Tom.
I love the very fact that your column exists. I truly believe that while objective statistics can't tell the whole story, they are most
often a better indicator of what is good and bad about a team than subjective measures. Saying a team "plays great defense" is
interesting. Citing points against per game and its national ranking is convincing. And I think there's a lot of room in college basketball for this; especially with so many teams and the fact that you must somehow adjust for strength of schedule. For example, instead of points against per game, maybe a good defensive measure would be to see how low a team
holds opponents below their scoring average. Of course, you might want to adjust for tempo of game, etc. But all in all, I love the concept of the column.
That said, you haven't done any statistical analysis. Your article is full of one game performances and "the best and worst." You highlight performances with tiny sample sizes and glorify the extremes. You only point out statistical anomalies. I know there haven't been a lot of games played, but you have a history of college basketball to work
with. You can at least highlight who's had the best offense, highest point differentials and try to come up with a methodology for evaluating teams.
If you want to highlight Jim Phelan's teams' free throw shooting, do
so, but maybe also dig up how his teams have done in his career. Is
this an anomaly? Current players? Or is it Phelan? There's enough
history at Mount St. Mary's for that. If you want to highlight Kevin
Braswell's bad performances, highlight how the Hoyas would be better off
having him dump it off into the post instead of shooting so much.
Great stuff, Alan. Box Scores and More was originally conceived as an "anomalies" column, somewhat on the order of Jayson Stark's baseball Week in Review. Somewhere along the way, we decided to cross Jayson's "box score line of the week" approach with Rob Neyer's brand of sabermetric analysis.
You'd be surprised how little analytical research data is really available in college basketball (at least for now). For instance, very
few team media guides list year-by-year shooting performances (or any other year-by-year data other than leading scorers, rebounders, etc.). I'm not every sure Jim Phelan could produce the free throw percentage of every one of his teams. However, these are exactly the kinds of questions we want to ask -- and answer -- over time. Just as newspapers never used to publish on-base and slugging percentage or expanded baseball box scores, we must raise the bar for what data becomes commonplace and
expected in our sport.
For starters, we thought a review of every single box score for one season -- along with a simultaneous dialogue with serious fans -- would identify the statistical categories to pursue over the long haul. And much of the "extreme data currently presented is, in fact, based on fairly large sample sizes. To make any assertion based upon, say, 1,000 games played is a powerful statement. When the number is 10,000, which it will be, we may really have something.
All of which is not to say there will be not be serious analysis in the short term. I just LOVE, for instance, your suggestion of points
allowed measured against average points scored by the opposition. But this is a case where the sample size really is too small (at least for the current season). As soon as it looks like enough teams have played
enough games, I'll break that one down.
Way to keep us on our toes!
I think it's in poor taste that you bad-mouth players. Do you think that shooting woes are what they wished for on their 20th birthday? Try writing something useful and meaningful for a change instead of putting people down. Remember: These are kids in school and some are actually trying to get an education. Don't ever forget that!
Whoa, Nellie! To say a player is missing a lot of shots isn't saying that guy is a bad person, flunking out, committing felonies or joining the Electoral College. It may be saying that he isn't helping his team win as much as the mainstream media would have you believe. The flip side is that other players may be modest scorers who actually have a
greater impact on wins and losses.
As long as we keep score of games, and as long as we openly publish field goals, field goal attempts, steals, turnovers and points, the
basketball public is going to form opinions on the individual players behind those numbers. Over the long haul, I'd like to think this column will create MORE informed opinions, not LESS. I have never booed a college player, and I never will, but that doesn't mean we're not allowed to count.
Education is a two-way street.
In your ESPN column on stats, you stated: "UMass didn't lose to Holy Cross on Nov. 25 because 76 free throws were attempted. The Minutemen lost because high-scoring guards Monty Mack and Shannon Crooks were a combined 5-for-27 from the field. But we're guessing there would be a negative correlation between the number of in-game stoppages and the teams' overall shooting percentages."
A caveat: Unless I am mistaken, a person fouled in the act of shooting who misses the shot is credited with a miss. In this case, since most (or at least many) shooting fouls don't result in a basket, then we expect some level of negative correlation between whistles and shooting percentage, simply because of the causation.
I'm actually not really sure that this is how the stats work. Maybe it's like baseball sacrifices, which don't count as at-bats. But if
not, you'll need to be careful not to capture the direct relationship when trying to measure the indirect "breaking the flow of the offense" effect.
I'll assume this isn't Andy Katz writing, who would know better. Missed field goals attempts resulting in a shooting foul do NOT count as a miss. The player is such a situation can only be credited with a "make" (should the shot drop). It's no wonder so many guys heave funky shots when bumped. They've got nothing to lose. And our original point stands.
I've enjoyed your column each time I've read it on ESPN.com and I thought you might be the person to explore a simple question that puzzles me. Why doesn't the NCAA compile national men's basketball statistics until the middle of December? By the time its first stats come out, we're already two months into the season, especially for teams
playing in exempted tournaments (not the case at Lafayette).
As a men's hoops SID, I find it strange, to say the least, because I-A and I-AA football stats are available every week of the season. I don't think that the NCAA as a whole has a problem with limited staffing (others may suggest that just the opposite is true), but I do understand that the staffing in the stats department at the NCAA may not be
reflection of the entire organization itself.
I've thought about calling and asking straight out, but it has to be tough to say, "so why aren't you guys doing this or that?" If someone called and asked me that, I'm sure I wouldn't be thrilled. I did call, however, to find out when the first report is scheduled to be available.
It's not that I'm a big stats freak or anything (for a Sports Info guy, I would be considered only mildly interested in stats), but there are some notes that would be of interest to our media and fans that I can only speculate about. For example, we had a guy with 11 assists in our opener. Where did that rank him in the nation during the first week and
how about now? If he led the country even for a week, it's a great note
for a small school (even for a big school). Others have come up, but
there's nowhere to find out until December.
I wondered if you have ever asked or know why there aren't any stats.
Anyway, good luck with your column in future.
Philip J. LaBella
Men's Basketball SID/Publications Coordinator
Lafayette College
I'm sure the NCAA's response would have something to do with sample size and that fact that some teams have played close to 10 games by now and others only two or three. Yet that only applies to team statistics. Individual superlatives can (and should) be recognized on an ongoing basis. Ironically, in my daily box scores review, I have focuses much more on team superlatives (as they will prompt more significant analysis down the line) than on individual superlatives. Next year, I guess we'll have to do that, too (at least in the early season). Like you, I anxiously await the first "official" NCAA release, if only because our household has an all-new, late-night lingo. "Not tonight, honey, I have a box score headache!"
Joe Lunardi is a regular in-season contributor for ESPN.com. He is also
contributing editor of the Blue Ribbon College Basketball Yearbook,
www.collegebaskets.com. Write to Joe at jlunardi@home.com.
|
|
ALSO SEE
Chart wrap: Joe Lunardi
Box scores and more
Nov. 30: Email Banter
|
|
